This paper describes the use of a new, fast optical imager for the macroscopic photoluminescence lifetime imaging of long decay emitting samples. The integration, image acquisition, and analysis procedures are described, along with the preparation and characterization of the sensor materials for the imaging and the application of the imager in studying biological samples.
This paper presents a new photoluminescence lifetime imager designed to map the molecular oxygen (O2) concentration in different phosphorescent samples ranging from solid-state, O2-sensitive coatings to live animal tissue samples stained with soluble O2-sensitive probes. In particular, the nanoparticle-based near-infrared probe NanO2-IR, which is excitable with a 625 nm light-emitting diode (LED) and emits at 760 nm, was used. The imaging system is based on the Timepix3 camera (Tpx3Cam) and the opto-mechanical adaptor, which also houses an image intensifier. O2 phosphorescence lifetime imaging microscopy (PLIM) is commonly required for various studies, but current platforms have limitations in their accuracy, general flexibility, and usability.
The system presented here is a fast and highly sensitive imager, which is built on an integrated optical sensor and readout chip module, Tpx3Cam. It is shown to produce high-intensity phosphorescence signals and stable lifetime values from surface-stained intestinal tissue samples or intraluminally stained fragments of the large intestine and allows the detailed mapping of tissue O2 levels in about 20 s or less. Initial experiments on the imaging of hypoxia in grafted tumors in unconscious animals are also presented. We also describe how the imager can be re-configured for use with O2-sensitive materials based on Pt-porphyrin dyes using a 390 nm LED for the excitation and a bandpass 650 nm filter for emission. Overall, the PLIM imager was found to produce accurate quantitative measurements of lifetime values for the probes used and respective two-dimensional maps of the O2 concentration. It is also useful for the metabolic imaging of ex vivo tissue models and live animals.
O2 is one of the key environmental parameters for living systems, and knowledge of the distribution of O2 and its dynamics is important for many biological studies1,2,3. The assessment of tissue oxygenation by means of phosphorescent probes4,5,6,7,8 and PLIM9,10,11,12,13 are gaining popularity in biological and medical research3,9,14,15,16,17,18,19. This is because PLIM, unlike fluorescence or phosphorescence intensity measurements, is not affected by external factors such as probe concentration, photobleaching, excitation intensity, optical alignment, scattering, and autofluorescence.
However, current O2 PLIM platforms are limited by their sensitivity, image acquisition speed, accuracy, and general usability. Time-correlated single photon counting (TCSPC), combined with a raster scanning procedure, is frequently used in PLIM and fluorescence lifetime imaging microscopy (FLIM) devices20,21,22. However, since PLIM requires a long pixel dwell time (in the millisecond range), the time of image acquisition is much longer than what is required for FLIM applications20,22,23. Other techniques, such as gated CCD/CMOS cameras, lack single photon sensitivity and have low frame rates20,24,25,26. Moreover, the existing PLIM systems are mostly used in the microscopic format, while macroscopic systems are less common27.
The TCSPC-based PLIM macro imager28 was set up to overcome many of these limitations. The design of the imager was greatly facilitated by the use of a new opto-mechanical adapter, Cricket, which has the following: i) two C-mount adapters, which provide easy coupling of the camera module on the back side and objective lens on the front side; ii) an internal housing for an image intensifier and a power socket for the latter on the outer side of the Cricket; iii) an internal space behind the front-side C-mount adapter where a standard 25 mm emission filter can be housed in front of the intensifier; and iv) a built-in light collimating optics with ring regulators, which allow optical alignment/focusing between the lens and the camera to produce crisp images on the camera chip.
In the assembled imager, the camera module is coupled to the back side of the Cricket adapter, which also houses an image intensifier consisting of a photocathode followed by a microchannel plate (MCP), an amplifier, and a fast scintillator, P47 phosphor. A 760 nm ± 50 nm emission filter is fitted inside the Cricket, and an objective lens, NMV-50M11'', is attached to the front side C-mount adapter. Finally, the lens and the camera are aligned optically with ring regulators.
The role of the intensifier is to detect incoming photons and convert them into fast bursts of light on the camera chip, which are registered and used to generate emission decays and lifetime images. The camera module comprises an advanced TCSPC-based optical sensor array (256 pixels x 256 pixels) and a new generation readout chip29,30,31,32,33, which allow the simultaneous recording of the time of arrival (TOA) and the time over threshold (TOT) of photon bursts at each pixel of the imaging chip with a time resolution of 1.6 ns and an 80 Mpixel/s readout rate.
In this configuration, the camera with the intensifier has single-photon sensitivity. It is data-driven and based on the speedy pixel detector readout (SPIDR) system34. The spatial resolution of the imager was previously characterized with planar phosphorescent O2 sensors and a resolution plate mask. The instrument response function (IRF) was measured by the imaging of a planar fluorescent sensor under the same settings as used for all the other measurements. The lifetime of the dye of around 2.6 ns was short enough for it to be used for the IRF measurement in PLIM mode. The imager can image objects of up to 18 mm x 18 mm in size with spatial and temporal resolutions of 39.4 µm and 30.6 ns (full width at half-maximum), respectively28.
The following protocols describe the assembly of the macro imager and its subsequent use for mapping the O2 concentration in biological samples stained with the previously characterized near-infrared O2 probe, NanO2-IR35. The probe is a bright, photostable, cell-permeable O2-sensing probe based on platinum (II) benzoporphyrin (PtBP) dye. It is excitable at 625 nm, emits at 760 nm, and provides a robust optical response to O2 in the physiological range (0%-21% or 0-210 µM of O2). The imager is also demonstrated to characterize different sensor materials based on Pt(II)-porphyrin dyes. Overall, the imager is compact and flexible, similar to a common photographic camera. In the current setup, the imager is appropriate for different widefield PLIM applications. Substituting the LED with a fast laser source will further improve the performance of the imager and could potentially enable nanosecond FLIM applications.
All the procedures with animals were performed under authorizations issued by the Health Products Regulatory Authority (HPRA, Ireland) in accordance with the European Communities Council Directive (2010/63/EU) and were approved by the Animal Experimentation Ethics Committee of the University College Cork.
1. Sample preparation
2. PLIM imaging setup
3. Image acquisition
4. Data analysis
For ex vivo imaging applications, fragments of intestinal tissues were stained by the topical application of the NanO2-IR probe on the serosal side of the tissue. For deeper staining, 1 µL of the probe was injected into the lumen. In the latter case, the 0.2-0.25 mm thick intestinal wall shielded the probe from the camera. The two staining processes are demonstrated in Figure 2A.
The resulting intensity and PLIM images are presented in Figure 2B–G. The colors clearly reflect the difference in lifetime values and, hence, the difference in oxygenation of the serosal and mucosal sides of the tissue. Figure 2C and Figure 2D refer to similar sets of tissues that were stained topically (Figure 2C) and intraluminally (Figure 2D). As expected, the tissues showed similar PLIM patterns. However, the intraluminal staining of the mucosal side of the tissue (Figure 2D) showed higher lifetime values, reflecting lower oxygenation in the inner surface of the intestine, compared to the topical staining of the serosal side (Figure 2C), which showed lower lifetimes. This reflects higher oxygenation in the outer surface of the intestine.
To examine the stability of the lifetime signals and respiratory activity of the intestine, a time-lapse PLIM analysis was conducted on the intraluminally stained samples over 2 h (Figure 2D-G). An increase in the lifetime values was observed between 15 min (54.4 µs ± 0.9 µs) and 2 h (61.1 µs ± 0.8 µs ) of the incubation, which reflects a reduction in O2 levels in the luminal part of the intestine. In addition, the low O2 levels after 2 h of incubation prove that tissue respiration continued for the whole period of dissection, preparation, staining, incubation, and imaging steps. Additionally, the change in the shape of the lumen observed in Figure 2D-G can be attributed to the distribution of the probe injected into the lumen, meaning the LT signal reflects the area where the probe is located.
To assess the future performance of the imager in vivo, an initial study on the imaging of hypoxia in grafted subcutaneous tumors grown in mice was carried out immediately after sacrifice (Figure 3). In this case, Balb/c mice were injected subcutaneously in the right flank with 200 µL of a mixture of CT26 cells, which were unstained or stained with a 0.05 mg/mL NanO2-IR probe. PLIM was performed on the seventh day of the tumor growth, and the mice were sacrificed just before the imaging (i.e., post-mortem). Images of the tumor areas in the mice on the seventh day of growth are shown in Figure 3A,B.
The intensity (left) and PLIM (right) images for the tumor areas in the mice on the seventh day of tumor growth (Figure 3B) demonstrate the performance and sensitivity of both the imager and the probe in such experiments (more detailed statistical data not shown). The animals were imaged for 20 s each. The left flanks of the mice were shaved and imaged as a blank. They produced no PLIM signals. In contrast, regions with the tumor and the probe produced stable lifetime signals of ~50 µs. The aim of this experiment was to evaluate the usability of the imager for in vivo studies with live animals and models of human disease. While this part showed preliminary qualitative results, a corresponding paper is under preparation, which provides a more thorough quantitative assessment.
Next, the imager was also demonstrated in the PLIM imaging of sensor materials based on Pt(II)-porphyrin dyes. The chemical structure of Pt-octaethylporphine dye (PtOEP) and technical details of the preparation of the corresponding sensor materials are described elsewhere6,39. The PtOEP-polystyrene-based phosphorescent solid-state sensor coatings deposited on a transparent polyester film (Mylar) as small spots were imaged by submerging the film in PBS buffer (air-saturated oxygenated state) or in PBS containing glucose and glucose oxidase, which provides a fully deoxygenated condition. The lifetime signals obtained in the oxygenated (25.6 µs ± 0.5 µs) and deoxygenated conditions (65.7 µs ± 1.5 µs) matched the results reported previously6.The intensity and PLIM images of the oxygenated and deoxygenated sensors are given in Figure 4A, B.
Figure 1: Experimental setup of the imager38. Reprinted with permission from Sen et al.38. Copyright: The Optical Society. Please click here to view a larger version of this figure.
Figure 2: O2 PLIM of the topically and intraluminally stained mouse intestine samples. (A) Illustration of the two staining methods using (B) phosphorescence intensity and (C) PLIM images of the intestine topically stained with the NanO2-IR probe. (D–G) Timelapse PLIM of the intraluminally stained intestine at 15 min, 30 min, 1 h, and 2 h after staining. The PLIM images are presented in a pseudocolor scale: the blue color corresponds to a lower lifetime, and the red color corresponds to higher lifetime values. Scale bars = 5 mm. Abbreviation: PLIM = phosphorescence lifetime imaging microscopy. Please click here to view a larger version of this figure.
Figure 3: O2 PLIM of grafted tumors in the right flanks of the mice. (A) Graphical illustration of tumor development in live animals. Phosphorescence intensity (left) and PLIM (right) images of tumor areas on the (B) seventh day of tumor growth. The PLIM image is presented in a pseudocolor scale: the blue color corresponds to a lower lifetime, and the red color corresponds to higher lifetime values. Scale bar = 5 mm. Abbreviation: PLIM = phosphorescence lifetime imaging microscopy. Please click here to view a larger version of this figure.
Figure 4: Phosphorescence intensity of the PtOEP-based sensor spot. Phosphorescence intensity (left) and PLIM (right) images of the PtOEP sensor spot in the (A) oxygenated and (B) deoxygenated states. The PLIM images are presented in a pseudocolor scale: the blue corresponds to a lower lifetime, and the red color corresponds to higher lifetime values. Abbreviations: PLIM = phosphorescence lifetime imaging microscopy; PtOEP = platinum octaethylporphyrin. Please click here to view a larger version of this figure.
The above protocols give a detailed description of the assembly of the new imager and its operation in the microsecond FLIM/PLIM mode. The TCSPC-based new generation Tpx3Cam camera, coupled by means of the opto-mechanical adaptor Cricket with the image intensifier, emission filter, and macro-lens, produces a stable, compact, and flexible optical module that is easy to operate. The imager was shown to perform well with a range of different samples and analytical tasks, which included the characterization of phosphorescent materials and live tissue O2 imaging. The near-infrared Pt(II)-benzoporphyrin-based cell-permeable soluble probe NanO2-IR and the red-emitting Pt(II)-porphyrin-based solid-state sensors were used in the imaging experiments. These sensors have been successfully used for the quenched phosphorescence detection of O2 due to their large Stokes’ shift, long emission lifetime, high sensitivity, fast response, and good photochemical stability.
These different O2-sensitive probes and solid-state coatings, when tested on the macro imager, produced results that are in agreement with previously published data40. The materials demonstrated good uniformity, contrast, and lifetime responses to changing environmental conditions, particularly in terms of temperature and oxygenation state. The comparison of these results with those produced on the confocal TCSPC-PLIM microscope shows that the new imager is accurate, has better-quality lifetime readings, and acquires PLIM images with high speed and sensitivity28,38,40.
The imager also demonstrated promising performance with phosphorescently stained biological samples of different types. Thus, detailed O2 concentration maps were produced for the samples containing suspensions of stained respiring cells, live postmortem animal tissue, whole organs, and grafted tumors. The imager has provided measurements of lifetime values from the surface and even at depths of up to 0.5 mm inside the tissue28,36,38.
Since phosphorescence lifetime values are strongly influenced by temperature41, it is necessary to implement tight temperature control of the imaged samples, such as by using a heated sample stage and/or an incubator chamber. While using UV excitation, it is important to carefully choose the sample holders, as many materials possess autofluorescence in this spectral region, which affects the true lifetime value of the sample. All the PLIM imaging in this study was performed at an LED power of 4 V and a pulse width of 50 ns for an integration time of 20 s; however, these parameters can be adjusted as required. Approximately 104,500 frames are recorded during the 20 s of image acquisition time. Lastly, it is important to perform the measurements in a dark chamber to avoid ambient light interference.
Thus, the PLIM imager can perform fast, quantitative lifetime measurements with macroscopic objects of significant size. While laser-scanning PLIM-TCSPS is possible, it would be too slow due to the long dwell times required for the oxygen-sensing dyes. This imager, however, is fast and can record all the pixels simultaneously. Further, intensity-based measurements can be influenced by the fluorescent/phosphorescent dye concentration, an unstable LED or laser intensities, photobleaching, the alignment of the optical components, and scattering from the samples. In contrast, the TCSPC-based macro imager is essentially independent of these factors. Therefore, it can perform accurate and calibration-free measurements of O2. The disadvantages of the camera include its relatively complex data processing and large data sizes (~2 Gb).
In the future, the imager can also be used not only to map O2 concentrations but also other chemical and biochemical parameters of tested samples, such as the pH and glucose dynamics, using the corresponding probes or sensors. This makes it a useful tool for physiological studies with various tissue and disease models. The imager can also be upgraded to operate in the nanosecond FLIM mode. However, this requires the replacement of the current LEDs with a faster excitation source (e.g., a ps laser). System operation in the dual PLIM/FLIM mode is also principally possible.
Altogether, the imager demonstrates simplicity and versatility, along with good functionality and operational performance in widefield TCSPC-PLIM applications. Commercially available wide-field imagers mostly perform intensity-based imaging, in which the phosphorescent intensities can be influenced by factors like photobleaching, measurement geometry, scattering from samples, etc. The current imager is based on lifetime imaging, which makes the measurements rather more quantitative than qualitative. Moreover, the integration of the Tpx3Cam optical camera with the Cricket adaptor and image intensifier provides single-photon sensitivity, simplicity, flexibility, and robustness of the measurements. Unlike other systems, it can be carried to the site of the experiments and rotated 360° based on the requirements of the sample and measurement task. With its current objective, which is easy to change for another lens, the imager can measure samples up to 18 mm x 18 mm in size in a matter of few seconds and with a high spatial resolution of 256 pixels x 256 pixels. This paper describes step by step procedures for how to prepare samples for testing and use the imager in the different PLIM applications.
The authors have nothing to disclose.
Financial support for this work from the Science Foundation Ireland, grants SFI/12/RC/2276_P2, SFI/17/RC-PhD/3484 and 18/SP/3522, and Breakthrough Cancer Research (Precision Oncology Ireland) is gratefully acknowledged.
627 nm LED | Parts Express | Can be replaced with different LED based on the excitation wavelength of the sensor. Used 390 nm LED for Pt-porphyrin dyes. | |
760 ± 50 nm emission filter | Edmund Optics | 84-788 | Can be replaced with different filter based on the emission wavelength of the sensor. Used 650 ± 50 nm bandpass filter for Pt-porphyrin dyes. |
Balb/c mice | Envigo, UK | Balb/c | |
Black box | Thorlabs | XE25C9/M | |
Cricket Adapter | Photonis | Cricket-2 | |
CT26 cells | ATCC | CT26.WT | https://www.atcc.org/products/crl-2638 |
DMEM | Sigma-Aldrich | D0697 | Other media can also be used |
ImageJ Software | ImageJ | Free Image analysis software. Can be downloaded from: https://imagej.nih.gov/ij/index.html | |
MCP-125 image intensifier with P47 phosphor screen | Photonis | PP0360EF | |
Mini dishes | Sarstedt | 83.3900.300 | 35 mm diameter |
Mylar plastic film, 75 micron | RS Ireland | 785-0795 | Othe plastic substrates can also be used |
NanO2-IR | home-made | n/a | The probe can be synthesised according to the published method 'Tsytsarev V, Arakawa H, Borisov S, Pumbo E, Erzurumlu RS, Papkovsky DB. In vivo imaging of brain metabolism activity using a phosphorescent oxygen-sensitive probe. J Neurosci Methods. 2013 Jun 15;216(2):146-51. doi: 10.1016/j.jneumeth.2013.04.005. Epub 2013 Apr 25. PMID: 23624034; PMCID: PMC3719178.' or provided by our lab. |
NMV-50M11” 50 mm lens | Navitar | Other lenses compatibel with C-mount adators can be used | |
Optical breadboard | Thorlabs | MB1836 | |
Petri Dishes | Sarstedt | 82.1472.001 | 92 mm diameter |
Power Supply | Tenma | 72-10495 | |
Pulse Generator | Tenma | TGP110 | |
Sophy | Amsterdam Scientific Instruments | n/z | Provided by ASI together with the Tpx3Cam |
Tpx3Cam | Amsterdam Scientific Instruments | TPXCAM | |
Tri2 Software | University of Oxford | n/a | Free Time Resolved Imaging software, can be downloaded from: https://users.ox.ac.uk/~atdgroup/index.shtml |
XYZ Translation Stage | Thorlabs | LT3 |