Microfluidic leukocyte isolation

Leukocyte isolation for microarray transcriptional profiling

A primary goal of the Program was to determine the human response to severe trauma or burn injury by transcriptome analysis of circulating blood leukocytes.  Sampling of blood from hospitalized patients is the most common method to obtain peripheral blood leukocytes for clinical care and clinical research. Blood is the compartment of choice for clinical diagnostics for a number of reasons.  Blood is easily accessible and relatively abundant for repeated sampling over time, and blood sampling is a routine procedure readily accepted by the patients themselves and family members who consent for the patient’s participation in the research study.  A blood sample is the essential source to acquire special subtypes of leukocytes whose concentrations can be enriched many times greater than usual for studying inflammation biology and immune responsiveness. Three specific leukocyte types that play key roles in the human body’s inflammatory and immune responses were studied in the Program and these include neutrophils, T lymphocytes (T-cells), and monocytes.

Early on in the Program, the investigators agreed that development of standard operating procedures (SOPs), protocols, and analytic methodologies for leukocyte isolation would require evidence-based confirmation of their precision, reproducibility, and analytic variance prior to introduction at the clinical sites.  SOPs and methodologies for all individual steps between blood sample collection and leukocyte isolation, the isolation of total cellular RNA from the leukocytes, and oligonucleotide microarray hybridization to the Affymetrix array platform were developed by consensus under the leadership of the scientists in the genomics and protein analysis/cell biology groups.  All SOPs and methodologies underwent vigorous alpha and beta testing at multiple laboratories before beta testing at the clinical sites enrolling the study patients.  Vigorous adherence to established SOPs was monitored using quality control and quality assurance procedures to insure high quality total RNA for gene expression profiling.

Standard macroscale leukocyte isolation

Total leukocytes (a heterogeneous mixture of all leukocyte cells) were collected from the “buffy coat” layer of whole blood samples (25 ml samples from most patients and 10 ml samples from pediatric patients when blood volumes were restricted) for global gene expression. Formed after density centrifugation of whole blood, the buffy coat is the thin layer between the clear plasma and the erythrocytes (red blood cells) containing most of the leukocytes and platelets. Our #G025 protocol provided plasma and total cellular RNA from mixed leukocytes in whole blood by initial centrifugation to recover the plasma fraction for plasma proteomic analysis, followed by lysis of the erythrocytes in the plasma-removed whole blood. Contaminating erythrocytes affect subsequent leukocyte expression so these are removed from the blood sample to recover a pure leukocyte population. Isolation of T lymphocytes and monocytes from the mixed leukocytes was performed using negative depletion procedures and antibody cocktails that took advantage of specific markers on the surface of these cells, usually described as cluster of differentiation (CD) markers. Adaptation of a readily available commercial cell preparation for isolation of human neutrophils provided a new and improved method to isolate neutrophils. Gentle extraction of the target cells was performed so as not to alter or activate the cells by the handling procedures during the cell separation steps.

These macroscale cell isolation techniques were tested, modified, and automated to produce SOPs for the clinical sites that provided quality total cellular RNA suitable for genome-wide expression analyses from isolated blood leukocytes and enriched subpopulations of neutrophils, T lymphocytes, and monocytes from healthy individuals and critically ill patients (Feezor et al, 2004; Cobb et al, 2005; De et al, JI, 2005; De et al, JII, 2005). Comparative transcriptome analyses of neutrophils, T lymphocytes, monocytes, and mixed leukocytes revealed unique expression patterns not seen in the whole-blood leukocytes, demonstrating a cell-specific response to severe trauma and burn injury (Laudanski et al, 2006; Cobb et al, 2005).  This finding was not surprising in that mixed leukocyte samples contain heterogeneous leukocyte subpopulations that can change quickly in the same sample, and these shifts in cell populations can interfere with analyses of changes in gene expression over time.

The leukocyte isolation methodologies dramatically reduced the variance in gene expression compared to traditional whole blood RNA measurements and increased the amount of genomic information obtained. The intra-sample and intra-individual variation in the genomic findings was markedly less than reported with other approaches and platforms, and artifacts that altered gene transcription could be controlled with strict adherence to the established SOPs. The macroscale isolation protocols were dependent on performing standard laboratory techniques (universal precautions, pipetting, density centrifugation, negative depletion) using centrally prepared reagents from Sample Collection Kits provided to each of the clinical sites.  These methodologies were robust enough to generate high quality genomic RNA from critically ill patients and sufficiently easy to perform by clinical nursing and laboratory personnel, but were time consuming and required large-volume blood samples.

Microfluidic approaches to isolating leukocytes from patient blood samples

Recent advances in microfluidic-based cell capture by the bioengineers at the MGH Center for Engineering in Medicine were highly promising for the application of their novel techniques to isolate specific cells from small-volume samples of complex cell mixtures (Murthy et al, 2004; Irimia et al, 2004; Sin et al, 2005, Toner and Irimia, 2005).   The Glue Grant investigators recognized the potential theoretical and practical advantages of microfluidic over macroscale methodologies in acquiring the blood leukocytes for analysis, and how the microfluidic-based approaches might facilitate the introduction of high throughput genomics in clinical medicine.  It had become evident from the scientific literature that gene expression profiling studies are more sensitive and powerful when enriched cell populations are the source of genomic RNA rather than mixed-cell populations.  The Glue Grant team agreed to supplant the macroscale techniques with microfluidic approaches for the isolation of mixed leukocytes, neutrophils, T lymphocytes, and monocytes to more precisely answer the biological questions posed in the grant application.

Microfluidic erythrocyte lysis

The multidisciplinary team designed, developed, and tested microfluidic lysis cassettes and protocols capable of isolating total leukocytes from buffy coat, based on lysis and separation of contaminating erythrocytes. Complete lysis of erythrocytes from whole blood was achieved with the microfluidic total leukocyte capture cassette in conjunction with ammonium chloride lysis, with nearly 100% recovery of leukocytes where the cells were exposed to the isotonic buffer solution for less than 40 seconds and without damage or activation of leukocytes (Sethu et al, 2004).  An improved protocol using deionized water in less than 10 seconds accomplished erythrocyte lysis and isolation of unactivated leukocytes from small blood volumes (Sethu et al, 2006).  Validation studies comparing the macroscale versus microfluidic techniques to isolate total leukocytes demonstrated that the microfluidic system surpassed macroscale lysis for both total and differential leukocyte recovery.  From the same volume of whole blood, the microfluidic cassette yielded 25% more leukocytes and many more T lymphocytes and monocytes.  In comparing the genome-wide expression patterns of total leukocytes isolated using the microfluidic cassette versus the macroscale isolation techniques, the investigators found no differences (Sethu et al, 2006).  In side-by-side comparison of RNA quantity, quality, and genome-wide expression patterns, sufficient amounts of total RNA were obtained for genome-wide expression analysis from 0.5 milliliters of whole blood.  Leukocyte expression patterns from samples processed using microfluidic and macroscale methodologies showed that there was less variability between samples derived from the two methods than between samples from different patients within clinical groups (Russom et al, 2008).

Microfluidic leukocyte cell separation and total cellular RNA recovery

With the success of microfluidic leukocyte isolation techniques in the clinical setting, the program next sought to develop microfluidic cell capture cassettes to isolate neutrophils, T lymphocytes, and monocytes.  Separation of leukocytes into different subtypes from whole blood is technically challenging and very time consuming, usually requiring multiple step dextran-Ficoll gradient isolation and milliliter volumes of blood.  To complicate matters, leukocytes are very sensitive to external perturbations and can be easily activated during the isolation process.  And different leukocyte subtypes contain much less RNA per cell than others, thus requiring increased cell numbers for an equivalent quantity of total cellular RNA. The Glue Grant investigators set the design targets to include capture of 95% pure homogeneous populations of neutrophils, T lymphocytes, and monocytes yielding more than 50 nanograms of total RNA and more than 10 micrograms of total protein for each of the three cell types from a whole blood sample.  A target cell capture time of less than 10 minutes was established to minimize the likelihood of leukocyte activation.

The multidisciplinary team of bioengineers, biologists, and clinicians worked together to develop and refine the physical devices and analytical protocols for microfluidic cell separation and sample processing for neutrophils, T lymphocytes, and monocytes.  The Glue Grant analytic and clinical sites participated in alpha and beta testing of the microfluidic cassettes being developed and optimized by contributing test samples from trauma and burn patients, patients with endotoxemia, and healthy individuals. Bioengineers optimized the capture antibody, flow conditions, and blood volumes in a series of design refinements to further improve captured cell purity, the number of captured cells for each cell subtype, and the quality/quantity of total cellular RNA isolated from the neutrophils, T lymphocytes, and monocytes.

The microfluidic neutrophil capture cassette was the first cassette implemented program-wide for the trauma and burn clinical studies following extensive testing, validation, quality control, and hands-on training.  The cassette contained a series of branched channels that were coated with CD66b, a cell-surface marker specific to neutrophils. Within a chamber with a height of 50 microns, blood flowed through each parallel capture channel and cells expressing CD66b antigen were specifically bound to the surface, with unbound cells washed away through the device outlet.  In 5 minutes, the cassette captured high enriched (greater than 95%) neutrophils directly from 150 microliters of whole blood in sufficient quantity and purity for genome-wide microarray and mass spectrometry-based proteomic analyses. In analyzing the neutrophil transcriptome response to trauma, the investigators were able to distinguish time-dependent signaling pathways in the patients.  This analysis represented the single, largest study of neutrophil genomics derived from clinical samples to appear in publication (Kotz et al, 2010).  A follow on study comparing dextran-Ficoll and microfluidic neutrophil isolation techniques demonstrated essentially equivalent high-quality total cellular RNA for genome-wide expression analysis (Warner et al, 2011).

The neutrophil capture device was modified to create the microfluidic T lymphocyte and monocyte capture cassettes.  Using CD2 as the antibody to capture T lymphocytes, performance testing of this cassette demonstrated sufficient high quality RNA for downstream microarray and proteomics analyses.  Isolation of monocytes proved to be more challenging due to their low numbers, particularly following burn injury, and the lack of a unique cell-surface protein that distinguishes monocytes from other more numerous leukocyte subtypes. Testing showed that CD36 was the best capture antibody for monocytes; however, CD36 is expressed on the majority of platelets as well. The bioengineers overcame this technical hurdle with the design, development, testing, and implementation of an on-line preprocessing methodology using hydrodynamic inertial focusing to allow for the rapid depletion of contaminating platelets (DiCarlo et al, 2007). This sized-based depletion technique separates the much smaller platelets from the larger sized monocytes.  Read more information about using inertial focusing to separate bioparticles in microchannels.  With sized-based depletion of platelets, the investigators can capture cells with purity exceeding 90% for each cell subtype and obtain high-quality RNA from the cell lysates.  Computational analysis of the transcriptome response of T lymphocytes and monocytes to severe trauma is ongoing.

Fast and efficient microfluidic capture of leukocytes

The microfluidic neutrophil, T lymphocyte, and monocyte capture cassettes have been integrated into a single, three-cell capture cassette and have undergone initial testing and validation.  The planned next steps include continued testing of on-chip modules for sample preparation of total cellular RNA for genomic analysis and sample preparation of peptide fragments for high-throughput proteomic analysis.   The Glue Grant participating investigators have demonstrated that genome-wide expression data from total leukocytes (mixed cells), neutrophils, T lymphocytes, and monocytes isolated using the microfluidic cell capture cassettes are no different than data generated by using macroscale methodologies. The advantages of microfluidics compared to macroscale techniques are many: lyses erythrocytes completely and rapidly, processes leukocytes without cell activation, provides pure leukocyte subtypes for genomic and proteomic analyses, and performs on-chip RNA and peptide fragment isolations.  The automated microfluidic cell capture technology can isolate leukocytes in minutes instead of hours using ten-fold less blood and can be performed by clinical research nurses and technicians at the patient’s point of care.

Relevant publications

Feezor RJ, Baker HV, Mindrinos M, Hayden D, Tannahill CL, Brownstein BH, Inflammation and Host Response to Injury, Large-Scale Collaborative Research Program. Whole blood and leukocyte RNA isolation for gene expression analyses. Physiol Genomics. 2004 Nov 17;19(3):247-54. PubMed PMID: 15548831

Cobb JP, Mindrinos MN, Miller-Graziano C, Calvano SE, Baker HV, Xiao W, Inflammation and Host Response to Injury Large-Scale Collaborative Research Program. Application of genome-wide expression analysis to human health and disease. Proc Natl Acad Sci U S A. 2005 Mar 29;102(13):4801-6. Epub 2005 Mar 21. PubMed PMID: 15781863; PubMed Central PMCID: PMC555033

De AK, Miller-Graziano CL, Calvano SE, Laudanski K, Lowry SF, Moldawer LL, Remick DG Jr, Rajicic N, Schoenfeld D, Tompkins RG. Selective activation of peripheral blood T cell subsets by endotoxin infusion in healthy human subjects corresponds to differential chemokine activation. J Immunol. 2005 Nov 1;175(9):6155-62. PubMed PMID: 16237112

Laudanski K, Miller-Graziano C, Xiao W, Mindrinos MN, Richards DR, De A, Moldawer LL, Maier RV, Bankey P, Baker HV, Brownstein BH, Cobb JP, Calvano SE, Davis RW, Tompkins RG. Cell-specific expression and pathway analyses reveal alterations in trauma-related human T cell and monocyte pathways. Proc Natl Acad Sci U S A. 2006 Oct 17;103(42):15564-9. Epub 2006 Oct 10. PubMed PMID: 17032758; PubMed Central PMCID: PMC1592643

Sethu P, Anahtar M, Moldawer LL, Tompkins RG, Toner M. Continuous flow microfluidic device for rapid erythrocyte lysis. Anal Chem. 2004 Nov 1;76(21):6247-53. PubMed PMID: 15516115

Sin A, Murthy SK, Revzin A, Tompkins RG, Toner M. Enrichment using antibody-coated microfluidic chambers in shear flow: model mixtures of human lymphocytes. Biotechnol Bioeng. 2005 Sep 30;91(7):816-26. PubMed PMID: 16037988

Murthy SK, Sin A, Tompkins RG, Toner M. Effect of flow and surface conditions on human lymphocyte isolation using microfluidic chambers. Langmuir. 2004 Dec 21;20(26):11649-55. PubMed PMID: 15595794

Irimia D, Tompkins RG, Toner M. Single-cell chemical lysis in picoliter-scale closed volumes using a microfabricated device. Anal Chem. 2004 Oct 15;76(20):6137-43. PubMed PMID: 15481964

De AK, Roach SE, De M, Minielly RC, Laudanski K, Miller-Graziano CL, Bankey PE. Development of a simple method for rapid isolation of polymorphonuclear leukocytes from human blood. J Immunoassay Immunochem. 2005;26(1):35-42. PubMed PMID: 15754803

Toner M, Irimia D. Blood-on-a-chip. Annu Rev Biomed Eng. 2005;7:77-103. Review. PubMed PMID: 16004567; PubMed Central PMCID: PMC3779643

Sethu P, Moldawer LL, Mindrinos MN, Scumpia PO, Tannahill CL, Wilhelmy J, et al. Microfluidic isolation of leukocytes from whole blood for phenotype and gene expression analysis. Anal Chem. 2006 Aug 1;78(15):5453-61. PubMed PMID: 16878882

Di Carlo D, Irimia D, Tompkins RG, Toner M. Continuous inertial focusing, ordering, and separation of particles in microchannels. Proc Natl Acad Sci U S A. 2007 Nov 27;104(48):18892-7. PubMed PMID: 18025477; PubMed Central PMCID: PMC2141878

Russom A, Sethu P, Irimia D, Mindrinos MN, Calvano SE, Garcia I, Inflammation and Host Response to Injury Large Scale Collaborative Research Program. Microfluidic leukocyte isolation for gene expression analysis in critically ill hospitalized patients. Clin Chem. 2008 May;54(5):891-900. PubMed PMID: 18375483; PubMed Central PMCID: PMC4011019

Kotz KT, Xiao W, Miller-Graziano C, Qian WJ, Russom A, Inflammation and the Host Response to Injury Collaborative Research Program. Clinical microfluidics for neutrophil genomics and proteomics. Nat Med. 2010 Sep;16(9):1042-7. PubMed PMID: 20802500; PubMed Central PMCID: PMC3136804

Warner EA, Kotz KT, Ungaro RF, Abouhamze AS, Lopez MC, Cuenca AG, et al. Microfluidics-based capture of human neutrophils for expression analysis in blood and bronchoalveolar lavage. Lab Invest. 2011 Dec;91(12):1787-95. PubMed PMID: 21931299; PubMed Central PMCID: PMC3957199

Contact

Mehmet Toner, PhD	Kenneth Kotz, PhD	Lyle Moldawer, PhD	Ronald Tompkins, MD, ScD
617-724-5336	617-726-3447	352-265-0494	617-726-3447

Visit the Glue Grant program website

See Less