Overview
It was abundantly clear to us that a large gap between patients and clinical practice existed in the realm
of Chronic kidney disease (CKD), but the challenge was discovering the most effective way to bridge that
gap. Our team’s goal was to create a meaningful solution to the current issues in CKD diagnostics, and to
do so, we had to ensure that our biosensor was safe, realistic, and that it would actually make a
meaningful difference in the lives of CKD patients. As aspiring undergraduate scientists, it is not
difficult to get lost in the science of synthetic biology without having a clear vision. To ensure our
vision of creating a frequent-monitoring and accurate biosensor could be realized, we integrated human
practices in every step of our project. From project ideation, to synthetic biology and protein construct
design, to the development of our physical biosensor and how it would be used by patients.
Chronic kidney disease is a complex disease, and of similar complexity are the practical, ethical, and
social implications of creating a biosensor to address CKD. So, we constantly engaged with professors,
doctors, dieticians, researchers, patients, students, the general public, and experts in patent law
throughout the lifecycle of our project. Despite being limited in our possibilities of engagement given the
ongoing pandemic, we made full use of virtual platforms to gain as much insight as possible into the
development of Velcrion, our transdermal biosensor to detect critical CKD metabolites. By augmenting the
feedback given to us from experts in a variety of fields, we not only gained valuable towards our project,
but also gained the opportunity to explore different aspects of science such as intellectual property,
sustainability, implementation, cost-effectiveness, and societal impact.
To ensure we integrated human practices into our project in the most effective manner possible, we followed
a three-staged protocol for successful implementation: 1. Identify, 2. Design, and 3. Analyze. The cyclical
nature of this protocol (Figure 1) allowed us to successfully identify the needs of CKD patients, design
our biosensor solution through intensive discussions with experts, analyze our work by integrating feedback
of professionals and then use that feedback to restart the cycle.
Figure 1. Overview of the integrated human practices cycle.
1. Identify
Prior to beginning any biosensor protein or physical biosensor prototype design, we consulted with several
experts in varying fields of healthcare to properly fill our gaps in knowledge about chronic kidney
disease (CKD). These gaps in knowledge were as follows: What compounds are the most important to monitor
in CKD patients? How often are these analytes monitored? What issues do CKD patients face in their daily
lives, and how could a biosensor help solve them?
To fill these gaps in knowledge we knew it was imperative to meet with healthcare professionals who
frequently interacted with CKD patients. Prior to beginning working on our project, we met with a
nephrologist, dietician, and an internist and geriatrician.
Meetings with Dr. Rachel Holden, MD
We met with Dr. Rachel Holden to gain knowledge and further our understanding about CKD. She told us about
the current gap between CKD patients and the clinic in relation to diagnostics, and how blood is only
sampled for patients approximately once every 6 weeks. Through our frequent discussions we learned how
ineffective this slow frequency of blood sampling is for effective monitoring of CKD metabolites. This
formed the premise of our project; creating a biosensor that allows for frequent, daily sampling of CKD
metabolites.
Next, we discussed what specific analytes we should focus on monitoring. Initially, our team only planned
on measuring phosphate, however after our discussions with Dr. Holden, we learned that would also be
extremely beneficial to monitor (1) potassium, (2) parathyroid hormone (PTH), (3) fibroblast growth factor
23 (FGF-23), and (4) glucose. We learned that it was important to monitor these analytes for the following
reasons: (1) elevated levels of potassium (hyperkalemia) can interfere with the electric signals of the
heart, potentially leading to arrhythmias, cardiac arrest, and death, (2) alterations in levels of PTH
(hyper- or hypo-parathyroidism) can deteriorate bone, (3) elevations in levels of FGF-23 can increase the
risk of cardiovascular-related mortality in CKD patients, and (4) many CKD patients are diabetic, therefore
providing patients with the ability to monitor their glucose (alongside the previous analytes) could
immensely increase the quality of life of these individuals.
We also discussed techniques to monitor CKD metabolites. As our vision was to create a biosensor that was
as minimally invasive as possible, our initial biosensor plans were to measure analytes either via saliva
or sweat. We learned that detecting phosphate, potassium, PTH, FGF-23, and glucose in the saliva would be
very difficult as it would not represent physiological concentrations in circulation. While sweat could
work, we learned it would not be ideal as topical creams would need to be applied on the patients’ skin,
likely causing irritation. After our discussions with Dr. Norris (later in this section), we decided a
microneedle-based approach would be the best option – which we confirmed with Dr. Holden as well.
Meetings with Jenny Munroe, RD
Our next discussions were with Jenny Munroe, a registered dietitian at our local hospital. In these
meetings we covered a range of topics, including (1) the benefit of measuring phosphate, potassium, PTH,
FGF-23, and glucose, (2) how diet plays an extremely important role in the lives of many CKD patients, and
(3) the utility of creating a phone application coinciding with our biosensor. We learned that it would be
extremely beneficial to measure the aforementioned analytes as oftentimes, dietitians and nephrologists
don’t have access to the data due to resource and sampling limitations. Therefore, a biosensor that
frequently monitors CKD analytes would be very valuable. We then leaned about the importance of diet in the
lives of CKD patients. As kidney function is greatly diminished, patients must be cognizant of their
phosphate, potassium, and calcium intake to avoid any direct mineral imbalances, or indirect hormonal
alterations. Lastly, we discussed the utility of creating a mobile application that would collect and store
the patients’ analyte data. We learned that patients could immensely benefit from knowing the levels of
analytes in their system real-time, especially after a meal. As an example, it is very common for patients
to be unaware of phosphate values in many of the foods they eat, therefore having the ability to know the
phosphate in their system just 30 minutes after a meal could provide them with valuable insight towards
which foods they should avoid.
Meetings with Dr. Mireille Norris, MD
We then met with Dr. Mireille Norris, an internist and geriatrician, to learn more about the feasibility of various analyte-collecting methods. Our initial research was invested into sweat-based collection methods, so we discussed this with Dr. Norris. She provided us insight towards the dermatological conditions present in many CKD patients, with references to malfunctioning sweat glands, xerosis cutis (skin dryness), among other conditions. Therefore, we learned that a sweat-based collection method that would likely use a topical cream to induce sweating would not be ideal for our biosensor. Our discussions then progressed towards suitable alternatives, with us landing on a microneedle-based approach. Microneedle-based biosensors use a series of microscopic ‘teeth’ to penetrate the skin, allowing detection of interstitial fluid between cells (1). As the minerals, and hormones we want to detect are distributed systematically, interstitial fluid would serve as a suitable bodily fluid for their detection at physiologically relevant levels which could be correlated to circulating values. Microneedles are said to be non-invasive, in that patients feel no pain when using them for drug delivery, or in our case, fluid collection and detection. After several meetings with both Dr. Norris and Dr. Holden, we concluded that our biosensor would use microneedles as it would be the most efficient, accurate, and most importantly, safest option for CKD patients.
Takeaways
1. The blood of CKD patients is only sampled once every 6 weeks, making this data very difficult to use and
effectively interpret.
2. A comprehensive CKD biosensor should measure phosphate, potassium, parathyroid hormone, fibroblast
growth factor 23, and glucose.
3. Knowledge of these analytes can be incredibly useful for guiding treatments, and tailoring diets to CKD
patients.
4. An accompanying phone app could greatly improve the day-to-day lives of CKD patients.
5. Many CKD patients have dermatological conditions, therefore an interstitial fluid-based microneedle
approach would be ideal for safe measurement.
2. Design
Meetings with Dr. Allingham, PhD
Once we had met with several healthcare experts to identify the needs and problems of the CKD population, we sought to meet with structural biology experts to design the solution. Our meetings with Dr. Allingham were frequent and occurred all throughout the summer. We discussed a wide range of topics, from colorimetric and absorbance-based detection of analytes, linker systems for FRET, and protein-protein immobilization. These meetings were instrumental in designing our hallmark fluorescent biosensor protein construct, which incorporates FRET-based fluorescence and a protein-immobilization system utilizing E / K coiled-coils. We decided upon FRET fluorescence to be used as our method of analyte detection as it would ensure a large dynamic linear range, and very high sensitivity over other techniques of measurement. Further, it would be easy to genetically-encode FRET fluorophores into our constructs, promoting accessibility and recreation of our protein designs.
Meetings with Dr. Adams, PhD
Dr. Adams is an expert in the development of cardiovascular diseases, with a strong emphasis on how chronic kidney disease greatly accelerates cardiovascular disease development. Our meetings with Dr. Adams were pivotal in determining the pathophysiology behind chronic kidney disease, and the importance of measuring our analytes. We learned about (1) the consequences of high phosphate levels and how it may lead to vascular calcification, increasing a CKD patients’ risk of encountering cardiovascular events, (2) the importance of circulating potassium and calcium levels, and how alterations in these levels can lead to abnormal hormonal secretion and vitamin D metabolism, (3) the regulatory roles of parathyroid hormone (PTH) and fibroblast growth factor 23 (FGF-23), and how the levels of these hormones can be altered in an individual with CKD, consequently further progressing the disease. We also learned about which bodily fluids to measure for appropriate correlation of circulating analyte levels and concluded that interstitial fluid measurement would be ideal for our biosensor. These discussions corroborated our findings with Dr. Holden and Dr. Norris.
Meetings with Dr. Campbell, PhD
Beyond his vast knowledge of structural biology and X-ray crystallography, Dr. Campbell is also an expert with molecular modelling – specifically molecular dynamics. As we weren’t able to access a lab throughout the summer, it was imperative that we explored advanced modelling techniques to demonstrate a ‘proof-of-concept’ of our system working. This was accomplished through nanoscale molecular dynamics simulations in the software GROMACS. However, GROMACS is not an easy platform to use, so we frequently met with Dr. Campbell to help us with learning the basics of the software, and all of our (plenty) troubleshooting requests. He also assisted us with developing our bioinformatics software MutaGuide, which can be found under our Model page.
Meetings with Dr. Davies, PhD
Our meetings with Dr. Davies were extremely relevant to determining the ideal method of protein-protein immobilization. We discussed many possibilities, including the use of SNAP-tags, or chemical site-specific immobilization. Many questions came to mind: which could be incorporated the easiest? Most cost-effective? Most stable and protease resistant due to the pH and protein composition of interstitial fluid? After going back-and-forth in several meetings, we concluded that a protein-protein immobilization would be best, specifically, the use of an E/K coiled-coil. This method of protein-immobilization was chosen as it would be cost-effective, easily accessible, and repeatable for mass-manufacturing purposes, stable under physiological conditions, and modular. Having a modular immobilization system was especially important to us as we wanted to ensure any binding protein / FRET combination could be used with our design, which we accomplished through our E/K coiled-coil development.
Takeaways
1. Forster resonance energy transfer (FRET) should be used for analyte-detection given its large dynamic
linear range and sensitivity.
2. The pathophysiology behind chronic kidney disease is complex, and a multi-variate biosensor approach
should be taken.
3. Molecular dynamics simulations can provide valuable insight towards the stability of our protein
constructs.
4. The E/K coiled-coil protein immobilization system would be the most suitable option for our biosensor.
3. Analyze
Graduate Students
Throughout the summer we met with Nolan and Byron, former iGEMers, every week to share updates and results. These meetings were instrumental in furthering the wetlab and drylab’s understanding of their work, and troubleshooting problems related to protein construct design, or biosensor modelling.
Business and Innovation Experts
Our meetings with these individuals will be further explored in the Entrepreneurship section, however they also played extremely valuable roles in analyzing our biosensor in a real-world setting. From our discussions we learned:
1. When pitching our project, we must take cost of manufacturing our biosensor, and the economic burden of chronic kidney disease into account. This led us to researching into the predicted cost to manufacture our physical biosensor, the electrical components inside it, the cost of our genetically engineered protein constructs, and how our biosensor compares economically to current gold standards in CKD diagnostics.
2. It is essential to validate our biosensor design (specifically our microneedle array) with industry and government experts. Although this was difficult to do given the current pandemic, we have noted it as a crucial next step.
3. Think about broadening our scope. Currently, our biosensor is very focused on CKD, however these experts suggested that we should investigate applying our modular biosensor technology towards other diseases.
4. Safe aggregation of data through our app is of immense importance. One of our biggest takeaways was how to design a safe application that would protect the privacy of CKD patients, and what resources and regulations we should refer to throughout the process. We have incorporated much of the feedback we received into our app design and have noted down what resources we should refer to in the future prior to making our app public.
5. We learned a lot of about the process of protecting intellectual property. We were informed about the different types of patents, what steps to take when intellectual property has been disclosed, and what laws and regulations surround patents. We also learned about the patent filing process, such as how to perform a patent search and what query skills are required to do so effectively. These experts also informed us on the patent application procedures, and how different countries have different rules that we must pay attention to. It was suggested that we record the protein construct technology we have developed in Velcrion as an invention, and file for a provisional patent. However, we were urged to be cautious if we were to apply, as our project has only been demonstrated to be functional theoretically. Although our in-silico results suggest our fluorescent biosensor protein constructs would be stable and functional, we must get into a lab and generate concrete results in vitro if we want to be successful in gaining approval for a provisional patent. Therefore, this is a crucial future direction for our team to advance our project.
Queen's Chemical Engineering
We also met with professors and graduate students from Chemical Engineering in the Faculty of Engineering and Applied Science. Within these meetings we focused on the design of the physical biosensor, including: (1) integration of the protein construct with the physical components within the biosensor casing, (2) most-effective design of the biosensor casing, and (3) design and manufacturing of the microneedle array.
We learned about:
1. The feasibility and costs of incorporating gold to immobilize our biosensor. As gold is required in the physical casing of the biosensor to immobilize our protein construct via cysteine chemistry, we were concerned about how feasible it would be to use this technology on a larger scale. We were assured that, despite sounding expensive, the gold required for this kind of immobilization is very cheap and could be incorporated in the manufacturing of many biosensors without a significant economic burden.
2. Many of our initial biosensor casing design problems were associated with how to most-effectively incorporate a laser, and sensor for FRET fluorescence without creating an overly large and impractical casing. Our discussions helped us design our revolver casing, which will only use one laser and sensor for FRET, but will have rotating chambers representing the different analyte-detection regions of the biosensor.
3. Transfer methods of interstitial fluid from the microneedle patch to the chambers of our revolver biosensor design. Our discussions informed us on using a microneedle array, containing 5 unique microneedle regions on the patch that would correspond to different chambers of the revolver. We learned that a pump would not be needed to extract the interstitial fluid, and that the fluid would enter the chambers on its own when latched onto the chamber. We conducted fluid dynamics modelling to confirm this. This also informed our disposable patch design to ensure no contamination would occur in the microneedles.
4. We learned that manufacturing microneedles can be quite difficult, but that our iterated design would be promote an easier manufacturing process (more on this in the prototype section). We also learned that in a scenario of mass-production, which would be required for our disposable microneedle patches, the costs would be minimized.
References
1. Imperial College London. "Microneedle biosensor accurately detects patient's antibiotic levels in real time." ScienceDaily. ScienceDaily, 30 September 2019.