Menu

Close

In vivo glucose imaging in multiple model organisms with an engineered single-wavelength sensor

Jacob P. Keller, Jonathan S. Marvin, Haluk Lacin, William C. Lemon, Jamien Shea, Soomin Kim, Richard T. Lee, Minoru Koyama, Philipp J. Keller, Loren L. Looger

Preprint posted on March 08, 2019 https://www.biorxiv.org/content/10.1101/571422v1

How to visualize glucose concentrations in vivo: the Looger lab has engineered a new family of single-wavelength glucose sensors to unravel the biology of glucose in models such as tissue culture, Drosophila and zebrafish.

Selected by Stephan Daetwyler

Context

Glucose is one of the most important molecules of life. As a product of photosynthesis, it is a major organic compound that serves as an energy store, fuel for metabolic engines, and constituent of other molecules. Moreover, it is the major form of transport of carbohydrates from one cell to another in animals. Its importance is also highlighted by its tight regulation. In humans, an intricate system involving the pancreas, brain, liver, gut, adipose and muscle tissue acts together to regulate glucose levels via various hormones such as insulin and glucagon, neurotransmitters and cytokines [1]. Disturbances of this interplay lead to metabolic disorders such as type 2 diabetes mellitus [2]. Diabetes is a major cause of blindness, kidney failure, heart attacks, stroke and lower limb amputation affecting about 422 million people in 2014 [3]. Techniques to measure glucose concentrations in vivo are therefore of high clinical and scientific importance. While many glucose sensors have been established ranging from finger-prick to intravenous implantable devices [4], techniques relying on fluorescence are the most promising for basic research and the clinics [5].

Fluorescence microscopy offers very sensitive and non-invasive measurements. Most existing fluorescence-based glucose sensors, however, suffer from lack of specificity [6], targeting [7], or modest fluorescence responses [8]. This makes their application to imaging studies in living animals and plants with high spatial and temporal resolution difficult. Recently, a novel glucose sensor relying on circularly permuted yellow fluorescent protein combined with a bacterial periplasmic glucose / galactose-binding protein was introduced that can address the above issues [9]. However, its application has only been demonstrated in E. coli cells. In their new preprint, Keller et al. introduce a novel family of single-wavelength sensors and demonstrate their value for studies of glucose concentrations in tissue culture, Drosophila and zebrafish model systems.

Key findings

1. Engineering of a family of genetically encoded glucose sensors

In this new preprint, a family of genetically encoded glucose sensors with a high signal-to-noise ratio, fast kinetics and affinities from 1 mM to 10 mM has been introduced. At the heart of the new sensor, named iGlucoSnFR, is a circularly-permuted green fluorescent protein (cpGFP) that has been inserted into a glucose binding protein (GBP) from the thermophile bacterium T. thermophilus (Tt). Upon glucose binding of the GBP, the configuration of the sensor changes to yield higher fluorescence signal (Fig. 1). To complement the iGlucoSnFR for ratiometric measurements and/or cell typing, the glucose-insensitive red fluorescent protein mRuby2 has been fused to the C-terminus of iGlucoSnFR (iGlucoSnFR-mRuby2). Both variants have been established as cytosolic, secretable and membrane-bound versions. Using protein engineering, mutations in the linker and binding regions of the GBP have been introduced to optimize the affinity and fluorescence gain (ΔF/F) of the sensor. iGlucoSnFR shows high affinity to glucose (6.5 mM) compared to other sugars such as galactose (20 mM) and 2-deoxy-glucose (45 mM), and applicability at pH values ranging between pH 6.0-9.5.

Figure 1: Schematic of iGlucoSnFR sensor. Upon binding to glucose (orange), the glucose sensor iGlucoSnFR changes its configuration to yield higher fluorescent signal (dF/F = 2.32). The sensor consists of a circularly-permuted green fluorescent protein (cpGFP, green) inserted into the glucose-binding protein (GBP, light and dark blue) from the thermophile bacterium T. thermophilus.

 

2. Demonstration of the glucose sensor in various applications

The novel glucose sensors have been applied to study neuron/glia co-cultures, larval Drosophila central nervous system explants, and zebrafish larvae. The experiments show that the new sensor can be easily targeted to specific populations of cells to visualize changes in glucose levels.

Importance

Applications of this novel glucose sensor to various biological questions are imminent. With this sensor in hand, spatiotemporal dynamics of glucose trafficking, maintenance and regulation on scales ranging from whole organisms to intracellular levels can be discerned. The opportunity to express this sensor in specific cell populations and specific intracellular components will open new ways to address long-standing questions in fields as diverse as development, behavioral science, neuroscience, metabolism and biology of disease such as cancer and diabetes.

Future directions and questions

  • Have the authors acquired high-resolution images of single cells with subcellular resolution? If yes, have the authors found interesting subcellular localization patterns of glucose? Along these lines, the authors also report the establishment of membrane-bound and secreted versions of their glucose sensors. How do they compare to the cytosolic sensor discussed in the preprint?
  • Can the authors elaborate why circularly-permuted variants of fluorescent proteins such as GFP are particularly suited for the construction of sensors?
  • Have the authors thought of or already performed long-term measurements of how glucose levels change during development in general, and specifically in different regions? A particularly exciting question would be to understand how onset of blood perfusion in a tissue changes local glucose levels.
  • Glucose monitoring in patients with diabetes is still often relying on a finger-prick glucometer. Do the authors of the study envision whether their novel sensor could help diabetes patients by providing novel tools to accurately measure glucose levels?

References

[1]       Röder PV, Wu B, Liu Y, Han W. Pancreatic regulation of glucose homeostasis. Experimental & Molecular Medicine 2016;48:e219–e219. doi:10.1038/emm.2016.6.

[2]       DeFronzo RA, Ferrannini E, Groop L, Henry RR, Herman WH, Holst JJ, et al. Type 2 diabetes mellitus. Nature Reviews Disease Primers 2015;1:15019.

[3]       World Health Organization. Diabetes Fact Sheet n.d. https://www.who.int/news-room/fact-sheets/detail/diabetes (accessed March 28, 2019).

[4]       Oliver NS, Toumazou C, Cass AEG, Johnston DG. Glucose sensors: a review of current and emerging technology. Diabetic Medicine 2009;26:197–210. doi:10.1111/j.1464-5491.2008.02642.x.

[5]       Wang H-C, Lee A-R. Recent developments in blood glucose sensors. Journal of Food and Drug Analysis 2015;23:191–200. doi:10.1016/j.jfda.2014.12.001.

[6]       Mandal DK, Bhattacharyya L, Koenig SH, Brown RD, Oscarson S, Brewer CF. Studies of the Binding Specificity of Concanavalin A. Nature of the Extended Binding Site for Asparagine-Linked Carbohydrates. Biochemistry 1994;33:1157–62. doi:10.1021/bi00171a015.

[7]       Ge X, Tolosa L, Rao G. Dual-Labeled Glucose Binding Protein for Ratiometric Measurements of Glucose. Anal Chem 2004;76:1403–10. doi:10.1021/ac035063p.

[8]       Deuschle K, Okumoto S, Fehr M, Looger LL, Kozhukh L, Frommer WB. Construction and optimization of a family of genetically encoded metabolite sensors by semirational protein engineering. Protein Sci 2005;14:2304–14. doi:10.1110/ps.051508105.

[9]       Hu H, Wei Y, Wang D, Su N, Chen X, Zhao Y, et al. Glucose monitoring in living cells with single fluorescent protein-based sensors. RSC Adv 2018;8:2485–9. doi:10.1039/C7RA11347A.

 

Tags: drosophila, glucose, sensor, zebrafish

Posted on: 5th April 2019

Read preprint (No Ratings Yet)




  • Author's response

    Jacob Keller and Loren Looger shared

    • Have the authors acquired high-resolution images of single cells with subcellular resolution? If yes, have the authors found interesting subcellular localization patterns of glucose? Along these lines, the authors also report the establishment of membrane-bound and secreted versions of their glucose sensors. How do they compare to the cytosolic sensor discussed in the preprint?

    We have looked only briefly at subcellular glucose, since we wanted to focus first on what other researchers might find most useful, but we agree that subcellular glucose imaging/measurements are fascinating. We have done a bit, though: in a previous paper {PMID: 27716484}, we expressed iGlucoSnFR in the lumen of the endoplasmic reticulum (ER), and showed that ER membranes are permeable to glucose, likely via ER-localized glucose transporters. There is much left to be learned about sub-cellular glucose trafficking, e.g. into the ER, mitochondria and axon terminals.

    To really address subcellular glucose trafficking, something like a lattice light-sheet microscope would be best, since high-resolution, volumetric time-lapse imaging would really bring out the most from the experiments.

    Regarding the differentially localized versions, we did not explore their use extensively, but did confirm that the extracellular version is successfully targeted to the membrane in cultured cells and in larval zebrafish mosaics, and also confirmed that the secreted version appears in the media of cultured cells. Admittedly, more work is definitely warranted on those two versions. We think that the secreted version in particular might work well in the CNS as a sensor of extracellular CSF glucose: with exchangeability and a higher amount of sensor-volume compared to the membrane-localized version, photobleaching would likely be significantly reduced, and the entire CNS should be homogeneously filled with sensor.

    • Can the authors elaborate why circularly-permuted variants of fluorescent proteins such as GFP are particularly suited for the construction of sensors?

    When the protein is circularly permuted in this fashion, the new termini end up immediately adjacent to the chromophore. This allows conformational changes in the PBPs to perturb the chromophore’s local environment, thus modulating its fluorescence properties. This trick has now been used to make a large number of sensors.

    • Have the authors thought of or already performed long-term measurements of how glucose levels change during development in general, and specifically in different regions? A particularly exciting question would be to understand how onset of blood perfusion in a tissue changes local glucose levels.

    Regarding long-term measurements, one has to appreciate that sensor expression levels can change over time, introducing confounding signals. This can to some extent be countered by performing ratiometry with the mRuby2-tagged version, but it is significantly harder, generally, to keep experimental subjects stable over long periods of time. With sufficient care, however, it should be possible.

    The question of blood perfusion, we agree heartily, would be an interesting one, and welcome you or others to try it. The transgenic flies have been deposited at Bloomington. The DNA constructs and AAV viruses have been sent to Addgene. The transgenic fish are freely available from the HHMI Janelia Research Campus.

    • Glucose monitoring in patients with diabetes is still often relying on a finger-prick glucometer. Do the authors of the study envision whether their novel sensor could help diabetes patients by providing novel tools to accurately measure glucose levels?

    Yes, we have considered this, and think that with the array of affinities that we have and a bit of standardization, a very accurate, precise, and rapid assay could be developed. We should point out that there has been a great deal of previous work in diabetic glucose monitoring, and this is tricky to get right. One interesting avenue is that since the biosensors are totally genetically encoded, patients’ own cells or tissues can be harnessed to measure glucose concentrations in real time. This could facilitate studies of patient-specific differences in glucose metabolism and trafficking and improve our understanding of metabolic disorders, potentially allowing diabetes management strategies to be tailored for specific patient populations.

    Have your say

    Your email address will not be published. Required fields are marked *

    This site uses Akismet to reduce spam. Learn how your comment data is processed.

    Sign up to customise the site to your preferences and to receive alerts

    Register here
    Close