Abstract
Introduction
Peripheral venous blood is a preferred biospecimen for research as one can obtain genomic, cellular, and soluble components from different cell types (i.e., leukocytes) and a multitude of soluble factors including components of the clotting cascade, cytokines, and other immune factors, as well as a variety of other secreted proteins. Peripheral blood is also a valuable specimen source for biobanking in that it minimizes the number of biospecimens that need to be collected from an individual subject to obtain a large number and variety of banked biosamples.
There are various processing methods available for separation of whole blood into its components. Such methods include depletion of red cells from the leukocyte fraction as well as enrichment of peripheral blood mononuclear cells (PBMCs) from whole blood. In addition, one can isolate plasma (the liquid part of blood) from either process as a source of soluble proteins and factors. Originally, such methods were manual in nature (e.g., construction of density gradients), time consuming, and frequently expensive. Furthermore, outcomes from such methods were often practitioner dependent, with more experienced personnel obtaining superior results to those found with novice applicants (as might be expected when harvesting cells from gradient interfaces).
As demand has increased for the use of blood specimens in research and biobanking, a variety of commercially available techniques1–4 are now available that essentially eliminate outcomes based on user experience with greater reproducibility, higher throughput, and less expense. However, the effects of these different processing methods on plasma and PBMC recovery are largely unknown.
In this study, we have directly compared the technical performance of each processing approach in a side-by-side manner using the same blood biospecimen obtained from multiple subjects processed with three of the most commonly used methodologies: the cell preparation tube (CPT) vacutainer, the SepMate™ centrifuge tube, and the Leucosep™ centrifugation device. We evaluated cell viability, nucleated cell recovery, cell differentials, and plasma protein recovery after each approach for each subject sample. It was observed that there were significant differences between the techniques in terms of cell differentials and plasma protein recovery, but not viability. Thus, care must be taken in terms of the methodological approach utilized for blood processing as it relates to downstream use of the biospecimen.
Materials and Methods
Collection of peripheral blood
Peripheral blood (50–60 cc) was collected from all subjects (all donors were healthy males and females between 20 and 40 years of age; N = 6–8) by phlebotomy of a peripheral vein using 10 cc heparinized vacutainers. The blood from each individual donor was pooled and then divided into three portions for processing by three methods: a CPT vacutainer (Becton Dickinson), a SepMate-50 (StemCell Technologies) device, and a Leucosep-50 (Grainier) device.
All samples were obtained with written consent from the donors according to the instructions from the local institutional review board at the University of Arizona.
Isolation of human PBMCs using CPT vacutainers
Blood was processed according to the manufacturer's instructions. 5 In brief, blood was centrifuged in the CPT vacutainer for 15 minutes at 1500 relative centrifugal force (800-1200 × g) (rcf). Plasma (1 mL each) was placed into labeled CryoVials (three aliquots) and frozen at −80°C immediately. The buffy coat (on top of the polyester gel inside the vacutainer) was collected into 15 mL tubes and centrifuged at 500 rcf for 10 minutes, and then washed with isotonic saline.
Cells were then counted using a Cellometer® Auto 2000 (1:1 dilution with acridine orange/propidium iodide [AO/PI]) in preparation for cell differentials and cryopreservation. Some cells were resuspended in 2 mL ice-cold dimethyl sulfoxide (DMSO) cryomedia and aliquoted into CryoVials (no <10 million cells per vial). Cells were stored in a liquid nitrogen (LN2) dewar as quickly as possible after freezing using a controlled-rate freezer.
Isolation of human PBMCs using the SepMate-50
Room temperature (15 mL) Sigma Histopaque®-1077 medium was carefully added into a SepMate-50 tube by carefully pipetting it through the central hole of the SepMate-50 insert. 6 Whole blood diluted 1:1 with Hank's balanced salt solution (HBSS) with no calcium and magnesium was transferred into the 50-mL conical centrifuge tube. The diluted blood sample was layered by carefully pipetting it down along the side of the tube and onto the Histopaque-1077 layer.
The blood was then centrifuged at 1100 × g (1800 RPM) for 20 minutes at room temperature (25°C), with the brake turned on. The top layer (interface) containing the plasma and enriched mononuclear cells (MNCs) was transferred into a fresh 50-mL conical tube, after first removing the plasma. The enriched MNCs were washed with HBSS. The Cellometer Automate 2000 was then used for cell counting. Cell and plasma aliquots were cryopreserved as already described.
Isolation of human PBMCs using the Leucosep-50
Room temperature (15 mL) density gradient medium (Sigma Histopaque-1077) was added into a Leucosep-50 tube. 7 The Leucosep tube was then centrifuged to move the density gradient medium to the bottom of the tube, below the filter. Whole blood was transferred into the Leucosep tube without dilution. The tube was then centrifuged at 1200 × g for 10 minutes at room temperature (25°C), with the brake turned on. The top layer (interface) that contained the plasma and enriched MNCs was poured off into a fresh 50 mL conical tube.
The Leucosep-50 tube was washed with 20 mL of phosphate buffered saline (PBS) by pipetting along the inside walls of the tube. The cumulative cell suspension was centrifuged at 500 × g for 5 minutes at room temperature. The cell pellet was resuspended in 5–10 mL of PBS. Cell counts were performed using the Cellometer Automate 2000. Cell and plasma aliquots were cryopreserved as already mentioned.
Analytical parameters
Before and after processing the peripheral blood samples by each of the methods, the biospecimens were analyzed for cell viability, total nucleated cell (TNC) counts, and cell differentials (by use of a Cellometer Automate 2000 and a Beckman-Coulter AcT diff 2 hematology analyzer). In addition, plasma obtained by each processing method was analyzed for total protein content using the Nanodrop-2000 analyzer.
Cryopreservation and thawing of PBMCs
Isolated PBMCs (10 × 106) were cryopreserved in 1.2 mL of Hyclone cryofreeze media (10% DMSO final) using a controlled rate freezer (programmed to freeze at 1°C/minute) and stored in the vapor phase of a LN2 dewar (−198°C) until thawed for analysis.
Cryopreserved PBMCs were thawed using a “cold–thaw” technique. The cold–thaw method involved placing thawed cells in cold (4°C) complete medium with 10% fetal bovine serum after rapid thawing with agitation at 37°C. Cell populations were then washed and resuspended in complete medium for analysis.
Thawed PBMCs were analyzed based on the processing method before freezing, and after thawing for cell viability, total number of cells, and cell differentials.
Operation of the Cellometer cell counter (by Nexcelom)
Cell counts and viability were determined using 20 μL of cells (1 × 106) and 20 μL of AO/PI solution provided by the manufacturer. The mixed AO/PI and cell suspension were then loaded into the counting chamber for analysis. Data were collected for cell viability, number of cells, and morphology.
Statistical analysis
Statistical analyses were performed using analysis of variance (ANOVA) with the SPSS program to analyze the effect of the PBMC separation method on individual outcome measures. p-values <0.05 were considered statistically significant.
Results
Peripheral blood was collected from healthy volunteer subjects and equally divided between each of the three processing methods (CPT vacutainer, SepMate, and Leucosep). Before and after processing the blood according to the manufacturer's instructions, the end product was evaluated for cell viability, recovery of TNC count, cell differentials, and total plasma protein content.
The cell viability (Table 1) and the TNC recovery for each of the subject's biospecimens were not significantly affected by any of the different processing methods (p > 0.05). However, the percentages of cell types (MNCs vs. polymorphonuclear cells) observed with the CPT vacutainer method differed significantly when compared with the other two methods (Table 1). The percentage of lymphocytes and monocytes (MNCs) recovered from CPT vacutainers (83.1%) was significantly less as compared with the SepMate and Leucosep methods (95.3% and 93.8%, respectively). This difference was significant by ANOVA statistical analysis (p = 0.0042).
Effect of Processing Method on Peripheral Blood Mononuclear Cell Recovery
Peripheral blood was harvested as described and processed by each of the indicated methods. N: number of independent experiments (six donors were used for viability, TNC recovery, and plasma protein measurements, whereas eight donors were used for the MNC and PMN recovery experiments). Viability was determined by AO/PI staining using a Cellometer. TNC recovery was determined by AO/PI staining using a Cellometer. Pre-TNC and post-TNC values are expressed as × 106. Cell differentials were determined using a hematology analyzer and recovery (% Lymph+Monos [MNC] vs. PMN) is expressed as a percentage. Total protein content in the recovered plasma was measured using the Nanodrop-2000.
p < 0.0042.
p < 0.0048.
p < 0.05.
AO/PI, acridine orange/propidium iodide; CPT, cell preparation tube; MNC, mononuclear cell; PMN, neutrophils; TNC, total nucleated cell.
Cells recovered from processing with a CPT vacutainer contained a much higher percentage of neutrophils than those blood samples processed by the SepMate or Leucosep methods; an average of 15.6% neutrophils in the final cell suspension as compared with the 4.4% from the SepMate method; and the 5.6% of neutrophils from the Leucosep method (N = 8), with a p-value of 0.0048 (Table 1).
An analysis of plasma obtained after processing of whole blood from each of four independent donors (N = 6) was performed (Table 1). It was observed that the total protein content from samples processed with the CPT processing method was much lower than plasma obtained with the other two processing methods (56.4 mg/mL; below the reported range of 60–80 mg/mL for plasma 8 ). The SepMate processing method showed a plasma total protein level nearly double the published range (135 mg/mL), but Leucosep processing had an average plasma total protein content of 77.8 mg/mL, within the published range. 8
Discussion
Research laboratories as well as academic biobanks depend on a reliable and reproducible means to obtain representative cellular samples from peripheral blood donors. Often the same biospecimen is used as a source of viable cells, plasma, and genomic material. Originally peripheral blood leukocytes were obtained by diluting blood specimens 1:1 with PBS or isotonic saline and then carefully layering the samples on top of a Lymphocyte Separation Media density gradient (e.g., LSM 1077). After centrifugation to remove red cells and a variable percentage of neutrophils from the sample, the enriched PBMCs (primarily MNCs) were delicately harvested from the gradient interface.
However, this approach was user dependent, being significantly affected by the user's length of experience and manual dexterity. In an effort to make this process more reproducible and less dependent on the length of the user's experience to obtain a satisfactory outcome, various commercial products have been introduced over the past several years, including CPT vacutainers, SepMate, and Leucosep 50 mL tubes. Each of these devices involves an in-place gel or filter to separate different cellular compartments during centrifugation. The SepMate and Leucosep devices also employ density gradient medium to further enrich for specialized subpopulations such as MNCs.
At reasonable costs, one only needs to load the blood (diluted or not) into the device, centrifuge the samples for a short period of time, and then pour off the end product (plasma and cells), eliminating buffy coat harvesting and its associated technical requirements. In addition, each of these methods is amenable to high-throughput processing and biobanking of large numbers of samples per day that minimizes labor costs.
Although a priori one might hypothesize that each of the methods would produce comparable and suitable outcomes, to our knowledge we have not seen a direct side-by-side comparison of these three approaches.1–4 Most studies have examined TNC recovery but neither cell differentials nor plasma protein recovery has been reported. A comparison has been made by other investigators of the CPT versus Ficoll and SepMate methods, and the CPT versus Lymphoprep methods1–4 but not a side-by-side examination of the three methods used in this study.
Initial analysis showed that all three blood processing methods seemed comparable, although the CPT vacutainer processing method retained more granulocytes (neutrophils) than the other two methods. The SepMate and Leucosep processing methods isolated primarily lymphocytes and monocytes (i.e., MNCs). This finding may be an important consideration when the number of cells to be aliquoted for cryopreservation is critical for downstream assays, as freezing of neutrophils may reduce the overall viability of the sample and the number of usable cells recovered after thawing. Neutrophils are very difficult to freeze and thaw with high viability at the end of the process, not to mention the intracellular granule enzymes that can be released, and that may damage bystander cells.9,10
In addition, for the purposes of plasma isolation, the CPT and Leucosep or SepMate method was expected to have been comparable in terms of total protein content, yet the CPT method produced consistently lower total protein levels than the Leucosep or SepMate method, even though the blood in the CPT vacutainer was not diluted before centrifugation. It may be that part of the sample (i.e., protein) is being retained by the gel insert but that speculation awaits further investigation.
Interestingly, the Leucosep method also uses undiluted blood for its cellular isolation, and it also results in a lower protein concentration than does the same blood processed with the SepMate device using diluted blood, although within published parameters. This observation could impact the ability to assay plasma obtained in this manner for its contents such as cytokines or scarce proteins and may even provide an underestimate of the overall plasma concentration.
In conclusion, three processing methods in common use to obtain cellular and soluble components from whole peripheral blood in a high-throughput manner are comparable, but not identical. All the methods are economical (although the CPT vacutainers are the most expensive), easily producing a large number of viable cells in a timely manner, with significant advantages over manual approaches. However, only the CPT vacutainer approach produces what could be considered a true buffy coat with an end product representative of all leukocyte populations normally found in whole blood. The other two approaches produce a highly enriched mononuclear population, ideal for immunological studies.
This difference must be considered when choosing an approach based on downstream applications. Similar considerations must also be kept in mind when isolating plasma for downstream analysis of proteins due to dilution of the plasma. For example, blood processed with the SepMate and Leucosep approaches produces plasma most similar to plasma found with whole blood that had been simply centrifuged to remove its cellular components, whereas the protein content of plasma obtained after CPT processing is significantly reduced in protein content and is outside reported parameters, even though all methods produced comparable MNC recovery.
Thus, SepMate and Leucosep plasma might be ideal for the measure of plasma cytokines by enzyme-linked immunoassay, whereas CPT plasma might underestimate the concentration of proteins found in the biospecimen. If plasma recovery is essential at the same time as MNC recovery, it might be best to simply collect an additional blood sample dedicated for this purpose, if feasible.
Footnotes
Acknowledgments
The authors thank the willing volunteers who participated in this study.
Author Disclosure Statement
The authors have no conflict of interests of any kind to declare.
Funding Information
The study was not grant funded.