This protocol details a high-throughput automation-compatible method to isolate human peripheral blood mononuclear cells for biobanking and other purposes.
Peripheral blood mononuclear cells (PBMCs) are a heterogeneous population of monocytes and lymphocytes. Cryopreserved PBMCs have stable viability in long-term storage, making them an ideal cell type for many downstream research purposes, including flow cytometry, immunoassays, and genome sequencing. Typically, PBMCs are isolated via density gradient centrifugation, however, it is a low-throughput workflow that is difficult and costly to scale. This article presents a high-throughput workflow using a magnetic bead-based PBMC isolation method that is quick to implement. Total cell concentration, viability, and population distribution with PBMCs obtained using density gradient isolation were compared, and cell viability and proportion of cell types were comparable for both techniques. Isolated PBMCs demonstrated over 70% viability up to 9 days after blood collection, although yield decreased by half after 5 days compared to PBMCs processed within 24 h of collection. In summary, this article describes a PBMC protocol that utilizes a bead-based approach to adapt to a high throughput workflow and demonstrates that both manual and automated bead-based methods can increase processing capacity and provide flexibility for various budgets.
Peripheral blood mononuclear cell (PBMC) isolation is a technique that separates and isolates lymphocytes and monocytes from other whole blood constituents. PBMCs are a versatile cell type used for numerous applications including, but not limited to, immunotherapy, vaccine development, target or biomarker identification, and antibody/small molecule drug development1,2. These cells may be isolated from healthy or diseased individuals and can be used immediately in downstream processes or cryopreserved for future research3. In some instances, the downstream purpose is known, while in others, as is common in biobanking, PBMCs are isolated and stored for future unspecified applications4.
Density gradient centrifugation is the traditional technique for isolating PBMCs5,6,7 from whole blood, utilizing the differential separation of the constituent cell types based on cell density during centrifugation. While there may be some variation in this method, whole blood is generally diluted with phosphate-buffered saline (PBS), layered over a density gradient medium in a specialized or standard centrifuge tube, and then spun. Four distinct layers result: the top plasma layer is enriched with platelets, a thin PBMC layer is above the density gradient medium, and finally, the bottom layer consists of red blood cells (RBCs) and granulocytes. Although this method has previously been termed 'the gold standard'8, there are limitations for scale-up, such as lengthy processing time, centrifuge capacity, difficulty in aliquoting other blood products (i.e., plasma and RBCs), and being laborious to automate. While automation is possible for this method9, it does require comprehensive programming of a liquid handler (with a centrifugation module to be fully automated) and would remain a lengthy process.
Henceforth, an alternate workflow that uses immunomagnetic bead separation with either an eight-holder magnet for manual processing or an instrument for fully automated processing is presented. This method utilizes an antibody cocktail that is added to cells and binds unwanted cell populations, in this case, platelets, granulocytes, and RBCs. These unwanted populations are subsequently removed by magnetic separation, leaving monocyte and lymphocyte populations in the negative fraction that are ready for downstream processing10. This negative selection method is faster than positive selection methods which require additional steps to remove the antibody and magnetic bead complex from the PBMCs. Negative selection is additionally advantageous as it has been described as a way to preserve cell functionality11,12.
Quality control blood specimens and in-house generated data (such as cell count, viability and age of specimen at processing) were obtained from consented research studies at the NSW Health Statewide Biobank, Royal Prince Alfred Hospital (HREC approval: 2019/ETH06776). Processed (i.e., plasma depleted), unscreened, and de-identified adult human blood samples were collected in EDTA tubes. These quality control blood specimens were processed less than 72 h after collection and were used for PBMC isolation experiments comparing the density gradient and bead-based methods. For the density gradient separation method used in the representative results, see the procedure in Supplementary File 1. The details of the reagents and the equipment used for this study are listed in the Table of Materials.
1. Whole blood preparation
2. Buffy coat collection
3. PBMC purification – Manual method
NOTE: Up to 8 samples per magnet holder can be processed by a single operator.
4. PBMC purification – Automated method
NOTE: Up to 4 samples can be processed by a single automated PBMC instrument. Multiple instruments can be run in parallel by a single operator.
5. Cell counting
6. Cryopreservation preparation
7. Statistical data analysis
The proportions of lymphocytes, monocytes, neutrophils, eosinophils, and basophils pre- (i.e., in whole blood) and post-PBMC isolation were measured when PBMCs were isolated using the bead-based or density gradient separation method. Proportions of lymphocytes and monocytes were significantly enriched by both methods (Figure 2A,B). Additionally, proportions of neutrophils, eosinophils, and basophils (i.e., granulocytes) were significantly decreased in PBMCs isolated by the bead-based method (Figure 2C–F). There were no significant differences in the proportions of lymphocytes, monocytes, neutrophils, eosinophils, basophils, and granulocytes between the density gradient and bead-based methods (Figure 2). Percentage recovery was additionally calculated for the cell types listed above (see Supplementary Figure 1). Mean cell viabilities were over 95% for both methods, and were not significantly different (Figure 3A). PBMCs were also counted using a size range (diameter) of 8 µm to 50 µm, where the density gradient separation method produced approximately two times higher total cell counts (Figure 3B).
Next, the automated bead-based method was compared to the manual method. No significant differences were identified for cell viability (Figure 4A) or total cell counts (Figure 4B) using the manual or automated method. The time to process these 8 samples was compared, including hands-off time requiring no human intervention. The total processing time of 8 samples was 43 min vs. 57 min for the manual vs. the automated method. The automated method included 35 min of hands-off processing (Figure 4C).
The bead-based workflow of PBMC isolation (with whole blood and plasma aliquoting) was piloted over 9 months, whereby 1410 PBMC samples were isolated from unscreened human blood using the manual platform for the first 820 participants and then using the automated method for the next 590 participants. Processing was performed either on the same day or up to 10 days post-collection as per the study acceptance criteria, with delays in processing largely due to sample transit times. With either the manual or automated method, PBMCs processed within 24 h of the collection had the highest viability and recovery (Figure 5). Mean cell viability was >90% for PBMCs processed within 5 days and >70% for PBMCs processed within 10 days (Figure 5A). Mean PBMC yields decreased by 50% after 5 days compared to specimens processed within 24 h (Figure 5B).
Figure 1: Schematic workflow of the bead-based PBMC isolation protocol from buffy coat. Please click here to view a larger version of this figure.
Figure 2: Proportions of lymphocytes, monocytes, neutrophils, eosinophils, basophils, and granulocytes in whole blood pre- and post-PBMC isolation by density gradient and bead-based methods. Percentage of lymphocytes (A), monocytes (B), neutrophils (C), eosinophils (D), basophils (E), and granulocytes (F) in matched whole blood pre-PBMC isolation (squares) and in PBMCs post isolation via the density gradient (open circles) or bead-based (closed circles) method (n = 8). ns = not significant, and *p < 0.05 as determined by One-Way ANOVA (A and B) or Friedman's test (C–F). Please click here to view a larger version of this figure.
Figure 3: Viability and total cell count of PBMC isolated by density gradient and bead-based methods. (A) Mean (± SD) cell viability for samples processed using density gradient (squares, dark bar) vs. bead-based (circles, open bar) methods for 8 paired samples. ns = not significant by paired t-test. (B) Total cell count for samples processed using density gradient (squares, dark bar) vs. bead-based (circles, open bar) methods for 8 paired samples. *p < 0.05 determined using a paired t-test. Please click here to view a larger version of this figure.
Figure 4: Comparison of manual and automated bead-based methods. (A) Mean (± SD) cell viability for samples processed using manual (squares, dark bar) vs. automated (circles, open bar) bead-based methods for 8 paired samples. ns = not significant by Mann-Whitney test. (B) Total cell count for samples processed using manual (squares) vs. automated (circles) bead-based methods for 8 paired samples. ns = not significant by paired t-test. (C) Processing time for 8 samples processed using manual vs. automated methods, including hands-on (light bar) and hands-off time periods (dark bar). Please click here to view a larger version of this figure.
Figure 5: Viability and yield of PBMC isolated from the bead-based method either manually or using automation according to the time between sample collection and PBMC storage. (A) Mean (± 95% CI) cell viability for participant samples processed using manual (dark bar) vs. automated (open bar) bead-based methods. Table detailing participant numbers. ns = not significant. na = not applicable. *p < 0.05 by Tukey's multiple comparison's test. (B) Total cell count for participant samples processed using manual (dark bar) vs. automated (open bar) bead-based methods. Table detailing participant numbers. ns = not significant. na = not applicable. *p < 0.05 by Tukey's multiple comparison's test. No value for Automated Day 10 due to the replicate number being lower than 3. Please click here to view a larger version of this figure.
Supplementary Figure 1: Figure of differential cell percentage recovery with the density gradient and bead-based methods. The percentage was calculated using absolute cell counts pre-isolation in whole blood and post-isolation cell suspension using the density gradient method (squares, dark bar) and bead-based method (circles, open bar) (n = 8). ns = not significant, and *p < 0.05 as determined by unpaired t-test. Please click here to download this File.
Supplementary File 1: Density gradient protocol. Please click here to download this File.
Supplementary File 2: Preparation of 0.1 M EDTA, pH 8.0 in PBS. Please click here to download this File.
Supplementary File 3: Detailed procedures performed to generate the representative data. Please click here to download this File.
Human PBMCs are versatile cell types used for numerous assays; however, isolation throughput is often a limitation in many laboratories, including biobanks16. Previously, the NSW Health Statewide Biobank isolated PBMCs using the density gradient method. Automation was investigated for the density gradient separation method to increase processing capacity, but barriers to implementation were identified, including (i) the cost of a liquid handler with a centrifugation unit to be fully automated, with the added requirement for a HEPA unit to produce a sterile product, (ii) trained personnel for programming, and (iii) time required for protocol testing. Therefore, this study explored alternative methods and identified the commercially obtained Human PBMC kit that could be used for manual and automated processing. Sterility is ensured as the equipment required for processing fits within a standard (length of 1.2 m) biological safety cabinet. This paper details modifications made to the manufacturer's recommended protocol14, to increase efficiency and lower reagent costs without sacrificing quality. Additionally, the manufacturer's protocol was extended to detail steps to biobank specimens for future research, including whole blood (step 1.2), plasma (step 2.2), and RBC (step 2.5) aliquoting from the original blood tube, as well as cell counting and cryopreservation.
In this study, three PBMC isolation protocols were compared: density gradient separation, manual, and automated bead-based isolation. Modifications to the manufacturer's manual bead-based protocol were made, including the removal of buffy coat dilution prior to PBMC isolation, and elimination of the requirement for break OFF for centrifugations, allowing investigators to follow an adaptable, cost-effective, and high throughput PBMC isolation protocol. First, eight matched whole blood samples were used to isolate PBMCs to compare the density gradient separation and manual bead-based technique. Importantly, cell population distributions, cell viability, and recovery of PBMCs were not significantly different between the two methods compared, as shown in Figure 2A-F, Figure 3A, and Supplementary Figure 1, respectively. In the representative data, cell counts were higher for the density gradient separation method using the trypan blue exclusion method but not when a hematology cell analyzer was used. The PBMC cell type settings on the cell counter employed a cell diameter range of 8-50 µm, and therefore, counts will include PBMCs as well as granulocytes (approximately 12-15 µm in diameter) when using the trypan blue exclusion method17. While the hematological cell counter offered higher specificity than the trypan blue exclusion method, some recovery calculations were higher than 100%, reflecting the instrument's margin of error (see Supplementary Figure 1). It is, therefore, recommended that investigators apply a combination of cell counting techniques when comparing the yields from PBMC isolation protocols, as most techniques do not offer both specific and highly sensitive differential cell counts. Further, assays were not performed to compare the functional activity of PBMCs yielded from either technique, which is a limitation of our analysis.
Next, the manual and automated bead-based methods were compared, and no significant differences between PBMC yield or viability were identified from 8 matched samples (Figure 4A,B). Cell populations were not individually compared, as the same antibody isolation cocktail was used for both methods. Importantly, the hands-on time to process 8 samples was reduced from 43 min to 22 min using the automated protocol (Figure 4C). While throughput, prevention of technician burnout, and consistency of specimen processing are ensured using automation, the cost of reagents and consumables are significantly higher at 3-4 fold more than the manual bead-based method. This is after making modifications to the manufacture's protocol to use 1 mL of buffy coat (from a whole blood volume of 10 mL) instead of the recommended range of 2-5 mL (from a whole blood volume of 10 mL and/or greater). If budget is a limitation, opting for the manual method can reduce processing time by ~25% while maintaining reagent and consumable costs comparable to those of the density separation method. It is recommended to process no more than 8 samples at a time per technician to ensure adequate staggering of samples (~30 s/sample) within the 5 min incubation periods (steps 3.7, 3.11, and 3.14).
In both the manual and automated bead-based methods, the removal of the buffy coat is a critical step to ensure optimal PBMC isolation. It is important to note that a buffy coat rather than whole blood is used in this protocol as the volume of reagents are based on starting material volumes10. Effectively removing the entire buffy coat volume can be technically challenging. Initially, this method detailed removing 0.5 mL buffy coat, however it was increased to 1 mL to improve recovery. To ensure adequate and consistent recovery of buffy coat, detailing this process is important in protocol documentation and training. It is recommended to swirl a pipette tip whilst aspirating the buffy coat while taking care not to aspirate too many RBCs from the layer below (step 2.3). It is important not to saturate the antibody cocktail that binds unwanted cells (i.e., granulocytes and red blood cells), which can impact yield and purity10. To minimize RBCs collected during the buffy coat extraction, the pipette tip should be inbetween the plasma and buffy coat layer. The volume of the buffy coat can be increased from 1 mL; however, care must be taken as described above to ensure that no more than 10% of the collected volume contains RBCs. Alternatively, an automated liquid handler can be used to collect buffy coats consistently18. However, the dedicated hours required to calibrate and troubleshoot liquid handling instrument protocols, especially considering the expense, may not be feasible for most laboratories.
The transition and application of the automation bead-based protocol was essential given the NSW Health Statewide Biobank's aim to process 23,000 PBMCs over the next 3 years. Here, it was demonstrated that PBMCs can be isolated from ACD tubes up to 9 days from collection with mean viabilities >70%. While yields were optimal 24 h post-collection, processing within this timeframe is not always feasible as samples may need to be transported. It was shown that PBMCs isolated with either the manual or automated bead-based method may have yields of >3 x 105 cells/mL whole blood when isolated within 4 days of collection and >1 x 105 cells/mL whole blood when isolated within 10 days of collection. Low numbers for samples over 5 days are noted as a limitation of this analysis. Further, data was not separated based on participant ages, genders, and clinical history, as this information was not made available. Although a comparative analysis is needed to examine the effects of delays in processing times for both methods, it has previously been reported that delays in PBMC isolation using the density gradient method decrease cell quality and significantly increase RBC contamination19,20. In addition, granulocyte proportions increase if processing is delayed, and hence, specimens of similar "ages" should be grouped for downstream analyses20,21,22. Of note, this experiment was performed using blood anticoagulated with acid citric dextrose (i.e., trisodium citrate, citric acid, and dextrose); however, yield and/or proportion of cell types may vary if other anticoagulants are used23; therefore investigators are advised to choose an appropriate anticoagulant based on the intended downstream PBMC analyses.
In summary, a protocol for PBMC isolation using magnetic beads that is adaptable to high throughput workflows is detailed to fulfill the requirements for scaling without compromising cell viability. Both the manual and automated methods can be optimized to produce specific cell concentrations by changing starting and resuspension volumes. The NSW Health Statewide Biobank has transitioned from extracting ~60 PBMCs per month using the traditional density gradient separation technique to ~300 PBMCs per month using this automation-compatible bead-based method. The authors' next aim is to use the automated platform to process up to 1200 PBMCs per month and further compare PBMCs isolated by both magnetic bead-based (manual and automated) and density gradient techniques to guide the implementation of this technique to other laboratories with a particular focus on biobanks.
The authors have nothing to disclose.
The NSW Health Statewide Biobank gratefully acknowledges support from NSW Health Pathology, the NSW Office for Health and Medical Research, and the Sydney Local Health District. Additionally, the authors gratefully acknowledge Omico and other research studies supported by the NSW Health Statewide Biobank for granting permission to publish in-house generated data and to use unused specimens for research purposes. The authors thank Professor Jennifer Bryne (NSW Health Pathology, University of Sydney) for critical leadership and discussions. Figure 1 was created with BioRender.com.
Cell cryopreservation media CS10, 100 mL (CRYOSTOR) | StemCellTM Technologies | 07930 | |
Class II Biological Safety Cabinet | Thermo ScientififcTM | 51033311 | |
CoolCell 1 mL FX | BioTools | BCS-407P | This is the control rate freezing container used. |
Distilled Water | Bacto Laboratories | 561832 | |
DxH 500 Hematology Analyzer | Beckman Coulter Life Sciences | B40601 | Referred to as external automated cell counter. |
EasyEightsTM EasySepTM Magnet | StemCellTM Technologies | 18103 | |
EasySepTM Direct Human PBMC Isolation Kit | StemCellTM Technologies | 19654 | Kit includes the magnetic bead tube and the cocktail mix tube |
Ethylenediaminetetraacetic acid | Sigma-Aldrich Pty Ltd | E6758-500G | Instructions to make 0.1M EDTA solution from EDTA salt is located in supplemental file 2. |
LymphoprepTM Density Gradient Medium | StemCellTM Technologies | 7851 | |
Megafuge ST4 Plus Centrifuge | Thermo ScientififcTM | THR75009903 | |
Orion Star A211 pH meter electrode | Thermo ScientififcTM | STARA2110 | |
Orion™ ROSS Ultra™ Glass Triode™ pH/ATC Combination Electrodes | Thermo ScientififcTM | 8302BNURCA | |
Phosphate buffered saline (PBS), solution, 1X, 500ml | Life Technologies Australia Pty Ltd | 10010023 | |
Prism | GraphPad | ||
RoboSepTM Buffer 1X | StemCellTM Technologies | 20104 | Software used for statistical analysis. |
RoboSepTM-S | StemCellTM Technologies | 21000 | Fully automated cell separator instrument. |
RoboSep™ Filter Tips | StemCellTM Technologies | 20125 | |
SepMateTM-50 (IVD) tubes | StemCellTM Technologies | 85460 | IVD – In vitro diagnostics. Also known as SepMateTM-50 tubes |
Vi-CELL XR Cell Anlayzer | Beckman Coulter Life Sciences | Internal automated cell counter. Instrument obsolete and no longer available for purchase (as of December 31, 2022). Alternative instrument is the ViCell BLU Cell Viability Analyzer (Product no. C19196). | |
Vi-CELL XR Quad Pack Reagent Kit | Beckman Coulter Life Sciences | 383722 |
.