Impact of Time From Blood Draw to Peripheral Blood Mononuclear Cell (PBMC) Processing and Cryopreservation on Functional Pathway Activity As Measured by Single Cell Network Profiling (SCNP) Assays

Abstract 4922

Cryopreserved peripheral blood mononuclear cells (PBMCs) are routinely used in biomarker development studies. Mutiple pre-analytic parameters related to blood draw, processing, and cryopreservation can impact the quality of PBMC samples used in functional assays. Single cell network profiling (SCNP) is a multi-parameteric flow cytometry based approach that measures intracellular signaling activity in response to extracellular modulators. Preservation of cell viability and functionality is therefore critical to the performance of the SCNP assay. In other immunological assays, such as the ELISpot assay, the length of time from blood draw to PBMC cryopreservation has been shown to be a critical parameter affecting assay performance. In this study, the effect of time from sample collection to cryopreservation on functional pathway activation was assessed by comparing SCNP assay readouts in paired PBMC samples processed within 8 or 32 hrs from blood draw.

40 mLs of peripheral blood was obtained for 20 donors (10 male/10 female, 60–83 yrs) at the Stanford Blood Center. Half of the sample volume from each donor was processed within 8 hrs of blood draw [Day 1 (D1)], and the remainder left at 25C overnight [Day 2 (D2)]. For D2 samples, PBMC isolation and cryopreservation was initiated 24 hrs from the processing start time of the corresponding D1 sample. For the SCNP assay, samples were thawed, modulated for 15 mins with 12 immunomodulatory stimuli (interferons, interleukins, TLR ligands, etc.), fixed, and permeabilized. Permeabilized cells were stained with fluorochrome-conjugated antibodies recognizing cell surface markers or intracellular signaling molecules (pStat1, pStat3, pStat5, pS6, pNFκB, pAkt, and pErk). 38 signaling nodes (readouts of modulated signaling) were measured in 7 distinct immune cell subsets (monocytes, B cells, NK cells, naïve/memory helper T cells, and naïve/memory cytotoxic T cells).

Analysis of paired PBMC samples revealed that D1 and D2 samples had no significant difference in the percentage of healthy cells (measured by the percentage of cleaved PARP⁻ cells) and no difference in subpopulation frequencies (as a percentage of parent populations) for the majority of the 7 subsets examined. A numerically small but statistically significant decrease in the percentage of healthy cells in D2 compared to D1 samples was observed for B cells, NK cells, and naïve helper T cells (mean difference 5.6%, 5.8%, and 3.2% respectively, p<0.05) while the monocyte subset was the only one to show a significant decrease (9.4%, p<0.05) in frequency (as a percentage of parent) on D2. Similar intracellular signaling pathway modulation responses were observed for D1 and D2 samples (Fig. 1A), although the majority of nodes displayed lower modulated responses in D2 samples (10.0% mean decrease between D1 and D2). Importantly, a good correlation (Spearman r>0.5) between D1 and D2 was observed for the majority (63%) of responsive signaling nodes. Within each dataset, inter-node correlation coefficients were calculated to generate immune signaling network maps. Comparing these maps showed good agreement between the correlations measured within each dataset [mean difference of −0.01 between inter-node correlations across days (D1 mean correlation 0.21, D2 mean correlation 0.20)] demonstrating biological consistency between the 2 datasets in the structure of the immune signaling network. Further, 13 age-associated differences (p<0.05) in immune signaling responses were identified in the D1 dataset and the majority of these remained significant in the D2 dataset (p<0.05). For example, several cytokine signaling responses within naïve cytotoxic T cells had a significant decrease with age in both datasets (Fig. 1B).

Fig. 1.

View largeDownload slide

A) Signaling ranges for nodes within naïve cytotoxic T cells for D1 (blue boxplots) and D2 samples (green boxplots). B) Cytokine signaling responses within the naïve cytototixic T subset with significant age-associations in both datasets.

Fig. 1.

View largeDownload slide

A) Signaling ranges for nodes within naïve cytotoxic T cells for D1 (blue boxplots) and D2 samples (green boxplots). B) Cytokine signaling responses within the naïve cytototixic T subset with significant age-associations in both datasets.

Close modal

These results demonstrate that blood samples processed the day following blood draw provide meaningful information on functional pathway activation using the SCNP assay and support the identification of statistically significant associations with clinical variables such as age. This is a critical observation since, in the clinical setting, overnight shipping of patient samples to the lab performing the test may be required.