A Novel Technology for Multiplex Gene Expression Analysis Directly from Whole Blood Samples Stabilized at Ambient Temperature Using an RNA-Stabilizing Buffer  Chang Hee Kim, Majid Abedi, Yenbou Liu, Sree Panuganti, Francisco Flores, Kevin R. Shah, Hannah Catterall, Krishna S. Morampudi, Robert Terbrueggen  The Journal of Molecular Diagnostics  Volume 17, Issue 2, Pages 118-127 (March 2015) DOI: 10.1016/j.jmoldx.2014.11.002 Copyright © 2015 American Society for Investigative Pathology and the Association for Molecular Pathology Terms and Conditions

Figure 1 CLPA measures gene expression directly from blood by the following steps: 1) stabilization of blood in DxCollect buffer; 2) the chemical ligation of S- and L-probes bound to the RNA by using TS1 and TS2; 3) the streptavidin-conjugated paramagnetic bead-capture of ligated probes that are hybridized to the RNA target by using biotinylated TC probes; 4) PCR amplification with URP and UFP in which each product is designed to have a unique size by varying the lengths of the SS1 and the SS2 domains; and 5) detection of the size-differentiated, ligated probe amplification products by using a CE instrument. B, biotin; CE, capillary electrophoresis; CLPA, chemical ligation-dependent probe amplification; L, leaving; M, paramagnetic beads; RFU, relative fluorescence unit; S, sulfur; SS, spacer sequence; TC, target capture; TS, target sequence; UFP, universal forward primer; URP, universal reverse primer. The Journal of Molecular Diagnostics 2015 17, 118-127DOI: (10.1016/j.jmoldx.2014.11.002) Copyright © 2015 American Society for Investigative Pathology and the Association for Molecular Pathology Terms and Conditions

Figure 2 RNA extracted from DxCollect blood stored at room temperature for up to 2 weeks is stable. The mean and CVs of RIN scores for RNA isolated from DxCollect and PAXgene blood from six healthy donors at different times at room temperature are depicted. Data shown are from the following time points: 1 day (T0), 7 days (T1), and 14 days after collection (T2). The percentages above the bars represent %CVs. Data are expressed as means ± SD. RIN, RNA integrity number. The Journal of Molecular Diagnostics 2015 17, 118-127DOI: (10.1016/j.jmoldx.2014.11.002) Copyright © 2015 American Society for Investigative Pathology and the Association for Molecular Pathology Terms and Conditions

Figure 3 Relative gene expression is stable through the titration of DxCollect-stabilized blood. A: Raw RFUs are responsive to the sample titration. B: Gene expression is stable when signals are normalized by the geometric mean of MRPS5 and MRPS18A. Data are expressed as means ± SD. RFU, relative fluorescence unit. The Journal of Molecular Diagnostics 2015 17, 118-127DOI: (10.1016/j.jmoldx.2014.11.002) Copyright © 2015 American Society for Investigative Pathology and the Association for Molecular Pathology Terms and Conditions

Figure 4 Gene expression measured with CLPA on DxCollect-stabilized blood samples is correlated (Spearman rank correlation) to gene expression measured with CLPA on RNA purified from matched blood samples. Sample input was 50 ng of RNA and 16.7 μL of DxCollect-stabilized whole blood. Signals were the ratios of gene-specific signal to the geometric mean of MRPS18A and MRPS5 mRNA signals. Measurements were performed on samples from three donors with five replicates per donor. A strong correlation (ρ = 0.80) with a P < 0.05 between the data points was observed. CLPA, chemical ligation-dependent probe amplification. The Journal of Molecular Diagnostics 2015 17, 118-127DOI: (10.1016/j.jmoldx.2014.11.002) Copyright © 2015 American Society for Investigative Pathology and the Association for Molecular Pathology Terms and Conditions

Figure 5 Chemical ligation-dependent probe amplification dynamic range is from 3 to 4 logs as observed by the linearity of response from 5.8 to 2.8 log10 of PCNA RNA concentrations (in amol/L) (A) and the linearity of response from 5.7 to 1.4 log10 of FDXR RNA concentrations (in amol/L) (B). The geometric normalized data for FDXR (A) and PCNA (B) are plotted as the log10 of RNA concentration versus log10 of the normalized signal. The geometric mean of MRPS18A and MRPS 5 mRNA signals were used for normalization. The limit of detection for FDXR is 50 amol/L as shown by the lowest point in B whereby the assay shows a linear response. This limit of detection is equivalent to 300 copies of RNA. The Journal of Molecular Diagnostics 2015 17, 118-127DOI: (10.1016/j.jmoldx.2014.11.002) Copyright © 2015 American Society for Investigative Pathology and the Association for Molecular Pathology Terms and Conditions

Figure 6 Large changes in the expression of a high signaling gene (GAPDH) do not affect the signaling of the other genes in the multiplex CLPA reaction. All gene signals were normalized to the geometric mean of signals for MRPS5 and MRPS18A. The CLPA test was performed on a DxCollect-stabilized blood sample with 30 cycles for the PCR step. Varying amounts of a nonamplifiable competitor oligonucleotides targeting GAPDH mRNA were added. The GAPDH mRNA signal decreases, whereas the signals from other genes remain constant. Change in signal is shown relative to the standard 66.7 times GAPDH attenuation (R2 = 0.999, slope = 0.972). Data shown are from the mean of two donors (five replicates per donor). CLPA, chemical ligation-dependent probe amplification. The Journal of Molecular Diagnostics 2015 17, 118-127DOI: (10.1016/j.jmoldx.2014.11.002) Copyright © 2015 American Society for Investigative Pathology and the Association for Molecular Pathology Terms and Conditions

Figure 7 Measurement of gene expression changes that occur on ex vivo irradiation and comparison with TaqMan RT-qPCR. Probe signals were normalized to GAPDH mRNA signals. Blood samples from six donors were either ex vivo irradiated or mock irradiated and cultured for 24 hours as described in Materials and Methods. The samples were stabilized with the addition of DxCollect buffer and either analyzed by chemical ligation-dependent probe amplification or purified to extract RNA which was analyzed by TaqMan RT-qPCR. The probes that were tested by both TaqMan RT-qPCR and CLPA show high concordance and similar variability (SD) in detecting the gene expression fold changes between the 2-Gy and 0-Gy samples. CLPA, chemical ligation-dependent probe amplification; RT-qPCR, quantitative real-time RT-PCR. The Journal of Molecular Diagnostics 2015 17, 118-127DOI: (10.1016/j.jmoldx.2014.11.002) Copyright © 2015 American Society for Investigative Pathology and the Association for Molecular Pathology Terms and Conditions