Preparative-scale protein separations have always been critical to the advancement of the life sciences. Among preparative-scale separation techniques, isoelectric trapping (IET) promises efficient separations and high production rates. This dissertation focuses on the improvement of two aspects of preparative-scale IET protein separations: the instrumentation used and the monitoring of the separation. The first aspect (preparative-scale) is the IET device: the improvement of a multicompartmental electrolyzer (MCE) to increase the efficiency and production rate of IET separations. The redesign focused on three major areas: (1) the sealing system, (2) the configuration of the liquid flow path, and (3) the cooling system. The second aspect (analytical-scale) is the monitoring of the IET separation: the design and manufacture of durable surface-modified capillaries which provide controlled, variable anodic and cathodic electroosmotic flow (EOF) to help develop, plan, and monitor the IET separations.

Preparative-scale protein separations have always been critical to the advancement of the life sciences. Among preparative-scale separation techniques, isoelectric trapping (IET) promises efficient separations and high production rates. This dissertation focuses on the improvement of two aspects of preparative-scale IET protein separations: the instrumentation used and the monitoring of the separation. The first aspect (preparative-scale) is the IET device: the improvement of a multicompartmental electrolyzer (MCE) to increase the efficiency and production rate of IET separations. The redesign focused on three major areas: (1) the sealing system, (2) the configuration of the liquid flow path, and (3) the cooling system. The second aspect (analytical-scale) is the monitoring of the IET separation: the design and manufacture of durable surface-modified capillaries which provide controlled, variable anodic and cathodic electroosmotic flow (EOF) to help develop, plan, and monitor the IET separations.

The classical isoelectric trapping (IET) technique, using the multicompartment electrolyzer (MCE), has been one of the most successful electrophoretic techniques in preparative-scale protein separations. IET is capable of achieving high resolution discrimination of proteins, by isolating proteins in between buffering membranes, in their isoelectric state. However, due to the inherent nature of the IET process, IET has suffered several shortcomings which have limited its applicability. During a classical IET separation, a protein gets closer and closer to its pI value, thus the charge of the protein gets closer and closer to zero. This increases the likelihood of protein precipitation and decreases the electrophoretic velocity of the protein, thus making the separation very long. Furthermore, the problems are aggravated by the fact that the instrumentation currently used for IET is not designed to maximize the efficiency of electrophoretic separations. To address these problems, a new approach to IET has been developed, pH-biased IET. By controlling the solution pH throughout the separation, such that it is not the same as the protein’s pI values, the problems of reduced solubility and low electrophoretic migration velocity are alleviated. The pH control comes from a novel use of isoelectric buffers (also called auxiliary isoelectric agents or pH-biasers). The isoelectric buffers are added to the sample solution during IET and are chosen so that they maintain the pH at a value that is different from the pI value of the proteins of interest. Two new pieces of IET instrumentation have been developed, resulting in major improvements in protein separation rates and energy efficiency. A variety of separations, of both small molecules and proteins, have been successfully performed using the pH-biased IET principle together with the new instrumentation.

Isoelectric trapping (IET) in multicompartment electrolyzers (MCE) has been widely used for the electrophoretic separation of ampholytic compounds such as proteins. In IET, the separation occurs in the buffering membranes that form a step-wise pH gradient in the MCE. Typically, buffering membranes have been made by copolymerizing acrylamide with Immobiline compounds, which are acidic and basic acylamido buffers. One major problem, however, is that these buffering membranes are not stable when exposed to high concentrations of acid and base due to hydrolysis of the amide bonds. Poly(vinyl alcohol)-based, or PVA-based, membranes were made as an alternative to the polyacrylamide-based membranes since they provide more hydrolytic and mechanical stability. Four mid-pH, PVA-based buffering membranes that contain single ampholytes were synthesized. These buffering membranes were used to trap small molecular weight pI markers for up to three hours, and were also used in desalting experiments to remove strong electrolytes from a solution of ampholytes. Additionally, the membranes were used in IET experiments to separate mixtures of pI markers, and to fractionate the major proteins in chicken egg white. The membranes did not show any degradation when stored in 3 M NaOH for up to 6 months and were shown to tolerate current densities as high as 16 mA/cm2. In addition, six series of PVA-based membranes, whose pH values can be tuned over the 3 < pH < 10 range, were synthesized by covalently binding aminodicarboxylic acids, and monoamines or diamines to the PVA matrix. These tunable buffering membranes were used in trapping experiments to trap ampholytes for up to three hours, and in desalting experiments to remove strong electrolytes from a solution of ampholytes. These tunable buffering membranes were also used in IET experiments to separate proteins, some with pI values that differ by only 0.1 pH unit. The tunable buffering membranes did not show any signs of degradation when exposed to 3 M NaOH for up to 3 months, and could be used in IET experiments with current densities as high as 20 mA/cm2. These tunable buffering membranes are expected to broaden the application areas of isoelectric trapping separations.

Isoelectric trapping (IET) in multicompartment electrolyzers (MCE) has been widely used for the electrophoretic separation of ampholytic compounds such as proteins. In IET, the separation occurs in the buffering membranes that form a step-wise pH gradient in the MCE. Typically, buffering membranes have been made by copolymerizing acrylamide with Immobiline compounds, which are acidic and basic acylamido buffers. One major problem, however, is that these buffering membranes are not stable when exposed to high concentrations of acid and base due to hydrolysis of the amide bonds. Poly(vinyl alcohol)-based, or PVA-based, membranes were made as an alternative to the polyacrylamide-based membranes since they provide more hydrolytic and mechanical stability. Four mid-pH, PVA-based buffering membranes that contain single ampholytes were synthesized. These buffering membranes were used to trap small molecular weight pI markers for up to three hours, and were also used in desalting experiments to remove strong electrolytes from a solution of ampholytes. Additionally, the membranes were used in IET experiments to separate mixtures of pI markers, and to fractionate the major proteins in chicken egg white. The membranes did not show any degradation when stored in 3 M NaOH for up to 6 months and were shown to tolerate current densities as high as 16 mA/cm2. In addition, six series of PVA-based membranes, whose pH values can be tuned over the 3 < pH < 10 range, were synthesized by covalently binding aminodicarboxylic acids, and monoamines or diamines to the PVA matrix. These tunable buffering membranes were used in trapping experiments to trap ampholytes for up to three hours, and in desalting experiments to remove strong electrolytes from a solution of ampholytes. These tunable buffering membranes were also used in IET experiments to separate proteins, some with pI values that differ by only 0.1 pH unit. The tunable buffering membranes did not show any signs of degradation when exposed to 3 M NaOH for up to 3 months, and could be used in IET experiments with current densities as high as 20 mA/cm2. These tunable buffering membranes are expected to broaden the application areas of isoelectric trapping separations.

Sepsis is a “life-threatening organ dysfunction caused by a dysregulated host response to infection” that has a widespread impact on human life around the world. It affects more than 1.5 million people, killing at least 250,000 each year in the US alone and affects 90,000 people annually, with estimated mortality rates of up to 30% in Canada. Our understanding of the different biochemical pathways that in the progression of sepsis has improved patient care for sepsis patients. One part of patient care is the use of biomarkers for patient prognosis that draws on the full range of relevant and available information to model the possible outcomes for an individual. Numerous biomarkers have been studied for patient prognosis that includes Procalcitonin (PCT), C-reactive protein (CRP), TNF-α, cfDNA, protein C and PAI 1. Using a panel of multiple biomarkers provided more accuracy in patient prognosis than using individual biomarkers and one such panel that was proposed used cfDNA, protein C, platelet count, creatinine, Glasgow Coma Scale [GCS] score, and lactate. Commercial, low cost POC techniques were available for the measurement of all biomarkers besides cfDNA and protein C. The objective of this doctoral thesis was chosen to develop low cost, microfluidic devices for the measurement of protein C and cfDNA using nonspecific fluorescence dyes that would enable the eventual integration of the systems and improve patient prognosis. The measurement of protein C in plasma required the separation of protein C from interfering proteins in plasma. This was done through the development of a two-stage separation process that included the development of tunable agarose isoelectric gates for separating proteins using their isoelectric point and the miniaturization of immobilized metal affinity chromatography and its extension to Barium for the selective binding of proteins using their chemical affinity. This was performed in a xurographically fabricated chip to reduce costs and enable the use of geometric focusing of the electric field to enable the operation of the device at a lower applied voltage. The challenges faced with cfDNA were different due to the different characteristics of the material and less interference from plasma. The requirement was to measure the total cfDNA content with minimal cost in comparison to currently available techniques. This was achieved through the development of thread microfluidic devices that showed the use of thread for automated aliquoting of samples by controlling length and twists of the thread. Preconcentration and use of external apparatus was avoided by showing that thread could be used to amplify fluorescence response to a range that was sufficient for the measurement of cfDNA in sepsis patients. A portable fluorescence imaging setup was developed for this purpose and was used in demonstration for the measurement of cfDNA in plasma with sufficient resolution. In conclusion, we developed technologies for rapid and low-cost measurement of protein C and cfDNA using xurographic and thread-based microfluidics that may serve as valuable in improving patient prognosis. / Thesis / Doctor of Philosophy (PhD) / Sepsis is a major reason for hospitalization and cause of death in hospitals worldwide. Its treatment is highly time sensitive with each hour of delay in diagnosis causing a significant increase in chances of death. Due to the wide range of symptoms that can be caused by sepsis, its diagnosis uses a scoring method that relies on the expertise of the onsite doctors and nurses increasing their workload. A more objective system for detection requires the measurement of the quantities of different biomarkers in blood. Biomarkers are proteins present in plasma that change in quantity due to the body’s reaction to sepsis. Several of these biomarkers have been identified and studied for their use in both diagnosing the presence of sepsis and in predicting the outcome with the current treatment plan. In this PhD study, we chose two of these biomarkers – circulating free DNA (cfDNA) and protein C and developed low-cost techniques for rapidly measuring their concentration in blood plasma. To do this, we made microfluidic devices with techniques that use low-cost materials such as plastic sheets and threads.The device for the measurement of protein C required separating it from many other proteins in plasma. We showed that a device fabricated from stacked plastic sheets and integrated with agarose gels could be used for the measurement of protein C in plasma with sufficient resolution to help with treating septic patients at a cost of less $5 per device. Similarly, we showed that a device that integrated threads with plastic sheets could be used for measuring the quantity of cfDNA in plasma in a portable format within 15 minutes. Overall, we developed tools for rapid measurement of two biomarkers of sepsis using low cost device that cost under $5 to run and could led to improving the quality of care for sepsis patients.

Isoelectric trapping (IET) has become an accepted preparative-scale electrophoretic separation technique. However, there are still a number of shortcomings that limit its utility. The performance of the current preparative-scale IET systems is limited by the serial arrangement of the separation compartments, the difficulties in the selection of the appropriate buffering membranes, the effect of Joule heating that may alter separation selectivity and a lack of methods for the determination of the true, operational pH value inside the buffering membranes. In order to bolster the current membrane pH determination methods which rely on the separation of complex ampholytic mixtures, a fluorescent carrier ampholyte mixture was synthesized. The use of a fluorescent mixture allows for a reduced load of carrier ampholytes, thereby reducing a possible source of error in the pH determinations. A mixture of carrier ampholytes tagged with an alkoxypyrenetrisulfonate fluorophore was shown to have suitable fluorescence and ampholytic properties and used to accurately determine the pH of high pH buffering membranes under actual IET conditions. In a more elegant solution to the difficulties associated with pH determinations, a method utilizing commercial UV-transparent carrier ampholytes as the ampholyte mixture to be separated was developed. By using commercial carrier ampholytes and eliminating the need to synthesize, purify, and blend fluorescently tagged ampholytes, the new method greatly simplified the determination of the operational pH value of the buffering membranes. In order to address the remaining limitations, a new system has been developed that relies on (i) parallel arrangement of the electrodes and the collection compartments, (ii) a directionally-controlled convection system for the delivery of analytes, (iii) short anode-to-cathode distances, (iv) short intermembrane distances, and (v) an external cooling system. This system has been tested in four operational modes and used for the separation of small molecule ampholytic mixtures, for the separation of protein isoforms, and direct purification of a target pI marker from a crude reaction mixture.