Single-cell imaging of α and β cell metabolic response to glucose in living human Langerhans islets

Abstract

Here we use a combination of two-photon Fluorescence Lifetime Imaging Microscopy (FLIM) of NAD(P)H free/bound ratio in living HIs with post-fixation, immunofluorescence-based, cell-type identification. FLIM allowed to measure variations in the NAD(P)H free/bound ratio induced by glucose; immunofluorescence data allowed to identify single α and β cells; finally, matching of the two datasets allowed to assign metabolic shifts to cell identity. 312 α and 654 β cells from a cohort of 4 healthy donors, 15 total islets, were measured. Both α and β cells display a wide spectrum of responses, towards either an increase or a decrease in NAD(P)H free/bound ratio. Yet, if single-cell data are averaged according to the respective donor and correlated to donor insulin secretion power, a non-random distribution of metabolic shifts emerges: robust average responses of both α and β cells towards an increase of enzyme-bound NAD(P)H belong to the donor with the lowest insulin-secretion power; by contrast, discordant responses, with α cells shifting towards an increase of free NAD(P)H and β cells towards an increase of enzyme-bound NAD(P)H, correspond to the donor with the highest insulin-secretion power. Overall, data reveal neat anti-correlation of tissue metabolic responses with respect to tissue insulin secretion power.

Direct optical nanoscopy unveils signatures of cytokine-induced β-cell structural and functional stress

“…In our experimental system, in fact, cytokine-induced insulin secretion is also associated with a cell metabolic status that recalls the one previously measured under glucose induced insulin secretion 57 . In both cases, the stimulus induces a neat increase of the ratio between protein-bound and free NAD(P)H lifetimes, classically interpreted as an increase in the oxidative-phosphorylation activity of the cell 54,57 . Based on the current model of β-cell function, the shift from low to high bound/free NAD(P)H lifetime ratios well agrees with the need to rapidly increase the ATP/ADP ratio by oxidative phosphorylation to then stimulate the cascade of biochemical reactions leading to insulin secretion.…”

“…In fact, FLIM on NAD(P)H signal is a label-free approach used to interrogate cell metabolism which exploits the discrimination of the free and protein-bound forms of NAD(P)H molecules based on their difference in auto-fluorescence lifetime 52,53 . In previous works by some of us, the method was used to test β-cell responses to glucose stimulation both in INS-1E cells 40 and in human-derived islets 54 . Here NAD(P)H lifetime imaging was performed in maintenance conditions (glucose concentration of 11 mM), and under pulsed glucose stimulation (shift in glucose concentration from 2.2 mM to 16.7 mM), both in control and cytokine-treated cells.…”

Machine-Learning-guided recognition of α and β cells from label-free infrared micrographs of living human islets of Langerhans

“…To overcome current limitations, here we implemented a machine-learning-based approach for the recognition of α and β cells directly from labelfree infrared micrographs of living and intact Langerhans islets derived from humans. It exploits the label-free microscopy dataset recently generated by some of us in the effort to study the metabolic response of human islets to glucose stimulation 19 . Data consist of micrographs of live-islet auto uorescence stimulated at 740 nm by multiphoton excitation and measured both in intensity and lifetime in the 420-460-nm optical window, which is dominated by NAD(P)H emission and lipofuscin signals.…”

Phasor Identifier: A Cloud-based Analysis of Phasor-FLIM Data on Python Notebooks

“…Fluorescence Lifetime Imaging Microscopy (FLIM) has revolutionized the field of biological sciences by enabling the investigation and quantification of a multitude of molecular quantities and processes: exemplary applications encompass the measurement of intracellular parameters (e.g. metabolism [1][2][3][4][5][6][7], temperature [8][9][10][11], viscosity [12-2 Materials and Methods…”

“…To analyze FLIM data, lifetime decay curves obtained from microscopy experiments are typically fitted with a variety of mathematical models. While some commercial closed-source software packages (such as those provided by Becker & Hickl, PicoQuant or Leica) and a few open-source options [6–8] were developed to this aim, the analysis of large datasets in a quantitative, easy, fast, and interactive manner remains challenging. In this regard, the phasor approach to FLIM data emerged as a valuable analytical tool.…”