Particle Motion Analysis in Pancreatic Islets Isolation using a Quadrupole Magnetic Flow Sorter by Venkata Sunil Kumar Sajja A dissertation submitted to the Graduate Faculty of Auburn University in partial fulfillment of the requirements for the Degree of Doctor of Philosophy Auburn, Alabama May 09, 2011 Keywords: Islet Isolation, Diabetes, Quadrupole Magnetic Sorter, Particle Tracking Velocimetry Copyright 2010 by Venkata Sunil Kumar Sajja Approved by Thomas R. Hanley, Chair, Professor of Chemical Engineering Paul W. Todd, Chief Scientist, Techshot Inc. James F. Leary, Professor of Biomedical Engineering, Purdue University Robert P. Chambers, Professor of Chemical Engineering Elizabeth Lipke, Assistant Professor of Chemical Engineering ii Abstract Pancreatic islet transplantation offers a viable option to achieve permanent metabolic control in Type 1 diabetes patients. However, large quantities of pure viable donor islet cells are necessary for transplantation. Using currently available islet isolation methods multiple donor organs are required to achieve successful transplantation, and there is a demand for an isolation method with high islet yield and viability. Additionally, with porcine xeno-islet cell transplantation providing much hope, improving the porcine islet isolation process has become a worthwhile endeavor. This dissertation is the summery of the work aimed to develop a Quadrupole Magnetic Sorter to isolate pancreatic islets from exocrine tissue. Computational Fluid Dynamics (CFD) simulations were used (Chapter 2) to predict the flow patterns, pressure drop and nonspecific crossover in a newly designed QMS flow channel for the isolation of pancreatic islets of Langerhans. Simulation results were compared with the theoretically and experimentally determined results to validate the CFD model. CFD simulations were employed to compare performance of two models of QMS flow channels with differing splitter positions. Results of the simulations were used to show that one design gives up to 10% less nonspecific crossover than another and this model can be used to optimize the flow channel design to achieve maximum purity of magnetic particles. Magnetic isolation is a promising method for separating and concentrating pancreatic islets for transplantation in Type 1 Diabetes patients. Continuous magnetic islet sorter was designed to overcome the restrictions of current purification methods that result in limited yield, viability and purity of the isolated islets. The performance of the islet sorter depends on the iii resulting speed of the islets in an applied magnetic field, a property known as magnetophoretic mobility. Essential to the design and operation of the magnetic sorter is a method to measure the magnetophoretic mobilities of magnetically infused islets. Magnetic particle tracking velocimeter (MPTV) was developed to measure (Chapter 3) the magnetophoretic mobility of particles up to 1000 microns in diameter. Velocity measurements are performed in a well- characterized isokinetic magnetic energy gradient using video imaging followed by analysis of the video images using a computer algorithm that produces histogram of absolute mobilities. Mobility distributions obtained indicated that magnetized islets have sufficient mobility to be captured by the proposed sorting method, with this result confirmed in test isolations of magnetized islets. To achieve islet isolation with high purity and yield Quadrupole Magnetic Sorting (QMS), a single cell separation method, is being modified for the isolation of pancreatic islets (Chapter 4). Islets are infused with 4.6?m Dynabeads? and separated continuously with QMS. Results from 10 porcine pancreas isolations indicated possibility of infusing islets with magnetic beads and isolating them continuously by reducing the exposure time of islets to enzymes. QMS isolated islets showed good morphology compared to standard COBE isolated islets and the Oxygen Consumption Rate (OCR) per DNA measurements confirmed the viability of the islets after isolation. QMS isolation can save the culture time and help to eliminate the mechanical stress due to centrifugation on the islets. Nude mice transplantation results confirmed Dynabeads do not affect the functionality of the islets. vi Acknowledgements I would like to express my sincere gratitude towards Dr. Thomas R. Hanley for giving me an opportunity to work with him on this research. I also would like to sincerely thank him for his consistent support, warm encouragement I would like to extend my thanks to Dr. Paul W. Todd for his support, guidance, thoughtful ideas, and also for serving on my committee. I would like to acknowledge Dr. James F. Leary, Dr. Robert P. Chambers and Dr. Elizabeth Lipke for serving on my committee and university reader Dr. Christopher J. Easley. I would like to express my sincere thanks to David J. Kennedy and Byron Guernsey from IKOTECH, LLC for their help and support. I would like to thank Mark Deuser and John Vellinger for giving me opportunity to work at Techshot, IN. I express my earnest appreciation toward the members of Techshot for their kind help, especially Alan Constance and Bill Johnson. I also want to thank Dr. Klearchos Papas and his group from University of Minnesota, MN, Mike Taylor and his group at Cell Tissue Systems and Robert McCarthy and his group at Vitacyte, IN group for providing facilities and help with isolations. I would like to express deep gratitude and gratefulness to my father Mr. Seshagiri Rao Sajja for being a constant source of inspiration and motivation, mother Mrs. Nagamani Sajja for her enduring love and immense moral support and family members Sunanda, Picheswara Rao and Sai Sreeja. My most special thanks belong to my wife Saroja for her love, sacrifice and subject discussions. vi Table of Contents Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ii Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iv List of Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix List of Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1 1.1 Diabetes and Transplantation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1.2 Recordi Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 1.3 Purification of Islets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 1.4 Density Gradient Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 1.5 Magnetic Separations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 1.6 Quadrupole Magnetic Sorting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .9 1.7 Magnetic Particle Tracking Velocimetry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 1.8 Magnetic Beads. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 1.9 Research Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 1.10 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .16 2. Computational Fluid Dynamics Simulation of a Quadrupole Magnetic Sorter Flow Channel: Effect of Splitter Position on Nonspecific Crossover . . . . . . . . . . . . . . . 23 2.1 Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .23 2.2 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 2.3 Materials and Methods . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 vi 2.3.1 QMS and Separation Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 2.3.2 Experimental Measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 2.3.3 Computational Fluid Dynamics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 2.3.4 Governing Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 2.3.5 Discretization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 2.3.6 Solvers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 2.3.7 Pressure-Velocity Coupling Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36 2.3.8 Geometric Modeling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37 2.3.9 Boundary Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41 2.3.10 Simulation Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41 2.4 Results and Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .41 2.4.1 CFD Simulations: Flow Analysis and Pressure Drop Predictions . . . . . . . 41 2.4.2 Nonspecific Crossover . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 2.5 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51 2.6 Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51 2.7 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52 3. Application of Magnetic Particle Tracking Velocimetry to Quadrupole Magnetic Sorting of Porcine Pancreatic Islets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 3.1 Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 3.2 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56 3.3 Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .58 3.4 Materials and Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61 3.4.1 Particles and Viscous liquid . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61 3.4.2 QMS system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62 vii 3.4.3 MPTV . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62 3.4.4 Magnet Assembly . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62 3.4.5 Flow system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63 3.4.6 Imaging System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63 3.4.7 Analysis of video data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63 3.4.8 Experimental . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65 3.5 Results and Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .66 3.6 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .78 3.7 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .80 4. QMS Isolation of Porcine Islets of Langerhans . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83 4.1 Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .83 4.2 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .84 4.3 Materials and Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86 4.3.1 QMS and separation theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86 4.3.2 Magnetophoretic Mobility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88 4.3.3 Pancreas Procurement and labeling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88 4.3.4 MRI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89 4.3.5 Histopathology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90 4.3.6 Pancreas Digestion and Islets Isolation . . . . . . . . . . . . . . . . . . . . . . . . . . . 90 4.4 Islet Quality Assessment . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91 4.4.1 Oxygen Consumption Rate (OCR) Assay . . . . . . . . . . . . . . . . . . . . . . . . . 91 4.4.2 DNA Quantification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91 4.4.3 Nude Mouse Bioassay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92 viii 4.5 Results and Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .92 4.5.1 Cross-over studies of QMS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92 4.5.2 Isolation of Islets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98 4.5.3 Magnetophoretic Mobility Measurements . . . . . . . . . . . . . . . . . . . . . . . . 100 4.5.4 Histology of Pancreas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105 4.5.5 Quality assessment of Islets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105 4.6 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .113 4.7 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .114 5. Summary and Future Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .118 Appendix A: QMS Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120 Appendix B: MPTV Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127 Appendix C: Fluent Simulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130 xiii List of Figures Figure 1.1 Ricordi chamber used for islet isolation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 Figure 1.2 Flow diagram of the islets isolation process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 Figure 1.3 Steps involved in Islet Transplantation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 Figure 1.4 Insulin independence, insulin dependence or absence of fasting C-peptide post last infusion islets alone recipients . . . . . . . . . . . . . . . . . . . . . . . .10 Figure 1.5 Schematic diagram of the quadrupole magnetic cell sorter (QMS) . . . . . . . . . . . . . 11 Figure 1.6 Cell tracking velocimeter (CTV) showing coordinate orientation . . . . . . . . . . . . . . . 13 Figure 1.7 Magnetophoretic Mobility Histograms for selected microbeads, 50nm ? 10?m . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 Figure 2.1 Schematic diagram of the quadrupole magnetic cell sorter (QMS) . . . . . . . . . . . . . . 27 Figure 2.2 Geometry drawings of Prototype I and Prototype II flow channels . . . . . . . . . . . . . . 31 Figure 2.3 Flow chart for segregated solver used in this study . . . . . . . . . . . . . . . . . . . . . . . . . 38 Figure 2.4 Unstructured triangular mesh on the faces of QMS flow channel . . . . . . . . . . . . . . . 39 Figure 2.5 Unstructured mesh along with the boundary layer on the inlet pipes and channel face . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40 Figure 2.6 Velocity (m/s) contours in x plane at different distance from flow distributor to the end point of the splitter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 Figure 2.7 a) Velocity (m/s) magnitude contours in the flow channel at z = 1.1875? (flow is in positive x direction. b) Velocity profile of the fluid at the middle point of the channel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 Figure 2.8 a) Pressure (Pa) contours in the flow channel at z = 1.1875? b) Comparison of the pressure drop from CFD simulations and empirical correlations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 xiii Figure 2.9 Comparison of calculated and observed nonspecific crossover as a function of outlet flow ratio at a total flow rate of 400 ml/min and inlet flow ratio of 0.25 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48 Figure 2.10 Comparison of the nonspecific crossover as a function of total flow rate at an inlet and outlet flow ratio of 0.25 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49 Figure 2.11 Nonspecific crossover as a function of particle concentration in the sample at a total flow rate of 400ml/min, inlet and outlet flow ratio of 0.25 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50 Figure 3.1 Schematic diagram of the quadrupole magnetic cell sorter (QMS) . . . . . . . . . . . . . . 60 Figure 3.2 A simplified diagrammatic representation of an MCTV system . . . . . . . . . . . . . . . . 64 Figure 3.3 Mobility histogram given by MPTV for Dynabeads . . . . . . . . . . . . . . . . . . . . . . . . . 67 Figure 3.4 Histograms of Dynabeads in each fraction of the QMS output . . . . . . . . . . . . . . . . . 68 Figure 3.5 Particle tracks developed by MPTV. Red dots are the particles tracked and black spots are disturbance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69 Figure 3.6 Magnetophoretic mobility histogram of BSI magnetic particles . . . . . . . . . . . . . . . . 71 Figure 3.7 Magnetophoretic mobility histogram of BSII magnetic particles . . . . . . . . . . . . . . . 72 Figure 3.8 Magnetophoretic mobility histogram of BSIII magnetic particles . . . . . . . . . . . . . . 73 Figure 3.9 Comparison of MPTV predicted fractional recovery of BSI particles in the three outlet fractions of the QMS with experimental results . . . . . . . . . . . 75 Figure 3.10 Comparison of MPTV predicted fractional recovery of BSII particles in the three outlet fractions of the QMS with experimental results . . . . . . . . . . . 76 Figure 3.11 Comparison of MPTV predicted fractional recovery of BSI particles in the three outlet fractions of the QMS with experimental results . . . . 77 Figure 3.12 a) Magnetophoretic mobility histogram of pancreatic islets isolated with QMS. b) MPTV predicted b fraction c) a fraction d) wall fraction of the islets at a total flow rate of 400ml/min and Ra? = 0.25 and Ra = 0.6 . . . . . . .79 Figure 4.1 Schematic diagram of the quadrupole magnetic cell sorter (QMS) . . . . . . . . . . . . . . 87 Figure 4.2 Crossover of the acinar tissue at different outlet flow ratios for constant inlet flow ratio and total flow rate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95 xiv Figure 4.3 Crossover of the acinar tissue at different inlet flow ratios for constant outlet flow ratio and total flow rate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .96 Figure 4.4 Crossover of the acinar tissue at different tissue concentrations at constant inlet flow ratio, outlet flow ratio and total flow rate . . . . . . . . . . . . . . . 97 Figure 4.5 Magnetophoretic mobility histogram of digested tissue sample taken before sending it through QMS for isolation . . . . . . . . . . . . . . . . . . . . . . . . . . . 101 Figure 4.6 Magnetophoretic mobility histogram of digested tissue sample taken from Negative (a) fraction of QMS isolation . . . . . . . . . . . . . . . . . . . . . . . . . . . 102 Figure 4.7 Magnetophoretic mobility histogram of digested tissue sample taken from Positive (b) fraction of QMS isolation . . . . . . . . . . . . . . . . . . . . . . . . . . . .103 Figure 4.8 Magnetophoretic mobility histogram of digested tissue sample taken from wall fraction of QMS isolation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .104 Figure 4.9 MRI of the control connecting/duodenal lobe (above) and the experimental splenic lobe (below), in which infused MP resulted in well-distributed hypointense regions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..106 Figure 4.10 Representative low and high magnification micrographs of an islet located in the experimental splenic lobe (distal region), illustrating minimal accumulation of magnetic particles in the acinar tissue (Fig. 1A, H/E) and significant accumulation within the islet (Fig. 1B, H/E) and magnetic particles within capillaries of an islet located in the proximal splenic lobe, near the site of infusion at the celiac trunk (Fig. 1C, insulin) . . . . . . . . . . . . . . .107 Figure 4.11 Pictures of the islets taken under microscope at 20X magnification a) islets isolated with QMS b) islets isolated with COBE . . . . . . . . . . . . . . . . . 108 Figure 4.12 Stimulation index for OCR measurements from control and infused islets from three separate isolations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110 Figure 4.13 Stimulation index for average OCR measurements from control and infused islets from three separate isolations at day0 and day7 . . . . . . . . . . 111 Figure A.1 Photograph of the QMS setup with fluid bags . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120 Figure A.2 Components of the QMS. Boxed labels are permanent QMS components, unboxed labels are consumable items replaced after one use . . . . . . . . . . . . . . .121 Figure A.3 Screnshot of the QMS software used to control the QMS operation . . . . . . . . . . . .122 Figure A.4 QMS set up for pancreas isolation at University of Minnesota laboratory . . . . . . . 123 xv Figure A.5 Absorbance detector traces for P810 isolation run. The QMS was turned on at time t=0 minutes. The turbidity sensor data logging was activated at t=42 minutes [1]. The digestion process was switched to recovery mode and the QMS began operations, with tissue leaving the flow channel at t=48 minutes [2]. The tissue concentration with the Ricordi chamber upright peaked at approximately t=72 minutes [3]. The Ricordi chamber was inverted to maximize the recovery at t=92 minutes [4]. The tissue exiting the inverted Ricordi chamber peaked at t=102 minutes [5]. The isolation was terminated at approximately t=114 minutes [6] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124 Figure B.1 Photograph of the MPTV assembly . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127 Figure B.2 Photograph of the MPTV set up . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128 Figure B.2 Screenshot of the IKOVISION used to run the MPTV . . . . . . . . . . . . . . . . . . . . . . 129 Figure C.1 Summery of the particle flow simulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .132 Figure C.2 Particles trajectories calculated in Fluent simulations . . . . . . . . . . . . . . . . . . . . . . . 133 xiii List of Tables Table 4.1 Total flow rates used in crossover studies and corresponding ISS and OSS values. Tissue which enters in between core and ISS has to cross the OSS to reach the positive fraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94 Table 4.2 Purity and yields of the islets from all the experiments conducted with four porcine pancreas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99 Table A.1 Summary of the Porcine isolation experiments . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125 1 1. Introduction 1.1 Diabetes and Transplantation Diabetes is a group of metabolic diseases that cause hyperglycemia. The consequences of diabetes include retinopathy (loss of vision), nephropathy (renal failure) and peripheral neuropathy with risk of amputation. The majority of diabetes cases fall into two categories: In type 1 diabetes, the cause is deficiency of insulin secretion. Type 1 diabetes accounts for 5 to 10 per cent of diagnosed diabetes. In type 2 diabetes the cause is combination of resistance to insulin action and deficit insulin secretion (Gavin et al., 2003). According to the reports presented by CDC in 2008, 24 million Americans, 8 per cent of the population, are diabetic (CDC, 2008). Diabetes costs the US economy $132 billion per year, $92 billion in direct medical costs (CDC, 2002; Paul et al., 2003). It has been suggested that insulin independence can be achieved in diabetes with islet transplantation (Markmann et al., 2003). Since the discovery of insulin in 1921, insulin therapy has saved many lives of type 1 diabetics. Insulin treatment cannot fully prevent chronic complications, and intensive insulin treatment to improve metabolic control has paralleled an increased risk of severe hypoglycemia (Berney et al., 2001; Oberholzer et al., 1999). The English surgeon Watson Williams was the first to attempt transplantation of pancreatic fragments in 1893 from a sheep to a boy. The first rodent pancreas transplantation was performed in 1966 along with kidney transplantation at the University of Minnesota (Ballinger et al., 1972). The International Pancreas Transplant Registry reports 17,000 whole pancreas transplants performed in the US up to 2004. Success after one 2 year varied from 72 to 84 per cent. Although whole pancreas transplantation is the most reliable means of restoring full metabolic control, it involves complex surgery. With only 5,000 cadaver organs available in any given year, transplantation is not available to treat all diabetes patients (Markmann et al., 2003). Islet transplantation offers a viable option to insulin therapy for improved metabolic control and to full organ transplantation with lower cost and risk. Successful islet transplantation can result in insulin independence (Berney et al., 2001). The recent dramatic improvement in the success rate of the islet transplantation in humans has prompted considerable interest for more widespread application of this methodology (Shapiro et al., 2000). The first clinical islet transplantation in humans was attempted in 1985 at Washington University Medical School in St. Louis (Ricordi et al., 1988). The process involved maceration of the organ into cell fragment broth, from which islets were purified by a density gradient method as islets are slightly less dense than acinar tissue. Islets constitute only 1-2 % of the pancreas volume and their isolation and purification are stressful mechanical and enzymatic procedures that can inflict significant damage, which may be further amplified by prolonged times of warm and cold ischemia with human islets (London et al., 1994; Robertson et al., 1998). One limiting factor to the islet transplantation is the need for the large numbers of islets, obtained from more than one donor pancreas per recipient. 1.2 Recordi Method The most important development for islet isolation was the automated method introduced by Dr. Camillo Ricordi. Ricordi?s method utilized a refined mixture of enzymes (collagenase) produced by the bacterium Clostridium hystoliticum. Ricordi disconnected the organ from the duodenum, and a canula was inserted into the main pancreatic duct which serves as the outlet for digestive enzymes produced by the pancreatic exocrine. With all the accessory ducts clamped, 3 the organ was distended by collagenase injected into the main duct. The dilated organ was placed in a conical chamber, the ?Ricordi Chamber (Figure 1.1)?, through which the heated enzyme solution was circulated. A ball mill using either glass or ceramic marbles was used to liberate islets from the tissue (Ricordi et al., 1988). Flow diagram of the Ricordi digestion process is shown in figure 1.2. The output from the Ricardi Chamber contains islets with some exocrine tissue. Large volumes of contaminated acinar tissue within islet auto grafts have been associated with portal hypertension, hepatic infarction, splenic bleeding and death. Therefore, it is necessary to purify islets before transplantation (Sulaiman et al., 2006). A diagram of the steps involved in the isolation and transplantation of pancreatic islets is shown in figure 1.3. 1.3 Purification of Islets Insulin independence can be achieved in type 1 diabetes by transplanting 10,000 Islet Equivalents (IEQ) per kilogram recipient weight of purified islets (Sakuma et al., 2008). To achieve high yields with required purities, many islet purification techniques have been developed for isolation of islets from exocrine tissue either by targeting islets or by targeting exocrine tissue, e.g. hand picking of islets, fluorescence-activated sorting by staining islets with neutral red (Gray et al., 1989; Jindal et al., 1994), or destruction of non-islet tissue by laser energy (Brunicardi et al., 1994). All these techniques were successfully employed to purify rat islets but failed to scale up to the human tissue digest volume. Islets or acinar tissue has been labeled with magnetic beads coated with monoclonal antibodies and then isolated with bar magnet (Muller-Ruchhoiltz et al., 1987; Winoto-Morbach et al., 1989; Winoto-Morbach et al., 1989a; Soon-Shiong et al., 1990; Winoto-Morbach et al., 1994; Davies et al., 1994; Davies et al., 1996). Since islets make up only one to two per cent of 4 Figure 1.1: Ricordi chamber used for islet isolation (www.biorep.com). 5 Figure 1.2: Flow diagram of the islets isolation process (Ricordi, 1992). 6 Figure 1.3: Steps involved in Islet Transplantation (Sulaiman et al., 2006). 7 pancreas, negative selection is difficult, requiring effective labeling of 98 per cent of the tissue. In addition, negative selection removes only the acinar cells, leaving ductal and other unidentified cell types remaining after pancreatic digestion. Positive selection of the islets offers the advantage over negative selection and the presence of beads in islets does not alter islet functionality (Nandigala et al., 1997). 1.4 Density Gradient Method While several methods of the islet isolation have been developed, the most successful method is density gradient centrifugation. In this method, tissue digestate is centrifuged and re- suspended in a dense medium. Less dense media are layered on top and the gradient is fed to a COBE 2991 cell processor. After several minutes of centrifugation, islets are separated from the interface between two discontinuous gradient layers (Samejima et al., 1998). The current clinical practice using on density gradient centrifugation does not achieve the necessary yield and purity of islets as this process relies on the small density difference between islets and acinar tissue. Density gradient cannot be standardized because tissue density and size distribution vary with each pancreas isolation. Tissue size is a function of the degree of enzymatic digestion, over which there is no direct control. In addition, density gradients have a limited ineffectiveness for digests in which islets have not been completely disconnected from the surrounding acinar tissue (London et al., 1998; Soon-Shiong et al., 1989). Exocrine tissue present in purified islets, such as lymph nodes, vascular tissue and ductal fragments, are major stimuli for induction of the immune response that results in acute rejection (Mitsukazu et al., 1986). Although some studies indicate that islets should be transplanted unpurified, transplantation of purified islets should be preferred to achieve appropriate revascularization of transplants (Heuser et al., 2000). 8 1.5 Magnetic Separations There is a major disparity between the number of available organ donors and recipients in need. Using current protocols 1 to 4 pancreas may be required to restore normal blood sugars in a single recipient. This is in large part due to variations in islet recovery and potency and the substantial losses in islet yield during organ procurement and storage as well as islet isolation and purification (Hering et al., 2002). Additionally, with porcine xeno-islet cell transplantation providing much hope (Hering et al., 2006), improving the porcine islet isolation process has become a worthwhile endeavor. Figure 1.4 shows the insulin independence achived in the patients who received islets transplantation in last decade (Michael et al., 2009). Immunomagnetic cell separation provides a highly attractive alternative to density?dependent methods for islet purification. The versatility of magnetic fields makes them a useful candidate for biological separations. Magnetic separations were initially developed to isolate glomuruli from murine kidneys by infusing iron oxide particles (Joyce Gauthier et al., 1988). Since islets of Langerhans have a similar angioarchitecture to mouse glumeruli, islets can be purified using magnetic force (Pinkse et al., 2004). The use of Dynabeads for islet purification was first reported in 1989. Magnetic particles coated with antibodies impart mobility to selected cells and allow separation by a magnetic field. Magnetic fields have been used to select islets and separate them from the acinar tissue (Winoto- Morbach et al., 1989b; Davies et al., 1995). The degree of purification achieved in rats nearly reached hand selection quality (Muller-Ruchholtz et al., 1987). By positive selection, many unwanted tissue elements, such as lymph nodes, arteries and other ductal fragments, would not contaminate the purified islets. Unlike the density gradient method, magnetic separation subjects islets to little mechanical stress and requires less time for separation. Magnetic separations are 9 effective in any medium and are easily scalable. The use of quadrupole magnetic separation increases the purity and yield of islets compared to use of simple magnet (Davies et al., 1994). 1.6 - Quadrupole Magnetic Sorting The QMS technology is based on a process known as split-flow lateral transport thin separation, also referred to by the acronym SPLITT, which is a subset of Field Flow Fractionation (FFF) technology. SPLITT was introduced by J.C. Giddings (Giddings et al., 1985) to extend the capabilities of FFF to the separation of colloids, macromolecules, and particulate materials (Martin et al., 1992). Quadrupole magnetic sorting (QMS) has been extensively modeled and tested and proved successful for single cell separation. QMS is a high-throughput, high-gradient, continuous magnetic cell separation system. QMS was initially designed for the positive separation of immunomagnetically labeled single cells from nonmagnetic cell population. Figure 1.5 provides a diagrammatic view of the QMS sorting mechanism. QMS employs three subsystems for operation: a fluid flow channel, a magnetic field for particle separation and pumps to regulate flow into and out of the flow channel. A sample containing labeled and unlabeled cells enters the flow channel (a?) with carrier buffer (b?). Separated labeled cells (b) and unlabeled cells (a) leave the flow channel at the bottom. Separation is a function of many factors including the entering flow rates, the total flow rate and the degree of magnetization of cells. Cell labeling may be quantified by the magnetophoretic mobility (Sun et al., 1998; Williams et al., 1999; Hoyos et al., 2000; McCloskey et al., 2003; McCloskey et al., 2003a; Moore et al., 2004; Tong et al., 2007; Jing et al., 2007). 10 Figure 1.4: Insulin independence, insulin dependence or absence of Fasting C-peptide post last infusion of islets alone recipients (Michael et al., 2009). 11 Figure 1.5: Schematic diagram of the quadrupole magnetic cell sorter (QMS). Qb? Qa? Qa Qb MA GN ET MA GN ET Tr an sp ort La mi na Tr an sp ort La mi na ri ro riss ross MA GN ET MA GN ET Tr an sp ort La mi na Tr an sp ort La mi na 12 1.7 ? Magnetic Particle Tracking Velocimetry Magnetophoretic mobility of magnetically labeled cells can be measured by a Magnetic Particle tracking velocimetry (MPTV). Figure 1.6 illustrates the MPTV system MPTV leverages current technologies in video microscopy, computer processing speed and finite element analysis for magnetic fields to measure the induced motion of cells and particles in the highly characterized magnetic and gravity field. The motion of the particles in the known field is then translated into a characteristic parameter called magnetophoretic mobility in a magnetic field and sedimentation rate in a gravity field. MPTV measurements involve videotaping the movement of immunomagnetically labeled particles through a known viscosity medium and magnetic susceptibility in a well-defined magnetic energy density gradient. The velocity of each particle along with its location within the magnetic energy gradient is recorded. From this information, magnetophoretic mobility of each particle is calculated. A significant advantage of this method over other techniques is that larger numbers of cells can be processed in very little time (Moore et al., 2004; Chalmers et al., 1999; Zhang et al., 2002; Chalmers et al., 1999a). 13 Figure 1.6: Magnetic Particle Tracking Velocimeter (MPTV) showing coordinate orientation. Computer Sample Buffer Observation Channel Magnet Pole Pieces Lens Camera Vacuum pump 14 1.8 Magnetic Beads Several manufactures like SpherotechTM, MiltenyiTM, InvotrogenTM produce magnetic beads from 50 nm to 10 ?m in size for biological separation applications. Some of the magnetic beads available for separations are tested for their magnetophoretic mobility. After considering the differences in the vasculature of the islets and acinar tissue Dynabeads? were selected to use for pancreas infusion to isolate islets from acinar tissue. Figure 1.7 shows the histograms of the magnetophoretic mobilities of different magnetic beads from 50nm to 10?m in size and Dynabeads shows the highest mobility among the all beads tested. 1.9 Research Objectives The application of the QMS system for isolation of islets can lead to greater efficiencies and lower costs in the islet transplantation for Type 1 diabetics. The overall objective of this research is to design and characterize a flow channel and develop optimized conditions for islet isolation using QMS. The specific objectives are as follows: ? To develop a reliable QMS process for isolation of human islets from acinar tissue for transplantation. ? To design and operate an improved MPTV to determine magnetophoretic mobilities of islets. ? To optimize flow channel operation using computational fluid dynamics to simulate the QMS flow channel to study the flow patterns and pressure variations. 15 Figure 1.7: Magnetophoretic Mobility Histograms for selected microbeads, 50nm ? 10?m (data provided by David J. Kennedy of IKOTECH, LLC) 16 1.10 References Ballinger WF, Lacy PE. 1972. Transplantation of intact pancreatic islets in rats. Surgery 72:175- 186. Berney T, Buhler L, Caulfield A, Oberholzer J, Toso CH, Alejandro R, Cooper DKC, Ricordi C, Morel PH. 2001. Transplantation of islets of Langerhans New developments. Swiss Med Wkly 47-48:671-680. Brunicardi FC, Oh Y, Shevlin L, Suh E, Kleinman R, Stein E, Lipaz G, Plant DV, Imagawa D, Fetterman HR. 1994. Laser destruction of human nonislet pancreatic tissue. Transplant Proc 26(6):3354-3355. CDC. 2008. Number of people with diabetes increases to 24 million, in Press Release, D.o.h.a.h. services, Editor, Center for Disease Control. CDC. 2002. National Health Interview Survey. Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, Division of Diabetes Translation. Chalmers JJ, Zhao Y, Nakamura M, Melnik K, Lasky L, Moore L, Zborowski M. 1999. An Instrument to Determine the Magnetophoretic Mobility of Labeled, Biological Cells and Paramagnetic Particles. Journal of Magnetism and Magnetic Materials 194:231-241. Chalmers JJ, Haam S, Zhao Y, McCloskey K, Moore L, Zborowski M, Williams PS. 1999a. Quantification of Cellular Properties from External Fields and Resulting Induced Velocity: Cellular Hydrodynamic Diameter. Biotechnology and Bioengineering 64:509- 518. 17 Davies JE, Robertson GS, Swift S, Chamberlain J, Bell PR, James RF, London NJ. 1994. Use of a quadripole magnet significantly improves immunomagnetic islet purification. Transplant Proc 26(2):649-650. Davies JE, Winoto-Morbach S, Ulrichs K, James RF, Robertson GS. 1996. A comparison of the use of two immunomagnetic microspheres for secondary purification of pancreatic islets. Transpntation 62(9):1301-1306. Davies JE, James RF, London NJ, Robertson GS. 1995. Optimization of the magnetic field used for immunomagnetic islet purification. Transplantation 59(5):767-771. Gavin JR, Alberti K, Davidson M, DeFronzo R, Drash A, Gabbe S, Genuth S, Harris M, Kahn R, Keen H, Knowler W, Lebovitz H, Maclaren N, Palmer J, Raskin P, Rizza R, Stern M. 2003. Report of the expert committee on the diagnosis and classification of diabetes mellitus: Committee Report. Diabetes Care 26(S1):S5-S20. Giddings JC. 1985. A System Based on Split-Flow Lateral-Transport Thin (SPLITT) Separation Cells for Rapid and Continuous Particle Fractionation. Sep Sci Technol 20:749-768. Gray DW, G?hde W, Carter N, Heiden T, Morris PJ. 1989. Separation of Pancreatic islets by Fluorescence-Activated Sorting, Diabetes 38(suppl 1):133-135. Hering BJ, Matsumoto I, Sawada T, Nakano M, Sakai T, Kandaswamy M, Sutherland DER. 2002. Impact of two-layer pancreas preservation on islet isolation and transplantation. Transplantation 74:1813-1816. Hering BJ, Wijkstrom M, Graham ML, et al. Prolonged diabetes reversal after intraportal xenotransplantation of wild-type porcine islets in immunosuppressed nonhuman primates. Nat Med 2006; 12 (3): 301. 18 Heuser M, Wolf B, Vollmar B, Menger MD. 2000. Exocrine contamination of isolated islets of Langerhans deteriorates the process of revascularization after free transplantation. Transplantation 69(5):756-761. Hoyos M, Moore LR, McCloskey KE, Margel S, Zuberi M, Chalmers JJ, Zborowski M. 2000. Study of magnetic particles pulsed-injected into an annular SPLITT-like channel inside a quadrupole magnetic field. J Chromatogr A 903(1-2):99-116. http://www.biorep.com/RicordiChamber-1#images/ptos/diabetes/ricordi-chamber-gal2.jpg (09/17/2010). Jindal RM, McShane P, Gray DWR, Morris PJ. Isolation and Purification of Pancreatic 1994. Islets by Fluorescence Activates Islet Sorter, Transplantation Proceedings 26(2):653. Jing Y, Chalmers JJ, Zborowski M. 2007. Blood progenitor cell separation from linical leukapheresis product by magnetic nanoparticle binding and agnetophoresis. Biotechnol Bioeng 96(6):1139-1154. Joyce Gauthier V, Mart Mannik. 1988. A method for isolation of mouse glomeruli for quantitation of immune deposits. Kidney International 33:897-899. London NJM, Swift SM, Clayton HA. 1998. Isolation, culture and functional evaluation of islets of Langerhans. Diabetes & Metabolism 23:200-207. London NJ, Robertson GS, Chadwick DR, Johnson PR James RF, Bell PR. 1994. Human pancreatic islet isolation and transplantation. Clin Transplant 8(5):421-459. Markmann JF, Deng SP, Huang X, Desai NM, Velidedeoglu EH, Lui CY, Frank A, Markmann E, Palanjian M, Brayman K, Wolf B, Bell E, Vitamaniuk M, Doliba N, Matschinsky F, Barker CF, Naji A. 2003. Insulin independence following isolated islet transplantation and single islet infusions. Annals of Surgery 237(6):741-751. 19 Martin M, Williams PS. 1992. Theoretical Advancement in Chromatography and Related Separation Techniques; Dondi, F., Guiochon, G., Eds.; Kluwer Academic Publishers: Dordrecht, The Netherlands 51:513- 580. McCloskey KE, Chalmers JJ, Zborowski M. 2003. Magnetic cell separation: Characterization of magnetophoretic mobility. Analytical Chemistry 75(24):6868-6874. McCloskey KM, Moore LR, Hoyos M, Rodrigues A, Chalmers JJ, Zborowski M. 2003a. Magnetic cell separation is a function of antibody binding capacity (ABC). Biotechnol Prog 19(3):899-907. Michael A, Hering B. 2009. CITR annual report. Mitsukazu Gotoh, Takashi Maki, Susumu Satomi, Janis Porter, Anthany P Monaco. 1986. Immunological characteristics of purified pancreatic islet grafts, Transplantation 42(4):387-390. Moore LR, Milliron S, Williams PS, Chalmers JJ, Margel S, Zborowski M. 2004. Control of Magnetophoretic Mobility by Susceptibility-Modified Solutions as Evaluated by Cell Tracking Velocimetry and Continuous Magnetic Sorting. Analytical Chemistry 76:3899- 3907. Muller-Ruchholtz W, Leyhausen G, Petersen P, Schubert G, Ulrichs K. 1987. A simple methodological principle for large scale extraction and purification of collagenase- digested islets. Transplant Proc 19(1):911-915. Nandigala P, Chen TH, Yang C, Hsu WH, Heath C. 1997. Immunomagnetic isolation of islets from rat pancreas. Biotechnol Prog 13:844-848. Oberholzer J, Triponez F, Lou J, Morel P. 1999. Clinical islet transplantation: A review. Annals of the New York Academy of Sciences 875:189-199. 20 Paul H, Tim D, Plamen N. 2003. Economic costs of Diabetes in the U.S. in 2002, Diabetes Care 26:917-932. Pinkse GGM, Steenvoorde E, Hogendoorn S, Noteborn M, Terpstra OT, Bruitin JA, De Heer E. 2004. Stable transplantation results of magnetically retracted islets: a novel method. Diabetologia 47:55-61. Ricordi C, SCHARP dw, Lacy PE. 1988. Reversal of diabetes in nude mice after transplantation of fresh and 7 days cultures (24oC) human pancreatic islets. Transplantation 45: 994-996. Ricordi C, Lacy PE, Finke EH, Olack BJ, Scharp DW. 1988. Automated method for isolation of human pancreatic islets. Diabetes 37:413-420. Ricordi C. The automated method for islet isolation. In Pancreatic Islet Cell Transplantation 1892-1991: One Century of Transplantation for Diabetes. Ricordi C, Ed. Austin, TX, R.G. Landes Company 1992, p.99-112. Robertson GS, Dennison AR, Johnson PR, London NJ. 1998. A review of pancreatic islet autotransplantation. hepatogastroenterology 45(19):226-235. Sakuma Y, Ricordi C, Miki A, Yamamoto T, Pileggi A, Khan A, Alejandro R, Inverardi L, Ichii H. 2008. Factors That Affect Human Islet Isolation, Transplantation Proc 40:343-345. Samejima T, Yamaguchi K, Iwata H, Morkawa N, Ikada Y. 1998. Gelatin density gradient for isolation of islets of Langerhans. Cell Transplant 7(1):37-45. Shapiro AM, Lakey JR, Ryan EA, Korbutt GS, Toth E, Warnock GL, Kneteman NM, Rajotte RV. 2000. islet transplantation in seven patients with type 1 diabetes mellitus using glucocorticoid-free immunosuppressive regimen. N Engl J Meg 343(4) 289-290. 21 Soon-Shiong P, Heintz R, Terasaki P. 1989. Identification of novel blood groupreactive monoclonal antibodies cytotoxic to human acinar cells but not islets. Transplantation Proceedings 21(1):2622-2623. Soon-Shiong P, Fujioka T, Terasaki P, Heintz R, Lanza RP. 1990. Islet purification by a novel immunomicrosphere cell depletion technique. Transplant Proc 22(2):780-781. Sulaiman AN, James Shapiro AM. 2006. Advances in pancreatic islet transplantation in humans, Diabetes, Obesity and Metabolism 8:15-25. Sun L, Zborowski M, Moore L, Chalmers JJ. 1998. Continuous, Flow-Through Immunomagnetic Cell Separation in a Quadrupole Field. Cytometry 33:469-475. Tong X, Xiong Y, Zborowski M, Farag SS, Chalmers JJ. 2007. A Novel High Throughput Immunomagnetic Cell Sorting System for Potential Clinical Scale Depletion of T Cells for Allogeneic Stem Cell Transplantation. Exp Hematol 35(10)1613-1622. Williams PS, Zborowski M, Chalmers JJ. 1999. Flow Rate optimization for the Quadrupole Magnetic Cell Sorter. Anal Chem 71:3799-3807. Winoto-Morbach S, Ulrichs K, Leyhausen G, Muller-Ruchholtz W. 1989. New principle for large-scale preparation of purified human pancreas islets. Diabetes 38(Supp 1):146-149. Winoto-Morbach S, Leyhausen G, Schunke M, Ulrichs K, Muller-Ruchholtz W. 1989a. Magnetic microspheres (MMS) coupled to selective lectins: a new tool for large-scale extraction and purification of human pancreatic islets. Transplant Proc 21(1):2628-2630. Winoto-Morbach S, Krout OS, Heiser A, Ulrichs K, Muller-Ruchholtz W. 1994. Lectin binding to acinar tissue for complete magnetophoretic purification of porcine pancreatic islets depends on the composition and pH of the incubation medium. Transplant Proc 26(2):646-648. 22 Winoto-Morbach, Ulrichs K, Hering BJ, Leyhausen G, Muller-Ruchholtz W. 1989b. Methods in Islet transplantation research, Hormone and Metabolic Research, Supplement 25:51-54. Zhang H, Nakamura M, Comella K, Moore L, Zborowski M, Chalmers J. 2002. Characterization/Quantification of the Factors Involved in the Imparting a Magnetophoretic Mobility on Cells and Particles. European Cells and Materials 3(2):34- 36. 23 2. Computational Fluid Dynamics Simulation of a Quadrupole Magnetic Sorter Flow Channel: Effect of Splitter Position on Nonspecific Crossover 2.1 Abstract In the Quadrupole Magnetic Sorter (QMS) magnetic particles enter a vertical flow annulus and are separated from non-magnetic particles by radial deflection into an outer annulus where the purified magnetic particles are collected via a flow splitter. The purity of magnetically isolated particles in QMS is affected by the migration of nonmagnetic particles across transport lamina in the annular flow channel. Computational Fluid Dynamics (CFD) simulations were used to predict the flow patterns, pressure drop and nonspecific crossover in a newly designed QMS flow channel for the isolation of pancreatic islets of Langerhans. Simulation results were compared with the theoretically and experimentally determined results to validate the CFD model. CFD simulations were employed to compare performance of two models of QMS flow channels with differing splitter positions. Results of the simulations were used to show that one design gives up to 10% less nonspecific crossover than another and this model can be used to optimize the flow channel design to achieve maximum purity of magnetic particles. Keywords: Quadrupole Magnetic Sorter, Computational Fluid Dynamics, Annulus Flow, Nonspecific Crossover 24 2.2 Introduction Insulin independence can be achieved in type 1 diabetes patients by transplanting purified pancreatic islets of Langerhans (Sakuma et al., 2008). Isolation of the islets is the important step in preparing the islets for the transplantation. To achieve high yields with required purities, various islet purification techniques have been developed for isolation of islets from exocrine pancreas tissue either by targeting islets or by targeting exocrine tissue (Soon-Shiong et al., 1990; Winoto-Morbach et al., 1989), e.g. hand picking of islets, fluorescence-activated sorting by staining islets with neutral red (Gray et al., 1989; Jindal et al., 1994), or destruction of non- islet tissue by laser energy (Brunicardi et al., 1994). All of these techniques were successfully employed to purify rat islets but failed to scale up to the human tissue digest volume. Present human islet isolation processes rely on the density gradient centrifugation method which depends on small density difference between islets and acinar tissue. The density gradient centrifugation process, however, has been identified as a potential source of islet mass loss and islet damage due to mechanical stress associated with centrifugation and prolonged exposure to proteolytic enzymes (Samejima et al., 1998; Pinkse et al., 2004). Quadrupole magnetic sorting (QMS) is a well-designed and tested technology for magnetic separations of single cells, e.g., stem cells (Moore et al., 2001; Nakamura et al., 2001; Lara et al., 2002) and has been successfully applied to several cell separations and purifications (Sun et al., 1998; Tong et al., 2007; Jing et al., 2007). When QMS, developed for single cell separation, was used for isolation of islets, very low purity was achieved with good yields. Low purity is due to the size difference between single cells and islets (Shenkman et al., 2009). Islet diameters are distributed between 50 to 500 microns whereas single cell size is around 1 to 10 microns. It has been observed that the mechanical stress applied in QMS does not affect islet 25 functionality after isolation (Shenkman et al., 2009). To overcome limitations associated with the commonly used density gradient method and to develop a technique that yields sufficient number of islets to reduce the donor to recipient ratio to 1:1, we redesigned (Kennedy et al., 2007) and tested high capacity QMS to isolate porcine islets of Langerhans. In QMS isolation procedures particles of interest labeled with magnetic micro particles are isolated from unlabeled particles in a magnetic field while flowing through an annular flow channel. Labeled particles migrate radially outward in a quadrupole magnetic field and leave the channel on its outer periphery as a purified fraction. The presence of any unlabeled particles in isolated magnetically labeled particles is known as nonspecific crossover. The occurrence of nonspecific crossover is of particular importance in the isolation of islets as the islets constitute only one to two percent of the entire pancreas. Thus, even one percent of nonspecific crossover could reduce the purity of islets by 50 percent. There are many factors that contribute to nonspecific crossover in QMS, such as hydrodynamic forces, particle concentration in the sample, total flow rate, flow ratios, splitter imperfections (Williams et al., 2003; Williams et al., 2003a; Williams et al., 2008) and splitter position. The effects of splitter position, sample concentration, total flow rate and flow ratios were tested in this study. Computational fluid dynamics software was used to predict the fluid flow pattern throughout the flow channel and nonspecific crossover. 2.3 Materials and Methods 2.3.1 QMS and Separation Theory The QMS technology is based on a process known as split-flow lateral transport thin separation, which is a subset of Field Flow Fractionation (FFF) technology that separates immunomagnetically labeled islets based on their magnetophoretic mobility (m). 26 Magnetophoretic mobility is defined as the velocity of a particle per unit magnetic energy gradient. A QMS system consists of three essential components: flow channel, magnet and pumps. Separation takes place in the flow channel which consists of a cylindrical core that is concentric with an external cylindrical shell (Figure 2.1). The sample, consisting of magnetically labeled islets along with unlabelled tissue particles, enters at a? close to the core through a specially designed inlet: carrier fluid, enters the channel at b? adjacent to the outer wall. Both flows merge at the end of the inlet splitter and continue to flow at each side of a virtual surface, known as inlet splitting surface (ISS) at distance riss from the center of the core. ISS can be seen in the Figure 2.1 close to the core and remaining at the same distance along the fully developed laminar flow in the annulus region of the channel. The outer flow rate Qb? is maintained higher than the inner flow rate Qa? to keep the sample flow near the core. A second virtual surface known as the outer splitting surface (OSS) at a distance ross from the center of the core separates the two outlets which are separated by the outlet splitter. The distance between ISS and OSS, ross-riss, is known as the Transport Lamina. Magnetically labeled particles in the sample migrate radially and cross the transport lamina and leave in the positive fraction b. The unlabeled and weakly labeled particles which cannot cross the OSS will exit in the negative fraction a. For an ideal separation in QMS with laminar flow, the transport lamina thickness should be zero as the unlabeled particles would not be expected to cross the transport lamina. However, magnetically labeled islets are generally not uniform in magnetization. A range of magnetization is exhibited due to variation in the number of magnetic particles that enter the vascular structure of islets. To completely isolate pure labeled particles, it is necessary to create conditions that allow the least mobile particles to migrate across the thickness of the transport lamina. The 27 Figure 2.1 : Schematic diagram of the quadrupole magnetic cell sorter (QMS). Qb? Qa? Qa Qb MA GN ET MA GN ET Tr an sp ort La mi na ri ro riss ross Non-Magnetic Particle Magnetic Particle MA GN ET MA GN ET Tr an sp ort La mi na 28 thickness of the transport lamina changes with changes in inlet and outlet flow rate ratios and change in splitter position. The particle trajectory in the annulus is described by the integral equation; ?? += k k m s z dkku vkvdz 1 )( )( 0 (1) where )(kv is the velocity profile of the fluid in the annulus, sv is the Stokes sedimentation velocity of the particle in the fluid, um(k) is the radial magnetophoretic velocity of the particle and k is defined as 0r rk = (2) Here is defined as the radial coordinate from the center of the flow channel core and 0r is the radius from the center to the wall. Upon the integration of eq. (1) and simplification we can get 1 0 1 0 2 1 11 00 )ln( )1( ),(. 2 ASm k kvr kA kkI mSr Qz mm s imm +?= pi (3) where mm is called the magnetophoretic mobility of the particle, 0mS is the magnetic force at the inner surface of the outside wall of the flow channel and sv is sedimentation velocity of the particle. 1A , 2A and ),( 11 kkI are functions used to simplify the calculations and are given by )1( 221 AkA i ?+= (4) i i k kA 1ln 1 2 2 ?= (5) k kkAkkkkI 1])(ln22ln4[),( 2 2 2 11 +?= (6) 29 The radial position of the ISS, KISS is calculated using the following integration factor (Kennedy et al., 2007): ISS i k kISSi kAkkAkkkkI ]ln22[),( 2 2 2 2 42 2 ?+?= (7) Similarly, the position of the OSS can be calculated using eq. (7) by replacing kISS with kOSS. 2.3.2 Experimental Measurements The flow channel used for the experiments was called ?prototype I?. Nonspecific crossover experiments were performed with nonmagnetic particles, Cultisphers?, macro-porous gelatin-coated micro-carrier beads with diameter range of 200 to 380 ?m, similar in size and density to islets and fragments of exocrine tissue in pancreas digests. The sample consisting of cultisphers was introduced into the a? inlet flow stream and carrier buffer was introduced into the b? inlet flow stream by dual head Watson-Marlow peristaltic pumps, and the positive fraction outlet was controlled by a peristaltic pump while leaving the negative fraction outlet to exit at atmospheric pressure to maintain equilibrium of flow in the flow channel. Experiments were carried out at total flow rates of 250, 300 and 400 ml/min with inlet flow ratio (Qa?/Q) of 0.25 and at different outlet flow ratios (Qa/Q) 0.25, 0.3, 0.5 and 0.7. Experiments were conducted at different sample concentrations to study the effect of particle concentration in the sample on nonspecific crossover. Turbidity sensors are connected to positive and negative fraction outlets to detect absorbance. Nonspecific crossover was calculated from the areas of the absorbance peaks. 2.3.3 Computational Fluid Dynamics The commercial CFD code, FLUENT, was used for the simulations, and FLUENT?s preprocessor, GAMBIT, was used to generate flow-field meshes. An unstructured mesh consisting of tetrahedral volumetric elements in 3D is used in the entire domain. Velocity inlet 30 boundary conditions with velocities normal to the plane of the channel inlets with fluid properties corresponding to water at 298K are set at the inlets, outflow boundary conditions corresponding to fully developed flow were used at the outlets and the no-slip condition is applied to all walls. The geometry used for the first case is based on the quadrupole magnetic separation channel for islets, ?prototype I?. The length of this flow channel is 254 mm with inner diameter of the outer wall of 6.03 mm and core diameter is 5.08 mm. Splitters with radius 5.588 mm and thickness 0.14 mm are used at inlet and outlet. A specially designed flow distributor with diameter 5.96 mm is placed in the b inlet. For the second case (prototype II), the channel length and outer wall diameter are the same as for Prototype I, but the core diameter is 5.08 mm and inlet splitter diameter 5.58 mm with thickness 0.14 mm and outlet splitter diameter 5.84 mm with thickness 0.14 mm. The two QMS separation flow channels analyzed in the study are shown in Figure 2.2. 2.3.4 Governing Equations For flow in the QMS flow channel, FLUENT, a commercial CFD package, is utilized to solve conservation equations for mass and momentum. FLUENT is based on a set of partial differential equations maintaining conservation of mass and momentum. These equations describe the convective motion of the fluid. The law of conservation of mass applied to an infinitesimal fixed control volume generates the continuity equation in differential conservation form: 31 a? b? a b a? b? a b Prototype I Prototype II INLET SPLITTER CORE FLOW ANNULUS OUTLET SPLITTER Figure 2.2: Geometry drawings of Prototype I and Prototype II flow channels. 32 0. =?+?? ? Vt ?? (10) The first term in the Equation 10 represents the rate of increase in the density and the second term represents the rate of mass flux passing out of the control volume. For a steady incompressible flow, ? = constant and 0=?? t? which reduces Equation 10 to 00. =??+??+??=? ? z u y u x uorV (11) Newton's second law applied to a fluid passing through an infinitesimal control volume yields the momentum equation: ijVVVt pi?? .).()( ?=?+? ? ??? (12) The first term in the Equation 12 represents the rate of increase of momentum per unit volume in the control volume. The second term represents the rate of momentum lost by convection through control surface. The forces applied by external stresses such as normal and shearing stresses which can be represented by the stress tensor ?ij. The stress tensor can be written as k k ij i j j i ijij x u x u x up ? ?? ? ?+ ? ?+?= ???pi 3 2)[( (13) Substituting equation 13 into equation 12 and simplifying produces the Navier - Stokes equations: ??? ? ??? ? ? ?? ? ?+ ? ? ? ?+??= ? k k ij i j j i j ij x u x u x u xvpDt VD ???? 3 2)( (14) For a Cartesian coordinate system, Equation 14 can be separated into three scalar equations. The conservation of momentum in the x-direction is given by 33 zyxx p Dt Du zxyxxx ? ?+ ? ?+ ? ?+ ? ??= ???? (15) The conservation of momentum in the y-direction is given by zyxy p Dt Dv zyyyxy ? ?+ ? ?+ ? ?+ ? ??= ???? (16) The conservation of momentum in the z-direction is given by zyxz p Dt Dw zzyzxz ? ?+ ? ?+ ? ?+ ? ??= ???? (17) The viscous normal stresses are given by ??? ? ??? ? ? ?? ? ?? ? ?= z w y v x u xx 23 2 ?? (18) ??? ? ??? ? ? ?? ? ?? ? ?= z w x u y v yy 23 2 ?? (19) ??? ? ??? ? ? ?? ? ?? ? ?= x u y v z w zz 23 2 ?? (20) and the viscous shear stresses are given by ??? ? ??? ? ? ?+ ? ?== y u x v yxxy ??? (21) ?????? ??+??== zuxwzxxz ??? (22) ??? ? ??? ? ? ?+ ? ?== z v y w zyyz ??? (23) The first law of thermodynamics applied to a fluid passing through an infinitesimal, fixed control volume yields the following energy equation: 34 ( ) ( ) ( ) ( )wpzvpyupxTVkhVht ????????????????=???????+?? ???? .. ?? ( ) ( ) ( )zxyxxx uzuyux ??? ??+??+??+ ( ) ( ) ( )zyyyxy uzuyux ??? ??+??+??+ ( ) ( ) ( )zzyzxz uzuyux ??? ??+??+??+ (24) For most problems in gas dynamics, it is possible to assume an ideal gas, defined as a gas whose intermolecular forces are negligible. An ideal gas obeys the ideal gas equation of state: RTp ?= (25) In this study, the ideal gas equation of state is used to calculate the density for the compressible gas flow with change in pressure. The continuity, x-momentum, y-momentum, z-momentum and energy equations are solved using FLUENT to simulate incompressible, laminar, steady-state flow. 2.3.5 Discretization The process of converting partial differential equations into a set of algebraic relations that can be solved using a computer is called discretization (Anderson, 1999). Discretization involves two main steps: (1) converting the continuum partial differential equations into algebraic relations and (2) converting the continuous physical domain into nodes, volumes, or elements where the algebraic equations will be solved. Discretization of partial differential equations can be accomplished in many different ways. Among those, the two types of discretization used to solve problems in fluid mechanics and heat transfer are finite differences (discretization of partial differential form of equations) and finite volumes (discretization of equations in the integral form). 35 In the finite difference approach, the continuous problem or domain is discretized so that the dependent variables are considered to exist only at discrete points. In the finite volume method, the conservation principles are applied to a fixed region or space known as the control volume. Using the finite volume approach, either control volumes are established first and grid points are placed at the center of the volumes (cell-centered method) or grid points are established first and then the boundaries of the control volume are fixed (vertex-centered method) (Tannehill et al., 2004). The finite volume method using the cell-centered approach is used in this study to convert the governing equations into an algebraic form that is then solved numerically. A variety of methods are used for discretizing the governing equations using the finite volume method. In this study, first-order and second-order accurate upwind methods are used. The cell-centered method is used, solving the flow properties at the cell centers; the values at the interfaces (used to obtain cell fluxes) are obtained by interpolating the cell-centered values with an upstream direction bias. This method is called an upwinding scheme. 2.3.6 Solvers FLUENT was used in this study to simulate flows through the QMS flow channel. This section discusses the FLUENT solver options that are used in this study. Governing equations for conservation of mass, momentum, energy, and chemical species are solved using cell- centered control volume based segregated solver. Using a segregated solver, the governing equations are solved sequentially (i.e., segregated from one another). Several iterations of the solution loop are performed before obtaining a converged solution. The segregated solver basic steps, shown in Figure 2.3, are: 36 1. Initial guess of pressure, velocity and species transport quantities, such as mass flux, is made. 2. The discretized Navier-Stokes equations (i. e., momentum equations) are solved using the guessed values of P*, u*, v*, etc. In this step coefficient to determine the fluxes through the cell/boundary faces by conduction/convection etc. The flow variables (p, u, v, w, etc.) for the entire domain (for each grid node) are obtained at the end of this step. 3. Once the flow properties are calculated, the continuity equation is next verified. If the continuity equation is not satisfied, pressure and velocity values (u, v, w) are corrected using pressure correction equation. SIMPLE algorithm is used for pressure-velocity coupling. 4. Once the flow properties are obtained, other discretized transport equations are solved. The initial guess for the transport quantities are depicted in step 1. After solving the transport equation, we get the transport properties at each node 5. In this step the flow and transport variables, obtained in step 3 and 4, at each node are compared with values of previous iteration and residuals are calculated using L2 Norm. If the residuals are less than prescribed tolerance limit, the solution is final. If the residual is higher than the tolerance limit, steps 1 through 5 are repeated until the convergence criteria are satisfied. The initial guess values are replaced with values of flow and transport variables obtained from the current iteration. 2.3.7 Pressure-Velocity Coupling Method A pressure-velocity coupling is used to derive an equation for pressure from the discrete continuity equation. The SIMPLE algorithm is used to obtain a relationship between velocity and pressure corrections to enforce mass conservation and thus obtain the pressure field. 37 2.3.8 Geometric Modeling GAMBIT, a commercial pre-processor, is used to generate geometric models used in this study. FLUENT is used to solve the discretized governing equations. For post-processing the results obtained from FLUENT, TECPLOT are used. GAMBIT is used to create a volume mesh in the fluid domain. An unstructured mesh consisting of tetrahedral volumetric elements in three dimensions is used in the entire domain except at the walls, where a boundary layer mesh consisting of prismatic volumetric elements is used. Figure 2.4 shows a sample mesh consisting of triangular elements on the splitter and sample inlet. Figure 2.5 shows the unstructured mesh on the top inlet face along with the prismatic boundary layer mesh on the inlet pipe faces. The effect of using a boundary layer mesh around the fibers and particles is also examined by considering a case with an unstructured mesh in the entire domain. Pressure drops on the structured and unstructured mesh around the fibers and the particles varied by up to two to eight per cent; moreover, the unstructured mesh results had significantly larger numerical error as estimated by a grid refinement study. The accuracy of the fully tetrahedral mesh results could be increased by refining the mesh, but the additional refinement is expensive in both computation cost and solution time. As a result, a prismatic boundary layer mesh is used around the fibers and the particles to reduce the computation cost and to improve numerical accuracy. 38 Figure 2.3: Flow chart for segregated solver used in this study. 39 Figure 2.4: Unstructured triangular mesh on the faces of QMS flow channel. 40 Figure 2.5: Unstructured mesh along with the boundary layer on the inlet pipes and channel face. 41 2.3.9 Boundary Conditions Three different boundary conditions used in this study are no-slip, velocity inlet, and outflow. For laminar flows with zero velocity at the wall, the no-slip boundary condition is enforced at the walls. Velocity inlet boundary conditions are used to define the flow velocity along with all flow properties at the inlet and in some cases at outlet. Outflow boundary conditions are used to model flow exits where the details of the flow velocity and pressure are not known prior to solution of the flow problem. At outflow boundaries, all of the necessary boundary information is extrapolated from the interior. 2.3.10 Simulation Procedure An iterative solution approach is employed where the solution is advanced in pseudo- time until the steady state equations are satisfied to a specified tolerance. At each iteration the governing equations are solved using a segregated solver. A finite volume technique is used to solve the discrete form of the governing equation on the computational grid. The standard SIMPLE algorithm is used in which the momentum equations are first solved for the velocity components. A pressure correction is then determined that drives the velocity field towards satisfying the mass conservation equation. Iterative convergence is assessed by monitoring the L2 norms of the steady-state residuals. The convergence criterion is set to 10-7 for all conservation equations. 2.4 Results and Discussion 2.4.1 CFD Simulations: Flow Analysis and Pressure Drop Predictions One important feature of the flow channel is the design of the flow distributor placed at the carrier buffer inlet (b?) to obtain circumferentially uniform flow in the flow channel. Uniform flow around the core is important to maintain laminar flow in the annulus to yield 42 minimum crossover and maximum isolation. Improvements made to the flow distributor design were verified by CFD analysis. Figure 2.6 shows the flow distributor design with flow profile at different distances downstream from the distributor. Incoming fluid enters the distributor from two branches. Direct flow through the distributor from these branches is prevented by the lack of notches at the two inflow locations. The simulations show the uniform flow developed by the time fluid reaches the splitter end point. CFD simulations are performed using the generated flow-field meshes in order to predict the pressure drop and analyze the details of the flow in the flow channel. A plane is extracted at the center of the domain (i.e., Z=3.01625 mm) and displayed as a colored ring for the case with total flow rate 400 ml/min with inlet and outlet flow ratio 0.25. The velocity magnitude contours are shown in Figure 2.7a for sample (a?) velocity 0.05 m/s and buffer (b?) velocity 0.15 m/s. Careful examination of the calculated fluid velocity profiles revealed that the flow was fully developed within a relatively short distance of the splitter edge for every ratio of inlet and outlet flows. Figure 2.7b shows the velocity profile at the middle of the channel. Pressure magnitude contours are shown in Figure 2.8 along with the pressure gradient found from CFD simulations plotted against total flow rates compared with the theoretically calculated pressure drop using the Haigen-Poiseuille equation for annular pipes. Figure 2.8b shows that the pressure drop predicted using CFD is in good agreement with theoretical calculations with pressure drop increasing with the increase in total flow rate. 43 Figure 2.6: Velocity (m/s) contours in x plane at different distance from flow distributor to the end point of the splitter. (m/s) 44 2.4.2 Nonspecific Crossover The new design of the flow channel is tested for nonspecific crossover at different total flow rates ranging from 250 ml/min to 400 ml/min, input flow ratios (Qa?/Q), outlet flow ratios (Qa/Q) ranging from 0.25 to 0.7. When flow rates less than 250ml/min were used, cultisphers settled in the tubing and clogging of the inlet was observed. The effect of the particle concentrations in the sample and total flow rate on the crossover was examined. Some of these experiments were also conducted with pig pancreas (results not presented) and the same results as with cultisphers were observed. CFD simulations were conducted for all of these cases using the discrete phase model and compared with the experimental results. Nonspecific crossover (Sb) is calculated as the ratio of non-magnetic particles leaving in positive collection (Nb) to the particles leaving in both negative (Na) and positive collection. ba b b NN NS += (26) Figure 2.9 shows the nonspecific crossover of Cultisphers at different outlet flow ratios and at fixed inlet flow ratio of 0.25 and total flow rate of 400ml/min. These experiments were conducted with the flow channel design prototype I, in which the inlet and outlet splitters have the same 5.58mm diameter. A clear decrease in nonspecific crossover with increase in outlet flow ratio is observed because the increase in outlet flow ratio increases the transport lamina thickness. Figure 2.9 also presents the comparison of nonspecific crossover obtained from experiments with prototype I and CFD simulation-predicted crossover for prototype II. Good 45 Figure 2.7: a) Velocity (m/s) magnitude contours in the flow channel at z = 1.1875? (flow is in positive x direction. b) Velocity profile of the fluid at the middle point of the channel. 46 Figure 2.8: a) Pressure (Pa) contours in the flow channel at z = 1.1875? b) Comparison of the pressure drop from CFD simulations and empirical correlations. 47 agreement is found between experimental results and CFD simulation-predicted values with same prototype flow channel with less than 10% deviation. This deviation might be due to the negligence of particle-particle interaction in CFD modeling. CFD-predicted crossover for prototype II is compared with that of CFD predictions and experimental values with Prototype I. Crossover is slightly less for prototype II when compared to prototype I. The diameter of the outlet splitter is increased in prototype II which increases the transport lamina thickness and reduces the number of nonmagnetic particle that leave in the positive fraction. Straight line in the figure 2.9 represents the ideal flow of the fluid without solid particles. Figure 2.10 shows the crossover of nonmagnetic particles at different total flow rates in the Prototype I channel at a fixed inlet and outlet flow ratio of 0.25. Experimental crossover values increase with decreasing total flow rates as the hydrodynamic lift forces move the particles away from wall at low flow rates which helps particles to cross the transport lamina and leave in the positive fraction. Crossover obtained from CFD simulations for different total flow rates are under predicted by 5 to 10 percent for prototype I. Comparison of calculated crossover values at different total flow rates for the two prototypes shows a decrease in crossover for prototype II due to the increase in transport lamina thickness resulting from the outlet splitter position. Figure 2.11 shows crossover values at different sample concentrations at a fixed total flow rate of 400ml/min. Outlet flow ratio is maintained equal to inlet flow ratio at 0.25. Increasing cultisphers concentration increases the particle-particle interaction which results in more nonspecific crossover. Higher concentrations of particles during experiments also clogged the flow paths in the sample inlets. CFD simulations predict decreased crossover with the new design prototype II when compared with the prototype I. 48 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.0 0.1 0.2 0.3 0.4 0.5 0.6 S b Qa/Q Experimental-Prototype I CFD - Prototype-I CFD - Prototype-II Ideal Figure 2.9: Comparison of calculated and observed nonspecific crossover as a function of outlet flow ratio at a total flow rate of 400 ml/min and inlet flow ratio of 0.25. 49 300 320 340 360 380 400 0.06 0.07 0.08 0.09 0.10 0.11 0.12 0.13 0.14 0.15 S b Q, ml/min Experimental - Prototype I CFD - Prototype I CFD - Prototpe II Figure 2.10: Comparison of the nonspecific crossover as a function of total flow rate at an inlet and outlet flow ratio of 0.25. 50 0.10 0.15 0.20 0.25 0.30 0.05 0.06 0.07 0.08 0.09 0.10 0.11 S b Cultispheres Concentration, g/L Experimental - Prototype I CFD - Prototype I CFD - Prototype II Figure 2.11: Nonspecific crossover as a function of particle concentration in the sample at a total flow rate of 400ml/min, inlet and outlet flow ratio of 0.25. 51 2.5 Conclusions CFD simulations of flow pattern performed on the new design of QMS flow channel confirm circumferentially uniform flow development around the annular channel. Quantitative agreement between experimental measurements of nonspecific crossover and prediction based on CFD modeling of the fluid flow was shown. For all flow conditions, crossover predicted by CFD simulations was found to be slightly lower than experimentally observed results. This difference may be due to the contribution of crossover from other factors such as particle lift and particle interactions. Diffusion was not considered in the CFD modeling. The good agreement between experimental and CFD predicted results allowed the performance of simulations with different channel prototype models to develop a design to minimize nonspecific crossover. Though increasing the diameter of the outlet splitter decreases nonspecific crossover the diameter needs to be optimized based on the yield of magnetically labeled particles with change in splitter diameter. Therefore future CFD simulations must include the migration of the magnetic particles. 2.6 Acknowledgements This research was funded in part by the U.S. Department of Health and Human Services under SBIR grant 5R44DK072647-03 from the National Institute and Digestive and Kidney Research (NIDDK) awarded to Techshot, Inc., Greenville, Indiana, USA. 52 2.7 References Anderson JD. Computational Fluid Dynamics: The Basics with Applications. McGraw Hill, Inc., 1995. ISBN 0-07-001685-2. Brunicardi, FC, Oh Y, Shevlin L, Suh E, Kleinman R, Stein E, Lipaz G, Plant DV, Imagawa D, Fetterman HR. 1994. Laser destruction of human nonislet pancreatic tissue. Transplant Proc 26(6):3354-3355. Gray DWR, Gohde W, Carter N, Heiden T, Morris PJ. 1989. Separation of Pancreatic islets by Fluorescence-Activated Sorting. Diabetes 38(suppl 1):33-135. Jindal RM, McShane P, Gray DWR, Morris PJ. 1994. Isolation and Purification of Pancreatic Islets by Fluorescence Activates Islet Sorter. Transplant Proc 26(2):653. Jing Y, Chalmers JJ, Zborowski M. 2007. Blood progenitor cell separation from clinical leukapheresis product by magnetic nanoparticle binding and magnetophoresis. Biotechnol Bioeng 96(6):1139. Kennedy DJ, Todd P, Logan S, Becker M, Papas KK, Moore LR. 2007. Engineering quadrupole magnetic flow sorting for the isolation of pancreatic islets. Journal of Magnetism and Magnetic Materials 311:388-395. Lara O, Nakamura M, Zborowski M and Chalmers JJ. 2002. Negative depletion cell sorting using a quadrupole magnetic cell sorter. Eur. Cells Mater. 3:62-64. Moore LR, Rodriguez AR, Williams PS, McCloskey K, Bolwell BJ, Nakamura M, Chalmers JJ and Zborowski M. 2001. Progenitor cell isolation with a high capacity quadrupole magnetic flow sorter. J. Magn. Magn. Mater. 225:277-284. 53 Nakamura M, Decker K, Chosy J, Comella K, Melnik K, Moore LR, Lasky LC, Zborowski M and Chalmers JJ. 2001. Separation of Breast Cancel Cell Line from Human Blood Using a Quadrupole Magnetic Flow Sorter. Biotechnol. Prog. 17:1145-1155. Pinkse G, Steenvoorde E, Hogendoorn S, Noteborn M, Terpstra OT, Bruijn JA, and De Heer E. 2004. Stable transplantation results of magnetically retracted islets: a novel method. Diabetologia 47:55. Sakuma Y, Ricordi C, Miki A, Yamamoto T, Pileggi A, Khan A, Alejandro R, Inverardi L, Ichii H. 2008. Factors That Affect Human Islet Isolation. Transplant Proc 40:343-345. Samejima T, Yamaguchi K, Iwata H, Morkawa N, Ikada Y. 1998. Gelatin density gradient for isolation of islets of Langerhans. Cell Transplant 7(1):37- 45. Shenkman RM, Chalmers JJ, Hering BJ, Kirchhof N and Papas KK. 2009. Quadrupole Magnetic Sorting of Porcine Islets of langerhans, Tissue Engineering: Part C 15(2):147-156. Shenkman RM, Godoy-Silva R, Papas KK and Chalmers JJ. 2009. Effects of Energy Dissipation Rate on Islets of Langerhans: Implications for isolation and Transplantation, Biotechnology and Bioengineering 103(2):413-423. Soon-Shiong P, Fujioka T, Terasaki P, Heintz R, Lanza RP. 1990. Islet purification by a novel immunomicrosphere cell depletion technique. Transplant Proc. 22(2):780-781. Sun L, Zborowski M, Moore LR, Chalmers JJ. 1998. Continuous, Flow-Through Immunomagnetic Cell Separation in a Quadrupole Field. Cytometry 33:469. Tannehill JC, Anderson DA, Pletcher RH. 2004. Computational Fluid Mechanics and Heat Transfer. Taylor and Francis, second edition, ISBN 1-56032-046-X. 54 Tong X, Xiong Y, Zborowski M, Farag SS and Chalmers JJ. 2007. A Novel High Throughput Immunomagnetic Cell Sorting System for Potential Clinical Scale Depletion of T Cells for Allogeneic Stem Cell Transplantation. Exp Hematol 35(10):1613. Williams PS, Moore LR, Chalmers JJ and Zborowski M. 2003. Splitter Imperfections in Annular Split-Flow Thin Separation Channels: Effect on Nonspecific Crossover. Anal. Chem. 75:1365-1373. Williams PS, Decker K, Nakamura M, Chalmers JJ, Moore LR and Zborowski M. 2003a. Splitter Imperfections in Annular Split-Flow Thin Separation Channels: Experimental Study of Nonspecific Crossover. Anal. Chem. 75: 6687-6695. Williams PS, Hoyos M, Kurowski P, Salhi D, Moore LR and Zborowski M. 2008. Characterization of Nonspecific Crossover in Split-Flow Thin Channel fractionation. Anal. Chem. 80:7105-7115. Winoto-Morbach S, Ulrichs K, Leyhausen G and Muller-Ruchholtz W. 1989. New principle for large-scale preparation of purified human pancreas islets. Diabetes. 38(Supp 1):146-149. 55 3. Application of Magnetic Particle Tracking Velocimetry to Quadrupole Magnetic Sorting of Porcine Pancreatic Islets 3.1 Abstract Magnetic isolation is a promising method for separating and concentrating pancreatic islets of Langerhans for transplantation in Type 1 Diabetes patients. We are developing a continuous magnetic islet sorter to overcome the restrictions of current purification methods that result in limited yield, viability and purity of the isolated islets. In Quadrupole Magnetic Sorting (QMS) islets are magnetized by infusing superparamagnetic microbeads into islets? vasculature via arteries that serve the pancreas. The performance of the islet sorter depends on the resulting speed of the islets in an applied magnetic field, a property known as magnetophoretic mobility. Essential to the design and successful operation of the QMS is a method to measure the magnetophoretic mobilities of magnetically infused islets. We have adapted a magnetic particle tracking velocimeter (MPTV) to measure the magnetophoretic mobility of particles up to 1000?m in diameter. Velocity measurements are performed in a well-characterized uniform magnetic energy gradient using video imaging followed by analysis of the video images with a computer algorithm that produces a histogram of absolute mobilities. MPTV was validated using surrogate magnetic agarose beads and subjecting them to QMS. Mobility distributions of labeled porcine islets indicated that magnetized islets have sufficient mobility to be captured by the proposed sorting method, with this result confirmed in test isolations of magnetized islets. Keywords: Particle tracking velocimetry, magnetic flow sorter, pancreatic islets isolation, magnetic particles. 56 3.2 Introduction Magnetic isolation of islets is gaining in popularity due to its ease of use, speed and selectivity. Current islet isolation procedures depend on centrifugation in density gradients. They are limited in their throughput and separation efficiency, and they subject islets to potentially detrimental physical stresses (London et al., 1998; Soon-Shiong et al., 1989). To address the problems with throughput and efficiency, a continuous flow quadrupole magnetic islet separator was developed (Kennedy et al., 2007). Essential to the design and operation of such a device is a method to measure the magnetophoretic mobilities of magnetically infused islets. The magnetophoretic mobility of a particle indicates how responsive the particle is to an applied magnetic field. Paramagnetic particles have positive susceptibility and move toward increasing magnetic field intensity. Islets are rendered paramagnetic by infusing the donor pancreas with paramagnetic microspheres such as Dynabeads?. Beads capture efficiency varies with the age of the donor and pancreas weight and affects the magnetophoretic mobility of the islets. Several different techniques have been used to measure the magnetophoretic mobility of paramagnetic particles, including the use of visual microscopic observation of the movement of the labeled particle with stopwatch and magneto-cytometry (Gill et al., 1960; Davis et al., 1993). Magnetic Particle Tracking Velocimetry (MPTV) measures cell surface antigen counts through the quantification of antibody binding capacities (ABC) (McCloskey et al., 2001; Melnik et al., 2001; McCloskey et al., 2000; McCloskey et al., 2001a) and can be used to characterizes magnetophoretic mobilities of nano and micro-particles (Haefeli et al., 2002; Zhang et al., 2002; Chalmers et al., 1999) and living cells (Chalmers et al., 1999). Two important features of the 57 MPTV apparatus set it apart from other magnetic tracking technologies: 1) the use of an isodynamic (i.e. constant-force) magnetic field as indicated by ?Sm = constant?, and 2) capacity to track a large number of individual particles, up to 10,000, in a fraction of an hour. Both features are essential for high accuracy and precision of the MPTV analysis (Sridhra Redy et al., 1996; Moore et al., 2004; Zborowski et al., 2003). Those features are at the basis of the competitive advantage of the MPTV over other magnetic motion analyzers described in the literature that are based on single-particle analysis in high-gradient magnetic fields which are difficult to measure(Watarai et al., 2001; Wanabe et al., 2004; Suwa et al., 2004). MPTV leverages current technologies in video microscopy, computer processing speed and finite element analysis for magnetic fields to measure the induced motion of cells and particles in the highly characterized magnetic and gravity field. The motion of the particles in the known field is then translated into a characteristic parameter called magnetophoretic mobility in a magnetic field and sedimentation rate in a gravity field. MPTV measurements involve videotaping the movement of immunomagnetically labeled particles through a known viscosity medium and magnetic susceptibility in a well-defined magnetic energy density gradient. The velocity of each particle along with its location within the magnetic energy gradient is recorded. From this information, the magnetophoretic mobility of each particle is calculated. A significant advantage of this method over other techniques is that large numbers of cells or particles can be processed in a very short time. Present MPTV devices developed to measure magnetophoretic mobility of cells have problems with flow cells and data analysis software. They also have fixed amount of magnetic field strength and are only able to measure mobilities of single cells. The MPTV described here in measures magnetophoretic mobilities and sedimentation rates of cells and large particles like 58 islets with reliable software and variable magnetic field strength. The software developed for MPTV also gives the optimum parameters to use with QMS to achieve isolation with maximum purity and yields. This paper explains the features of a refined MPTV system with results obtained with large particles including porcine pancreatic islets of Langerhans. 3.3 Theory The following analysis applies to the use of MPTV applied to the isolation of magnetically labeled pancreatic islets of Langerhans by quadrupole magnetic sorting. The magnetophoretic mobility (?m) of a particle is defined as the ratio of the velocity of the particle in the magnetic field, uf, to the magnetic field energy gradient, Sm. m f m S u=? (1) Any population of magnetically labeled particles exhibits a statistical distribution of ?m based on several characteristics. Mobility of the pancreatic islets depends on the number of magnetic Dynabeads? infused into each islet. The following equation relates the Dynabeads and islet properties to the magnetophoretic mobility of an immunomagnetically labeled islet. i iD m D Vn pi? ?? 3 ?= (2) Where nD is the number of Dynabeads, ?? is the difference in magnetic susceptibility between the medium and the beads, Vi is the volume of a bead, ? is the viscosity of the medium, Di is the diameter of the islet. The magnetophoretic mobility distribution measured using MPTV is used to predict the QMS output fractions based on the defined flow rate parameters. QMS is a split flow type continuous magnetic sorter developed to separate single cells by labeling specific cells with magnetic beads. QMS was improved to isolate Dynabeads infused islets from exocrine tissue 59 based on their magnetophoretic mobility (?m) (Kennedy et al., 2007). Figure 3.1 describes the basic QMS mechanism for isolation of magnetic cells from nonmagnetic cell populations. QMS can be operated continuously with labeled islets with unlabeled tissue entering at inlet a? (sample) and carrier buffer at inlet b? (buffer) and isolated islets exiting the system at outlet b (positive), unlabeled tissue at outlet a (negative), some of the tissue which are labeled high enough to reach the wall of the flow channel will stick to the wall and can be collected at end of the isolation process and is considered as Wall (W) fraction. The boundary between inlet flows is called as Inner Splitting Surface (ISS) and the boundary between outlet flows is called as Outer Splitting Surface (OSS). Distance between ISS and OSS is called the Transport Lamina. The first critical mobility whereby a particle entering the flow channel at the ISS just reaches the OSS and is eluted into the b (positive) fraction is identified as ?m0 and is defined by 0 0 2 1 1 00 0 )ln( )1( ),( 2 m ISSOSSs i OSSISS m m LS vr A I LSr Q ll l ll + ?= pi? (3) Where il is the radial position given by ii rr=l and OSSISSo lll ,, correspond to wall, ISS and OSS. L is distance between two splitters. 1A , 2A and ),( 11 llI are functions used to simplify the calculations and are given by )1( 221 AA i ?+= l (4) i iA l l 1ln 1 2 2 ?= (5) l llllll 1])(ln22ln4[),( 2 2 2 11 AI +?= (6) The radial position of the ISS, ISSl is calculated using the following integration factor (Kennedy et al., 2007): 60 Figure 3.1: Schematic diagram of the Quadrupole Magnetic cell Sorter (QMS) Sample Inlet (a?) Buffer Inlet (b?) Positive Fraction (b) NegativeFraction (a) Magnet Non-Magnetic Particle Magnetic ParticleISS OSS Magnetic ForceT ra ns po rt La mi na Splitter Tr an sp or t L am ina 61 ISS i AAI ISSi lllllllll ]ln22[),( 2222422 ?+?= (7) Similarly, the position of the OSS can be calculated using eq. (7) by replacing ISSl with OSSl . The second critical mobility whereby a particle entering at the wall of the core reaches the OSS and exit with the b fraction is identified as ?m1and is defined by 0 0 2 1 1 00 1 )ln( )1( ),( 2 m iOSSs i OSSi m m LS vr A I LSr Q ll l ll + ?= pi? (8) The third critical mobility whereby a particle entering at the ISS reaches the wall of the shell and may be trapped in the flow channel is identified as ?m2 and is defined by 0 0 2 1 1 00 2 )ln( )1( ),( 2 m ISSos i oISS m m LS vr A I LSr Q ll l ll + ?= pi? (9) The final critical mobility whereby a particle entering at the core wall reaches the wall of the shell and remains on the wall is identified as ?m3 and is defined by 0 0 200 3 )ln( 2 m ois m m LS vr ALSr Q ll+= pi? (10) All the particles with m