DEVELOPMENT OF A PAPER BASED LATERAL FLOW DEVICE FOR THE DETECTION OF IÎ±IP VIA COMPETITIVE ELISA

This research has the focus on developing a paper-based device capable of sequentially flowing three separate fluids. Different assembly methods are investigated and compared to develop a more streamlined approach for the prototyping and testing of biological reagents. A faster method of fabrication is developed followed by an investigation of the potentials to detect the presence of Inter α Inhibitor Protein (a biomarker associated with sepsis). This is primarily achieved by applying the techniques of an Enzyme Linked Immunosorbent Assay (ELISA) onto a paper based platform. The research presented shows the steady progression toward and optimization of a lateral flow device capable of achieving the above and eventually presenting the capability of a competitive ELISA to detect IαIP. This platform allows for the possibility to generate results within a fraction of an hour and furthermore with smaller fluid volumes as compared to current bench top tests, such as ELISA which is typically conducted in 96 well plates.

I also have to thank my advisor, Dr. Mohammad Faghri, who has been a continued source of guidance, motivation and support throughout this year. I certainly will not forget the drive, determination, and persistence that he has instilled into me in my ambitions to complete this project. These are characteristics that I will take with me during my studies in Germany and further onto my career.
I owe Dr. Constantine Anagnostopoulos a great gratitude as he was always able to provide direction and solution during even the busiest points of the semester.
I have to admire his sense of composure and calmness in the realm of problem solving, as he always discussed ideas in a logical manner; in doing so he was often able to point out likely sources of error and suggestions for next steps. Our      The first chapter of this master's thesis serves to introduce point of care devices and describe the current state of their development.

Point of Care Devices an Introduction
Point of Care (POC) devices are medical devices developed for use outside of laboratory settings and are most commonly applied to disease detection. These devices are fast becoming a popular area of research as they offer improvements in 2 scalability, cost, detection times, and access [3]. Although POC devices are typically associated with resource limited settings, the benefits they offer are proving to be advantageous for a variety of applications including: hospital rooms, emergency response scenarios, at home, and military settings [1]. The common characteristic that is sought after POC diagnostics is "simplicity"; which is defined by the Food and Drug Administration in table 1-1:  Is a fully automated instrument or a unitized or self-contained test.  Uses direct unprocessed specimens, such as capillary blood (fingerstick), venous whole blood, nasal swabs, throat swabs, or urine.  Needs only basic, non-technique-dependent specimen manipulation, including any for decontamination.  Needs only basic, non-technique-dependent reagent manipulation, such as "mix reagent A and reagent B."  Needs no operator intervention during the analysis steps.  Needs no technical or specialized training with respect to troubleshooting or interpretation of multiple or complex error codes.  Needs no electronic or mechanical maintenance beyond simple tasks, e.g., changing a battery or power cord.  Produces results that require no operator calibration, interpretation, or calculation .  Produces results that are easy to determine, such as 'positive' or 'negative,' a direct readout of numerical values, the clear presence or absence of a line, or obvious color gradations.  Provides instructions in the package insert for obtaining and shipping specimens for confirmation testing in cases where such testing is clinically advisable.  Has test performance comparable to a traceable reference method as demonstrated by studies in which intended operators4 perform the test. If a reference method is not available for a test you are proposing for waiver, please contact OIVD to discuss your proposed plan prior to submitting your application.  Contains a quick reference instruction sheet that is written at no higher than a 7th grade reading level.
In an effort to meet these criteria, corporations have begun to develop and market POC tests which make a simpler device including or exclude instrumentation for detection [5].

Instrumented Devices
 Permanent integrated instruments-Designed for high-throughput to return results quickly and accurately within a laboratory setting. However, they cannot be considered "simple tests" due to the trained personnel needed. [6]  Permanent Instruments using disposable components-Included in this category are disposable components on which results are developed to be processed by an analyzer (ie. handheld device or laptop). The analyzer may also serve as a source of power to move fluids within the system, for thermal control, or for detection purposes [5]. These can interpret multiple tests over a short period of time with the aid of a processor. Additionally, results may be transmitted to doctors via internet for further examination. This is an ideal scenario for POC settings, as it brings the expertise of a hospital to the frontlines of disease. The downside is that the processor can drive up the cost of such a test, and may require special components for repair when damaged. [6]

Un-Instrumented (Autonomous) Devices
 Pure disposables-Usually develop a detection scheme on a disposable platform (ie. paper) which integrates many inexpensive components and reagents. A major drawback of many disposable platforms is that they lack the sensitivity that lab tests provide and cannot provide quantitative results.

POC Microfluidic Devices
Although POC tests have also been developed in many forms to utilize advantages of different materials, the greater majority do not fall into the category of microfluidic devices [7]. Microfluidic devices, allow for a manipulation of small volumes of fluid, which reduces the consumption of reagents and of sample fluids.
This means smaller sample fluid volume is needed from patients; depending on the sample needed this could cause discomfort to the patient upon extraction. Also, for detection of disease in infants, larger volumes of samples may not be possible.
Infants cannot provide as high a volume of blood as compared to an adult.
Microfluidic devices may further be manipulated by an Instrumented Device, which can incorporate pumps and channels to direct fluid toward analysis sites in the device. Channels can be made from glass, silicon, and polymers (ie. polydimethylsilaxane PDMS). The integration of pumps and chips increases complexity and correspondingly, increase the chance for mechanical malfunction [8].
In a resource-limited setting replacement parts may be hard to find.
In turn, as more research groups are focusing on the development of POC diagnostics, they are turning to paper-based or fibrous-based materials. These materials are generally cheaper to manufacture and fall within the constraints for a successful POD device: Affordable, Sensitive, Specific, User-friendly, Rapid and robust, Equipment-free, and Deliverable to end users (ASSURED) [8; 7].

POC Diagnostic Applications
Concerning Global Health Issues, POC devices have been the turn-to solution as they are rapid, simple, and inexpensive, making them more readily accessible [5].
POC devices are designed to detect the presence of any chemical substance or compound that can help to indicate the presence of disease, infection, or other abnormality. This is called an analyte. Most POC tests utilize blood, but depending on the application, saliva or urine may also be used. Arguably, saliva is the most preferred solution for POC testing, as it is the least invasive and is least hazardous to both patients and health care workers; the problem is that it may have a varying consistency of analyte concentration when compared to that in blood or urine [5].
Depending on the target, any of these fluids may be used in a range of diagnostic applications, including but not limited to: Virus detection, diabetes, bacterial Infection, pregnancy, allergy, agriculture and disease [6; 8; 9; 10].

Dipsticks
One of the simplest paper-based mechanisms is a Dipstick Assay or a Testing  These test strip or pH strips are pieces of paper that are treated with various concentrations acid-alkali solutions. After being exposed to a test sample, these dried reagents react with the in-sample hydrogen ions (H + ) to develop a color change.
Depending on the number of H + different color reactions will occur, which can be quantified by comparison to an included chart. [8] Urine Test strips can detect metabolic products such as leukocytes, nitrates, urobilinogen, protein, pH, haemoglobin, specifi Gravity, ketornes, bilirubin, and glucose. Diabetes, renal and urinary tract issues, liver diseases or haemolytic disorders are common applications [12].

Lateral Flow Assays
Another class of paper-based tests is lateral flow assays use fibrous materials to take advantage of their autonomous flow characteristics (capillarity). Additionally, they integrate various biological reagents to create the necessary reaction schemes for detection of virus, or infection. Lateral flow immunoassays (LFIA) have become widely accepted for a broad range of diagnostic needs. LFIA generally consist of a sample pad, a conjugate pad, an absorbent pad and a detection area. The detection area produces two indicator lines (test line and a control line) when a sample runs through it. The absorbent pad provides driving force through capillarity to pull a sample fluid along the test strip [8]. A membrane is incorporated ahead of the absorbent pad to serve as the testing platform, where diagnostic schemes are developed. Capture molecules such as enzymes or antibodies may be isolated on this membrane; which eventually help to develop the detection scheme [8]. Just before the membrane, lays a conjugate pad which stores detection molecules, which would provide signal or color change at the end of the test. These labeled-molecules have a high affinity to bind the target. Eventually, the label will provide the visual proof for presence of a target analyte.
As a sample fluid is applied, it passes first through the conjugate pad, where the target analyte binds to labeled antibodies; this process develops the top half of what is commonly called an 'Assay-Sandwich'. This complex is then pulled by capillary flow toward the membrane, which has on it anchored molecules designed only to capture the top half of the sandwich compound. When this biding occurs the Assay-8 Sandwich is complete and consists of an anchored capture molecule, an analyte, and a labeled detection molecule. The accumulation of such complexes forms the test line on the device and is the basis for any Lateral Flow Immunosorbent Assay (LFIA) [13]. Most commonly, this type of reaction scheme utilizes antibodies and to capture a virus and develop diagnostic results.

Biology of a Lateral Flow Immunosorbent Assay
Antigens and antibodies are the two major biological components necessary to effectively run an immunoassay. The term immunoassay, combines the word "immuno", referring to an immune response, and "Assay", referring to a test; thus, an immunoassay combines antibodies (generated in correlation with an immune response) and antigens (responsible for the immune response) to develop a diagnostic test.
Antigens are foreign or toxic entities that elicited a response by an immune system. This response is characterized by a development of antibodies, which bind to foreign entities (virus or bacteria), effectively marking them for elimination.
Antibodies, also known Immunoglobulin (Ig), are large multifunctional proteins created by the immune system to identify and to help eliminate antigens. As glycoproteins they are composed of four polypeptide chains which covalently link to one another through disulfide bonds to form the characteristic Y-shape [14]. As depicted by the below figure, antibodies are made from two light chains and two heavy chains. The antibody structure is sub-divided into two regions: variable region 9 and a constant region. these are discussed.

Antibody Description Complexes
IgA Can be found in areas that line the gut, respiratory tract, and urogenital tract.

IgD
Serves as an antigen receptor on B-cell, unexposed to antigens. B-Cells are part of the immune system; they are responsible for making immunoglobulins.

IgE
Involved in the histamine release associated with allergens, also protects against parasitic worms IgG Most commonly found isotope and is responsible for majority of antibody-based immunity.

IgM
This is responsible for the immune response during early stages of immune system development (before sufficient IgG is created) IgG is the most common isotope found in human serum, it represents 85% of the antibodies that can be found in serum [17].
Antibodies may also be present in Polyclonal and Monoclonal forms. The primary difference resides in these forms inherent ability to bind to epitopes.
Epitopes are binding sites on a surface of an antigen to which antibodies, B-Cells, or T-cells bind. This region is also called a determinant or antigenic determinant. There may be many different epitopes on the surface of an antigen available for an antibody [18].
Antibodies are isolated within sheep, rabbit, or goats. An immune response is elicited by injecting these animals with an antigen (this may be done a number of times). After an immune response develops, blood is extracted and serum is isolated.
This serum which contains the molecules created in response to an antigen is called antiserum; it may contain a mixture of antibodies, each may have an affinity to bind to a different epitope on the antigen (see below) [18]. On the other hand, monoclonal antibodies are more specific, in that they will bind only to a specific epitope on an antigen (see below). They are isolated from a hybridoma cell line, which are tumor cells that produce many copies of a particular antibody. Tumor cells can produce antibodies indefinitely and consistently [18]. By taking advantage of this chemistry and binding affinity between antibodies and antigens, diagnostic tests can be developed.

Enzyme Linked Immunoassay (ELISA)
As it stands, Enzyme Linked Immunoassay (ELISA) is the gold standard for antigen detection within a laboratory setting. It offers superior sensitivity and specificity and is robust and easy-to-use, but trained person is needed to effectively run a test.
An ELISA is conducted by placing a sample fluid onto a functionalized surface to make it easier for antibodies to bind. As the solution incubates, the antibodies' affinity to a particular analyte helps to develop antibody-antigen complexes. After a certain amount of time, the surface is then washed to remove any unbound proteins, 12 and a second fluid containing detection antibodies is added. As mentioned previously, detection antibodies are labeled with an indicator that eventually induces a signal. In the case of ELISA, detection antibodies are labeled with an enzyme. When an "ELISA Sandwich" complex is formed it includes an anchored antibody, the analyte and labeled detection antibody. Following a second wash step to further remove any unbound proteins, a substrate is added. The substrate interacts with the enzyme bound to the detection antibody to create a measureable indication through color change or electrochemical signal [5]. The figure below provides a visualization and comparison between the reaction schemes used in a Lateral Flow Assay and an Enzyme Link Immunosorbant Assay.

13
A major challenge of adapting ELISA to a POC platform is the need for multiple fluid steps. Additionally many of the necessary reagents needed for ELISA become ineffective if stored at room temperature. None the less, the superior sensitivity of ELISA is the driving force for many research groups in their endeavors to bring ELISA into a Lateral Flow Setting.

Lateral Flow Assay Reaction Schemes
Lateral Flow Assays may be conducted within two formats or reaction schemes, through a Non-Competitive Assay (or "Sandwich Assay" as described in section 1.4) or through a Competitive Assay. It is important to note that solid phase immunoassays have played a large role in assay development in the last 70 years.
Solid Phase refers to the solid surface (i.e. microtiterplate or a membrane) on which proteins (ie. capture antibodies) may be adhered. This surface functionalization is the key functional unit within an ELISA reaction scheme.

'Sandwich Assay' or Non-Competitive Assay
The sandwich principal was briefly described in section 1.4, but a review and comparison will serve to thoroughly translate the concept. The first step in a noncompetitive set up is to functionalize a surface for the anchoring of a capture antibodies. The surface may be within a 96 well micro-titer plate or similarly may be a strip of nitrocellulose, the most common detection area in paper based devices.
Once the surface has been functionalized, a particular capture antibody may be anchored through incubation in a microtiter well or by dropping a small volume of 14 solution onto the fibrous membrane. A sample containing analyte may then be added to develop the bottom of a so called 'Assay Sandwich.' A wash step is then introduced by adding a buffer solution to remove any unbound complexes.
Subsequently, a solution containing labeled antibodies is added, these will complete the assay sandwich and furthermore will serve to provide the necessary indicator for binding activity and thus the presence of analyte. After analysis, higher signal intensities will be indicative of a greater presence of detection antibodies and thus a higher concentration of the analyte. The below figure further illustrates this process [19].
 Step 1: Sample is applied after capture antibodies have been immobilized  Step 2: Sample fluid is removed or washed to remove unbound proteins  Step 3: Labeled detection antibodies are added  Step 4: Unbound detection antibodies are removed or washed away  Step 5: "Assay Sandwich" is completed and signal analysis may be conducted.

Competitive Assay
The Sandwich or non-competitive assay may not always be a favorable choice due to the limitation that smaller particles (compared to antibodies) may not elicit the same binding affinity. These particles include analytes such as drugs or environmental pollutants; in these cases, a competitive assay may be used.
Antibodies specific to a particular target antigen are anchored under the same methods for a non-competitive assay. The major difference between the schemes resides in the sample fluid. Within a competitive platform, detection antibodies are not used, but rather the sample (which allegedly contains an analyte) is prepared with a labeled version of the analyte in question. When the sample containing defined quantities of labeled-analytes and un-labeled analytes is introduced to a predefined set of capture antibodies, they compete for limited binding sites. Another difference in the competitive scheme is that signal intensity is inversely proportional to the concentration of target analyte. When there are fewer analytes present, more labeled analytes will bind to the capture antibodies producing a stronger signal [19].
The following figures illustrate the two scenarios, one with higher concentration of labeled analyte and one with a lower concentration of labeled analyte in relation to the target analyte.  Step 1: Sample containing labeled and un-labeled analytes is added a platform containing capture antibodies  Step 2: Sample is removed or wash away to remove unbound analytes  Step 3: Signal analysis is conducted.

Assays and Antibody Use
When designing an assay it is important to consider the clonality of antibodies to choose an appropriate capture protein and for detection. As mentioned previously, monoclonal antibodies may bind to only one epitopes or binding site on a particular analyte. Within the "Assay Sandwich" the capture antibody is generally a variation of a monoclonal antibody, for the reason that it is desirable to have the highest possible specificity, capturing only the target analyte. By using an antibody that can recognize and bind only to epitopes on a specific analyte, the chance to capture erroneous proteins is reduced.
Once the detection antibody has been captured, the desire remains to generate the highest possible signal. Thus, polyclonal antibodies are used as the detection antibody in order to increase the chances for binding. Since polyclonal antibodies can bind to different epitopes on an analyte, thus the chances for binding are increased subsequently increasing the signal intensities that can be generated.

Common Detection Methods
Detection antibodies may have a range of labels providing the visual confirmation of target analyte presence. The label may come in different forms and may produce direct or indirect signals. The most popular labelling schemes are listed in the table below. Alkaline phosphatase (ALP), Horseradish peroxidase (HRP), beta-Galactosidase An ELISA utilizes an enzyme as a detection method because of the color change that occurs when a substrate is introduced. Enzymatic reaction schemes generally offer a signal with high contrast leading to a favorable signal-to-noise ratio and also have low limits of detection [22].

Sepsis an Introduction, Definition, and Motivation
The principals above explain the basic components and basic mechanisms that may be applied to point of care devices for to disease detection. In the past ELISA has been applied to HIV and to Malarial detection, but the proposal of this research two-fold, to bring ELISA to a paper based platform and to develop a test for the inflammatory disease sepsis.

Introduction to Sepsis
Sepsis has become a major health problem not only for developing countries, but also within industrialized societies [23]. Additionally, with a mortality rate precedence of 39.6%, sepsis ranks 13 th on the list of international classification of disease [24]. Within the US, there are more than 1,000,000 cases each year [25].
Sepsis is characterized by an inflammation in response to a foreign microbial infection. Although it is easy to recognize pathologically, it has long remained difficult to define and to diagnose. American physician, William Osler, once said that patients appear to die as a result of the body's response to an infection rather than from the infection itself [25].
If the early onset of symptoms is left untreated, the body's inflammatory response could progress to a state of shock and eventually could cause organ failure resulting in fatalities. Sepsis can develop from a number of sources including: pneumonia, blood-stream infections, intra-abdominal infections, urological infection and furthermore by surgical infection. The bacterial infection is the most common source, but symptoms can arise from fungi, virus, or even parasites [26]. A septic infection can start in the abdomen and urinary tract, but most often originates in the lungs with a 50% incidence [27].
Symptoms that arise from a septic infection are similar other infection, but may additionally include high fevers, hot/flushed skin, elevated heart rates, hyperventilation, altered mental status, swelling, and low blood pressure [28].

19
The medical community has struggled to define the illness holistically. In 2001, a system similar to the one that defines cancer progression was proposed. It included 'Predisposition, Infection, Response, and Organ Dysfunction' or PIRO [29].
This system has been updated to reflect the progression of sepsis from early onset to stages of potential mortality. See the below table for a more details. Severe Sepsis and one of the following:  Low systemic mean blood pressure  Need for dopamine

Refractory Septic Shock
Need for dopamine or norepinephrine or epinephrine to maintain average blood pressure above 60 mmHg Symptoms may progress quickly to a state of septic shock which has nearly a 50% rate of mortality. As the incidence of sepsis increases in hospital settings, early recognition and treatment are necessary to effectively combat the onset of lifethreatening symptoms. Although there has been much research on sepsis, it remains the most lethal affliction in the modern hospital [23].

Sepsis and the Role of Inter-α Inhibitor Protein
Sepsis has been difficult to diagnose because of the variety of sources that can be responsible for the onset of symptoms. This in turn makes it challenging to isolate and characterize a clear biomarker. Recent research has shown promise for such a biomarker with the protein called Inter-α-inhibitor-protein (IαIP), which has been correlated with many physiological and pathological activities including tumor invasion, metastasis, stabilization of extracellular matrix, inflammation, and wound healing [30] [31] [32]. IαIP falls within the category of plasma-associated serine protease inhibitors which are synthesized in the liver. These proteins can be been found in the synovial fluid within inflamed knee joints for example. Additionally, the levels of IαIP in blood serum have been shown to decrease by 20-90% in patients showing signs of sepsis. The levels of IαIP have proven to be inversely correlated with mortality. Studies have shown that the presence of neonatal sepsis results in a 62% decrease in the plasma levels of IαIP, which an average of 322 mg/L in non-septic patients decreases to a level of 121 mg/L in patients with neo-natal sepsis. [33].

Inter-α Inhibitor Protein and links to Mortality
In the past IαIP was thought only to inhibition of many proteases 1 , however a group headed by Dr. Yow-Pin Lim at Brown University has shown that it may be also involved with inflammatory response. His group conducted experiments to measure 1 Proteases are any enzyme that performs proteolysis or the breakdown of proteins into their composite components or amino acids, which are the building blocks for all proteins. the correlation between IαIP activity and LT-induced cytotoxicity 2 (lethal toxin induced cytotoxicity). Mouse peritoneal macrophages were induced with inflammatory ascites 3 . Results from this experiment are displayed below. Further studies have shown that adult patients with severe sepsis show 2 Cytotoxicity is a quality of a chemical or compound making it toxic to cells (ie. venom). 3 Ascites is the buildup of fluid in the space between lining of abdomen and abdominal organs. 22 decreased levels of IαIP and increased mortality rates. This was presented by Dr.
Yow-Pin Lim, who is a leading researcher in the field of IαIP.

Figure 1-11: Correlation between IαIP levels and mortality rate [35].
The reports from this research show that decreased levels of IαIP are correlated with higher chances for mortality. The data evaluated from 51 patients showed a range of 20-90% decrease in plasma IαIP levels compared to healthy patients. This phenomenon has also been supported by research from Balduyck et al [36]. Although IαIP may only be a precursor in the steps that induce inflammation and furthermore the onset of sepsis, it offers a method for treatment, and furthermore a target around which to develop an ELISA-based platform. More information about the inherent structure and role of IαIP is provided in the appendix.

Motivation and Objective Outline
The main aims of this thesis work are to develop a functional test that is capable of running a Lateral Flow Assay for the diagnostic application of sepsis. In 23 collaboration with Dr. Yow-Pin Lim's group a, competitive assay will be used to detect levels of IαIP. Further aims of this research are to create an autonomous device that may run multiple fluids.
For the purposes of providing a framework for this research, Chapter 1 has served to provide basic background knowledge for the technology and mechanisms at play within ELISA, LFA, and biological targets for septic diagnosis.
Chapter 2 elaborates on these basic principles introduced in chapter 1 and will also discuss the current state of paper based platforms through a literature review.

CHAPTER -REVIEW OF LITERATURE
Within this chapter the current state of research on paper-based technologies is reviewed along with their application to microfluidic devices and LFAs. To begin, the components that commonly make-up such devices are reviewed, specifically including research on preparation techniques, design advantages, and advances in fluid flow control. Following this, the applications to biological experiments including ELISA are further discussed with application to the biological target for sepsis.

Components of LFAs
The components and materials were briefly described in chapter 1 but a more detailed view of these materials will help the understanding Lateral Flow Test Strips (LFTS). Millipore is a major manufacturer of LFTS and they provide a comprehensive manual about the materials and mechanics of such devices. Figure 1-2 was provided in this manual, but its components are commonly used in many other test strips; another example is shown in the below figure.

Nitrocellulose
Again, the major components of a lateral flow test strip include a sample pad, a conjugate pad, a nitrocellulose membrane, and a wicking area. The most important functional unit of a LFD is the nitrocellulose membrane, as it allows for anchoring of capture proteins and serves as their detection area. The membrane plays an effect in reagent deposition, assay sensitivity, assay specificity, and consistency in test lines.
As mentioned nitrocellulose is the most common membrane and it uses electrostatic forces to anchor proteins. This interaction takes place between the strong dipoles of nitrate esters in the membrane and the strong dipoles of the peptide bonds in the proteins. Their chemical structure is provided through graphical images below. [38]  Depending on the interactions and molecules within a test strip, other materials may take advantage of alternate binding mechanisms.

Flow Rates in Nitrocellulose
As the membrane will be integrated into a LFTS, it is also important to understanding its flow properties of such a material. The speed at which fluids may flow through a material is referred to as 'capillary flow-rate'. Capillary flow is a 26 measured in units of cm/sec and is largely dependent on the material properties of the porous structure. Pores are the empty spaces between the fibers of the membrane through which fluids to pass. The pore size, pores size distribution, and porosity may change the flow characteristics. [38] Pore size is a measurement of the largest pore that may be found along direction of flow. Larger pore sizes will allow for higher flow rates, but smaller pore sizes may offer better filtration options. Membranes are often specified by their pore sizes and are on the μm scale [38] Pore distribution may be highly variable within the same nitrocellulose. This distribution of pore sizes varies from membrane to membrane and can play an effect in the capillary flow rate. [38] Porosity is the volume of air that is occupied by the three-dimensional space of the membrane. It is also referred to as the bed volume or the air volume; it is directly proportional to pore size and membrane thickness. [38]

Conjugate pad
Before coming into contact with the nitrocellulose membrane, a sample fluid must interact with proteins which are stored in a conjugate pad. As mentioned, this is an important component for storing the elements that would make up the top half of the assay sandwich. When sample flows to the conjugate pads, theoretically analytes will incubate with the stored detection reagents. Millipore describes the important characteristics for the conjugate pad:  Low non-specific binding-this property describes the ability of the material to release dried reagents. If detection antibodies are not allowed to flow to the nitrocellulose, no signal will be produced.
 Consistent flow characteristics-there must be a consistent deposit of reagent onto the membrane to produce a clear signal  Consistent bed volume-this refers to the amount of volume that may be contained in the material. Variable volumes may produce inconsistent signals.
 Low extractables-the material should not contain any particles that can interfere with the biology of a test or interfere with flow of the device.
 Good web handling characteristics-the materials must be easy to handle in order to be incorporated into the device.
Fiber glass filter paper is a common material used for the conjugate pad.
Other common materials used are described below:

Sample Pad and other materials
The materials used for the sample pad are cheapest and easiest to manufacture in the scheme of the LFTS. They can be made from glass fiber, cellulose fiber, or a filter paper.
The most important characteristic of a sample pad is to evenly and consistently distribute a sample fluid to the conjugate pad and subsequently to the detection zone. Sample pads may have other uses as well:  To increasing sample viscosity  To increase ability to solubilize detection reagents  To prevent non-specific binding of reagents to downstream materials  To modify sample to be more compatible with immune-complex [38] Additionally, woven meshes have been introduced to LFDs as they can more evenly distribute the sample volume. They also may be more expensive and difficult to process. [38] Just as with the conjugate pad, thickness, basis weight, tensile strength, extractable and of course dimensional variables can affect the flow and effectiveness of the device [38].
The absorbent area's main function is to create a gradient of force through capillarity to maintian flow of fluid.

Housing
A housing or casing is typically the last component included in a LFTS. It may 29 often be beneficial to the test, but are certainly not necessary. Housings or cases may offer better preservation capabilities for reagents, and can store fluids until they are needed for a test. The housing can also offer a platform on which to include labels or instructions to aid in the use of the device. Also instructions for interpreting test results may be provided. Lastly, the housing can provide a force to ensure better proper flow of fluids from one membrane to the next.

Fluid Flow Characteristics within LFD
Fluid flow and timing can be controlled though various mechanisms and design techniques. The fundamental theories that describe fluid flow through a medium along with methods to characterize LFD design and functions are explained below.

Fluid Flow Characteristic Equations
Edward Washburn pioneered the physics of capillary flow and followed in the footsteps of Richard Lucas. They each published papers that explain the physics of capillary flow. From a theoretical standpoint capillary flow can be explained by the Washburn Equation, which assumes a material to have parallel cylindrical tubes. [39] = √ 4

Equation 1: Washburn Equation
As water is the most commonly used fluid, its surface tension (γ) at room temperature is estimated to 0.0728 N/m and viscosity (η) is 1.002x10 -3 Ns/m 2 . This equation takes into account two assumptions: one, that the fluid traveling through a membrane originates from a non-limiting source, and two, that the material through which the fluid flows is dry.
Fully wetted flow, that means to say flow through a wet material, is described by Darcy's Law. Assumptions can be made for the viscosity of water at room temperature (1.002x10 -3 Ns/m 2 [41]), for the change in pressure ΔP which is constant with capillary flow (4,560 N/m 2 [42]), and for the permeability of paper (3x10 -13 m 2 [42]).
This leaves only the variable of the fluid channel geometry to be determined. This set of equations come in aid during the initial set up and design of a fluidic chip.

Controlling the Direction of Fluid Flow
The previous equations merely serve as a point of reference from which to build on and to develop the geometry to run a LFA. Different geometries may be often desirable to adequately control fluid flow within a LFD. There are many techniques which may be employed to fabricate paper-based devices to control fluid flow characteristics.
The foremost problem in this field has been to make paper hydrophobic to create the boundary to guide fluid flow. The fabrication methods for creating hydrophobic areas fall within two categories:  Selective hydrophobization -involves treating selective areas with a hydrophobic agent to prevent hydrophilic activity. Of course, areas that have not been treated by these agents may still channel fluids because they remain hydrophilic. This may be achieved through wax printing or ink jet printing.

PDMS
Using a negative, PDMS plastic may be allowed to cure to a mold that contains a necessary pattern for fluidic control.
• Cheap, easy to fabricate, devices are also flexible. ○ Barriers are not sharp, and cannot be put into high throughput fabrication.

Inkjet Printing
Patterns may be developed on the computer and simply printed on a desired paper platform.
• Cheap and simple to make patterns, good for prototyping ○ Need for a printer or other means to print channels, not suitable for mass production

Plasma Etching
By using plasma, hydrophilic areas may be created..
• Patterning is very cheap ○ Multiple masks and steps are needed to create hydrophobic area

Wax Printing
Using a printer wax may be used in a similar manner to ink in order to create the fluidic channel framework. Wax must be melted through depth of paper.
• May be able to create many devices in short amount of time. ○ Extra heating steps are need, and modified printer needed.

Wax Screen Printing
Using a screen wax may be selectively transferred to paper to develop channels • Simple fabrication process. ○ Low achieved resolution, and new screens must be made for each new pattern.

Wax Dipping
A metal part may be clapped to paper to block application of wax while paper is dipped into a wax solution.
• Fast fabrication and sharp channel resolution ○ A new metal part must be made for each new channel design.
Although the above are the most popular methods for fabrication cutting of paper to form channels is often used as well. A high level of precision, accuracy, and resolution may be achieved with the application of a laser cutter. Paper may then be combined or layered with other materials to increase stiffness and support for the fragile material. This has proven to be the simplest method for fabrication [45].
Additionally, wax printing has been more widely used since printing equipment is more affordable, easy-to-use, and is favorable for creating variations in fluid channel geometry. After printing wax onto paper and subsequent melting to create hydrophobicity through the entire thickness of the paper, edge resolution may be sacrificed. For prototyping and research, this method offers great advantages in the realm of PBD development as design adjustments for experimentation can be made quickly and easily.
These two methods can be combined to create more complicated patterns to expand capabilities of PBD by construction within a three dimensional framework.
This will be discussed in a following section.

Unlimited and Constricted Flow Characteristics
As described by the Washburn and Darcy equations, theory can predict flow characteristics, however, there are many variables (i.e. variation in the fibrous 34 structure of the paper, temperature and humidity) that may change how fluids flow in paper. Another prominent variable is fluid volume supplied to a channel. For example, the case of a limited or constricted fluid source may be modeled by a thinner channel geometry feeding into a thinner channel. An example has been provided below.

Figure 2-3: Fluidic Channels that Exhibit a Contraction along the Direction of Flow [42]
Channels that exhibit a limited source or an expansion show a slower fluid flow after the expansion, in comparison to a channel that has been designed with a constant width. The limiting source provides some resistance against the flow and prevents it from reaching a further distance.
On the contrary, when channels exhibit a contraction, quite a different phenomenon is observed. In the following figure, channels have been built to exhibit 35 a contraction along various points along the length of the channel.

Figure 2-4: Fluidic Channels that Exhibit an Expansion along the Direction of Flow [42]
The contraction in flow seems to have no effect on the distance that the fluid may travel along the channel. By a non-limiting source, the fluid may travel an equal distance or a slightly further distance along the direction of flow. As seen in the data, there is not a significant difference between the variations, and therefore it can be assumed that the contraction does not limit the fluid flow characteristics.
Furthermore, this supports the flow characteristics of the Washburn equation, as the relative slope of flow/time 1/2 in the thinner channel agrees with the flow in the thicker channel [42].
The resistance created by the case of expansion offers a potential to better control the direction and timing of flow. This for example may be used to create a one-way resistance, can be especially useful in a 3-D fabrication.

Designing Fluidic Circuits by Electrical Analogy
As LFDs on a paper platform have incorporated more complex designs, analogies have been made to electric circuit design. This has been a useful method with which to plan the functions and fluidic timings needed within a microfluidic device.
In the case of the expansion and contracting channels, these have been equated to resistors within electrical diagrams. Resistance to flow resulting from changes in channel width can be modeled using the above diagram and furthermore with Darcy's law along with Ohms law.
When resistances are taken in parallel, the resistances may be summed by the

New Components of Lateral Flow Devices
In addition to the geometric design of a fluidic circuit, advances have additionally be made in the realm physical construction, assembly techniques, and reagent pretreatment to offer more complex design solutions for microfluidic devices. In fact, moving construction of PBDs into a 3D format allows for more complex fluidic control.

3 Dimensional Fabrication of PBD
As PBDs have become a more researched topic, the complexity in construction and use has been developed. One of the first developments transitioned such devices from a 2 dimensional to a 3 dimensional assembly. A group at the University of Texas at Austin [46], published a method that utilized the principals of origami to translate fluids to different areas through a multilayered device. In a 2-D format, fluid flow only travels in one plane and is directed by the patterning of wax walls. In 3-D format, these 2-D patterns are layered upon each other to allow for the capillary flow to continue across multiple layers. Other research groups have adopted similar methods of fabrication, but have expanded this to suit the various needs of PBDs.

The Mechanical Switch
A group at Harvard University [47] has presented a method by which a PBD may be constructed with multiple layers to generate more complex flow patterns.
They present a button that prevents for a continual fluid flow until pressed. This "mechanical switch" or button takes advantage of a disconnection in paper channels to prevents fluid flow. This disconnect may be created by a wax barrier between two channels or also by an air gap. In the case of an air gap, the fluid will not flow until two paper portions have been forced together. This works very nicely in a 3D format as shown by the image below.

Figure 2-8: Construction and Cross-Sectional View of a PB Mechanical Switch [47]
In the above diagram, an air gap prevents fluid flow until a mechanical force connects the overlapping channels. When the area is pressed, the two channels become connected, thus allowing for fluid flow to continue.
Initial experiments showed that fluid would sometimes bypass the air gap, so two air gaps were incorporated into the design for more reliable function. An important component in the construction is the incorporation of a doublesided tape which is needed to adhere the two layers of patterned-wax. This layer of double-sided tape has gaps cut out in order to provide an area through which fluid may travel to the subordinating layer. Another proposed method to adhere each layer is the use of spray adhesive. Other devices take advantage of hydrophilic solutions or materials in order to prevent fluid flow through a particular area.

A One-Way Fluidic Valve
A novel technology for PBDs was developed by Hong Chen at the University of Rhode Island. His research serves as the basis for what is called a "one-way valve" [48]. This valve was developed on a paper platform which has been treated with hydrophobizing agent. Under most circumstances this treatment will not break down until a surfactant agent is introduced to the area. When a solution comes to a hydrophobic rendered barrier, it will not be able to pass that barrier. On the other hand, if the solution has the surfactant dissolved within, it may break down the hydrophobicity and continue to flow further along the microfluidic chip. The figure below helps to illustrate this. This concept may be incorporated into a two way circuit and in this layout follows diode notation borrowed from electrical engineering.
In the diagram above, paper is soaked with a compound called   Although paper disks were used in this layout, there are many other mediums which may be used to incorporate the valve into a device. For example, cellulose powder can be used as a paste to be spread into the cutouts of the double sided tape.
Aside from the steps to hydrophobize paper, other solutions may be needed to assure a proper outcome when biological components are incorporated into a paper based device.

Fabrication with Adhesive Spray
Another method for fabrication has been presented by a research group at Pennsylvania State University, who used a spray adhesive to assemble the multiple layers of each 3-D chip. By spraying the individual layers with 3M's Super 77™ Multipurpose Adhesive spray a higher throughput for the assembly was achieved [49]. The spray adhesive can replace the function of the double-sided tape which requires more time for assembly, needs to be precisely aligned, and requires a filling (paper or paste) to connect the functional paper layers of the device. This method for fabrication shows promise to save time and resources, especially in the prototyping and testing stages where many chips need to be fabricated and tested while certain parameters are optimized. The group reports that for every two sheets of paper that need to be adhered only one needs to be sprayed to leave 0.9mg of spray per cm 2 . To achieve this paper was sprayed for approximately 1 second at a distance of 24cm [49]. With this method, a 50% reduction in the time for fluid to travel through multiple layers was achieved in comparison to using double-sided tape and paste [49].
In a fabrication that included double-sided tape, there are many inconsistencies with alignment and more so with the connection between individual layers of the device. By switching to a spray adhesive many of these inconsistencies could be eliminated resulting in more consistent and predictable functionality. This would be especially significant when incorporating other components into the chip, such as one-way valves and biological assays.

Adding Biology to a PBD
As biology is integrated into a paper based platform, there are a few key preparation steps that need to be considered in order to produce a chip that results with the best possible result. In the following section methods to improve this resulting signal are discussed.

Surface Treatment
When comparing traditional ELISA run in a 96 well plate to ELISA on a paper based platform clear differences and advantages are presented. As mentioned previously, tests generally completed on paper need far less time and smaller fluid volumes for the necessary reagents. Furthermore, color enhancers or substrates may be used to create a signal that can be interpreted by the naked eye. This in turn cuts out the need for analysis equipment as needed by a traditional ELISA test. One of the major drawbacks of running such tests in paper is the lower Limit of Detection (LOD).
In fact, the LOD for ELISA run in a 96-well plate is about three times lower [42].
Research groups have developed a few methods with which to increase the value for experiments run on paper.
In the same way that 96 well plates can be treated with a compound to promote binding of proteins or enzymes to the surface, a compound that blocks further binding to that surface may be used as well. This step is referred to as a blocking step, it usually follows the anchoring of capture antibodies. Bovine Serum 47 Albumin (BSA) is a popular method for blocking areas in the 96 well plates which have not been occupied by capture proteins. BSA may also be used on paper, specifically on the nitrocellulose detection area in order to maximize the detection signal. Other methods include blocking of paper with amino (NH3) and carboxyl (COOH) altered silicon dioxide (SiO 2 ), have shown to increase the signal intensity on paper by a factor of 7 for specific concentrations of detection protein [50].
Washing steps are another method that improve signal by removing unbound proteins. This is commonly used for ELISA run in 96 well plates to remove the free labeled-proteins. On a paper platform, it is not possible to perform multiple washing steps because fluid cannot be removed from the testing setup. It is however possible to incorporate one wash step to clean up background noise on such a device so that more contrasting signal may result. A group at the University of Washington incorporated a wash of TRIS-buffered saline with Tween 20 in order to remove any unbound proteins [51]. Once a signal is obtained signal amplifiers can be used to make detection more readable for devices and for the naked eye.

Signal Amplification
The most preferred method for detection within a paper-based device is through visual confirmation; therefore a signal must be visible to the naked eye and must not rely on the computing power of a reader. The same group from Washington  By adding the gold nanoparticle to the label-antibody conjugate, a stronger more intense signal is generated. Below the illustration of this process, there are 5 test lines that represent various concentrations of the analyte. In this case, the gold nanoparticle increases the size of the detection particle and changes the color of detection from pink to purple. The limit of detection is shown to go below the range of 1 mIU/mL (million international units per milliliter). A following experiment showed, furthermore, how the signal intensity could be increased using the amplification particles along with a wash step. The results from this experiment at displayed in the chart in the right side of Figure 2-14. This signal was increased by 4fold when gold nanoparticles were used for detection.

Combining Technologies and Building off Previous Work
As research is conducted in the development of paper-based devices for the application to diagnostics, the techniques and technologies described above may be Point-of-Care Diagnostics [53].'

CHAPTER -METHODOLOGY
This chapter provides the methodology for LFA fabrication including a description of materials used, fabrication processes, necessary tools, and preparation of biologicals needed for a Competitive ELISA. Steps for optimizing and general protocols are also discussed.

Chip Fabrication Introduction
Designing a PBD for the application of an Assay requires 2 or more fluids.
When designing and fabricating a multi-fluidic device (one with more than 2 fluids) appropriate timing and controlled release of these fluids is key. All the PBDs designed under Professor Faghri at the University of Rhode Island utilize a filter paper printed with wax. This is done in accordance to the work by Hong Cheng, Wilke Föllscher and Roman Gerbers [53].
The design presented in the following pages was developed in conjunction with Hanno Teiwes [54] that takes design queues from the previous work by Gerbers However, instead of using surfactant paste and hydrophobic discs, a complete layer was dedicated for hydrophobic treatment, as it offered more control and a faster time for assembly. Below is an image of a previous chip design to serve as reference; 51 this image will be again called upon to describe ideal fluid flow sequence.

Chip Fabrication
Materials and software needed to fabricate the above device include: After a design is printed it must be cut using a CO 2 laser cutter (Epilog® Mini 24) to remove all the excess paper and to cut out the individual components or layers of the device. On an 8x10 piece of paper, it is possible to print many layers for incorporation into the final device, but depending on the dimensions of the design multiple sheets may be needed. Typically, 6 layers are printed on one sheet of 8x10 in Filter paper. In order to maintain precision when assembling the chips, the laser cutter is used to cut out the individual layers.

Wax Paper Melting Process
Since wax is only printed on one side of the paper it must be heated to melt the wax through to the other side of the sheet. This is done by placing the components wax-side-up in a vacuum oven for approximately 25 seconds at a temperature of 130°C-140°C. In all constructions the Isotemp® Model 280A vacuum oven from Fisher Scientific is used. This process allows the wax to carry through to the opposite side of the filter paper. See the below images for the melting process.

Figure 3-2: Process for Melting a Wax-Printed Layer
When, designing the chips it is important to note that wax will travel about 0.5mm toward the negative spaces. For example, if a 4mm circle is printed it will shrink down to a 3mm circle after melting. Also the longer a layer sits in the oven the further the wax will travel into the negative space. See below images for an example.

Figure 3-3: Effects of Melting on Wax-Paper Features
Also, it is important to note that the melting times are dependent on the size of the features, temperature of the oven, and placement within the oven. These conditions were the optimal standard, but there was some variation.
In addition to the filter paper, a blotting paper (Whatman Grade GB003 20x20cm by 0.8mm thick) and absorbent paper (Sterlitech® GA-55) are included to maintain connection between layers and create more consistent flow to waste areas.

Double Sided Tape
In order to adhere each layer of the device, double sided tape is used as previously mentioned. According to previous protocols, Ace Hardware's® doublesided Carpet tape (No. 50106) is the best option as it is inexpensive ($3.99 for a 1.5inx42ft roll). Double sided tape is by applying segments to Reynolds® parchment paper in order to cut and to handle more easily during chip assembly.
Placing double sided tape onto the Reynolds parchment paper Double sided tape and parchment paper. Two rows are double-sided tape are needed per layer.
Cutting of double sided tape in laser cutter.
Removing of unnessesary parts from double the cutouts.

Figure 3-4: Process for Preparing Double-Sided Tape
After the double sided tape has been cut out and cleaned up, it may be used to assemble the device.  The 1 st and 2 nd Tape layers also have circular cut outs in which to fill hydrophilic material; again this can be in the form of a cellulose based paste or filter paper cut outs.

Laser cutter settings
As mentioned earlier Corel Draw is used for all design and cutting with the CO 2 laser cutter. Depending on the material being cut, different power and speed settings must be used. Specifications for each type of material are listed below. Another important tool to aid in the cutting process is an alignment guide.
When printing designs, they may not sit in perfect alignment to the edges of the paper. For example, when designs are printed the edges may not sit parallel to the edges of the paper. So in effect, the laser cutter is aligned to the design rather than to the edge of the paper on which the design is printed  In the second image above, a preliminary cut is made. From the first cut, the whole 58 cute must move 1mm in the x direction and 0.5mm in the y direction. The top and left dashes in the alignment guide start at 0 then go to 1mm, 2mm, 3mm, 5mm; while the right and bottom dashes are marked starting at 0.5mm and go to 1.5mm, 2.5mm, 3.5mm, 4.5mm, 5.5mm.

Chip Assembly
Once all the materials needed have been prepared, chip assembly can begin.
The image below shows all the components needed for chip assembly. The above materials include:

Hydrophobic treatment & Paste Application
Within this assembly the most important functional unit is the hydrophobic layer, which holds fluids in reserve until the hydrophobic areas are bridged by surfactant (this chemical process is described in 2.3.3). This particular design incorporates a whole layer of wax-printed paper treated with hydrophobic solution.
Therefore, the surfactant needed to breakdown the hydrophobic barriers must be 61 placed in the layer below. This is done in the form of a cellulose paste which is mixed with a solution of surfactant. Surfactant paste is spread into the negative spaces of the 2 nd tape layer. To create surfactant paste a surfactant solution is first prepared by adding a 35 wt.% of Tween 20 in Ethanol. This solution is next diluted with ultra-pure water at the ratio of 0.25mL/mL. Once that is completed, the water-tween 20-Ethonal solution can be added to cellulose powder at the ration of 0.5g/mL (cellulose/solution). This mixture 62 is next homogenized using a spinner and can then be applied to the cutouts in the 2 nd Tape following its adherence to the 3 rd wax layer.
Hydrophilic paste is a mixture of the same ratio of cellulose to ultrapure water (0.5g/mL). The pastes may then be applied with a metal spatula or with a popsicle stick wrapped in paraffin film. The paraffin film helps to prevent absorption of surfactant solution into the wood. The paste then needs approximately 10 minutes to dry. It should also be noted that surfactant past should not be applied directly to the hydrophobic layer as it will immediately render the hydrophobic areas ineffective.
Rather, it can be applied to the 3 rd layer and allowed to dry.
Another option is to use hydrophobic, hydrophilic, and surfactant treated discs of filter paper which can be applied directly into the cut outs of the tape layers.
Each can be cut to the exact size of the cut outs in the tape layers.
As previously stated, the chips are assembled in batches of six. This is done to maximize the efficiency during assembly and furthermore, to have tests for the optimization of surfactant and ultimately of biological reagents.

Figure 3-11: Fluid Flow Characteristics by Roman Gerbers and Wilke Föllscher [53]
In the designs from Gerbers and Föllscher, three fluids are applied to the chip for sequential loading. These include a sample fluid containing the analyte for detection, a wash-fluid which is used to clean up the signal on the nitrocellulose by removing any unbound proteins, and a substrate which creates the final color change and indication for the presence of an analyte. The wash fluid is stored in the first reservoir and will not interact with the chip until the hydrophobic discs below are inactivated through fluid interaction in the trigger channel. The substrate is stored in the second reservoir and again will not be triggered until the hydrophobic areas 64 below are inactivated by the fluid flow in the trigger channel.

Detection Antibody:
Anti-Rabbit-IgG (AP) Capture Antibody: Anti-Rabbit-IgG (unlabeled) Analyte: Rabbit-IgG (1st) Wash-Fluid: SuperBlock® blocking buffer in TBS (2nd) Substrate: BCIP/NBT (3rd) This process is for an ELISA, however a chip to develop a competitive ELISA is needed. When designing and testing such devices, colored water are used for initial tests. After sequential flow is ensured, biological reagents may be used.

Protocol Development Chip
One of the first objectives of this project is to develop a chip on which to optimize various parameters. For example, during the initial stages of this research there were many boundaries encountered specifically with assembly, alignment, and surfactant & hydrophobic application. A less complex platform is designed to optimize many of the biological parameters independently from those of the fluidic valve. Also as there is always an interest to improve the assembly times for each chip, a platform is needed with which the assembly technique using spray adhesive could be studied.
In summary, a design was needed with which:  Surfactant and Hydrophobic treatments could be further optimized, ideally in a 1-valve set up  Spray adhesive assembly techniques could be investigated  Biologicals could be tested independently from valve optimization

Design
Given the above goals and constraints, a 1-valve design was developed.
Images of all components of an 8-batch chip along with each layer are provided below and dimensions are displayed in the appendix.    The assembly time and number of materials needed are the downfalls this assembly method. For example, when pastes are applied they must dry for at least 10 minutes. Additionally, during the assembly the double sided tape can be cumbersome to handle. For these reasons an interest developed to experiment with spray adhesive, introduced by G. Lewis et. al [49].

Spray Adhesive Method (5 layers)
In a spray adhesive assembly there are a few necessary changes. Since the double-sided tape layers are removed, a new wax layer is substituted on the bottom.
As shown below, the 3 rd layer is integrated and includes a smaller area for surfactant (the white circle on the left). Secondly, a complete layer of wax is needed to prevent fluid from flowing from the bottom layer or from the timing layer directly into the absorption pad. This was previously achieved by the 3 rd layer of double-sided tape.
Additionally, the geometries were slightly expanded in order to maximize the flow rate and decrease the resistance to flow from the fluid reservoir.  In the illustration above, the hydrophobic area is slightly cut in a semicircular manner and is then melted. This is done to ensure that hydrophobic solution does not spread into the surrounding wax area during application. Also, the surfactant area may be treated directly or it can be cut out and replaced by a surfactant treated disk (this option is shown in the illustration). The discs are treated with surfactant solution in a petri-dish and allowed to dry at room temperature.

71
These two steps ensure that surfactant solution does not travel into the surrounding wax area. If this occurred, a less than intended volume would be left within the designated surfactant or hydrophobic area.
Once all materials are prepared, the chip may then be assembled using 3M's super 77 multipurpose adhesive. The adhesive is applied by spraying each side of the layers once for 0.25 second at a distance of approximately 24cm.

Spray Adhesive Method (4 layers)
The possibility of removing yet one more layer is investigated. Instead of dedicating a whole layer to surfactant dots, the 3rd layer (from Figure 3-17) may be removed leaving the 4th layer (the connection layer) to be treated with surfactant directly.

Figure 3-19: 4 Layer Assembly with Spray Adhesive
The cut is also removed from the hydrophobic area in order to prevent premature flow of fluid from the reservoir through the gaps. The hydrophobic area and surfactant area are slightly enlarged to be able to contain the same volume of reagent as previous tests. The same process follows for applying the spray adhesive.
An image of all necessary materials is shown below. Although it is not shown in the image below, the 1 st layer also has cutouts, which would serve as the spaces under

Surfactant Application
which the nitrocellulose pads would fit.

Figure 3-20: Materials Needed for Spray Adhesive Assembly (4 layers)
In this assembly the number of needed materials decreases. As discussed in 4.3, the assembly times also dramatically decrease.

Re-optimization of Hydrophobic and Surfactant Application
With new assembly methods, the hydrophobic and surfactant treatments need to be optimized to function properly, consistently, and predictably. To optimize the volumes and treatments the setup in Figure 3 volumes to the 1mm circular cutouts and are allowed to dry for at least 12 hours at room temperature.

Hydrophobic Optimization (5 Layers)
Once the optimal wt.% is determined, the volumes of hydrophobic solution are varied. This is still done in accordance with previous procedures.

Multi-fluidic Sequential Flow
In order to fully justify the use of a spray adhesive assembly, a chip capable of sequentially flowing three fluids is needed. A chip is designed under similar parameters as the protocol development chip. This is done to run tests with the same hydrophobic and surfactant solution volumes and treatments.

Design of a Multi-Fluidic Chip
Using the methodology of a 4-Layer chip, the following design is developed (dimensions may be found in the appendix). The first layer incorporates an 'H' shaped cutout under which the nitrocellulose pad could fit. On the second layer 6 columns of red dots are added. These serve as a guide with which to precisely place nitrocellulose in for proper alignment with the layer above. Many dots are incorporated in order to handle a range of nitrocellulose geometries for 5-3mm widths and 7-4mm heights.

Controlling Backflow in the Second Layer
After tests are conducted the chips may be disassembled to investigate the flow characteristics and determine sources of error. Previously, it was discovered that many of the secondary fluids exhibited a backward flow within the 3 rd wax layer (timing/surfactant layer). Backward flow can be a problem especially when using biologicals. The fluid that returns backwards along the source path is lost and will not 77 be effective in the detection scheme. For example, if some volume of substrate is lost due to backward flow the resulting signal on the nitrocellulose may not be as distict.  [42]. An expansion, however, will slow the progressive flow of a fluid along a channel [42]. In a contraction, the wider channel serves as an unlimited source of fluid, however, during an expansion, the amount of fluid that can continue down a channel is limited by the volume held in the thinner channel [42].
This can be seen in section 2.2.3.
Using this characteristic, 'kinks' are incorporated into the secondary layer for two reasons. The foremost reason is to test the geometry's ability to restrict backward flow and secondly to test if the valve opening times are delayed. Designs for the 3 rd wax layer are shown below and dimensions have been provided in the appendix.

Enzymatic Activity Test
The protocol development chip is also designed as a platform on which to test the activity of biological interactions. Since IαIP is detected through enzymatic activity, a model experiment is developed to prove the feasibility of such as test. To show that the concept is possible, an enzymatic activity test is developed for the detection of Alkaline-Phosphatase (ALP). Alkaline Phosphatase is an enzyme that is responsible for removing phosphate groups from any type of molecule. These molecules can be nucleotides, proteins, or alkaloids. As its name suggests, ALP is effective in an alkaline environment [56]. This was the same substrate used by Roman Gerbers and Wilke Föllscher.

Alkaline Phosphatase Activity Test
Abcam® offers mouse monoclonal and polyclonal antibodies labeled with ALP After the fluids have been completely depleted and the nitrocellulose dries, the chips are scanned at a resolution of 1200 DPI on a HP Photosmart 6510 scanner and the signal is measured using the ImageJ software to calculate the intensity of each spot on the nitrocellulose.

Nitrocellulose Shape Variation
The previous test is repeated with variations in the shape of nitrocellulose.
After running the previous test, it was observed that there was a large amount of area on the nitrocellulose which was not covered with the label antibody; therefore no signal is produced on these areas. Also, these areas offer a path for the substrate to flow through without the chance to interact with ALP. In order to maximize the interaction between ALP and BCIP/NBT, the width of the nitrocellulose is varied along the middle from widths of 5.0, 4.0, 3.0, and 2.0mm. After running the test in the same manner, the chip is scanned and the signals are then processed with ImageJ. from the development of this platform may be seen within his thesis [54].

Reducing Background Signal
One of the problems encountered in the coinciding work of Hanno Teiwes was the strong background signal generated during the development of 'Sandwich' ELISA. Within both the 'Sandwich' ELISA and 'Competitive ELISA the signal is generated by the addition of the DAB substrate. However, a signal develops when DAB binds to HRP-Strep which binds to Biotin (the label that is attached to the detection antibody for 'Sandwich' ELISA and also to IαIP for Competitive ELISA).
This reaction scheme can be particularly problematic if there is extraneous HRP-Strep floating around in the background on the nitrocellulose. When DAB is introduced to the nitrocellulose, not only will it create a signal on the HRP bound to Biotin, but it will also create a signal on the free HRP that has not been washed off the Nitrocellulose. In an effort to clean up this background signal, wash steps are introduced to remove some of the free HRP.
To test this, the detection antibody, Biotinilated-R21, is spotted onto 1.

Competitive ELISA
When working toward a paper-based device that can capture and detect IαIP the variables must be isolated to be optimized. The first step is to determine the correct amount of capture antibody to use in order to capture enough protein to generate a signal. Second, an effective concentration for the labeled competitor is added to the chips at a volume of 60μL, followed by a 30μL PBS after the sample has been pulled through the device. Then the HRP in 1wt.% Milk-PBST is added at a volume of 30μL, followed by 30μL PBS after the HRP has been pulled through the device. At this point DAB substrate is added at a volume of 30μL to create the color reaction. The chips are scanned at 1200 DPI and the images are processed using the ImageJ software.

Optimization of B-IαIP Concentration
After determining the appropriate volume for the capture antibody, the next step is to determine the optimal concentration for the competitor protein, Biotinylated IαIP (B-IαIP).
The simplified chip was again assembled under the same protocol as in the previous experiment. The volume for the capture antibody is increased to a total of 2μL and is applied in 5μL increments to the 1.

87
The idea behind this experiment is to determine the corresponding concentration for the saturation point of the capture antibodies. When looking at the signal to concentration curve, this value lays just before the plateau. The concentration correlated with this value would offer an adequate competition within the competitive ELISA scheme.

Development of Standardized Curve
It was determined that a concentration of B-IαIP between 1:100 and 1:400 is sufficient for the competition of un-labeled IαIP. In an effort to develop a standardized curve for a competitive ELISA on paper, the test chips were assembled, Ideally, the test would include 3 or fewer fluids. To achieve a lower level of complexity, the HRP-Strep is added to sample fluid to incubate with B-IαIP. This is done for the 1:400 B-IαIP in 1wt.% Milk-PBST at the same ratio as previous experiments. For the 60μL sample, 30μL of HRP are added and mixed. This is then added to a prepared chip to determine if an adequate signal may be generated.

Complete Competitive Test
With the generation of a standardized curve and reduction of fluids, the next step is to develop the Competitive ELISA on a fully autonomous device. The 3-Fluid design may be used to develop a signal curve.

Protocol Development Chip (Double-Sided Tape)
Using the protocol described by Gerbers and Föllscher and the construction described in section 3.5.2 the following result was obtained.

Optimization of Surfactant
Three

Optimization of Hydrophobic Solution (5 Layers)
Following this test, optimization procedures were developed for the hydrophobic areas. According to the methodology of Gerbers and Föllscher [53], hydrophobic areas were treated with 8µl. With the optimal concentration of surfactant solution (5-10 wt.%), the amounts of hydrophobic solution were decreased in the range of 8, 6, 4, and 2µl. Three batches were assembled for each concentration for a total of 12 batches each with 8 individual chips or tests. Each 94 concentration of hydrophobic volume was also combined with 10 and 5 wt.% of surfactant to measure the full efficacy of each. During the procedure, the red fluid was added first, and allowed to sit for at least 5 minutes before adding the main sample. This was done to ensure that the hydrophobic valves held. The following chart shows the average opening times for each volume.

Figure 4-4: Valve Re-optimization for Surfactant and Hydrophobic Treatment in an Assembly with Spray Adhesive (5 Layers)
The data from this experiment is shown in the appendix. As can be seen in the chart above, there is no significant difference between 10 wt.% and a 5 wt.% surfactant solution on the opening times of the valves. This is especially apparent for

Optimization of Hydrophobic Solution (4 Layers)
Under the same methodology as the 5 Layer assembly, Hydrophobic treatments were tested in the range for 8, 6, 4, and 2µL. Surfactant solution was added directly to the third layer in 2 and 1 µL volumes at a 5 wt.% solution. Within this experiment no particular fluid is optimal, however the results may serve as a reference for designing future circuits which may be require more complex geometries.

Comparison of Time to Assemble Adhesive vs. Double-Sided Tape
When preparing all materials for chip assembly, the major components for each assembly method remain the same. However, the assembly times for each chip differ. The following charts describe the differences in assembly along with a breakdown of the time needed for each step in the assembly process.    As can be seen above, there are a few steps which are shared by both assembly methods. However, the overall number of steps needed with the double 99 sided tape method is 50% greater than the assembly methods of spray adhesive.
Also, the time needed is 150% more than with the spray adhesive. Again this is largely due to the number of materials, steps in assembly, and the drying times necessary for the double sided tape method.
Additionally, the double-sided tape tends to be cumbersome to handle and needs more time compared to simply spraying the components with adhesive.
Previously the double-sided tape restricted the dimensions of the overall chip. This was partly because the tape is 36mm wide. This would end up defining the overall length or width of a chip. If chips were designed any longer or wider than this width, then there would be the potential for fluid to travel along the seams of parallel tape strips which lay between the various layers of wax. With the spray adhesive, there are a greater range of dimensions and geometries that may be employed which do not have to conform to the size of the double sided tape.
The double sided tape cannot be overlapped, as it has a backing that needs to be removed prior to application. If the tape strips are overlapped, the removal of this backing will be obstructed.
Although multiple spraying steps are described in the chart above, these can be further reduced by spraying all the layers at once. For example, the images below provide some intuition for the spraying of multiple layers and their subsequent assembly. Chip) one can see distinct similarities. One major difference is the size at which this chip was developed. Traditionally, a batch of six chips was designed to be 69mm in height. This was primarily done due to the width of the double-sided tape. Since the new assembly is not restricted by this width, a larger design was developed to be 80mm in height for a batch of 6 chips. However, if these chips were to move into a scaled up production, appropriate sizes of double-sided tape could be made in order 101 that they don't become the limiting factor of the design. Still, within a mass fabrication setting double sided tape may pose problems in assembly.

Confirmation of Sequential Flow
The following images show a sequence of flow along with the times that were recorded when fluids arrived into the 1 st wax channel. Two separate chips are shown with sequential flow of a three fluids chip. µL. If this was increased to 4µL, the failure rate of the valves could decrease.

Controlling Backward flow
In an effort to reduce the amount of backward flow within the 3 rd wax layer the narrower sections or 'kinks' were added to the channels to restrict flow. Seen below, are the effects of adding these narrowed sections.  By using this design element, a higher level of control over the fluids is achieved. This element can be used effectively to control backward flow and additionally can be used as a delay for secondary and tertiary fluids. This design element may be combined with the ability to add different volumes of hydrophobic treatments to achieve even a higher level of control with in a multi-fluidic device.

Alkaline Phosphatase Activity
As a model for the eventual enzymatic test of IαIP, the following results were obtained from the ALP enzymatic activity test. When processing the dot intensities with Imagej, a value between 0 and 255 is obtained. This is the RGB scale with white registering at 255 and black at 0. With these in mind, the chip above would produce a downward sloping response curve. In order to produce and upward sloping curve, the data is corrected by subtracting from 255. This results in a white value of 0 and a black value at 255. Under these calculations the following chart was obtained. The activity response of ALP is plotted on a log-log scale in accordance with Macarron and Herztberg [58]. As can be seen in the above chart, the signal is a maximum with the concentrations of 0.5mg/ml and 0.1mg/ml. The lowest response is with 0.1e-6 and 0.1e-5 mg/ml with values at approximately 10 on the RGB scale.
The above chart follows log( ) = * log ( ) with k being the slope. For the points between 10 -7 and 10 -6 k=1, and for the sloping area between 10 -6 and 1 k is approximately 0.24 or ~1 4 ⁄ . This indicates that there is route of 4 in the sloped region, suggesting some level of signal saturation at higher concentrations of ALP.
The values obtained further show a low level of variance, as can be seen by the standard deviation bars. The signal to noise ratio is calculated by dividing the signal obtained from the purple spot by the signal from the clearest point on the nitrocellulose. The signal on the nitrocellulose was around 245 on the RGB scale and when it was inverted it came out to about 10. Again, the data is presented on a log-log scale in accordance with [58].
As mentioned previously, ALP and BCIP/NBT are one of the most widely used enzyme-substrate reactions within ELISA. One of the major reasons for this is that are stable and produce a high signal to noise of nearly 20, as seen by the above chart.

Nitrocellulose Varying Widths
When moving to a narrower width for the functionalized area of the  The following data shows the signal to noise ratio observed after a signal was returned from BCIP/NBT-ALP activity. When comparing this data to the previously obtained data there is a slight decrease in the Signal/Noise ratio. This value will change depending on if the nitrocellulose has dried completely or not. After examining the assembly, one may notice that nitrocellulose sits directly upon a black wax area from the 2 nd wax layer. When the nitrocellulose is wetted from a sample it becomes slightly transparent, therefore the intensity of the white color is not as strong. If the nitrocellulose is allowed to dry for 30-60mins prior to scanning and processing, the signal to noise ratio will increase as the contrast between the purple signal and the white background increases. Although the signal created by the substrate is not altered to a significant degree, there may be cases in which the narrowing of nitrocellulose plays an effect.
For example during the step when detection complexes flow over the nitrocellulose, more of these complexes may be captured within a thinner nitrocellulose which would have capture antibodies on 80-95% of the width.
The laser cutter provides some restrictions in this method. When cutting a nitrocellulose sheet, high speed and low power settings must be used so that the fibers are not damaged. With these settings, intricate cuts do not come out as cleanly. For example diagonal lines end up being perforations and require much time to separate from a nitrocellulose sheet after cutting.
In summary, this test reinforces the fact that it is possible to run biological experiments on a chip that has been assembled with spray adhesive. The favorable assembly times and ability to run biology within this platform present a convincing case for spray adhesive as a method for assembly.

Reducing Background Signal
After performing the various washes of HRP-Strep, the following signal/noise ratios were calculated with ImageJ. It can be seen from the above chart that as more washes are used during the addition of HRP and DAB, a higher signal/noise ratio is achieved. When adding a wash step at the end, after both HRP and DAB are added, the signal to noise ratio is doubled. A higher ratio may not have developed here due to the staining of the nitrocellulose that occurs when DAB is added to the chip. In preliminary tests, there did not seem to be a significant removal of the background noise when only a wash at the end was added. However, when adding a wash step between the addition of HRP and DAB, the signal/noise is increased 4&5 fold. There is not a significant difference between the '1 wash (middle)' and '2 washes', so it seems that only a single wash between HRP and DAB is enough to significantly reduce the background.
The results from the 0 wash (1.76) are slightly different from results obtained from Hanno Teiwes when he obtained a maximum of 1.2 for the signal/noise. This indicates that there may be other techniques to decrease the value of the background noise. After discussing this with Dr. Yow-Pin Lim, a suggestion was made that there may be excess biotin within the stock solution of the R21 detection antibody. As this excess biotin lingers on the nitrocellulose, it develops a signal as the substrate is added, resulting in a higher background signal.

Competitive ELISA Development
The following section focuses on the conduction of experiments working toward a functional and autonomous lateral flow device for the detection of IαIP through competitive ELISA. As mentioned in the methodology, many of the 111 optimization steps were conducted by pipetting fluids onto the main channel and allowing them to flow toward and through the nitrocellulose. An autonomous test was used only toward the end of the research.

Optimization of Detection Antibody (MAB 69.26)
The first major goal in developing a competitive ELISA was to maximize the signal that could be generated by optimizing the amount of capture antibody (MAB 69.26). As mentioned in the methodology, the antibody was added in separate 0.6μL volumes. This was done because a 7x4mm nitrocellulose could not handle higher volumes without overwhelming the nitrocellulose. By adding only 0.6μL the effective detection zone could be controlled by making it a clear dot that fit within the confines of the nitrocellulose geometry.

Optimizing Biotinylated-IαIP Concentration
The next step involved determining the concentration of B-IαIP, which would serve as a competitor for the un-labeled IαIP. The goal was to determine the appropriate saturation point for B-IαIP (the point where signal does not increase with increased concentrations) with which to dilute the un-labeled protein with. This point can be found corresponding to the values just before the plateau or saturation region on the signal vs concentration curve. As shown, by the following graphs, the point at which this occurs seems to be between 1:100 and 1:400. best chance for an un-labeled IαIP to effectively compete with a labeled protein.

Development of Standardized Curve
Since the optimal competitor concentration was determined to be between 1:200 or 1:400, dilutions of un-labeled IαIP were created and run through the chips.
As shown by the below figures the first signs of a standardized curve are seen within   From this data, it is clear that it is possible to detect the presence and further to quantify this presence using the competitive ELISA for IαIP on a paper-based platform. The next step is to move this test onto a fully autonomous platform in order to satisfy the requirements of a 'simple test' as developed by FDA (these are described in 1.2).

Removal of One Fluid
When moving toward the autonomous device it is important to minimize complexity where ever possible. When more fluids and thus more valves are introduced, the success rate of the chips falls. The following images show a comparison of two chips designed to capture IαIP. In an effort to try to minimize the number of fluids that need to be entered into this device, an experiment was The first two chips (in the upper left) tested the ability to capture B-IαIP and second two chips (in the upper right) tested the ability to capture B-IαIP which was prepared with HRP-Strep prior to its addition.
As can been seen by the above images, adding HRP-Strep to the prepared solution rather than to the chip achieves comparable results. Not only is a slightly stronger signal achieved, but also the signal to noise ratio is increased. Although this was calculated by ImageJ the difference can be seen by the naked-eye. This confirms that the HRP-Strep may be integrated into the sample fluid and furthermore that a 118 PBS wash can take its position in a 3-fluid circuit, when moving toward the development of an autonomous chip, Another finding of this experiment came as a result of changing the dilution fluid from 1wt.% Milk-PBST to just PBST. As the results were calculated this change yielded a much higher background signal. So it seems that the 1wt.% Milk-PBST is an integral component for reducing the background signal and thus increasing the signal/noise.

Complete Competitive ELISA
After determining that the competitive ELISA could be run with only three fluids, forgoing the necessity to design a more complex chip, the following results were achieved. The below table shows the signal created from each concentration ratio of IαIP to 1:400 B-IαIP in 1wt.% Milk-PBS solution. Again it is important to note here that PBS was used instead of PBST, as Tween 20 is the surfactant that is responsible for opening the valves. From the above images it can be seen that the signal remains until the 1:5 dilutions, at which point it begins to fade due to the higher concentration of unlabeled protein. The following chart was generated from data of the above images which was calculated with ImageJ. The curve again is reproduced and shows a decrease in signal at the same point as previously, at the 0.25-0.1 dilutions. Again this decrease of intensity occurs due to the higher number of un-labeled proteins relative to the single production labeled proteins. Within this test for samples containing a low level of un-labeled protein, signal/noise turned out to be slightly lower averaging around 2.
As can be seen by the above data, this paper based chip has the ability to run a competitive ELISA to quantitatively detect a concentration of IαIP within a sample.

Concentration of CP in B-IaIP
Complete Competitive ELISA 120 platform is operating at 3-5% of that range. In order to make this a viable product that range needs to be increased.

CHAPTER -CONCLUSION
During the course of this research the development of the Paper Based Diagnostic test was progressed in three particular areas. The first being the construction with spray adhesive, which drastically reduces the time for assembly and the number of necessary components. With a spray adhesive, the tests became more reliable and easier to assemble. It also allowed for greater possibilities in geometry as the dimensions were no longer restricted by the widths of double-sided tape.
The second innovation was through addition of 'kinks', which added further control for timing and reducing loss of secondary fluids. This technique show cases the broad spectrum of possibilities that are available for a paper based device.
The third innovation and likely the most impactful came through the development of a competitive ELISA as well as a complete ELISA by Hanno Teiwes. By developing a standard curve for the detection of IαIP through the competitive ELISA scheme, a new method for detection is presented. This method is simpler, easier to use, and of course faster than conventional laboratory tests. An important finding through this study was the addition of HRP-Strep into the sample fluid which allowed for a less complex device. This may also be beneficial to the research of Hanno Teiwes as his device can become less complex as well.

Recommendations for Future Work
Within this section directions for further research are discussed to further streamline the fabrication, reliability, and application of a PBD.

Increasing detection range
Increasing the detection range capabilities is next major challenge to be tackled in this project. Compared to competitive ELISA conducted in a 96 well plate, the paper based set up is operating at about 5% the detection capacity. Greater ranges may require optimization of competitor protein concentration, adjustment of sample and substrate volumes, and potential increases in detection protein volume on the nitrocellulose. Up to this point this work serves as a proof of concept and a foundation from which to build upon.
Another problem with this methodology lays in the fact that all reagents are pipetted by hand. As a result this causes great variances in the various sample concentrations and especially in the detection antibodies. When adding detection antibodies with a pipette at low volumes (0.5 µL) it is impossible to be accurate. This inaccuracy then carries over into the signal generation which then has variances.
When examining the data of  As a visual to compare the current state of this PB Competitive ELISA, Figure   5-1 shows the detection capabilities for the PB Device vs the current bench top set up used by Dr. Yow-Pin. As mentioned currently this test is can detect only higher concentration of IαIP of 1:5. Perhaps great advancements could be made through the précising that could be created by a robot. Such an instrument could greatly improve the consistency and reproducibility of detection ranges for this research.

Enzymatic Inhibitory Assay
As discussed in the second chapter, the enzyme trypsin is well known to interact with IαIP. Early in this research, experiments were conducted to test the possibility of developing an inhibitory assay for the detection of IαIP. Currently, this can be done in-solution with along with the substrate BAPNA. Preliminary test on paper yielded weak results due to the light yellow color change that is created by the BAPNA substrate. One option may be to investigate the chance to use another substrate such as ALP, which is shown to have a very high signal/noise in a paper based setup. If such a test could be applied to a paper-based device, there may be the chance to have lower limits of detection as compared to current bench-top test.

New Substrates
One of the weaker components of this research was the signal returned as a result of using DAB and HRP-Strep. After repeated experimentation the background signal proved too high, even with multiple wash steps. In order for this platform to be competitive with the current bench top set-ups, a higher signal/noise ratio needs to be determined. Perhaps this could be developed by using other substrates that create higher signal intensities. If biotin was instead conjugated with ALP, a BCIP/NBT substrate could be used, which has been shown to generate signal/noise in the range of 17-20 at the highest concentrations. There may also be the chance to add amplifiers as discussed in chapter 2, which could lead to better signal generation.

Construction
Another area that has a great frontier yet to be explored is in the construction of paper based devices as they pertain to competitive ELISA.

Miniaturization
In particular the miniaturization could improve this device in many aspects.
The foremost improvement would come by the reduction of cost. As of now, each sheet of filter paper costs nearly $2. On this one sheet of filter paper, there is only enough room to print one batch of 6 chips (with all the necessary layers). A miniaturization would drive down the cost per chip. This would also offer the opportunity to run more optimization tests over a short period of time.
Subsequently, smaller geometries may be able to have a higher yield of success.

Making Valves more Reliable
A large area of development may come in refining the valve technology. As it stands a 60% yield was achieved within this research, which needs to be improved if such a device were to go into mass production. This is one of the most important components to a multi-fluidic device as it offers the chance to run ELISA and other more complex tests than are currently available for PBDs. As biologicals are introduced into chips integrated with valves, further optimization and reconfiguring may also need to be refined. 126

Nitrocellulose Contact with Main Channel
One of the biggest problems seen in the development of a competitive assay was the connection between the main fluid channel and the nitrocellulose pad. Often there was not a strong enough connection to ensure consistent fluid transfer from the main channel through the nitrocellulose and further into the waste pad. This is the most important functional component in the device and it is imperative to the results to ensure a consistent reagent flow. Usually, a small amount of pressure was applied to the overlapping area of the main channel and the nitrocellulose which would speed along the flow. Perhaps this pressure could come from an enclosure or housing.

IαIP Structure and Physiological Pathways
Inter-α-inhibitor or Inter-α-trypsin inhibitor was isolated in the 1960's, and its function remained a mystery up until the research in the last two decades revealed much about its structure and physiological influences [32].
The IαIP is made from three polypeptides including two heavy chains and one light chain. Another major component is the bikunin end which is responsible for the inhibitory activity of IαIP. IαIP is synthesized in the liver by hepatocytes, which are the cells making up 70-80% of the liver's tissue. Hepatocytes are responsible for protein synthesis, storage, transformation of carbohydrates, synthesis of cholesterol, bile salts and phospholipids, and are also responsible for detoxification, modification and excretion of exogenous and endogenous substances.
Initially the protein family was thought be involved with protease inhibition, however, it has been shown that proteases are more efficiently inhibited by their respective physiological inhibitors [60]. Research has shown that IαIP rather acts as a 128 shuttle that captures and transports protease inhibitors to the physiological inhibitors [61]. The bikunin end has been shown to also inhibit granzyme K, which is found in T-cells and is involved with cell-mediated cytotoxic defenses [62]. Other functions of bikunin included mitogenic activity, inhibitory activity of hyaluronidase, prevention of premature delivery, ischemia stress and renal failure, and suppression of pancreatitis, colitis and arthritis [63]. The research that was involved in these findings was mostly focused on bikunin, but another phenomenon has been shown with relation to this protein. [59] In other areas of the body, bikunin has been present in forms both with one heavy chain and without any heavy chains. More functions are described with relation to hyaluronan (HA), a glycosaminoglycan which contributes to cell proliferation and migration, and also has shown functions with knee joint lubrication.
HA has been found also in complex with serum-derived hyaluronan-associated protein (SHAP). HA is bound to SHAP though an ester bond. [59]

SHAP-HA, IαI, and inflammatory diseases
The SHAP-HA complex has additionally been found more highly expressed in patients with rheumatoid arthritis. The complex has been found within synovial fluid, which is needed for lubricated joints particularly in the knee. It has been suggested to play a role in the inflammatory response, by regulating the adhesive properties of leukocytes, or white blood cells. Leukocytes are also responsible for defending the body against foreign bodies and infectious diseases.