<%BANNER%>

Toward Developing a Bioartificial Pancreas

Permanent Link: http://ufdc.ufl.edu/UFE0043893/00001

Material Information

Title: Toward Developing a Bioartificial Pancreas Optimization of a Bioartificial Pancreas and Development of a Tcho Calibration Curve with an Inductively Coupled Implantable Coil
Physical Description: 1 online resource (81 p.)
Language: english
Creator: Neelam, Srujana
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2011

Subjects

Subjects / Keywords: bioartificial -- pancreas
Biomedical Engineering -- Dissertations, Academic -- UF
Genre: Biomedical Engineering thesis, M.S.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Type 1diabetes (T1D) is an autoimmune disease that is most often diagnosed in children, adolescents and young adults. In T1D, the immune system attacks and destroys the insulin producing cells of the pancreas, leading to an inability for proper insulin secretion in response to elevated blood glucose levels. The bioartificial pancreas is a cell-based device made from insulin-secreting cells entrapped in biocompatible components, and designed to replace the insulin secreting cells of the endocrine pancreas. Once implanted, its function is to sense glucose, and secrete insulin as necessary. This device may have potential as a long term treatment for T1D. To predict failure of the implant before it stops regulating blood glucose levels, the implanted device needs to be monitored for structural integrity, metabolic activity, longevity and function, with least inconvenience to the patient. It has been demonstrated in the past that nuclear magnetic resonance imaging and spectroscopy can be used to monitor the implant. To improve signal-to-noise (S/N) and enhance the detection of cells, an approach using an implantable, inductively coupled coil system was used. Total choline (TCho) is a combination of various metabolites such as phosphocholine, glycerol-3-phosphocholine and free choline. These metabolites are components of the cell membrane which can be used to indicate viable cells. Therefore, TCho can be used as a marker to determine the total number of viable cells and monitor the implant over time. The two fundamental goals of the project are to: optimize a manufacturing technique for the bioartificial pancreas; and develop a TCho calibration curve for NMR detection which can predict cell viability. Various alginates and gelling methods were tested to encapsulate the insulin secreting cells. This bioartificial pancreas was tested in vitro and in vivo for its structural integrity and glucose consumption/ insulin secretion. A manufacturing method was found that allowed for implants to regulate blood sugars in diabetic mice. Using NMR techniques, a TCho calibration curve was developed. 1H NMR spectroscopy detection was achieved by inductive coupling between the implantable coil and a surface coil in an 11 Tesla magnet using the LASER pulse sequence. This calibration curve will be used to determine the number of cells in the implanted construct and predict the construct failure. This periodic monitoring will help determine the functionality of the implant and be useful to determine if the body can maintain normoglycemia with the implant or needs replacement with a new implant.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Srujana Neelam.
Thesis: Thesis (M.S.)--University of Florida, 2011.
Local: Adviser: Keselowsky, Ben.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2013-12-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2011
System ID: UFE0043893:00001

Permanent Link: http://ufdc.ufl.edu/UFE0043893/00001

Material Information

Title: Toward Developing a Bioartificial Pancreas Optimization of a Bioartificial Pancreas and Development of a Tcho Calibration Curve with an Inductively Coupled Implantable Coil
Physical Description: 1 online resource (81 p.)
Language: english
Creator: Neelam, Srujana
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2011

Subjects

Subjects / Keywords: bioartificial -- pancreas
Biomedical Engineering -- Dissertations, Academic -- UF
Genre: Biomedical Engineering thesis, M.S.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Type 1diabetes (T1D) is an autoimmune disease that is most often diagnosed in children, adolescents and young adults. In T1D, the immune system attacks and destroys the insulin producing cells of the pancreas, leading to an inability for proper insulin secretion in response to elevated blood glucose levels. The bioartificial pancreas is a cell-based device made from insulin-secreting cells entrapped in biocompatible components, and designed to replace the insulin secreting cells of the endocrine pancreas. Once implanted, its function is to sense glucose, and secrete insulin as necessary. This device may have potential as a long term treatment for T1D. To predict failure of the implant before it stops regulating blood glucose levels, the implanted device needs to be monitored for structural integrity, metabolic activity, longevity and function, with least inconvenience to the patient. It has been demonstrated in the past that nuclear magnetic resonance imaging and spectroscopy can be used to monitor the implant. To improve signal-to-noise (S/N) and enhance the detection of cells, an approach using an implantable, inductively coupled coil system was used. Total choline (TCho) is a combination of various metabolites such as phosphocholine, glycerol-3-phosphocholine and free choline. These metabolites are components of the cell membrane which can be used to indicate viable cells. Therefore, TCho can be used as a marker to determine the total number of viable cells and monitor the implant over time. The two fundamental goals of the project are to: optimize a manufacturing technique for the bioartificial pancreas; and develop a TCho calibration curve for NMR detection which can predict cell viability. Various alginates and gelling methods were tested to encapsulate the insulin secreting cells. This bioartificial pancreas was tested in vitro and in vivo for its structural integrity and glucose consumption/ insulin secretion. A manufacturing method was found that allowed for implants to regulate blood sugars in diabetic mice. Using NMR techniques, a TCho calibration curve was developed. 1H NMR spectroscopy detection was achieved by inductive coupling between the implantable coil and a surface coil in an 11 Tesla magnet using the LASER pulse sequence. This calibration curve will be used to determine the number of cells in the implanted construct and predict the construct failure. This periodic monitoring will help determine the functionality of the implant and be useful to determine if the body can maintain normoglycemia with the implant or needs replacement with a new implant.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Srujana Neelam.
Thesis: Thesis (M.S.)--University of Florida, 2011.
Local: Adviser: Keselowsky, Ben.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2013-12-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2011
System ID: UFE0043893:00001


This item has the following downloads:


Full Text

PAGE 1

1 TOWARDS DEVELOPING A BIOARTIFICIAL PANCRE AS: OPTIMIZATION OF A BIOARTIFICIAL PANCRE AS AND DEVELOPMENT O F A TCHO CALIBRATION CURVE WITH AN INDUCTIVELY COUPLED IMPLANTABLE COIL By SRUJANA NEELAM A THESIS PRESENTED TO THE GRADUATE SCHOOL OF T HE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2011

PAGE 2

2 2011 SRUJANA NEELAM

PAGE 3

3 To mom and d ad

PAGE 4

4 ACKNOWLEDGMENTS I am really fortunate to have Dr. Nicholas E. Simpson as my Principal Investigator. I would like to express my gratitude for his expert guidance, support and encouragement throughout the project work and research. I especially thank Dr. Huadong Zeng for training me in performing NMR studies. I also would l ike to thank my committee members, Dr. Benjamin Keselowsky, Dr. Thomas Mareci and Dr. Johannes van Oostrom, for understanding and supporting me. I thank Barbara Beck, Garrett Astary, and Kelly Jenkins for their assistance in teaching me how to build coils construct the phantom, and properly set up animals for NMR studies. Most importantly, I want to thank the entire Simpson research lab for all their assistance with this project. Although the work presented in Chapter 3 was essentially performed by me a lone, much of the work outlined in Chapter 2 would have been impossible to complete without the team effort of this group. To this, I want to especially thank Mark Beveridge for helping me with cell culture and animal work, and lab mates Michelle Corrado and Ansuya Deosaran for their help with the table top experiments to optimize the bioartificial pancreas, manufacturing of the constructs and their assistance with the surgical implantations, and monitoring of the animals post implantation. Their efforts have made this thesis possible.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .................... 4 LIST OF TABLES ................................ ................................ ................................ ................................ 8 LIST OF FIG URES ................................ ................................ ................................ .............................. 9 ABSTRACT ................................ ................................ ................................ ................................ ........ 11 CHAPTER 1 DIABETES: THE CASE FOR A BIOARTIFICIAL PANCREAS ................................ .... 13 1.1 Overview ................................ ................................ ................................ ............................... 13 1.1.1 Pancreas ................................ ................................ ................................ .......................... 13 1.1.2 Diabetes ................................ ................................ ................................ .......................... 13 1.1.3 Diagnosis ................................ ................................ ................................ ........................ 15 1.1.4 Various Treatments and their Drawbacks ................................ ................................ .... 15 1.2 Bioartificial Pancreas ................................ ................................ ................................ ........... 17 1.3 Non invasive Monitoring Methods for Implants ................................ ................................ 19 2 OPTIMIZING THE BIOARTIFICIAL PANCREAS ................................ ........................... 22 2.1 Introduction ................................ ................................ ................................ ........................... 22 2.2 Background ................................ ................................ ................................ ........................... 22 2.2.1 Insulin Producing Cells ................................ ................................ ................................ 23 2.2.2 Beta Tumor Cells (TC) ................................ ................................ ................................ 23 2.2.3 Biomaterials ................................ ................................ ................................ ................... 24 2.2.4 Alginate ................................ ................................ ................................ .......................... 25 2.2.5 PDMS ................................ ................................ ................................ ............................. 27 2.3 Aim of the Study ................................ ................................ ................................ .................. 27 2.4 Materials and Methods ................................ ................................ ................................ ......... 27 2.4.1 Cell Culture ................................ ................................ ................................ .................... 27 2.4.2 Cell Entrapment in Beads ................................ ................................ ............................. 28 2.4.3 Entrapment in Macroconstructs ................................ ................................ .................... 29 2.4.4 Animal Preparation and Handling ................................ ................................ ................ 31

PAGE 6

6 2.4.5 Glucose Consumption ................................ ................................ ................................ ... 32 2.4.6 Histology ................................ ................................ ................................ ........................ 32 2.4.7 Von Kossa Stain ................................ ................................ ................................ ............ 33 2.5 Results ................................ ................................ ................................ ................................ ... 33 2.6 Discussion and Future Work ................................ ................................ ............................... 44 3 DEVELOPMENT OF THE TCHO CALIBRATION CURVE ................................ ........... 50 3.1 Introduction ................................ ................................ ................................ ........................... 50 3.2 Implantable Coils ................................ ................................ ................................ .................. 53 3.3 Improving Coil Sensitivity ................................ ................................ ................................ ... 54 3.4 LASER Pulse Sequence ................................ ................................ ................................ ....... 55 3.5 Background ................................ ................................ ................................ ........................... 55 3.6 Aim of the Study ................................ ................................ ................................ .................. 57 3.7 Materials a nd Methods ................................ ................................ ................................ ......... 57 3.7.1 Implantable Coil ................................ ................................ ................................ ............ 57 3.7.2 Coating of the Implantable Coil ................................ ................................ ................... 57 3.7.3 Choline Samples ................................ ................................ ................................ ............ 58 3.7.4 Cell Culture ................................ ................................ ................................ .................... 59 3.7.5 RF Coil ................................ ................................ ................................ .......................... 59 3.7.6 Animal Work and Surgery ................................ ................................ ............................ 61 3.7.7 Preparation for NMR Measurements ................................ ................................ ........... 61 3.7.8 Shimming ................................ ................................ ................................ ....................... 62 3.7.9 NMR Imaging and Spectroscopy Measurements in vitro ................................ ........... 62 3.7.10 tet cells ................................ ................................ .... 64 3.8 Results ................................ ................................ ................................ ................................ ... 6 5 3.9 Discussion and Future Work ................................ ................................ ............................... 69 3.10 Conclusion ................................ ................................ ................................ .......................... 71 APPENDIX A PROTOCOL FOR MAKING ALGINATE BEADS ................................ ............................ 73 B PROTOCOL FOR MAKING BIOARTIFICIAL PANCREAS WITH 3% LVM BEADS IN 2% LVG ................................ ................................ ................................ .............................. 74

PAGE 7

7 C PROTOCOL FOR MAKING A BIOARTIFICIAL PANCREAS WITH 3% LVM BEADS ALGINATE MIXES ................................ ................................ ................................ ................ 74 D PROTOCOL FOR MAKING 3% LVM OR 2% LVG PLUGS ................................ ........... 75 E PROTOCOL FOR CASTING PDMS ................................ ................................ .................... 75 F PROTOCOL FOR MAKING AN IMPLANTABLE RF COIL ................................ ........... 75 G PROTOCOL FOR THE VON KOSSA STAIN ................................ ................................ .... 76 REFERENCES ................................ ................................ ................................ ................................ ... 77 BIOGRAPHICAL SKETCH ................................ ................................ ................................ ............. 81

PAGE 8

8 LIST OF TABLES Table page 2 1 Alginate properties ................................ ................................ ................................ ................ 26 2 2 Summary ................................ ................................ ................................ ................................ 44 3 1 Resonant frequencies of coils before and after the PDMS coating with their Q factors. ................................ ................................ ................................ ................................ ..... 61

PAGE 9

9 LIST OF FIGURES Figure P age 2 1 Structural formula of a GGMM guluronic mannuronic acid sequence in alginic acid ...... 26 2 2 Egg box model of alginate gel ................................ ................................ ............................... 26 2 3 Hemat oxylin/eosin stained cross sections of Variant 1 plugs A) plug 1 on Day 1, B) plug 2 on Day 1. ................................ ................................ ................................ ..................... 33 2 4 Hematoxylin/eosin stained cross sections of Variant 1 plugs A) plug 1 on Day 28, B) plug 2 on Day 28. ................................ ................................ ................................ ................... 34 2 5 I n vitro glucose consumption by the Variant 1 bioartificial pancreas. .......................... 34 4 2 6 G lucose response in vivo studies. ................................ ................................ ..................... 35 5 2 7 The above picture was taken under the microscope: a Variant 1 bioartificial pancreas in which the microbeads all turned white. ................................ ................................ ............ 36 2 8 Cross sections of Variant 1 bioartificial pancreas removed from the mouse peritoneal cavity.. ................................ ................................ ................................ ................................ ... 36 6 2 9 Hematoxylin/eosin stained cross sections of Variant 2 plugs.. ................................ ......... 37 7 2 10 Hematoxylin/eosin stained cross sections of Variant 2 plugs.. ................................ ........... 38 2 11 I n vitro glucose consumption by the Variant 2 bioartificial pancreas. .......................... 38 8 2 12 G lucose response in vivo studies. ................................ ................................ ....................... 40 2 13 The above picture was taken under the microscope: beads in a Variant 2 bioartificial pancr eas did not turned white with mineralization. ................................ ........................... 40 0 2 14 Cross sections of Variant 2 bioartificial pancreas removed from the mouse peritoneal cavity.. ................................ ................................ ................................ ................................ ..... 41 2 15 Cross sections of Variant 3 and 4 bioartificial pancreas removed from the mouse peritoneal cavity after 30 days. ................................ ................................ .......................... 42 2 2 16 I n vivo blood glucose response in diabetic mice to alginate implants. ............................ 43 3 1 A spinning nucleus in a magnetic field. ................................ ................................ ................ 51 3 2 Implantable c oil design and the bioartificial pancreas pl ug ................................ ............... 58 3 3 P hotographs of in vitro studies to generate the choline calibration curve. l. .................... 60

PAGE 10

10 3 4 NMR images and spectroscopy data obt ained with an inductively coupled implantable coil system. ................................ ................................ ................................ ..... 64 3 5 The graph is plotted between the choline concentration and ratio between area under choline peak and area under water peak.. ................................ ................................ ............. 66 3 6 The graph is a plot between the choline concentration and the ratio between the area under the choline peak and area under the water peak. ................................ ....................... 67 3 7 The graph is plotted between the choline concentration and ratio between area under choline peak and area under water peak. ................................ ................................ ............ 68 3 8 The graph is plotted between the cell number in volume of interest and ratio between area under choline peak and area under water peak.. ................................ ........................... 69

PAGE 11

11 Abstract o f Thesis Presented t o T he Graduate School o f The University o f Florida In Partial Fulfillment o f The Requirements f or The Degree o f Master Of Science TOWARDS DEVELOPING A BIOARTIFICIAL PANCREAS: OPTIMIZATION OF A BIOARTIFICIAL PANCREAS AND DEVELOPMENT OF A TCHO CALIBRATION CURVE WITH AN INDUCTIVELY COUPLED IMPLANTABLE COIL By Srujana Neelam December 2011 Chair: Benjamin G. Keselowsky Major: Biomedical Engineering Type 1diabetes (T1D) is an autoimmune disease that is most often diagnosed in children, adolescents and young adults. In T1D, the immune system attacks and destroys the insulin producing cells of the pancreas, leading to an inability for proper insulin secretion in response to elevated blood glucose levels. The bioartificial pancreas is a cell based device made from insulin secreting cells entrapped in bioco mpatible components, and designed to replace the insulin secreting cells of the endocrine pancreas. Once implanted, its function is to sense glucose, and secrete insulin as necessary. This device may have potential as a long term treatment for T1D. To p redict failure of the implant before it stops regulating blood glucose levels, the implanted device needs to be monitored for structural integrity, metabolic activity, longevity and function, with least inconvenience to the patient. It has been demonstrat ed in the past that nuclear magnetic resonance imaging and spectroscopy can be used to monitor the implant. To improve signal to noise (S/N) and enhance the detection of cells, an approach using an implantable, inductively coupled coil system was used. T otal choline (TCho) is a combination of various metabolites such as phosphocholine, glycerol 3 phosphocholine and free choline. These metabolites are components of the cell membrane which can be used to indicate

PAGE 12

12 viable cells. Therefore, TCho can be used as a marker to determine the total number of viable cells and monitor the implant over time. The two fundamental goals of the project are to: optimize a manufacturing technique for the bioartificial pancreas; and develop a TCho calibration curve for NMR d etection which can predict cell viability. Various alginates and gelling methods were tested to encapsulate the insulin secreting cells. This bioartificial pancreas was tested in vitro and in vivo for its structural integrity and glucose consumption/ ins ulin secretion. A manufacturing method was found that allowed for implants to regulate blood sugars in diabetic mice. Using NMR techniques, a TCho calibration curve was developed. 1 H NMR spectroscopy detection was achieved by inductive coupling between the implantable coil and a surface coil in an 11 Tesla magnet using the LASER pulse sequence. This calibration curve will be used to determine the number of cells in the implanted construct and predict the construct failure. This periodic monitoring will help determine the functionality of the implant and be useful to determine if the body can maintain normoglycemia with the implant or needs replacement with a new implant.

PAGE 13

13 CHAPTER 1 DIABETES: THE CASE FOR A BIOARTIFICIAL PANCREAS 1.1 Overview 1.1.1 Pa ncreas The pancreas is a relatively small but important organ of the digestive system located under the stomach and alongside the upper small intestine. It has both exocrine and endocrine functions. Its exocrine function is to release enzyme containing j uices into the digestive tract to help in the digestion of fats, proteins, and carbohydrates. Approximately 98% of the pancreas is used for this exocrine function whereas only 2% of pancreas functions as a part of endocrine system to regulate blood glucos e levels. This endocrine portion is scattered throughout the pancreas in small clusters of cells, called the Islets of Langerhans. cells secrete hormones called insulin and glucagon, respectively, which help regulate blood glucose levels. cells normally make up more than half the cell mass of the islets. 1 4 1.1.2 Diabetes The normal blood glucose level in humans is between 70 99 mg/dL 5 The body converts excess sugars to compounds it can store as fuel and use later. The hormone insulin helps in this conversion by signaling cells that the excess glu cose is available. Insulin increases the cellular uptake of glucose, inhibits the utilization of stored fuels, and increases the cellular production and storage of fuels such as glycogen and fat. If the body does not respond to the insulin hormone (e.g., lack of or problem with the insulin receptor), or if there is insufficient insulin secreted in response to rising blood glucose levels, blood glucose levels cannot be regulated and will remain elevated. This condition is called hyperglycemia, and it is a hallmark of diabetes mellitus. Long term hyperglycemia leads to health problems discussed below. One form of

PAGE 14

14 diabetes mellitus is caused by the immune destruction of the cells. When these cells are destroyed there is an absence of insulin which lead s to increased blood glucose levels. This is called type 1 diabetes mellitus (T1D), also known as juvenile or insulin dependent diabetes. This disease often occurs as a many year process, with the onset generally in children or adolescents. 1 4 6 8 Diabetes mellitus affects almost every organ in the body, and a major reason f or this is the effect of high blood glucose levels on the vasculature. The rationale for this profound effect is as follows. Cells in the body need a constant supply of oxygen, otherwise the cells may die. The heart pumps oxygenated blood to all parts o f the body through the arterial blood vessels which eventually branch out into the smallest vessels, called capillaries. The red blood cells (RBCs) in the blood carry the oxygen and deliver this oxygen to cells primarily at the capillary level. The deoxy genated blood continues back to the heart, where it is pumped to the lungs for oxygenation, and then back to the heart for another circuit. When the sugar levels in the blood go above normal, the glucose tends to bind to protein or lipids around them, i.e ., to the hemoglobin present enzymatic glycosylation. Due to glycation, the blood vessels stiffen and become fragile. This may prove to be fatal as the fragile and stiff blood vessels are susceptible to aneurisms. This stiffness does not allow the vessels to dilate well, thereby leading to high blood pressure. The stiffening of capillaries also makes it difficult for the RBCs to travel, reach tissues, and deliver the n ecessary oxygen. This eventually leads to death of tissue or decreases in organ performance. Some examples of the consequences of this process follow. One common condition in people suffering with diabetes, and a leading cause of blindness, is diabetic retinopathy. As capillaries in the eye in a diabetic become very fragile, they tend to burst leading

PAGE 15

15 to bleeding in front of the retina. As a result, the light does not fall on the retina and the vision is blurred. Eventually the person may lose his sig ht completely. Another example is how high blood sugar levels also damage the kidney. Due to the high levels of sugars in the blood that enters the kidney, the capillaries critical for the filtration function of the kidney are damaged. Over time, the ki dney is unable to regulate blood pressure or filter all the waste and toxic products in the body. This condition is diabetic nephropathy. Lastly, poor circulation in the extremities leads to an inability to heal, and a loss of nerve sensation (neuropathy ). These conditions make infections easy to start and difficult to treat, often resulting in amputation of limbs. 3 9 1.1.3 Diagnosis T1D can occur at any age, but generally it is observed in persons less than 30 years of age. The islet destruction is faster in children as compared to adults. 10 The symptoms of diabetes include increased thirst (polydipsia), frequent urination (polyuria ), itching (pruritis), tiredness, unexplained weight loss, leg cramps or pains, delayed healing of skin wounds, and recurrent infections of the skin, genitalia or urinary tract. 11 1.1.4 Various Treatments and their Drawbacks Insulin injections: this is the standard therapy in which insulin is injected subcutaneou sly several times a day, the amount and frequency of injection determined by frequent blood sugar monitoring and analysis. Unfortunately, the patient may forget or neglect injections, which will again lead to high glucose levels. Poor glycemic control in patients is common in diabetic patients. Moreover, even in well regulated diabetics, blood glucose levels are generally higher than normal, and over time, the devastating consequences as described above will occur.

PAGE 16

16 Because of these issues, open and clos ed loop insulin delivery systems, described below, have been developed. Open loop insulin infusion: in this method, insulin is delivered at a predetermined rate through an external insulin pump, which pumps small amounts of insulin subcutaneously into the patient, irrespective of the blood glucose levels. This is not ideal, as it may lead to hypoglycemia if insulin is released when the patient is having normal blood glucose levels, or hyperglycemia, if the glucose load is larger than the infused insulin c an handle. Therefore, it is still critical for the patient to monitor their blood sugars and adjust the insulin dosage appropriately. The patient can control or override the system when necessary, as before or after a large meal, or after exercise. The above methods, direct insulin injection or the open loop delivery system, may regulate blood glucose levels, but they still cannot reach the accuracy of a normal physiological regulation system. Therefore, a closed loop insulin delivery system, an artific ial pancreas approach, has been explored. Closed loop insulin pump therapy: this system uses feedback from a glucose sensor which measures the blood glucose levels of the patient and then releases the required amount of insulin. Technology forms the basi The major components of artificial pancreas are the glucose sensor, insulin pump, delivery catheter, computer program, battery and insulin storage unit. Although this is a very promising approach which is underg oing a great deal of attention today, the problems that can arise from this closed loop approach are due to component issues: sensor failure, pump failure, clogging of the catheter, computer or software malfunction, or loss of battery power. If any of the se problems arise, the patient cannot be regulated properly, and will experience potential life threatening hyperglycemia or hypoglycemia. Also, the insulin needs to be replenished regularly 3 which is a problem if the closed system is implanted.

PAGE 17

17 Other older biological approaches that have been used clinically with some success include whole organ transplantation, and more recently, islet transplantation. Whole organ transplantation: in this approach, the whole pancreas is transplanted from a donor to a recipient (TID patient). This procedure eliminates the need for exogenous insulin treatment, frequent blood glucose meas urement and extensive dietary restrictions, as the new pancreas behaves and performs normally. However, this approach comes with a whole new set of complications which include intra abdominal infections due to surgery, and significant complications caused by lifelong immunosuppressive therapy. Also the number of donor organs is severely limited, as the organs are harvested from cadavers. Islet transplantation: in this technique, only the insulin secreting islets from a donor cadaver pancreas are isolated and given to the recipient. These islets lodge into the liver after infusion through the portal vein via a catheter. However the drawbacks for this therapy are similar to those of whole organ transplantation: a single recipient needs two or more cadaver ic donors for a successful transplantation, and the receiver needs to undergo lifelong immunosuppressive therapy 3 12 1.2 Bioartificial Pancreas Another potential treatment for T1D is the bioartificial pancreas, an approach which avoids many of the issues of the therapies described above. The bioartificial pancreas has both a biological component and an artificial com ponent: insulin secreting cells entrapped in a biocompatible polymer. The biopolymer is a material or gel which allows for exchange of oxygen, carbon dioxide, insulin and other nutrients, but does not allow for larger molecules like antibodies and white b lood cells to reach the cells. In this device, the cells make and store insulin, sense glucose level of the body, and secrete appropriate amounts of insulin. Ideally, the material and implant site provides a good environment for the cells to perform the same function

PAGE 18

18 cells, and respond to blood glucose. The main advantages of a bioartificial pancreas are that its behavior is similar to normal physiological behavior, the patient need not be on immunosuppressive drugs, there i s no problem of donor shortage as the cells would come from tissue culture, and the device responds immediately 13 There are three main types of bioartificial pancreas that have been studied: vascular devices, macroencapsulation devices and microencapsulation devices. The vascular device consists of a hollow tubular structure surrounded by a chamber which holds the pancreatic islets or insulin secreting cells 12 The hollow fiber is used as a shunt b etween the artery and vein. The glucose containing blood flowing through the hollow tube stimulates the cells to release insulin into the blood. However, this method needs a surgical procedure involving the vasculature, which may further lead to other co mplications such as thrombosis. It is not presently being pursued as a viable option any longer. Microencapsulation is an approach in which islets or other insulin secreting cells are entrapped into polymer beads smaller than 2 mm in diameter. These bea ds can be injected into the body and need no difficult surgical procedure 22 However, although this approach is effective in treating animal models of diabetes, the main drawback of this treatment is that the beads cannot be retrieved in case of complica tions or failure. That renders this approach unusable for humans. Macroencapsulation of islets or other insulin secreting cells is an approach in which the cells are embedded into a polymer of larger size. This so called macroconstruct may be created in any shape, though cylinders or sheets are common. The advantage of a macroconstruct is that it can be retrieved for replacement or in case of complications. The disadvantages of this approach include possible cell necrosis towards the center due to diff usional problems causing a lack of nutrients and oxygen 13 14 Because there is always a need to monitor implants fo r their structural integrity,

PAGE 19

19 functionality, performance and cell viability 15 we are interested in devel oping a bioartificial pancreas which is functional and can be monitored noninvasively using NMR techniques. A detailed description of the biomaterials used to manufacture our macroconstruct is given in Chapter 2. Our NMR approach to monitor the implanted macroconstruct is detailed in Chapter 3, but below are various techniques that have been used to monitor implants. 1.3 Non invasive Monitoring Methods for Implants Tissue engineered constructs and acellular scaffold implants need to be monitored non invas ively to provide spatial and temporal information for in vivo experiments. There are different monitoring techniques which provide different spatial and temporal resolutions, have different depths of penetration, and have various strengths. These differe nt techniques can be used to monitor cell viability, metabolic pathways, protein protein interaction, gene expression or molecular processes. Here are brief descriptions of some of these techniques. Positron emission tomography (PET) and single photon em ission computed tomography (SPECT) techniques can be used to image radionuclide probes. PET and SPECT provide detailed spatial and temporal images by tracking emission from injected radionuclide probes. PET is the most common radionuclide modality used t o monitor cells because of its high sensitivity. PET imaging has also been applied to the study of beta cells, specifically alterations in beta cell mass through the development of TID, although the imaging resolution for the small percentage of beta cell s within the pancreas is a significant limiting factor 16 17 Charged coupled device (CCD) techniques detect light to form images 18 This approach has good thermal noise reduction and increased sensitivity. Another light approach is bioluminescence imaging techniques, which tra ck the firefly luciferase reporter gene expressed by implanted cells.

PAGE 20

20 The main goal of tissue engineering is to replace, repair or enhance biological function at the tissue or organ level. Cells or tissues are transplanted into a suitable polymeric matrix or scaffold. Also, it is important to monitor these tissue engineered constructs regularly, yet cause least inconvenience to the patient. A versatile monitoring technique that is arguably superior for observing tissue engineered constructs is nuclear ma gnetic resonance (NMR). NMR is a noninvasive technique that can monitor the tissue engineered constructs or tissue in the body both structurally and functionally 19 20 It is important to study how the biomaterials affect the functionality of the cells/tissue, and it is important to monitor the construct and predict the functionality, observe remodeling in vitro and i n vivo and determine its structural integrity. NMR is a tool that also can monitor the metabolic activity of the bioartificial constructs. NMR can be used to image and study surface and deep seated tissue with high resolution. Additionally, the cells i n the samples can be studied for their metabolic activities without any need for further genetic modification. Some tissue engineered constructs that have been monitored noninvasively using NMR technique are cardiac scaffolds, bone grafts and tissue engine ered cartilage. Following is a brief description for each. Cardiac scaffold: Patients who suffer from myocardial infarction have damaged heart wall tissue. To support this weak tissue, a biomaterial was added as a scaffold. To monitor the scaffold, in vivo NMR techniques were used. The study demonstrates that NMR can be used to determine the location and size of the scaffold and also monitor its degradation in vivo 21 Bone grafts: Bone formation studies can be performed using standard histology, but this requires complete destruction of the scaffold. NMR gives bone formation information non destructively and noninvasively. Magnetic resonance microscopy gives a high resolution three dimensional image of the graft. Therefore, tissue level information can be obtained 22 Tissue

PAGE 21

21 eng ineered cartilage: The damaged cartilage can be repaired by a scaffold containing chondrocytes. This study demonstrates that the MRI can be used to evaluate the cartilage repair as a function of time 23 The work here focuses on developing a bioartificial pancreas and monitoring it noninvasively using NMR imaging and spectroscopy. Chapter 2 describes the various materials and methods used to manufacture our bioa rtificial pancreas, and shows the results of its implantation into recipient diabetic mice. Advances in an NMR approach toward monitoring the bioartificial pancreas are discussed in Chapter 3.

PAGE 22

22 CHAPTER 2 OPTIMIZING THE BIOARTIFICIAL PANCREAS 2.1 Intro duction cells. The bioartificial pancreas consists of 3 main components: 1) insulin secreting cells; 2) biomaterial (alginate) to entrap the cells and; 3) PDMS ring which houses the impla ntable coil. The insulin cells are encapsulated in a biomaterial which acts as a selective barrier and allows exchange of nutrients, glucose and insulin between it and the host, but is impermeable to larger molecules like antibodies, the white blood cells, and gives partial immunoprotection 24 The PDMS ring which houses the implantable coil will hold the bioartificial pancreas. The advantage of this approach is that the recipient may not require immunosuppressive drugs, even if non syngeneic cells are used. tet cells. These were selected because they come from the animal models we use, we have experience with their culture, they behave similar to normal insulin secreting cells, and they have a tetracycli ne operon which may be useful for future studies in the event of excess proliferation. In this cell line, the tetracycline operon regulates expression of a proliferation oncogene (SV40 large T antigen). In the presence of tetracycline, the oncogene produc tion is shut off, and the cells undergo growth arrest. The biomaterial used to encapsulate these cells is alginate. Alginate is a widely used biomaterial with which our lab has experience. This material offers stability, and cells grow well in it, provi ded the right composition of alginate used. These components for the bioartificial pancreas were combined in four designs as described in detail below 25 26 2.2 Background The paper describing the first bioartificial pancreas was published by Lim and Sun 27 They encapsulated islets in small alginate beads using a microencapsulation technique which was

PAGE 23

23 pio neered by Thomas Chang 49 Later investigators used sheets, chambers and cylinders to entrap the cells, as described in Chapter 1. Various biocompatible materials have been used, such as agarose, polylysine, polyethyleneimine, etc. However, alginate was proved to be least immunoreactive 27 32 and is the material used in these studies. 2.2.1 Insulin Producing Cells The biological portion of the bioartificial p ancreas is comprised of insulin secreting cells. cells in the Islets of Langerhans are part of pancreatic endocrine system 1 However, al though these cells would be ideal for a bioartificial pancreas, there is difficulty in getting sufficient number of native islets or beta cells for a bioartificial pancreas, as they do not proliferate in culture. Therefore, islets must be obtained from ca davers. However, there are very few donors as compared to the number of diabetic people that require insulin secreting cells. Also, the pancreas needs a long and complex processing procedure to obtain the cells. In the process, some cells are destroyed, therefore more than one pancreas is often needed. Islets from other species, such as pig are commonly used in research, but the same issues of procurement ex ist. These drawbacks motivated scientists to develop modified insulin cell lines or genetically engineered stem cells. These ore appropriate for use as a substitute for human insulin secreting cells because these cell lines can be cultured and amplified in vitro 2.2.2 Beta Tumor Cells (TC) Transgenic mice carrying the hybrid insulin promoter simian virus 40 antigen gene that developed cells release proinsulin I and II and process them to insulin. However, the disadvantage of these

PAGE 24

24 cells is that they do not have a good insulin release profile: that is, th ey do not secrete insulin cells which release constitutively secrete insulin; they do n ot have a regulatory feedback system which can control the insulin release rate 33 34 These same researchers dev eloped another tet) which has better glucose response with more normal insulin release rate. Also, cell proliferation in this line is controlled by the integration of a bacterial tetracycline operon regulatory system (tet), as discussed above. This regul atory system is activated upon exposure to tetracycline, and when activated, the cell cycle is stopped, controlling the cell proliferation 35 When the cells were c ultured in the absence of tetracycline included in the culture media the proliferation was inhibited 35 This cellular feature might be useful in future studies, as described later. 2.2.3 Biomaterials The definition of a biomaterial is c omplicated, but it can be thought of as a material, natural or synthetic, that is compatible with the host, and assists the biological system. The materials can be used to treat, augment or assist an organ, tissue or body function. There are metallic, ce ramic and polymeric biomaterials. In our device, we use polymers. Polymers are materials which contain large macromolecules composed of many repeating units (monomers). The covalently bonded long chain molecules have localized electrons; therefore the p olymers tend to have weak thermal and electric properties. The macromolecules are either cross linked, or interact through weak van der Waals forces or hydrogen bonds, or by entanglement. In this project, the polymers used are alginate and PDMS.

PAGE 25

25 2.2.4 Al ginate Alginate is a naturally occurring non toxic copolymer, extracted from seaweed, marine brown algae, and also found in soil bacteria. It is non toxic and biocompatible, therefore used in pharmaceutical industry for cell immobilization and encapsulati on. Alginate is a binary L D mannuronic residues in various proportions, order and molecular weight. Alginate has 2 monomers, (1 4) D mannuronate (M) and L guluronate (G), which bind covalently to form a long linear chain polymer. The M residues and G residues are arranged randomly, giving rise to differences in physical and chemical properties between alginate varieties. A low viscosity, high guluronic acid content alginate is referred to as LVG, and a low viscosity, high mannuronic acid content alginate is referred to as LVM (see Table 2 1). Alginate dissolves in an aqueous solution, but gels in the presence of divalent cations. The interactions between the guluronic residues in the alginate and the cations lead to the gel formation. The type of divalent cation (e.g., calcium, barium), along with the length and frequency of consecutive guluronic acid residues, determine the strength of the resultant three dimensional networks. The mechanical strengt h, porosity, gel uniformity and biocompatibility depend on the composition of the alginate molecule (i.e., the ratio of mannuronic to guluronic acids, the frequency and size of guluronic acid blocks, the molecular weight and concentration of cations used t o form the gel). Alginates with high mannuronic acid content are softer, less porous and swell during gelling. The alginates with high guluronic acid content have stiff pores, maintain integrity longer, and do not undergo excess of shrinking or swelling during the gelling process 36 Figure 2 1 shows the structural formula of the G residues and M residues. When the alginate is exposed to a divalent cation solution such as calcium chloride, an alginate Ca ++ gel is formed described by the egg box model shown in Figure 2 2.

PAGE 26

26 Table 2 1. Alginate properties LVG LVM Guluronic acid content (%) 73 38 GG content (%) 56 18 Molecular Weight (Dal tons) 189,000 209,000 Viscosity (mPas) 156 193 Figure 2 1. Structural formula of a GGMM guluronic mannuronic acid sequence in alginic acid Figure 2 2. Egg box model of alginate gel

PAGE 27

27 2.2.5 PDMS Polydimethylsiloxane (PDMS) is a widely used organic po lymer in the field of tissue engineering and biomedical applications. It is a biocompatible polymer and is easy to handle. PDMS is made of a siloxane base with a cross linking agent. The two parts are mixed together and cured to form PDMS elastomer. Th e platinum bases in the curing agent act as the catalyst and aid in the formation of SiH bonds across the vinyl groups and Si CH 2 CH 2 Si linkages. These multiple bonds help in the formation of three dimensional crosslinks throughout the PDMS. It is a cle ar, colorless viscous liquid, and gels to a clear material. PDMS is a compound used in our studies to provide mechanical protection for a bioartificial pancreas. It also is chosen because it can house an NMR radiofrequency (RF) coil and insulate it from the salts and liquids in the body. 2.3 Aim of the Study The fundamental aim of this study is to develop a bioartificial pancreas to treat T1D. This bioartificial pancreas is expected to maintain cell viability, induce no immune response from the host, and maintain normoglycemia after implantation. The goals in this chapter are: 1) Optimize tet cells to the alginate; 2) Test different concentrations and combinations of BaCl 2 and CaCl 2 solutions as the gelling agents for the alginate; 3) Test the bioartificial pancreas in vivo as a method to treat T1D in a mouse model. 2.4 Materials and Methods 2.4.1 Cell Culture ory (Albert Einstein College of Medicine, Bronx, NY) 35 Cells were cultured in T flasks as monolayers, and fed every 2 days

PAGE 28

28 contains 20mM glucose and is supplemented with L glutamine to a final concentration of 6mM, 15% heat inactivated horse serum, 2.5% fetal bovine serum, and a 1% penicillin/streptomycin solution. Cell cultures are designed to take a week to grow to confluency. When confluent, the cells were collected from the T 175 flasks using 0.05% Trypsin (Sigma, St. Louis, MO). To reduce genetic drift, only cultures of passages 32 40 were used in these studies. 2.4.2 Cell Entrapment in Beads tet cells were encapsulated in 3% LVM to form beads using the procedure developed by Lim and Sun 28 The beads were made using an electrostatic bead generator (Nisco, Zurich, Switzerland) which pum ps the solution through a small aperture (needle) and places a charge on the droplet to force it off the needle. The beads formed were allowed to fall into a 0.275% BaCl 2 solution. The size of the beads was regulated by controlling these parameters: size of the needle; voltage applied; distance of the needle from the surface of the BaCl 2 solution; and the rate of flow through the needle. The diameter of the needle was 350 m, distance between the needle tip and surface of the solution was 2 cm and the fl o w rate was 0.3 mL /minute. The resulting beads were the size of the beads was approximately 500100m. The cell density was 1x10 7 per mL of alginate. The alginates for the studies were made by dissolving the powdered alginates (LVM: 62%/38% mannuronic/g uluronic acid content; and LVG: 27%/73% mannuronic/guluronic acid content) obtained from NovaMatrix (Oslo, Norway) into saline (0.85%NaCl) and allowing them to stir overnight. The next day the mixture was sterile filtered into a sterile centrifuge tube us ing a 0.22m syringe filter. The final concentration for these alginates are: LVM = 3% (w/v) and LVG = 2% (w/v).

PAGE 29

29 150l of beads contain 2x10 6 cells, which is the standard we maintained for both in vitro and in vivo experiments. The cells were encapsulate d into beads made of alginate with higher mannuronic acid residues (LVM). These beads were entrapped into a macroconstruct of various alginate and gelling solution, in a attempt to make the construct stiffer and regulate cell proliferation, if this became an issue. For a detailed description for making beads, refer to Appendix A. 2.4.3 Entrapment in Macroconstructs In this study, four types of macroconstructs were tested. In Variant 1, the 3% LVM beads were entrapped into 2% LVG to make a macroconstruct plug. In Variant 2, the beads were entrapped into a mix of alginates (2% LVG and 3% LVM in a 1:1 ratio). Two methods of entrapping cells without beads were also used to create the final construct. In Variant 3, the tet cells were entrapped directly into a 2% LVG macroconstruct plug. And lastly, in tet cells were entrapped directly into a 3% LVM macroconstruct plug. Variant 1 bioartificial pancreas prepared with LVM beads into a 2% LVG macroconstruct. C tet cells were mixed into 2% LVG. This mixture was gelled using the divalent cations Ba ++ and Ca ++ by washing the beads in 0.14% BaCl 2 and spinning the macroconstruct exposed to 1.1% CaCl 2 to form a plug. For a detailed description of the procedure, p lease refer to Appendix B. For in vitro studies, the plugs were placed in 6 well plates, and fed with 2mL of media every other day. Glucose samples were taken 24 hours apart every 7 days to measure the daily glucose consumption rate. One bioartificial p ancreas was removed every 7 days, fixed in 10% formalin and processed for histology. H/E stains were obtained from these fixed samples to check the proliferation of the cells, and unstained slides were used for the von Kossa staining, discussed later. Fo r in vivo studies the

PAGE 30

30 plugs were implanted into diabetic animals. Each animal was monitored for 30 days for their weight and blood glucose levels. On the 30 th day, each animal was euthanized and the plug was removed for histological studies. Variant 2 bi oartificial pancreas prepared with 3% LVM beads into a mixed alginate macroconstruct. Beads containing tet cells and formed with 3% LVM were mixed into an alginate mix (1:1 ratio of 3%LVM and 2% LVG). This mixture was then gelled using the divalent cations Barium and Calcium, by spinning the constructs in 0.275% BaCl 2 and in 1.1% CaCl 2 solutions for 20 mins. A detailed description of procedure is given in Appendix C. For in vitro studies, the plugs were placed i n 6 well plates and fed with 2mL of media every other day. Glucose samples were taken every 7 days to measure the glucose consumption rate. One bioartificial pancreas was taken every 7 days, fixed in 10% formalin and processed for histology. H/E stains were obtained to check the proliferation of the cells. For in vivo studies, the plugs were implanted into diabetic animals, and each animal was monitored for 30 days for weight and blood glucose levels. On the 30 th day, each animal was euthanized and the plug was removed for histological studies. Variant 3 bioartificial pancreas prepared by mixing cells directly into the 2% LVG plug Trypsinized tet cells were mixed into 2% LVG alginate such that the concentration of cells was 1x10 7 cells/mL The 2% LVG cells mix was poured into the construct. The gelling was achieved by spinning the construct in a 0.412% BaCl 2 solution for 30 minutes and t hen in a 1.1% CaCl 2 solution for another 30 mins. Please refer to Appendix D for detailed step by step procedure. For in vivo studies, the plug was implanted in diabetic animals and the animal was monitored for 30 days for weight and blood glucose levels

PAGE 31

31 Variant 4 bioartificial pancreas prepared by mixing cells directly into the 3% LVM plug tet cells were mixed into 3% LVM alginate such that the concentration of cells was 1x10 7 cells/mL The 3% LVM cells mix was poured into the construct. The gelling was achieved by spinning the construct in a 0.412% BaCl 2 solution for 30 mi nutes, and then in a 1.1% CaCl 2 solution for another 30 mins. Please refer to Appendix D for a detailed step by step procedure. For in vivo studies, the plugs were implanted into diabetic animals, and each animal was monitored for 30 days for weight and blood glucose levels. 2.4.4 Animal Preparation and Handling The in vivo experiments were conducted with female C3H/HeN mice weighing between 20 30gm, according to a protocol approved by the University of Florida Institutional Animal Care and Use Committee. The animals were made diabetic by a single tail vein injection of alloxan (75mg/kg). Alloxan, which is structurally similar to the glucose molecule, has the ability to kill the insulin producing beta cells in the pancreas. Therefore, to induce diabetes in mice, mice are fasted overnight, reducing the competing glucose, and then alloxan is injected. The beta cells take up the alloxan through the glucose transporter GLUT 2. The alloxan decomposes rapidly and forms hydrogen peroxide, cell death, and causing diabetes. The blood glucose should be >300mg/dL for 2 consecutive days to consider an animal diabetic. Once the animal is diabetic, it is ready to receive the implant. Before surgery, each animal was injected subcutane ously with 150L of buprenorphine for pain, and the abdomen was shaved. The surgeries were performed under general anesthesia by inhalation of 2% isoflurane in oxygen, and maintained throughout the surgery with the animal in a supine position. The constr uct was implanted in the peritoneal cavity by a midline celiotomy with a 1

PAGE 32

32 consists of antibiotics (100U/mL penicillin an d 100ng/mL streptomycin) and an immunosuppressant (100nM dexamethasone). The incision was closed using sterile, synthetic absorbable suture for the inside muscle and silk non absorbable suture for the outer skin. The animals were monitored for blood gluc ose levels using a blood glucose monitoring system (One Touch, LifeScan, Inc.) and weighed for 30 days. On the 30 th day, the animal was euthanized (cervical dislocation) and the bioartificial pancreas was retrieved. After retrieval, the implants were fix ed in 10% formalin (Fisher Scientific) and prepared for histology staining. 2.4.5 Glucose Consumption For in vitro experiments, the constructs were placed in a 6 well plate (Corning Inc orporated, Corning, NY) with 2mL DMEM medium, and maintained on a rocke r for 30 days inside a carbon dioxide (5%), temperature regulated incubator. Samples of the culture media were collected from each well at initial feedings (T0) and 24 hours later (T24) every 7 days for 30 days. Glucose levels in the samples were measure d with the Vitro DT60II bioanalyzer (Ortho Clinical Diagnostics, Rochester, NY). The glucose consumption rates were calculated by determining the difference in glucose levels of T0 and T24 and adjusting for volume and cell number at the initiation of the study. 2.4.6 Histology The samples were fixed in 10% formalin for 2 3 hours, then washed with 70% ethanol three times and stored in 70% ethanol at 4 C. The samples were embedded in paraffin, sliced and stained with Hematoxylin and Eosin. Unstained slid es were obtained for Von Kossa staining.

PAGE 33

33 2.4.7 Von Kossa Stain The unstained slides were tested for mineralization (calcification) with the Von Kossa calcium staining method. The Von Kossa stain is conducted to demonstrate calcium deposits in tissue. The method is based on the substitution of tissue calcium, bound to phosphates, by silver ions and the subsequent visualization of silver ions by hydroquinone reduction to metallic silver. Please refer to Appendix G for a detailed description for this proces s. 2.5 Results Variant 1 bioartificial pancreas prepared by entrapping LVM beads into a 2% LVG plug In vitro studies The images below show the Hematoxylin and Eosin stain which show the live cells. Initially the cells are mono dispersed and distributed t hroughout the bead (Figure 2 3). By day 28, the histological cross sections show that the majority of cell growth is closer to the surface or edge of the beads (Figure 2 4). A B Figure 2 3. Hematoxylin/eosin stained cross sections of Variant 1 plugs A) plug 1 on Day 1, B) plug 2 on Day 1.

PAGE 34

34 A B Figure 2 4. Hematoxylin/eosin stained cross sections of Variant 1 plugs A) plug 1 on Day 28, B) plug 2 on Day 28. Even though the cells grew well in the alginate, the glucose consumption rate (Figure 2 5) was n ot as expected. The results do not match what we have observed with tet cells grown as a monolayer or cultured in microbeads. Figure 2 5. in vitro glucose consumption by the Variant 1 bioartificial pancreas. The graph is a plot of the glucose consumption rate by Variant 1 plugs vs. the time in days (n=6). 1 1 3 5 7 9 0 5 10 15 20 25 30 GCR (nmoles/hr*10^5 cells) Time (days) Glucose Consumption of TC tet in a Variant 1 construct

PAGE 35

35 In vivo results Figure 2 6 shows the glucose response of diabetic animals implanted with the Variant 1 bioartificial pancreas (n=6). The animals were injected with alloxan on day 0 and then implanted with the bioartificial pancreas on day 2 which is indicated by the vertical blue dashed line. The horizontal blue band indicates the normal blood glucose level. The vertical green line is an indication for the termination of the experiment on the day 30. After the implantation the blood glucose levels of the animals d rop for couple of days but soon return to a hyperglycemic state. Additionally, there is wide variation in response, with some animals responding well, and others poorly. This is reflected in the large standard deviation error bars. Figure 2 6. G lucos e response in vivo studies. The graph is a plot of the blood glucose levels of the alloxan induced diabetic mice (n=6) treated with Variant 1 plug over 30 days. 0 100 200 300 400 500 600 0 5 10 15 20 25 30 blood glucose (mg/dl) days post i.v. alloxan Response (Variant 1 implant)

PAGE 36

3 6 Figure 2 7. The above picture was taken under the microscope: a Variant 1 bioartificial pa ncreas in which the microbeads all turned white. As the small beads in the constructs turned white under both in vitro and in vivo conditions (Figure 2 7), we conducted a Von Kossa stain to determine if this color change was due to calcium mineralization o f the beads. Figure 2 8 (A) is an H/E stain of the whitened bead in the construct. This shows no live cells. Figure 2 8 (B) shows the Von Kossa stain of the white beads, confirming that the beads underwent a mineralization. The black staining confirms the calcium mineralization: the pink staining reflects the nuclei of the few cells that are alive. A B Figure 2 8. Cross sections of Variant 1 bioartificial pancreas removed from the mouse peritoneal cavity. A) Hematoxylin/eosin stained cross section; B) Von Kossa stain for mineralization.

PAGE 37

37 To overcome the problems faced by Variant 1 bioartificial pancreas, a Variant 2 was designed to have reduced mineralization (which is choking off the entrapped cells) and thus a better glucose response. Numerous table top experiments were conducted to optimize the alginate mix and the cations that must be used for gelling this construct. We tested various LVG concentrations, and gelled the alginate with another divalent cation, barium. After several tests, we found th at the 3% LVM beads can be encapsulated in an alginate mix of 2% LVG and 3% LVM in a 1:1 ratio and gelled with a combination of 0.275% BaCl 2 and 1.1% CaCl 2 Variant 2 bioartificial pancreas prepared by entrapping LVM beads into a mixed 3%LVM and 2% LVG con struct In vitro results The images below (Figure 2 9) show the Hematoxylin and Eosin stain which show live cells on Day 1. The cells are mono dispersed and distributed throughout the bead. By Day 28, the cells grow in clusters, confirmed by the H/E stai n (Figure 2 10). A B Figure 2 9. Hematoxylin/eosin stained cross sections of Variant 2 plugs. A) plug 1 at Day 1, B) plug 2 at Day 1.

PAGE 38

38 A B Figure 2 10. Hematoxylin/eosin stained cross sections of Variant 2 plugs. A) plug 1 at Day 28, B) plug 2 at Day 2 8. Even though the cells grew in the alginate, the in vitro glucose consumption rate (Figure 2 11) was not as expected. The results again did not match with the normal glucose consumption tet cells, in which as the cells grow in number, the co nsumption rate rises. Figure 2 11. I n vitro glucose consumption by the Variant 2 bioartificial pancreas. The graph is a plot of the glucose consumption rate by Variant 2 plugs vs. the time in days (n=6). 4 2 0 2 4 6 8 10 0 5 10 15 20 25 30 GCR (nmoles/hr*10^5 cells) Time (days) Glucose Consumption of TC tet in Variant 2 construct

PAGE 39

39 Note that the cells are consuming little, if an y glucose in the media, suggesting gluconeogenesis. In vivo results The figure below (Fig 2 12) shows the blood glucose response of diabetic animals implanted with the Variant 2 bioartificial pancreas (n=6). The animals were injected with alloxan on day 0 and then implanted with the bioartificial pancreas on day 2 which is indicated by the vertical blue dashed line. The horizontal blue band indicates the normal blood glucose level. The vertical green line is an indication for the termination of the experi ment on the day 30. After implantation, the blood glucose levels of the animal drops for couple of days but returns to hyperglycemic state. The results obtained from Variant 2 bioartificial pancreas are similar to those obtained from the Variant 1 bioarti ficial pancreas; but the result was not what was desired, i.e., normoglycemia throughout the experiment duration. However, it was observed under the microscope that the beads in the mixed alginate plugs were not turning white over time (Figure 2 13), sugg esting that our approach to eliminate mineralization was successful, though the construct still did not work long.

PAGE 40

40 Figure 2 12. glucose response in vivo studies. The graph is a plot of the blood glucose levels of the alloxan induced diabetic mice (n=6 ) treated with Variant 2 plug vs. the time in days. Figure 2 13. The above picture was taken under the microscope: beads in a Variant 2 bioartificial pancreas did not turned white with mineralization. The Von Kossa staining was used to demonstrate that the Variant 2 bioartificial pancreas did not undergo mineralization due to calcium mineralization (Fig. 2 14). 0 100 200 300 400 500 600 0 5 10 15 20 25 30 blood glucose (mg/dl) days post i.v. alloxan Response (Variant 2 implant)

PAGE 41

41 A B Figure 2 14. Cross sections of Variant 2 bioartificial pancreas removed from the mouse peritoneal cavity. A) Hematoxylin/eosin stained cro ss section; B) Von Kossa stain for mineralization. The in vitro and in vivo studies for Variant 2 bioartificial pancreas, i.e., the mixed alginate plugs with LVM beads, demonstrate that even though they do not undergo mineralization due calcium and are str ucturally strong, they still do not help diabetic animals to maintain normoglycemia. Therefore, we further tested the alginate mixes and came up with another bioartificial pancreas which does not contain beads, but has the cells directly mixed into the al ginate and gelled using barium and calcium solutions as described earlier. These are termed Variant 3 and Variant 4 bioartificial pancreata. Comparision between Variant 3 and Variant 4 bioartificial pancreas In vivo results The Variant 3 bioartificial pa tet cells entrapped in a 2% LVG tet cells entrapped in a 3% LVM macroconstruct. These constructs were implanted into alloxan induced diabetic animals (n=6 for ea ch Variant). Figure 2 16 shows the comparison between blood glucose levels of animals with the Variant 3 bioartificial pancreas and animals receiving the Variant 4

PAGE 42

42 bioartificial pancreas. We found that the Variant 3 constructs (LVG) were found to be whit e when retrieved from the animal on the 30 th day. Also the glucose response of animals with Variant 3 bioartificial pancreas was not satisfactory. Even though some animals initially reach normoglycemia, they became hyperglycemic eventually. Whereas, the Variant 4 bioartificial pancreas rapidly brought the blood glucose levels down below 300mg/dl throughout the study, and well into the normal range, and with little variation. The retrieved Variant 4 plugs were found to be intact and did not turn white. Figure 2 15 shows the hemotoxylin/eosin stain of the variant 3 and variant 4 bioartificial pancreases. In variant 3, the cells do not grow in large clusters. Additionally, there are immune cells (fibrotic tissue) which line the outside of the construct. In variant 4, the cells grow in large colonies/ clusters toward the outer edge of the alginate implant. A) B) Figure 2 15 Cross sections of Variant 3 and 4 bioartificial pancreas removed from the mouse peritoneal cavity after 30 days. A) Hemat oxylin/eosin stained cross section of Variant 3; B) Hematoxylin/eosin stained cross section of Variant 4

PAGE 43

43 Figure 2 16 in vivo blood glucose response in diabetic mice to alginate implants. The red line plots animals treated with the Variant 3 plug (n= 9), and the green line represents the plot for animals treated with the Variant 4 plug (n=7). Normal mouse blood sugar is represented by the gray horizontal line, with the variation (s.d.) reflected by the blue shaded area.

PAGE 44

44 Table 2 2. Summary A pproach Rationale Results Variant 1 3% LVM beads entrapped in 2% LVG macroconstruct LVM: cells proliferate LVG: controls proliferation Poor control; mineralization of beads confirmed with Von Kossa staining Variant 2 3% LVM beads entrapped in mi xed alginate macroconstruct Prevent mineralization No mineralization; confirmed with Von Kossa staining; poor glucose control Variant 3 TC tet cells directly entrapped in 2% LVG macroconstruct Control: cells grow poorly in LVG: compare response wi th Variant 4 Poor glucose control when implanted into diabetic mice Variant 4 TC tet cells directly entrapped in 3% LVM macroconstruct To study the response when cells were entrapped in a 3% LVM macroconstruct Excellent control of blood glucose lev els 2.6 Discussion and Future Work tet cells had been entrapped into LVM alginate beads and injected into the peritoneal cavity of diabetic host mice. These would disperse throughout the cavity; therefore the retrieval of these beads in order to study them was very difficult. Moreover, it was nearly impossible to study their functionality and structural integrity non invasively. As discussed earlier, it is important to study the structural integrity, metabolic activity, or fun ction of cells entrapped in the beads. Therefore, to keep the beads together and allow for easier location, the beads were included in a macroconstruct which can be investigated and characterized. To enhance the ability of NMR to monitor these implanted cells, a coil can be implanted with the construct. Therefore, our present macroconstruct design allows for it to be

PAGE 45

45 surrounded by an RF coil to monitor the construct noninvasively through NMR means. Moreover, hyperglycemia for prolonged period leads to v arious health complications, as mentioned in Chapter 1. Therefore it was important to develop and optimize a bioartificial pancreas which not only keeps the animal alive but also brings the blood glucose levels to the normal range. Because these construct s had never been studied before, we cultured some in vitro to determine how they would grow over time. Surprisingly, the in vitro tet cells, grown as a monolayer or in alginate microbeads, expand in numbers over time, and their glucose consumption increases as the number of cells incr eases. However, in the Variant 1 bioartificial pancreas we constructed, the glucose was not being consumed over time, but actually being generated (Fig: 2 5). Histology of the constructs at the end of the experiment shows that the cells are not dying out but are growing in number, so the decrease in glucose consumption is not due to a loss of cells. This gluconeogenesis is unexpected, and warrants further study, as it suggests an epigenetic change of the cells, perhaps to a cell line more liver cell lik e. Mice induced with alloxan would die rapidly: at a 75 mg/kg dose, half an untreated were dead by day three, and all were dead by day five (data not shown). Animals that were implanted with the Variant 1 construct did not respond well, and most returned to a hyperglycemic state within 1 2 weeks of implantation, though they survived for the entire month implantation period (Fig. 2 6). Upon explantation, many of the constructs were seen to have a whiteness around the alginate beads. The Von Kossa stains ( Fig: 2 8), suggest that the whiteness noted is due to calcium mineralization around the beads, and this mineralization reduced transport to the cells, compromising their viability and causing the mice to return to a hyperglycemic state. The reason

PAGE 46

46 for thi s mineralization is unclear. We thought that the calcium was possibly coming from the outer LVG alginate, which has more guluronic residues to coordinate the calcium than LVM alginate. As the construct ages, the calcium can be released from the alginate. It also could be due to a slight dehydration of the construct, though we did not weigh the constructs upon removal to compare their change from the initial implant weight. In any event, to avoid this problem, we tried a different outer alginate for the construct, using a mix of 3% LVM and 2% LVG ; the Variant 2 bioartificial pancreas. Constructs removed at the conclusion of the study did not show the whiteness, and the Von Kossa stain of the Variant 2 plug showed that there was no calcium mineralization (Fig: 2 14). Therefore our approach using a mixed alginate avoided this problem. However, the Variant 2 construct could not control the blood glucose levels in the diabetic mice (Fig: 2 12), so another approach was needed that would combine the non mine ralization characteristic, and yield efficacious implantable constructs. After conducting several experiments, we came up with the Variant 3 and Variant 4 bioartificial pancreases. These models do not use beads at all, but entrap the insulin secreting ce lls directly into the construct alginate slab. Various table top experiments were conducted to help in optimizing the alginate which can be used as a scaffold to hold the cells in the construct and remain homogeneously gelled. The divalent cations used t o gel the alginate also play a major role in defining the structural integrity of the construct. The Variant 3 was made by entrapping the cells in 2% LVG, which we knew could gel into the construct shape desired, but was a negative control, of sorts, as t tet cell line is known not to thrive in the LVG matrix 24 The Variant 4 was made by entrapping the cells in 3% LVM, but a novel gelling process was needed to allow the alginate to gel completely and uniformly. The gelling agents

PAGE 47

47 and processes were the same for both the Variant 3 and Variant 4 constructs, and are discussed further in Appendix D. These studies demonstrate that we can achieve cures for T1D with the implanted construct when the materials and gelling processes are optimized. The compar ative study between the 2% LVG and 3% LVM plug shows that the LVM alginate has worked the best by bringing down the blood glucose levels of the alloxan induced diabetic animals and maintaining normoglycemia for at least 30 days. As predicted, the LVG cons truct helped to keeping the animal alive, but the animals remained hyperglycemic. Figure 2 15 demonstrates the poor cell viability within these LVG constructs, and shows that there is some immune response to the implant that can lead to poor oxygenation w ithin the construct, and choke off the implanted cells. This is expected, as the tet cells are known not to grow well when entrapped in alginate with a high guluronic residue content 24 The Variant 4 LVM bioartificial pancreas showed cellular growth in clusters, especially toward the outer edge, where the nutrition would be bette r. There was no evidence of any immune response. Both LVG and LVM plugs showed that the viable entrapped cells contained insulin (data not shown). These results indicate that the Variant 4 LVM plug can now be included in a PDMS coated implantable coil c onstruct for non invasive monitoring using NMR imaging and spectroscopy techniques. In the future we would like to study the mechanical properties of our bioartificial pancreas by testing their tensile strength. The tensile strength or the mechanical prop erties of the construct might change over time because of the interaction with the host, or due to the intraperitoneal environment. Hence, we will test the construct on the day it is made, day 0, and then again on the day it is retrieved from the animal, day 30. The experiment conducted until now have been under normal physiological temperatures, but the alginate may be sensitive to

PAGE 48

48 changes in temperature (in case of fever) or pH. Therefore, the construct should also be studied under different thermal an d pH conditions. These results can be used to determine the impact of these conditions on the structure, and perhaps give insight into how to avoid alginate weakening (e.g., intraperitoneal injections of divalent cation solutions; alternate gelling condit ions to strengthen the final construct, etc.). Longer termed (> 30 days) successful bioartificial pancreas implants also need to be studied after explantation through histological methods. If the cells in these long lived constructs grow too well in vivo the construct may be compromised, as the structure could weaken and break apart. Also, too many cells could lead to an excess of insulin secretion and consequent hypoglycemia, which is also a dangerous state for the recipient. Although we believe a robu st growth of cells will be rare due to limited nutrition supply in the peritoneal cavity, if this is the case, we can control the cell proliferation with the tetracycline operon included in our tet cell system. By injecting tetracycline or giving it orally, the cell growth can be stopped, because the cells have been designed to stop dividing upon tetracycline exposure. Also, studies need to be conducted to test the response of the bioartific ial pancreas in the condition of excess glucose levels by giving glucose bolus. The present blood glucose measurements are performed on animals fasted for 4 hours. To determine the response rate and insulin release characteristics of the implant, a diffe rent type of study is needed, a glucose stimulation test. In this test, a bolus of glucose will be given to the animal, and the blood sugars and insulin level of the blood measured often over a short period of time. Compared to a control mouse, we antici pate the response of the implant to be damped, as it will take time for the tet cells, and for the secreted insulin to reach the blood stream. Even if

PAGE 49

49 the implant exhibits a damped response, we want the implant to respond to the h igh level of sugar. Additionally tet cells. Therefore, there are no major histocompatibility protein differences between the cells and the host mouse that would induce an immune attack against the impl anted cells. This is an ideal situation, but not one representative of the approach needed for a widespread use of the bioartificial pancreas. Lastly, the diabetes we induced in these studies was due to the injection of alloxan. Normally, diabetes is no t caused chemically, but caused by an immune response which destroys the insulin secreting cells. Therefore, in the future, we will test the ability of our bioartificial pancreas to function in non syngeneic animals such as the NOD mice, which also develo ps T1D naturally, i.e., by an immune system attack on the native pancreas. This will allow us to determine how robust our approach is toward treating T1D.

PAGE 50

50 CHAPTER 3 DEVELOPMENT OF THE T CHO CALIBRATION CURV E 3.1 Introduction Nuclear magnetic resonance (NMR) is arguably the most advanced technique which can be used to both image and get spectroscopic data. NMR is a diagnostic and analytical tool that helps obtain noninvasive images, and acquire spectroscopic data to determine molecular structures, conc entrations of metabolic compounds, and nondestructive metabolic activities of various amino acids and proteins. Radiofrequency (RF) coils are used in NMR to transmit power to and receive signal from the object being studied. RF coils such as the surface coil, volume coil, and quadrature coil have good sensitivity for superficial tissues, but for deeper tissues, implantable coils have been developed which can be coupled with an external surface coil 37 It was demonstrated earlier that the inductively coupled implantable coil has improved sensitivity at high fields, and the signal to noise ratio (S/N) was much higher than that obtained with conventional surface co ils 37 Basic principle : An atom consists of a positively charged nucleus (containing protons and uncharged neutron) surrounded by negatively charged electrons. The components of the nucleus, possess angular momentum, or spin. These particles try to form pairs which cancel out their individual angular momentums, so not all nuclei have spin. The net nuclear angular momentum (I), is zero if a nucleus has an even number of both protons and neutrons that are paired. Many elements and their isotopes have nuclei with odd atomic mass or odd atomic number and therefore possess nuclear spin. Commonly used nuclei in biologic studies include 1 H, 13 C, and 31 P. Consider t he simplest nucleus, that of hydrogen, which is only a single proton. This nucleus has a spin of I=1/2. When a magnetic field Bo is applied, the proton will precess with a particular resonant frequency about the applied magnetic field (defined as the z d irection), as

PAGE 51

51 shown in Fig: 3 1 38 When there are many nuclei all rotating, they collectively create a net magnetization vector. Because the x and y axis components of so many nuclei cancel, one only has to consider the net magnetization along the z axis. Figure 3 1. A spinning nucleus in a magnetic field. If an RF wave is applied whose frequency matches the precessional freque ncy of the nuclei in the external magnetic field, then the nuclei absorbs this RF energy (resonance). This RF pulse can be thought to tip the net magnetization vector away from the z axis. Its angle will depend on the pulse length and strength. This mag netization vector does return back to the original orientation along the defined z axis, because any particle tends to relax or return to low energy state from a high energy state or excited state. The processes that achieve this relaxation are: longitudi nal (spin lattice) or T 1 relaxation, and transverse (spin spin) or T 2 relaxation. After the initial RF pulse to move the magnetization vector, these excited nuclei start to relax and the net magnetization vector induces a current in a receiver RF coil that surrounds the sample. This current is amplified and electronically modified into NMR data.

PAGE 52

52 Typically the proton resonant (or Larmor) frequency depends on the magnetic field applied. However, the Larmor frequency is also influenced by the chemical enviro nment, neighboring nuclei and electrons near the nuclei of interest. These influences can make the nuclei resonate at a slightly shifted different Larmor frequency. This phenomenon is called chemical shift and is the basis for MRS. The chemical shifts a re usually normalized by the magnetic field strength so that the shifts are independent of field. They are commonly expressed in ppm. Therefore, NMR spectroscopy can be used to detect the presence and concentration of a compound through the detection and quantification of resonances with chemical shifts associated with specific compounds. Proton ( 1 H) is the most sensitive nucleus for NMR both for intrinsic sensitivity and natural abundance since almost all metabolites contain proton. 1 H NMR spectroscopy is a very useful tool to quantify a large number of biological compounds in tissue. In our study, we have used 1 H proton spectroscopic techniques to measure choline metabolites. The reason for the interest in this compound is as follows. The phospholipi d synthesis pathways and degradation pathways reflect the membrane turnover of a cell. Choline containing compounds are involved in both these pathways; therefore detection of choline can be an indicator for a cell population. Importantly, this marker is rapidly lost once a cell dies, so the choline level can be a measure of the number of viable cells. Choline has a resonance at 3.2 ppm from a combination of different choline related metabolites such as phosphocholine, glycerol 3 phosphocholine and free choline studies are NMR visible, the choline detectedand quantified by NMR can determine the viability of the insulin secreting beta cells entrapped in the bi oartificial pancreas.

PAGE 53

53 Radiofrequency coils : RF coils are used for transmitting energy and receiving signals in NMR. The signal to noise ratio (S/N) in MRI can be improved with the help of radiofrequency coils. There are three categories of NMR coils: tra nsmit/receive coils, receive only coils and transmit only coils. Various types of RF coils are: Surface coil; Whole volume coil; Phase array coil; Solenoid coil; and Quadrature coil. The parameters that determine a good RF coil are: Signal to noise ratio quality factor, coil sensitivity, coil homogeneity, filling factor, and effective range. Coil technology has evolved, and the sensitivity and performance has increased as coil designs have moved from surface coils to quadrature coils. However, if the re gion of interest is a small deep internal structure, such as an implanted bioartificial pancreas, then the need for integrating an implantable coil arises, because this coil approach increases the MR measurement sensitivities and gives better spatial local ization. Therefore, in my work I used the implantable coil (loop gap resonator coil) inductively coupled to a surface coil. This system was compared to results obtained with a volume coil. 3.2 Implantable Coils In the past, many coils, like simple surfac e coils, quadrature surface coil systems, phased arrays of surface coils, volume coils etc., were used to obtain signal for imaging and spectroscopy. For further improvement in MR signal sensitivity researchers have developed an implantable coil which can be coupled with extracorporeal surface coil creating a single unit 39 This inductively coupled system also helps obtain a more localized and target specific signa l with increased S/N. In the past, implantable coils were used, but they had wires going through the skin in order to be connected to the amplification system and other electronic circuitry. In

PAGE 54

54 the event of complications related to infections through thi s wiring into the body, these implantable coils had to be removed through another surgical procedure. The inductively coupled implantable coil system can bypass all these complications and theoretically can be retained in the body for much longer durations This coil system has no internal wires that need external connections. Therefore, there is no portal for infection. In the inductively coupled system, there are two coils. One coil is coated with a biocompatible and insulating material and implanted in the body: this is called the implantable coil. The other coil is an external surface coil, and it is inductively coupled to the implanted coil 40 41 To operate, the two coils are tuned to the proper frequencies. The implantable coil is excited by the surface coil RF pulse, and the signal from the sample picked up by the implanted coil is transmitted to the ext ernal coil through inductive coupling. For our purposes, a simple loop gap resonator coil was developed as the implantable coil. It is coated with a biomaterial which will not attract any immune response when implanted in the animal. 3.3 Improving Coil S ensitivity These studies have improved sensitivity over previous work because of two important factors. One, the sensitivity in any system can be improved by increasing the magnetic field strength. When the magnetic field strength is increased, the energ y gap between the two energy levels will also increase leading to increase number of nuclei in the lower energy level. Therefore the probability of resonance will increase and a more intense signal will be obtained. For these studies we used an 11.1 Tesl a (T) magnet, whereas previous studies used a 4.7T magnet. The second technique used to improve the signal sensitivity is through inductive coupling of the implantable and surface coils. An implantable coil which fits the region of interest is coupled wi th a surface coil inductively without any external wires 37 42

PAGE 55

55 3.4 LASER Pulse Sequence To obtain signals from our system, we used the Localization by Adiabatic Selective Refocusing (LASER) pulse sequence. It is a single scan 3D localization method which is executed with adiabatic excitation and refocusing RF pulses based on the principle of frequency selective re focusing with pairs of Adiabatic Full Passage (AFP) Refocusing pulses. A single AFP pulse in the presence of a magnetic field gradient can achieve slice selective refocusing. The frequency modulation of the pulse induces a non linear B 1 and position depen dent phase across the slice which will lead to severe signal cancellation; however a second identical AFP pulse can refocus the nonlinear phase such that perfect refocusing can be achieved. Because frequency selective adiabatic excitation pulses remain el usive, the entire sample is excited with a nonselective adiabatic excitation pulse after which three pairs of AFP pulses achieve 3D localization by selectively refocusing three orthogonal slices. The advantages of the LASER technique over STEAM and PRESS sequences are twofold; namely, the method is completely adiabatic, and, by employing high bandwidth AFP pulses the localization can be extremely well defined, both in terms of minimal chemical shift displacement, as well as sharpness of the localization ed ges 43 46 LASER is not a regular spin echo method, but is more closely related to the CPMG multiple spin echo pulse train. LASER is an excellent method to d etect strongly coupled spin systems. 3.5 Background NMR has been previously used to study anatomical structures and also the functional anatomy. The field of tissue engineering has exploded in recent years, and the rapid development of tissue engineered c onstructs has created the need for better and more robust ways to noninvasively monitor them day to day. Because tissue engineered implants may be

PAGE 56

56 subjected to mechanical stress, thermal changes, immune responses, etc., it is important to monitor the impl ant regularly. NMR is a powerful tool that can look for certain markers to identify cellular viability and function, and use these measures to predict the construct failure before it actually fails. The noninvasive NMR technique to measure the viability o f cells in a bioartificial pancreatic construct was developed by Cherie Stabler 47 49 The method involved uses 1 H NMR spectroscopy to measure the total choline which is indicative of the number of viable cells within the construct. The signal from choline was used by Stabler to monitor the construct both in vivo and in vitro using a 4.7T NMR instrument with a surface coil. This technique could be used to meas ure a cell number of seven million and above. However, in our bioartificial pancreatic construct the number of cells is 2 million in total. Therefore it is necessary to increase the sensitivity of NMR to study the lower concentration of cells. It has bee n discussed earlier that the sensitivity of NMR measurements can be improved by increasing the magnetic field applied and/or by increasing the sensitivity of the RF coil. Nelly Volland had developed an implantable coil which can be inductively coupled wit h the surface coil 36 This implantable coil fits exactly with the bioartificial pancreas therefore the signal can be received from the region of interest. Also, this inductive coupling increases the sensitivity, as discussed above. The inductive coupl ing also obviates the need for external wiring that would pass through the skin and connected to a power source or amplifiers. These external wires increase the risk of infection therefore decreasing the duration for safe monitoring. Therefore, the induc tively coupled implantable coil has many advantages compared to the regular surface coil or the implanted coil. Nelly Volland proved that the sensitivity improved by 2 times for the implantable coils as compared to the surface coil. Also as the magnetic field strength increased

PAGE 57

57 from 4.7T to 11.1T the gain increased further by 2.4 times. The images obtained with the implantable coils had better contrast to noise, good S/N, and better sensitivity to nuclei like 31 P, 19 F, which may be explored in the future as markers of cellular function. 3.6 Aim of the Study The central goal in this chapter is to develop a calibration curve of TCho using the inductively coupled implantable coil developed by Nelly Volland. Cherie Stable had proved that by measuring TCho w e can actually quantify viable cells in a bioartificial pancreas. Therefore: 1) a calibration curve was developed with different concentrations of choline solutions; and 2) tet cell number. For all the measurements, water was used as an internal reference. 3.7 Materials and Methods 3.7.1 Implantable Coil The method to prepare the implantable coil was followed 36 A 2mm wide 202m thick copper wire was used to make the inductive loop gap resonator with a single fixed tuning capacitor (4.3pF or 4.7pF) to provide the desired resonance. The surface coil s work which could be tuned and matched to 470.75 MHz 37 It was tuned and matched to the frequency at which the implantable coil resonates when it is normally l oaded. 3.7.2 Coating of the Implantable Coil The implantable coil was coated with polydimethylsiloxane elastomer (PDMS). PDMS is an organic polymer which is biocompatible and has low dielectric properties. The coating thickness changes the resonance freq uency of the coil and quality factor (Q), such that the frequency and Q of the coil is reduced after coating with the PDMS layer. The coating procedure is followed from the PDMS casting procedure explained in Appendix E. After the

PAGE 58

58 PDMS was poured on to t he glass slab, the loop gap resonator coil was placed in the PDMS and pressed into it. After the coil is completely submerged, another glass slab was placed on top. The PDMS was cured overnight, and the construct was cut with a diameter of 20 mm using ho llow punches. Another 12mm punch in the center is made to house the bioartificial pancreas. Two more 1 mm thick PDMS sheets were prepared to make the top and bottom layers. The outer diameter for the top and bottom layers is same as the middle layer, i.e. 20mm, but the inner diameter is 10mm. The three layers are shown in the figure below. They are glued together with PDMS to finish the PDMS ring. The cell alginate mix is shown in Figure 3 2 (B). The dark blue dots represent the beads. A B Figure 3 2 A) Implantable coil design with PDMS coating; B) This bioartificial pancreas was made with 3% LVM beads entrapped in 2% LVG. 3.7.3 Choline Samples To obtain a choline calibration curve, choline solutions of different concentrations were prepared from cho line chloride, >98% (Alfa Aesar Company). 1mM, 10mM, 20mM, 50mM and 100mM solutions were prepared by mixing the choline chloride in distilled water. 0.2 mL of

PAGE 59

59 each of these solutions were filled in microcentrifuge tubes and fitted into the PDMS coated im plantable coil. This setup was used to obtain NMR spectroscopy data and develop a calibration curve. 3.7.4 Cell Culture tet Cells were cultured in T flasks as monolayers and fed with fresh medium supplemented with L glutamine to a final concentration of 6mM, 15% heat inactivated horse se rum, 2.5% fetal bovine serum, and a 1% penicillin/streptomycin solution every 2 days. 3.7.5 RF Coil The resonance frequency of the implantable coils with a 4.3 pF capacitor before the PDMS coating was ~508 512 MHz. After the coils were coated with 4mm thi ck PDMS, the frequency shift was ~ 6 8 MHz. The quality factor of the coated coil was ~300, and when it was introduced into a phantom gel, the Q decreased by 10 fold. After the coil is implanted in the animal or placed in a phantom, the resonant frequenc y shifts down by another ~10MHz. When this coil is inductively coupled with a surface coil, the resonant frequency splits into both co rotating and counter rotating modes. The frequency of the co rotating mode is not always at 470.75MHz. To bring the fr equency of the co rotating mode to 470.75MHz the tuning capacitor on the surface coil needs to be adjusted. Therefore, coils were made such that they resonate at a frequency approximately 40MHz higher than the desired resonant frequency.

PAGE 60

60 A B C D E Figure 3 3 photographs of in vitro studies to generate the choline calibration curve. A) PDMS coated implantable coil and PDMS construct without coil; B) PDMS coated implantable coil; C) surface coil placed over the implantable coil; D) set up of the s urface coil coupled with implantable coil; E) sample placed in the implantable coil.

PAGE 61

61 Table 3 1. Resonant frequencies of coils before and after the PDMS coating with their Q factors. Coil capacitor used (pF) Number of coils tested f(MHz) before coating f( MHz) after coating Q 4.3 6 508 502 301.19 4.7 6 520 512 303.08 3.7.6 Animal Work and Surgery The procedure followed for animal preparation and surgery was approved by University of Florida Institutional Animal Care and Use Committee. The animals used for the experiments were C 3 H/HeN mice which weighed between 20 30 gm. The NMR studies and surgeries were conducted under general anesthesia. The anesthesia gas mixture was a combination of 2% isoflurane in oxygen, and was inhaled by the mouse during the course of surgery or NMR measurement. The construct was prepared under sterile conditions and implanted into the peritoneal cavity of the recipient mouse under deep anesthesia. A 2 3cm long incision was made on the right side of the mouse abdomen while t he animal was placed in supine position. After the construct is implanted, the incision was sutured and closed. 3.7.7 Preparation for NMR Measurements Animals implanted with an implantable coil construct were taken for NMR measurements a day after the imp lantation surgery was performed. Each animal was placed in a supine position on the cradle and monitored with constant anesthesia supply (isoflurane 1.5% in oxygen). The surface coil was placed at 9mm distance from the surface of the mouse abdomen. The temperature and respiratory sensor were used to monitor the body temperature and respiratory rate of the mouse inside the bore. The respiration rate was maintained at 20 30 breaths/min and the skin temperature was maintained between 24 26 o C using a warm b lower placed inside the

PAGE 62

62 bore. The NMR measurements were conducted in the 11.1T 40cm clear horizontal bore magnet equipped with Agilent Console. 3.7.8 Shimming To obtain a homogenous field across the sample the magnet needs to be shimmed. Shimming is acco mplished by changing the applied current for the coils surrounding the probe. Small magnetic fields in the region of the sample are created when the current is applied, increasing or opposing the present magnetic field. A shim library from previous simil ar experiments was loaded to get a starting point for this process. Data acquisition and processing were done using Agilent VnmrJ software. The area under the spectroscopic peaks was analyzed and calculated using NUTS software (Acorn, Fremont CA). Follo wing are the methods and pulse sequences used to obtain images and spectroscopic data. 3.7.9 NMR Imaging and Spectroscopy Measurements in vitro NMR imaging and spectroscopy were performed using an Agilent system with an 11.1 T horizontal bore magnet operat ing at 470.74 MHz. The magnet was equipped with an Agilent console and 40cm horizontal bore Magnex self shielded gradients. The NMR signal was obtained by inductive coupling of a surface coil with an implantable coil. The distance between the implantabl e coil and surface coil was 9mm. After the coils were coupled, the surface coil was tuned and matched to 470.74MHz, and the gain maintained ~27dB. The RF coil was positioned in the isocenter of the magnet. The position of the coupled coil system was chec ked using a position pulse sequence on the magnet system. Only when the position of the coupled coil system was at the isocenter in all the three directions (axial, sagittal and coronal) was imaging and spectroscopy performed. A test

PAGE 63

63 scan with a single p ulse sequence was run to give the linewidth of the sample. If the linewidth is too large (>30), then manual shimming was conducted. When the linewidth <30, scout 1 H gradient echo images with TR=30ms, TE=5ms and acquisition times of 11.5sec were obtained to determine the position of the construct, as shown in Figure 3 4 (A). This figure shows the images of the sample in the coronal, sagittal and axial views. These images are used to select the slices for the gradient echo multislice pulse sequence (GEMS) The TR=82.9ms, TE=4ms and acquisition time is 10.6sec for 12 slices. Figures 3 4 (B) and (C) show 8 coronal slices obtained, each of which are 1 mm thick. These images were obtained to place the voxel in a region of interest for performing spectroscopy This pulse sequence was also used to obtain the axial slices shown in Figure 3 4 (D). Localized spectroscopy data was obtained from localized adiabatic selective refocusing pulse sequence (LASER). A three dimensional voxel of well defined position and size is selected whose spectrum is then collected and analyzed. For the measurements, voxels were selected which were 2x2x2mm and 3x3x3mm in dimension. These voxel sizes were selected because they were big enough to give TCho peak, and shimming obtained on these voxels was good (linewidth ~7 8). The LASER pulse parameters used for all the experiments were: TR=500ms, TE=40ms, number of averages=16 and receiver gain 2d B without water suppression. This unsuppressed water peak was used as an internal refe rence, and is shown in Figure 3 4 (E). All localized, water suppressed 1 H NMR spectra used a TR=3000 ms, TE=40 ms and the number of acquisitions was 256, collected at a constant receiver gain of 10 dB. Water suppression was achieved by a chemical shift s elective (CHESS) water suppression schema. The peak shown in Figure 3 4 (F) depicts a choline peak of 100mM concentration. The total time required for each water suppressed spectrum was 12min 56 sec. Spectral data was processed using the NUTS software. Time domain data were apodized with an

PAGE 64

64 exponential line broadening of 1Hz for choline solutions and 5Hz for data obtained from cells, Fourier transformation and baseline correction. tet cells tet cells were cultured in T flasks as monolayers and fed with fresh medium A confluent flask was trypsinized and centrifuged. The centrifuged cells were mixed with 3% LVM. 0.2mL of the cell/alginate mix was added into a microcentrifuge tube and taken for NMR studies. A B C D Figure 3 4. NMR images and spectroscopy data obtained with an inductively coupled im plantable coil system. A) SCOUT images of sample; B) GEMS images of sample slices; C) Volume of interest selected with the help of images obtained; D) side view of implantable coil with sample; E) water peak at 4.7ppm obtained with the LASER pulse sequenc e, 16 averages; F) TCho peak of a 100mM choline solution at 3.2ppm obtained with LASER pulse sequence, 256 aves.

PAGE 65

65 E F Figure 3 4 Continued 3.8 Results Choline solutions with different concentration were used to obtain TCho calibration curve by NMR spect roscopy. Measurements were done with both volume coil and inductively coupled implantable coil. The graph below (Figure 3 5) shows the calibration curve of choline obtained using a volume coil. It depicts the ratio of area under the choline peak with wa ter peak vs. the [choline]. The plot is linear, with an R 2 =0.9996. The volume of interested selected was a 3x3x3

PAGE 66

66 mm voxel, which was placed at the center of the sample. However, the choline peak obtained from the volume coil was very small, and it was v ery difficult to process the data. Figure 3 5 The graph is plotted between the choline concentration and ratio between area under choline peak and area under water peak. the data was obtained with a volume coil with a voxel size of 3x3x3mm. The TCho pe ak data is linear with an R 2 =0.9996, n=1. Using a surface coil inductively coupled to the implantable coil, we obtained another calibration curve with n=3 repetitions. The graph below (Figure 3 6) depicts the ratio of area under the choline peak with wate r peak vs. the [choline]. The plot is a linear relationship with an R 2 =0.9962. The volume of interested selected was a 2x2x2 mm voxel, at the center of the sample in the implantable coil. Here, a smaller voxel size was chosen because it was easier to sh im, thus avoiding issues caused by poor linewidth. It was easier to see the choline peak and process the data, even with a 1mM solution, and the signal to noise was better than that obtained with the volume coil. The standard deviation obtained was also small, showing that the calibration curve is reproducible. y = 0.0023x 0.0026 R = 0.9996 0 0.05 0.1 0.15 0.2 0.25 0 20 40 60 80 100 120 choline calibration curve choline volume coil Linear (choline volume coil) ratio of area of choline peak to the water peak concentration of choline in mM

PAGE 67

67 Figure 3 6. The graph is a plot between the choline concentration and the ratio between the area under the choline peak and area under the water peak. Data was obtained with an inductively coup led implantable coil with a voxel size of 2x2x2mm. The relationship is linear with R 2 =0.9962, n=3. Bars represent the standard deviation. Using the same coupled coil system, another calibration curve was obtained with n=1 repetition, but with a bigger v oxel size (3x3x3mm). Even though it was difficult to shim this volume, the signal obtained was much larger than the signal obtained from 2x2x2 voxel. The graph below (Figure 3 7) depicts the ratio of choline with water vs. the [choline]. The plot is a l inear relationship with an R 2 =0.9987. R = 0.9962 y = 0.0023x 0.0031 0.05 0 0.05 0.1 0.15 0.2 0.25 0 20 40 60 80 100 120 Choline calibration curve choline calibration curve Linear (choline calibration curve) ratio ofarea of choline peak to the water peak concentration of choline in mM

PAGE 68

68 Figure 3 7 The graph is plotted between the choline concentration and ratio between area under choline peak and area under water peak. The data was obtained using an inductively coupled implantable coil and a voxe l size of 3x3x3mm. The relationship is linear with R 2 =0.9987, n=1. However, more repetitions need to be performed to demonstrate the reproducibility of this calibration curve. Because we demonstrated that a surface coil inductively coupled to the implant ed coil can be used to obtain calibration curve with the choline solutions, we took the next step to correlate the cellular choline (TCho) content and the cell number, a method to determine the viability of cells. The calibration curve shown below (Figure 3 8) is obtained with LVM alginate. The mixture was poured into the implantable coil and NMR studies were performed. A graph was plotted with the ratio of area under TCho and water peaks vs. the cell number. We used a 3x3x3 mm voxel because we could not get a signal with the 2x2x2 voxel. y = 0.0019x 0.0019 R = 0.9987 0.000 0.050 0.100 0.150 0.200 0.250 0 20 40 60 80 100 120 choline concentration (mM) Choline calibration curve ratio ofarea of choline peak to the water peak

PAGE 69

69 The data obtained below is with two different cell numbers obtained with 2 repetitions. Experiments need to be conducted with more cell numbers and repetitions. Figure 3 8 The graph is plotted between the cell number in volume of interest and ratio between area under choline peak and area under water peak. The data was obtained with an inductively coupled implantable coil system with a voxel size of 3x3x3mm. 3.9 Discussion and Future Work The loop gap resonator coil coated with PDMS layer which can be used as inductively coupled implantable coil system was proven to have high sensitivity for proton NMR measurements by Nelly Volland. By using implantable coil which surrounds the bioartificial pancreas we can obt ain a more localized signal with better signal to noise. Cherie Stabler demonstrated in her studies that 1 H proton spectroscopy can be used to measure the TCho in cells entrapped in alginate microbeads; a version of a bioartificial pancreas. However, the data obtained by Cherie Stabler was without an internal reference which gave large standard 0 0.0005 0.001 0.0015 0.002 0.0025 0.00E+00 5.00E+03 1.00E+04 1.50E+04 2.00E+04 2.50E+04 Choline calibration curve 3x3x3 voxel choline calibration curve 3*3*3 voxel rel area of choline/rel area of water cell number in VOI

PAGE 70

70 deviations 25 This large standard deviation was due to the change in sensitivity of the coil with the changes in the environment and also because of the variance of shimming from experimental session to session. It is almost impossible to obtain the same shim every time the experiments are conducted. Hence, we need an internal reference, whose signal is always constant, such as water, and which can be compared w ith the sample signal, which is variable. Therefore, even if there are changes in the sensitivity of the coil leading to changes in signal intensity, the ratio of the signal from the sample to the signal from the reference ideally remains the same. This ratio can be plotted to obtain a calibration curve with minimum standard deviation. Our studies suggest this will be a much superior approach. In this study the coupled coil system was used to obtain calibration curve s from choline tet cells. The ratio of area under the TCho at ~3.2 ppm and the area under the water peak at ~4.7MHz was always the same for a particular concentration of solution when the RF gain, number of averages, TR and TE remai n constant. The first curve obtained with the volume coil demonstrates that the choline signal can be obtained using NMR spectroscopy, and the peak area changes with change in choline concentration. The next step was performed to prove that the surface c oil coupled inductively with the implantable coil can give a signal with better S/N and has better signal sensitivity. Different voxel sizes were chosen and spectroscopy was performed on them. It was observed that the 2x2x2 voxel can be shimmed better an d therefore has better S/N compared to the 3x3x3 volume. However the 3x3x3 voxel has better signal sensitivity due to the increased number of spins it includes. Therefore, more studies need to be performed to optimize a voxel size which has better S/N an d signal sensitivity. Another calibration curve was initiated tet cells mixed in 3% LVM alginate. Different concentration of cells were taken and plotted against the TCho signal obtained. It was

PAGE 71

71 difficult to obtain a good shim with a acceptable linewidth. The voxel size was 3x3x3 in these studies because the signal was too small in the 2x2x2 volume. In future studies, experiments need to be performed with more averages (n=512) and a longer TR (4000ms) to obtain better signal. Variations in cell numbers must also be included to create a meaningful and us eful calibration curve. These calibration curves can be used to non invasively monitor tet cells embedded in an implanted bioartificial pancreas. This approach will help monitor the viability and metabolic activity of these insulin secreting cells. We believe this approach will allow us to predict the impending failure of a construct, allowing necessary measures to be taken to replace the construct. Also, this approach may assist our development of better bioartificial pancreas designs by letting us optimize the best alginate mixture and gelling agents that allow for steady blood gluc ose regulation and construct longevity. In the future, we plan to induce cell death in the construct and monitor the change in TCho peak with time. This will be an important study to analyze how TCho peak changes with cell death and how long it takes befo re we actually see the change in TCho after the cell death. We also plan to perform NMR studies to detect immune cell infiltration into, and fibrotic growth around the construct using NMR techniques. These processes should be the most common in rejection of a construct. 3.10 Conclusion Bioartificial pancreata were manufactured with tet cells in different combinations of alginate varieties and gelling agents. The plugs were first tested in vitro for their structural integrity and glucose response followed by implantation into diabetes induced C 3 H/HeN mice. Diabetic animals were monitored for body weight and blood glucose level changes for 30 days

PAGE 72

72 following the implantation. The plugs were removed at the end of the experiments and studied with H/E stains and Von Kossa stains for cell proliferation and bead mineralization respecti vely. The results of these studies showed that a functional bioartificial pancreas could be designed if appropriate biomaterials and gelling conditions were met. The successful design (the Variant 4 bioartificial pancreas), was manufactured by entrapping the cells directly into in 3% LVM, using both barium and calcium salts to gel the material. This design worked as desired: i.e., long term normoglycemia was achieved in the diabetic animals. This construct is designed such that it can include a PDMS coat ed implantable coil. The implantable coil can be coupled inductively with a surface coil to obtain NMR data with improved S/N. Work here also demonstrates that a choline calibration curve can be generated. Moreover, by using an internal reference, the calibration should be superior to what we previously had determined. The future of the bioartificial pancreas now is brighter than before, and although there are many hurdles ahead, this approach is showing great promise towards controlling T1D.

PAGE 73

73 APPENDI X A PROTOCOL FOR MAKING ALGINATE BEADS To make 3 % alginate beads Add 30mg of powdered LVM alginate in 10mL of 0.85 NaCl (physiological saline) to make a 3% LVM mixture. Spin overnight with a magnetic stir bar. Filter the mixture into sterile centrifuge tu be under the hood using a 0.22m Prepare bead making apparatus Place the bead maker and the electrostatic bead maker under the hood Place a bowl with sterilized 0.275% BaCl 2 filled to exactly 2cm distance between the surface of the solution and the top of the vessel with a small magnetic stir bar inside the bead maker Attach the sterilized needle (diameter 0.35m) and tubing. Lower the armature such that the needle aligns with rim of the bowl. Close the plexiglass doors and set the voltage to ~6kV Set the syringe pump rate to 0.3 ml/min Set the syringe diameter to 12.06mm for a 5cc syringe and set the target volume. Cell preparation Remove cells from flasks by trypsinization, add DMEM, count cells and centrifuge them at 500rpm for 10mins. Suck out a ll the media, add PBS with Ca ++ and Mg ++ and centrifuge again at 500rpm for 15mins. Remove PBS, leaving cell pellet. Add 5m L of 3% LVM to 10x10 7 cells and mix well. Put this mixt ure into a 5mL syringe. Attach this syringe to the tubing attached to the bea d maker needle. Secure the syringe to the syringe pump and hit the run button and turn on the bead maker. Monitor and adjust the voltage if needed. When all the cell alginate mix from the syringe has formed into beads in the bowl turn off the pump and the bead maker. Remove the bowl and suck out all the 0.275% BaCl 2 Rinse the beads 6 times with 0.275% BaCl 2, and 6 times with PBS (no Calcium and Magnesium). Take a small sample and measure the size of the beads under the microscope. Rinse the armature, tubi ng and needle. Vacuum dry.

PAGE 74

74 APPENDIX B PROTOCOL FOR MAKING BIOARTIFICIAL PANCRE AS WITH 3% LVM BEADS IN 2% LVG Use the regular bead making protocol to make beads with 2x10 7 cells per mL of the 3% LVM. After the beads are rinsed remove all the PBS and add t he 2% LVG solution (volume of LVG added= 2x0.63xvolume of beads). Mix it well and make sure lumps are not formed Set up the multiple magnetic stirrers, with 6 magnets and PDMS formers in Petri dishes fixed on the magnets. Make sure all the magnets spin we ll. Dip filter papers in 1.1% CaCl 2 and place them in the Petri dish. Into the PDMS formers, place an empt y PDMS constructs. Pipette 0.2mL of the prepared mixture into the construct, filling it, and place another filter paper on top of it. Fill the Petri dish with 1.1%CaCl 2 and start spinning. Let it spin for 20 mins continuously. buffer before implanting into the animals. APPENDIX C PROTOCOL FOR MAKING A BIOARTIFICIAL PA NCREAS WITH 3% LVM B EADS ALGINATE MIXES Use the regular bead making protocol to make beads with 2*10 7 cells per mL of the 3%LVM. After the beads are made, rinse them 6 times with 0.14% BaCl 2 and then 6 times with PBS solution. Mix the beads with appropri ate of 3% LVM without letting bubbles form. Set up the multimagnet stirrer following the regular construct making protocol with stirring speed between 2 3. Dip the filter paper in 0.275% BaCl 2 and place the construct on top of it. Now usin g a 1mL syringe pour about 0.2mL of the 3% LVM beads mix into the construct. (Take care not to let bubbles form.) Place another filter paper dipped in 0.275% BaCl 2 and place it on top of the construct and start spinning. While spinning fill the dish with 0.275% BaCl 2 Let it spin for 40 mins. (check regularly to make sure the magnets are spinning uniformly). After 40 mins, stop the stirrer and suck out all the BaCl 2 from the dish. Add 1.1% CaCl 2 without disturbing the construct and let it spin again for 15mins. Take off th e top filter paper and lift the construct slowly and using a sterile forceps pop out the gelled plug very carefully.

PAGE 75

75 APPENDIX D PROTOCOL FOR MAKING 3% LVM OR 2% LVG PLU GS Set up the multi magnet stirrer, the PDMS holder, filter papers and the PDMS co nstruct as described above. Mix 1x10 7 trypsinized cells with 1mL of 3% LVM. Usin g a 1ml syringe pour about 0.2mL of the alginate cell mix into the construct. Place another filter paper dipped in 1.1% CaCl 2 on top of the construct, fill the Petri dish wit h the same solution and spin it for 35 mins. After 35 mins remove the solution and add 0.412% BaCl 2 and spin again for 30 mins. Remove the filter paper and pop the alginate plug out of the PDMS construct. APPENDIX E PROTOCOL FOR CASTING PDMS The curing agent should be mixed with the base for the curing process of the PDMS. The ratio of curing agent to the base should be 1:10. The two components should be mixed well until you are sure its uniform by agitating the mixture. This reaction will produce lots o f bubbles. The air bubbles form but must be broken to get a bubble free uniform sheet of PDMS. To degas the mixture the container must be placed in 14 psi vacuum for 15 mins. The bubbles rise to the surface. Remove the container from the vacuum and poke th e bubbles with a needle. After most of the surface bubbles are poked place the container in the vacuum again for 15 mins and remove. Repeat this until all large bubbles are gone. The calculated volume of PDMS which is bubble free is poured on to a glass sl ab. Place another glass slab on top of this cast at 1mm or 2 mm height whichever thickness is needed. Leave it over night as it can cure at room temperature. After the PDMS is cured it must be removed and separated from the glass slabs. Remove the top slap slowly and then take a spatula and separate the cured PDMS from the glass slab carefully. Cut the PDMS cast into desired circular sizes as shown in the figure. APPENDIX F PROTOCOL FOR MAKING AN IMPLANTABLE RF CO IL A 202m thick copper sheet was cut into 2mm wide strips. A circular wooden bar of 1.2cm diameter was used to get the desired coil diameter. Loop the 2mm wide copper strip around the wooden bar. Using a soldering iron, solder a 4.3pF capacitor to the copper strip. Measure the resonance frequen cy of the coil.

PAGE 76

76 APPENDIX G PROTOCOL FOR THE VON KOSSA STAIN Material needed: silver nitrate ultra pure grade (Acros), sodium thiosulfate, 98% extra pure, anhydrous (Acros organics), Aluminium sulfate hydrate, 97+% (Alfa Aesar), nuclear fast red, 94%, pu re (Acros Organics), Xylene, Ethanol. Deparaffinize and hydrate the slide Place the slide in 5% silver nitrate solution and expose it to 60 watt lamp for 1 hour. Place a silver soil on the opposite side of the lamp to reflect the light. Rinse the slide in distilled water 3 times. Place the slide in 5% hypo solution for 5 minutes. Wash the slide in tap water and then in distilled water Place the slide in nuclear fast red for 5 minutes Wash the slide in water Dehydrate the slide by washing it in ethanol and x ylene for 3 minutes each Clear and coverslip the slide.

PAGE 77

77 REFERENCES 1. Skelin M, Rupnik M, Cencic A. Pancreatic Beta Cell Lines and their Applications in Diabetes Mellitus Research. Altex Altern Anim Ex. 2010;27(2):105 113. 2. Sper ling M. Diabetes mellitus in children Philadelphia: Saunders; 2005. 3. Sperling M. Type 1 diabetes : etiology and treatment Totowa, N.J.: Humana Press; 2003. 4. Yoon JW, Jun HS. Autoimmune destruction of pancreatic beta cells. American journal of thera peutics. Nov Dec 2005;12(6):580 591. 5. Tseng KH. Standards of medical care in diabetes -2006: response to the American Diabetes Association. Diabetes care. Nov 2006;29(11):2563 2564; author reply 2564 2565. 6. Lebovitz HE, American Diabetes Association. Therapy for diabetes mellitus and related disorders 4th ed. Alexandria, Va.: American Diabetes Association; 2004. 7. Alberti KG, Zimmet PZ. Definition, diagnosis and classification of diabetes mellitus and its complications. Part 1: diagnosis and classi fication of diabetes mellitus provisional report of a WHO consultation. Diabetic medicine : a journal of the British Diabetic Association. Jul 1998;15(7):539 553. 8. Mathis D, Vence L, Benoist C. beta Cell death during progression to diabetes. Nature. Dec 13 2001;414(6865):792 798. 9. Nathan DM. Long Term Complications of Diabetes Mellitus. New Engl J Med. Jun 10 1993;328(23):1676 1685. 10. Atkinson MA, Maclaren NK. Islet cell autoantigens in insulin dependent diabetes. J Clin Invest. Oct 1993;92(4):1608 1616. 11. Diagnosis and classification of diabetes mellitus. Diabetes care. Jan 2005;28 Suppl 1:S37 42. 12. Piper M, Seidenfeld J, Aronson N. Islet transplantation in patients with type 1 diabetes mellitus. Evid Rep Technol Assess (Summ). Jul 2004(98):1 6. 13. Kizilel S, Garfinkel M, Opara E. The bioartificial pancreas: progress and challenges. Diabetes technology & therapeutics. Dec 2005;7(6):968 985. 14. Silva AI, Mateus M. Development of a polysulfone hollow fiber vascular bio artificial pancreas de vice for in vitro studies. Journal of biotechnology. Feb 5 2009;139(3):236 249.

PAGE 78

78 15. Constantinidis I, Long R, Jr., Weber C, Safley S, Sambanis A. Non Invasive monitoring of a bioartificial pancreas in vitro and in vivo. Annals of the New York Academy of Sc iences. Nov 2001;944:83 95. 16. Sweet IR, Cook DL, Lernmark A, et al. Systematic screening of potential beta cell imaging agents. Biochemical and biophysical research communications. Feb 20 2004;314(4):976 983. 17. Gambhir SS, Herschman HR, Cherry SR, et al. Imaging transgene expression with radionuclide imaging technologies. Neoplasia. Jan Apr 2000;2(1 2):118 138. 18. Yong J, Rasooly J, Dang H, et al. Multimodality imaging of beta cells in mouse models of type 1 and 2 diabetes. Diabetes. May 2011;60(5): 1383 1392. 19. Macdonald JM, Grillo M, Schmidlin O, Tajiri DT, James TL. NMR spectroscopy and MRI investigation of a potential bioartificial liver. NMR in biomedicine. Apr 1998;11(2):55 66. 20. Ross B, Bluml S. Magnetic resonance spectroscopy of the huma n brain. The Anatomical record. Apr 2001;265(2):54 84. 21. Stuckey DJ, Ishii H, Chen QZ, et al. Magnetic resonance imaging evaluation of remodeling by cardiac elastomeric tissue scaffold biomaterials in a rat model of myocardial infarction. Tissue enginee ring. Part A. Nov 2010;16(11):3395 3402. 22. Hartman EH, Pikkemaat JA, Vehof JW, Heerschap A, Jansen JA, Spauwen PH. In vivo magnetic resonance imaging explorative study of ectopic bone formation in the rat. Tissue engineering. Dec 2002;8(6):1029 1036. 2 3. Dhollander AA, Huysse WC, Verdonk PC, et al. MRI evaluation of a new scaffold based allogenic chondrocyte implantation for cartilage repair. European journal of radiology. Jul 2010;75(1):72 81. 24. Darquy S, Reach G. Immunoisolation of pancreatic B cel ls by microencapsulation. An in vitro study. Diabetologia. Oct 1985;28(10):776 780. 25. Constantinidis I, Rask I, Long RC, Jr., Sambanis A. Effects of alginate composition on the metabolic, secretory, and growth characteristics of entrapped beta TC3 mouse insulinoma cells. Biomaterials. Nov 1999;20(21):2019 2027. 26. Stabler C, Wilks K, Sambanis A, Constantinidis I. The effects of alginate composition on encapsulated betaTC3 cells. Biomaterials. Jun 2001;22(11):1301 1310. 27. Teramura Y, Kaneda Y, Iwata H. Islet encapsulation in ultra thin layer by layer membranes of poly(vinyl alcohol) anchored to poly(ethylene glycol) lipids in the cell membrane. Biomaterials. Nov 2007;28(32):4818 4825.

PAGE 79

79 28. Lim F, Sun AM. Microencapsulated Islets as Bioartificial Endocr ine Pancreas. Science. 1980;210(4472):908 910. 29. O'Shea GM, Goosen MF, Sun AM. Prolonged survival of transplanted islets of Langerhans encapsulated in a biocompatible membrane. Biochimica et biophysica acta. May 22 1984;804(1):133 136. 30. Fritschy WM, Wolters GH, van Schilfgaarde R. Effect of alginate polylysine alginate microencapsulation on in vitro insulin release from rat pancreatic islets. Diabetes. Jan 1991;40(1):37 43. 31. Iwata H, Amemiya H, Matsuda T, Takano H, Hayashi R, Akutsu T. Evaluation of microencapsulated islets in agarose gel as bioartificial pancreas by studies of hormone secretion in culture and by xenotransplantation. Diabetes. Jan 1989;38 Suppl 1:224 225. 32. Iwata H, Takagi T, Amemiya H, et al. Agarose for a bioartificial pancre as. Journal of biomedical materials research. Jul 1992;26(7):967 977. 33. Dionne KE, Colton CK, Yarmush ML. Effect of hypoxia on insulin secretion by isolated rat and canine islets of Langerhans. Diabetes. Jan 1993;42(1):12 21. 34. Dionne KE, Colton CK, Yarmush ML. Effect of oxygen on isolated pancreatic tissue. ASAIO transactions / American Society for Artificial Internal Organs. Jul Sep 1989;35(3):739 741. 35. Efrat S, Fusco DeMane D, Lemberg H, al Emran O, Wang X. Conditional transformation of a pancr eatic beta cell line derived from transgenic mice expressing a tetracycline regulated oncogene. Proceedings of the National Academy of Sciences of the United States of America. Apr 11 1995;92(8):3576 3580. 36. Martinsen A, Skjak Braek G, Smidsrod O. Algin ate as immobilization material: I. Correlation between chemical and physical properties of alginate gel beads. Biotechnology and bioengineering. Jan 5 1989;33(1):79 89. 37. Volland NA, Mareci TH, Constantinidis I, Simpson NE. Development of an inductively coupled MR coil system for imaging and spectroscopic analysis of an implantable bioartificial construct at 11.1 T. Magnetic resonance in medicine : official journal of the Society of Magnetic Resonance in Medicine / Society of Magnetic Resonance in Medici ne. Apr 2010;63(4):998 1006. 38. Balc M. Basic H and C NMR spectroscopy 1st ed. Amsterdam ; Boston: Elsevier; 2005. 39. Yung AC, Kozlowski P. Signal to noise ratio comparison of phased array vs. implantable coil for rat spinal cord MRI. Magnetic resona nce imaging. Oct 2007;25(8):1215 1221.

PAGE 80

80 40. Silver X, Ni WX, Mercer EV, et al. In vivo 1H magnetic resonance imaging and spectroscopy of the rat spinal cord using an inductively coupled chronically implanted RF coil. Magnetic resonance in medicine : offici al journal of the Society of Magnetic Resonance in Medicine / Society of Magnetic Resonance in Medicine. Dec 2001;46(6):1216 1222. 41. Wirth ED, 3rd, Mareci TH, Beck BL, Fitzsimmons JR, Reier PJ. A comparison of an inductively coupled implanted coil with optimized surface coils for in vivo NMR imaging of the spinal cord. Magnetic resonance in medicine : official journal of the Society of Magnetic Resonance in Medicine / Society of Magnetic Resonance in Medicine. Nov 1993;30(5):626 633. 42. Decorps M, Blon det P, Reutenauer H, Albrand JP, Remy C. An Inductively Coupled, Series Tuned Nmr Probe. Journal of Magnetic Resonance. 1985;65(1):100 109. 43. De Graaf RA. In vivo NMR spectroscopy : principles and techniques 2nd ed. Chichester, West Sussex, England ; H oboken, NJ: John Wiley & Sons; 2007. 44. Bolan PJ, DelaBarre L, Baker EH, et al. Eliminating spurious lipid sidebands in 1H MRS of breast lesions. Magnetic resonance in medicine : official journal of the Society of Magnetic Resonance in Medicine / Society of Magnetic Resonance in Medicine. Aug 2002;48(2):215 222. 45. Bolan PJ, Meisamy S, Baker EH, et al. In vivo quantification of choline compounds in the breast with 1H MR spectroscopy. Magnetic resonance in medicine : official journal of the Society of Ma gnetic Resonance in Medicine / Society of Magnetic Resonance in Medicine. Dec 2003;50(6):1134 1143. 46. Kassem MNE, Bartha R. Quantitative proton short echo time LASER spectroscopy of normal human white matter and hippocampus at 4 Tesla incorporating macr omolecule subtraction. Magnet Reson Med. May 2003;49(5):918 927. 47. Stabler CL, Long RC, Sambanis A, Constantinidis I. Noninvasive measurement of viable cell number in tissue engineered constructs in vitro, using 1H nuclear magnetic resonance spectroscop y. Tissue engineering. Mar Apr 2005;11(3 4):404 414. 48. Stabler CL, Long RC, Jr., Constantinidis I, Sambanis A. In vivo noninvasive monitoring of a tissue engineered construct using 1H NMR spectroscopy. Cell transplantation. 2005;14(2 3):139 149. 49. Co nstantinidis I, Stabler CL, Long R, Jr., Sambanis A. Noninvasive monitoring of a retrievable bioartificial pancreas in vivo. Annals of the New York Academy of Sciences. Jun 2002;961:298 301.

PAGE 81

81 BIOGRAPHICAL SKETCH degree with second rank for her overall performance in Biomedical engineering department from the reputed Osmania University, Hyderabad, India in the year 2005. Her research interest is concentrated in the field of tissue engineering and she continued he r research at the University of Florida as a masters student. For her excellent academic performance she was awarded with an achievement award.