How to Build a Time-Gated Luminescence Microscope

Dayong Jin; Yiqing Lu; Robert C Leif; Sean Yang; Megha Rajendran; Lawrence W Miller

doi:10.1002/0471142956.cy0222s67

. Author manuscript; available in PMC: 2026 Oct 2.

Published in final edited form as: Curr Protoc Cytom. 2014 Jan 2;67:2.22.1–2.22.36. doi: 10.1002/0471142956.cy0222s67

How to Build a Time-Gated Luminescence Microscope

Dayong Jin ¹, Yiqing Lu ¹, Robert C Leif ², Sean Yang ², Megha Rajendran ³, Lawrence W Miller ³

PMCID: PMC13628505 NIHMSID: NIHMS2212738 PMID: 24510771

Abstract

The sensitivity of filter-based fluorescence microscopy techniques is limited by autofluorescence background. Time-gated detection is a practical way to suppress autofluorescence, enabling higher contrast and improved sensitivity. In the past few years, three groups of authors have demonstrated independent approaches to build robust versions of time-gated luminescence microscopes. Three detailed, step-by-step protocols are provided here for modifying standard fluorescent microscopes to permit imaging time-gated luminescence.

Keywords: lanthanide, time-gated luminescence, microscopy, autofluorescence

INTRODUCTION

In cytometry, fluorescence is the predominant modality for imaging cellular or sub-cellular targets (Chan and Nie, 1998; Campbell et al., 2002). The most common cytometry technique applied to biological tissues and cells is fluorescence microscopy, which has been driven by the availability of target-specific spectral probes. For example, revolutionary work in genetically encoded fluorescent proteins (GFPs) was recognized by the Nobel Chemistry Prize in 2008. Higher spatial resolution is provided by multimodality laser scanning confocal microscopy (Peng et al., 2012) or “super-resolution” techniques (Huang et al., 2008; Enderlein, 2012; Liu et al., 2012). Analyzing a large population of cells, however, has demanded other approaches such as automated cell analysis by scanning cytometry (Kamentsky and Kamentsky, 1991; Bae et al., 2012), flow cytometry (Shapiro, 2003; Perfetto et al., 2004; Chattopadhyay et al., 2006; Grégori et al., 2011; Robinson et al., 2012), and automated imaging flow cytometry (Mirabelli et al., 2012). Most fluorescence probes, including organic dyes, fluorescent proteins, and nanoparticles, are sensitive enough to reveal sub-cellular information. However, biological substances are naturally autofluorescent under excitation by UV or visible light. Nonspecific, autofluorescent background can substantially increase the detection threshold, which will decrease the dynamic range of probe signals seen within cells or tissue specimens.

Short-lived autofluorescence from raw biological samples can effectively be excluded by pulsed excitation and time-delayed gated detection based on lanthanide probes (mainly Eu³⁺ and Tb³⁺ complexes), which display long-lived (msec-scale) luminescent emissions (Hemmilá and Mukkala, 2001; Yuan and Wang, 2006; Eliseeva and Bunzli, 2010; Rajapakse et al., 2010). The advantages of lanthanide probes have been demonstrated for high-contrast microscopic imaging (Hemmilá and Mukkala, 2001; Selvin, 2002; Petoud et al., 2003; Bunzli and Piguet, 2005; Yuan and Wang, 2006; Yung and Ponce, 2008; Rajapakse et al., 2010; Jin, 2011a,b; Jin and Piper, 2011). Absence of background also enables rapid scanning of single cells with both flow cytometry (Jin et al., 2007a,b, 2009; Leif et al., 2009) and scanning cytometry (Lu et al., 2011; J. Lu et al., 2012; Y. Lu et al., 2012) instruments and provides improved sensitivity and specificity. The recent development of highly luminescent lanthanide complexes (Petoud et al., 2003; Weibel et al., 2004; Deiters et al., 2009), methods to enhance lanthanide luminescence (Leif et al., 2006b), responsive lanthanide-based luminescent probes (Song et al., 2006; Thibon and Pierre, 2009a,b), functionalized lanthanide-ion nanocomposites (Makhluf et al., 2008), and nano-encapsulation of lanthanide-containing biolabels (Harma et al., 2001; Wu et al., 2008, 2009; Song et al., 2009) further enhance the potential of lanthanide-based cellular imaging. Efficient lanthanide complexes and protein-conjugated complexes are also commercially available from vendors such as Sigma-Aldrich and the Invitrogen division of Life Technologies (Jin, 2011b). The rapid progress in lanthanide probes and luminescence bio-assay and bio-imaging techniques has been described in recent, comprehensive reviews (Jin, 2011a).

The principle of time-gated luminescence (TgL) detection for cytometry or microscopy is straightforward. TgL detection requires target analytes to be labeled with a luminescent probe with an emission lifetime >10 μsec, more than 1000 times longer than the autofluorescence lifetimes (<0.01 μsec) typically associated with non-target organisms and organic or inorganic detritus of biological specimens. Referring to Figure 2.22.1, following pulsed excitation, autofluorescence from non-target material fades rapidly (within 0.1 μsec), while target luminescence is sustained over hundreds of msec. While background noise is blocked from detection throughout the period of excitation and for a short time delay (a few μsec) after excitation ceases, luminescence signals can be subsequently observed during the signal-collection window. To achieve the maximum TgL signal, it is necessary to optimize the duty cycle for the pulsed excitation, time delay, and gated detection. Because of the long lifetimes of lanthanide probes (on an msec scale), the temporal resolution required of TgL instrumentation is on the order of microseconds, and this can be easily achieved with a variety of hardware configurations.

Figure 2.22.1 — Schematic illustrating the principle of time-gated luminescence detection using a long-lived luminescence lanthanide complex as biolabel (T is time, τ is lifetime, and I is intensity). Reproduced from Jin and Piper (2011).

This unit describes the steps to set up a robust time-gated luminescence microscope (TgLM) according to three typical design concepts, which have been demonstrated by three groups independently. Accordingly, Basic Protocol 1 (by D. Jin and Y. Lu) describes TgL detection engineered via a modified mechanical chopper unit; Basic Protocol 2 (by R.C. Leif and S. Yang) describes an approach to modify a CCD camera for TgL detection, and Basic Protocol 3 (by L.W. Miller and M. Rajendran) illustrates a TgL detection system assembled from all off-the-shelf commercial products including a gated, intensified CCD camera (ICCD). The three basic protocols share a common excitation strategy in which pulsed, ultraviolet (UV) excitation is provided by commercial light-emitting diodes (LEDs). Basic Protocols 1 and 2 also provide construction details for building a UV LED illuminator for a microscope. A comparison specification sheet is provided as Table 2.22.1, so that users can choose one approach to meet different laboratory conditions and assay requirements. For the sake of practicality, this unit emphasizes the balance between TgLM performance, cost, complexity, and compatibility with existing fluorescence microscopes. The authors provide detailed information on the components they have used to build their microscope systems. Alternative components can be used if they have the same or better specifications than the ones used by the authors.

Table 2.22.1.

Comparison of Specifications for Time-Gated Luminescence Microscopes

Characteristic	Chopper (Basic Protocol 1)	Analog summation (Basic Protocol 2)	Image intensifier (Basic Protocol 3)
Minimum measurable lifetime	10 msec	100 msec and longer (depends on the switching time of the CCD)	Several nsec (depends on the switching speed of the image-intensifier system)
Capacity to change settings	By changing the chopper blades	Change pulse generator settings	Change electronics settings
Setup complexity	Requires dedicated optical-alignment skills	Moderate	Quicker
Image colors	True color	Monochrome	Monochrome
Cost	Very low	Manufacturing costs would be less than the present components of a conventional fluorescence microscope	Expensive
Disadvantages	Fixed ratio of on-off period to one chopper blade. Cannot measure fluorescence lifetimes and includes custom parts.	Only works with very limited number of cameras. Needs an internal amplifier. Cannot measure fluorescence lifetimes. Includes custom parts.	Limited red and near-infrared quantum efficiency and resolution

Open in a new tab

BASIC PROTOCOL 1 CONSTRUCTION AND USE OF A MECHANICAL CHOPPER TgL UNIT

Fundamentally, a TgLM consists of a pulsed illumination (excitation) source, a time-delayed gated detector, and a signal-synchronization unit. Accordingly, in this protocol, we first present steps for producing a robust excitation optics, which consists of a modified filter cube and an electronics controller box to drive an LED for rapid pulsed excitation; we then describe the steps for mechanical modification of a commercial chopper and assembly of a fast time-gating unit, which can be attached to a camera port of a commercial fluorescence microscope. At the end of this protocol, we describe steps to synchronize the pulsed excitation and gated detection. Refer to Jin (2011b) and Jin and Piper (2011).

Pulsing the UV LEDs for excitation

The development of TgLM imaging has matured over the last decade because of the availability of high-power UV LEDs. UV LEDs provide compact, low-cost, and simplified design for pulsed excitation compared to the bulky nitrogen laser (337 nm, which only has a 40-Hz repetition rate and a nsec-scale pulse length), the expensive solid-state frequency-tripled Nd:YAG laser (355 nm; up to 100 kHz), and the low-efficiency xenon flash-lamps (UV emission <1% of overall power emission; and low repetition rate of 100 Hz to 1 kHz with a pulse length of approximately 2 μsec). The small size of the UV LED permits an excitation filter and a condenser lens to be built within a fluorescence microscopy filter cube or to be mounted directly into the optical path, which is compatible with existing fluorescence microscopes.

Optical chopper unit

Time-gated detection has also been facilitated by the rapidly growing semiconductor industry, with the development of fast and highly sensitive cameras and detectors. The technology that enables time gating has evolved from mechanical chopper or ferro-electric shutter to gated CCDs and gated-intensifier CCDs. Mechanical choppers are relatively slow and less stable than electronic gating options, but they are low cost. This basic protocol describes a smart approach to accelerate the chopping speed as well as optimize TgLM signal collection efficiency. By engineering an economical chopper-based time-gating unit, we removed a major barrier to producing a low-cost TgL microscope allowing direct visual (naked-eye) observation of true-color TGL images in real time, which would enable practical utilization of TgLM to identify target analytes free of background in routine laboratory diagnostics (Vereb et al., 1998).

This protocol, as discussed below, has significant advantages including true-color imaging, suitable naked-eye inspection of the samples in background-free condition, and low cost, as well as compatibility with all commercial microscopy systems and cameras; however, it requires a significant number of customized mechanical/electrical components, which are not commercially available off-the-shelf items. The Advanced Cytometry Labs at Macquarie University maintains a research service to provide these items in Australia. Once the components are obtained, assembly, installation, optical alignment, and pulse synchronization will become relatively easy to carry out according to the following protocol steps.

Materials

The basic components for modification of a filter cube for the pulsed excitation design used in this protocol are shown in Figure 2.22.2:

A high-power UV LED, such as #NCCU033A, commercially available from Nichia Corp. at a cost of less than $100 (http://www.nichia.co.jp/en/product/uvled.html)
A UV excitation filter, such as a colored glass band-pass filter U330; 0.5-in. diameter) available from Edmund Optics (http://www.edmundoptics.com/)
A standard dichroic filter, such as 400dclp filter, >400-nm pass, from (http://www.chroma.com) within a fluorescence microscopy filter cube, such as a standard Olympus fluorescence filter cube used in this work
An aspherical lens, such as #C110TME from Thorlabs (http://www.thorlabs.com/)
This modification only requires one customized mechanical lens holder, as shown in Figure 2.22.2C. The holder can be machined from plastic or aluminum (in this work at cost of ~ $200) with internal screws for fine tuning the aspherical lens position.

The basic components for construction of a gated mechanical chopper unit are shown in Figure 2.22.3:

An enclosed version of high-speed chopper (C995 Optical Chopper) commercially available from, Terahertz Technologies (http://www.terahertztechnologies.com/) at a cost of ~$1000
A condenser lens (30 mm EFL, shown in Fig. 2.22.3B) available from Edmund Optics (http://www.edmundoptics.com/)
A standard eyepiece (×10) from commercial microscopes, such as Olympus microscopes
Other mechanical holders (shown in Fig. 2.22.3), customized by the Advanced Cytometry Labs at Macquarie University, are necessary to minimize the chopper vibration and maximize the optical alignment accuracy. These aluminum mechanical holders and adaptors are usually machined under a precisely programmed cutting machine, with cost at ~$4000 each job.
Stanford Research Signal Generator DG535, or other low-cost signal generator

Pulsed UV excitation

The 365-nm LEDs at power level of >250 mW in continuous wave (CW) mode are good enough to excite the most widely used europium complexes and some terbium complexes. Manufacturers have been improving the optical power of these LEDs, from 100 mW in 2004 to 250 mW, 350 mW, and, now, 450 mW in 2013! In fact, the optical power is not a critical issue anymore for the TgLM application, since even the 100-mW LED, after beam shaping within the microscope, can deliver optical power as high as 5 to 10 mW on the specimen. Instead, there are three additional key factors relating to these high-power UV LEDs that deserve extra attention (described as steps 2, 3, and 5) for the TgLM application.

The concept for a compact excitation design in this protocol is to pack the LED excitation optics consisting of the LED, an excitation filter, and the beam-collimating lens, into a commercial fluorescence filter cube, as shown in Figure 2.22.2.

1
Make an LED driver circuit to provide sufficient current while maintaining its rapid switching capability.

The LED operates in an efficient pulsed mode with a peak current of as high as 500 mA to 1 A. The critical question is how to switch the high current off rapidly in order to switch off the UV emission, which is essential for ensuring the efficient operation of the TgLM. With switching capability of <1 μsec, the example circuit design has been previously reported by the authors (reproduced here as Fig. 2.22.4 for further reference; Jin et al., 2007b). To control the LED pulser circuit, a TTL signal generator (such as Stanford Research DG535 unit with four channels) is needed.
2
Mount the LED chip on a heat-dissipating plate, which is a bare printed circuit board without solder mask, as shown in Figure 2.22.2A.

The high-power UV LEDs typically consume a lot of power at 500 mA to 1 A (bias voltage of 4 V) to emit high optical power, which requires that the LED be mounted on a heat sink.
3
Place a UV filter (such as an absorption-type Schott glass filter) in front of the LED to block its visible emissions, as shown in Figure 2.22.2. See Critical Parameters and troubleshooting for further information.
4
Use a high-NA condenser lens to shape the LED emission into a collimated beam.

The LED emits light at a large solid angle (>45°), which should be collimated before being reflected by the dichroic mirror to the microscope objective. An additional lens holder, the aluminum lens holder suggested in Figure 2.22.2A, should be machined with screw threads for fine tuning the lens focus.
5
Adjust the distance of the condenser to defocus the UV LED emission for a more uniform field of illumination at focus.

This fine tuning is achieved by turning the aspherical lens inside the customized aluminum lens holder. Alternatively, diffuser optics can be used.

If the defocusing step is omitted, the array of UV LED emitters generates an uneven illumination pattern, as shown in Figure 2.22.5. Defocusing or diffuser optics will drop the excitation power; however, these UV LEDs emit more than enough excitation, as discussed earlier.

Figure 2.22.4 — LED injection current generator circuit provides the LED with substantial current up to 2 A with rapid termination of the driving current pulse. Capacitor values are in μF and resistor values are in Ω. Three common diodes were connected in series, using their forward voltage drops to regulate the voltage difference between the base terminals of the two transistors. The UV LED was connected in series to the two parallel resistors R6 and R7 (equivalent to 1 Ω resistance), permitting the oscilloscope to monitor the LED injection current level (optional). Reproduced from (Jin et al., 2007b).

Figure 2.22.5 — The high-power UV LEDs consist of arrays of emitters, which generate an uneven illumination pattern on the specimen. (A) the latest-generation UV LED chip, 350 mW. (B) The first generation of UV LED, 100 mW. (C) The far-field projection of a UV LED image on a piece of paper.

Fast-gating chopper unit

The idea and prototype demonstration of optimizing a mechanical chopper suitable for time-gated detection were reported by the authors of this unit in 2011 (Jin, 2011b; Jin and Piper, 2011). The assembly of the time-gated luminescence chopper unit is displayed in Figure 2.22.3. Since mechanical choppers vibrate during operation, mounting the mechanical chopper onto a microscope requires some mechanical engineering for better stabilization purpose.

6
Select and/or modify the chopper blade for a desired duty ratio.

Commercial chopper blades have a typical duty ratio of 50%, which is not necessary for our application. Here, we here use the laser micromachining technique to remove every second outer-layer blade to produce a duty ratio of 1:3, so that at maximum speed there is a 100-μsec blocking time to prevent the pulsed excitation, and 300 μsec for time-delayed luminescence signal collection. The chopper manufacturer (Terahertz Technologies) provides such a modification service at a cost of ~$200 for each job. Care should be taken during blade modification to ideally keep the blade very flat, to obtain the best chopping performance during high-speed rotation.
7
Use a condenser lens (triplet lens assembly) with a tunable screwed holder to bring the chopper blade exactly into focus.

The beam size should be as small as possible. See Critical Parameters and Troubleshooting for more information.

The alignment can be done simply by putting a piece of paper on the blade and watching the focused transmission light from the microscope while tuning the condenser lens.
8
Attach a pinhole plate (with 1-mm-diameter pinhole) as shown in Fig. 2.22.3C) to the chopper enclosure.

This pinhole aperture will remove stray light and further increase the chopping accuracy and stability.

Ideally, this pinhole plate should be made of metal, and the hole should be tapered from outside towards the chopper blade so that a thicker plate can be used without blocking the emission light. The pinhole plate also reduces chopper noise and dust access to the chopper and internal optics.
9
Align the pinhole plate.

This can be done by adjusting the position of the chopper enclosure (the black cover plate), to which the pinhole plate is attached. Originally, there are four holes on its corners for fixing the position using screws. The holes can be enlarged by an extra 0.5 mm in diameter, so that it will provide additional flexibility to adjust the position of the cover plate, which can still be fixed by the same screws. The alignment can be simply done while transmission light is projected onto a piece of paper behind the pinhole.
10
Place an eyepiece after the chopper to project the gated image to a camera or the naked eye.

The longer the distance from the chopper to the eyepiece, the smaller the field of view and the larger the magnification. Therefore, this distance can be adjusted so that the camera images the entire field of view at highest magnification. Note that it is dependent on the effective sensing area of the camera chip.

Synchronization

6
Feed the chopper TTL signal train into the external trigger of a TTL signal delay generator (one channel is sufficient), such as a Stanford Research Signal Generator DG535, or other low-cost signal generator.

The pulse width (on time) should be 80 μsec or less to match the period that a blade can completely block the light.

We typically set the chopper duty ratio to 1:3 (by removing every second blade). There are a total of 15 blades left. If the chopper runs at its full speed, the chopping frequency should be around 2.5 kHz. This could be calculated as 100 μsec chopper on (light blocked) and 300 μsec chopper off (light transmitted). However, due to the sunrise-sunset effect (see Critical Parameters and Troubleshooting), there is typically an ~11 μsec rising/falling time between each switch, so that setting the LED at 80 μsec or less should guarantee that no light can pass through the chopper blade to reach the camera during the excitation period.
7
Use the signal generator to generate a phase-delayed TTL signal train to trigger the LED driver.
8
Choose an appropriate delay through the signal generator.

How does one find the best time delay? A simple experiment is to place a piece of paper on the objective. Since UV excites blue autofluorescence from paper, by watching the detector while changing the delays, until no fluorescence is observed behind the chopper (camera), the pulsed excitation should be observed behind the chopper blade. To completely avoid the excitation light passing through to the detector, we usually set the LED excitation pulse at less than 80 μsec to avoid the ~11-μsec chopping edges (sunrise-sunset efforts on chopping the 1-mm beam).

BASIC PROTOCOL 2 CONSTRUCTION AND USE OF CAMERA WITH ANALOG SIGNAL SUMMATION

Leif and his collaborators, in a series of papers, have developed inexpensive instrumentation that makes use of lanthanide luminescence, affordable for both research and clinical laboratories, as well as potentially useful for screening and/or diagnosis in less developed countries. These papers described the use of Nichia UV LEDs (Leif et al., 2006a) and demonstrated the previous digital summation technique (Leif et al., 2003, 2006a), which was to digitize each frame and add the result to the sum of the previous frames. In order to avoid numeric overflow, the previously used standard interline CCD camera was modified to increase the size of the internal data storage from 16 to 32 bits. This digital summation technique has a protracted acquisition time. The time required to obtain 2,500 images at a rate of approximately 10 frames per sec was 250 sec (Leif et al., 2006b). At this rate, the actual acquisition time for 250 frames was 250/10 (25) sec, or 50 times the sum of acquisition (1 msec) and signal local storage time (1 msec), which is 2 msec or a total of 0.5 sec. Digital summation of the frames also introduced a horizontal undulating background. These studies with digital summation motivated the search for another summation method. Analog image summation with a modified interline CCD camera was found to have much less noise and to have a much shorter acquisition time (Leif and Yang, 2010; Leif et al., 2012). Previously, analog summation had been used with a frame-transfer CCD that was optimized for nanosecond-scale lifetime measurements (Mitchell et al., 2002a). This CCD was replaced with an interline CCD (Mitchell et al., 2002b, 2007).

Materials

Europium FireRed 5- and 0.5-μm microspheres (Newport Instruments, part nos. EuFR5UM and EuFR0_5UM, respectively; http://www.NewportInstruments.com)

DNA-Check beads, 10 μm diameter (Beckman Coulter, part no. 6603488, https://www.beckmancoulter.com/)

The Nichia UV LED is the same as that described in Basic Protocol 1. As is shown in Figure 2.22.6A, a quartz condenser can be used to focus an LED that is substituted for the excitation lamp of a Leica fluorescent microscope. The UV LED is positioned close to the back of a LINOS condenser, which is attached to the excitation entrance of a fluorescence microscope (Leif et al., 2006a). The mounting system described below can be used to mount any small light weight object that needs to be centered including a LED or an optical fiber. Nichia can supply a heat sink attached to the LED for free.

A Leitz MPV II fluorescence epi-illuminated microscope or equivalent equipped with a 10× 0.25 NA, a 40× 0.65 NA, and an infinity-corrected objective high-ultraviolet-transmission UPL Fluorite 60 oil, NA 1.25, with aperture (Olympus part no. 1UB532), was employed to observe and to electronically photograph the beads

The UV fluorescence was excited at 365 nm with a similar UV LED from the same manufacturer, Nichia, that was employed Basic Protocol 1

The emitted light was observed through an Omega Optical (http://www.omegafilters.com/) PloemoPak cube UV DAPI, equipped with a 365-nm narrow-band-width excitation filter (Omega 365HT25) and a 400-nm beam-splitter (Omega 400DCLP02). The CCD optical path was optionally equipped with either a 619 nm narrow-band emission filter (Omega 618.6NB5.6) or a standard DAPI 450 nm emission filter (Omega 450DF65).

Qioptiq LINOS (http://www.qioptiq-shop.com/) Combi. Condenser (16/21.4 mm; part G063011000)

The parts for construction of the Condenser Mount and LED Mount were available at the precision instrument shop and the local hardware store

The basic components are shown in Figures 2.22.6 and 2.22.7. The Laserlab supply was used to power the UV LED in pulsed and continuous mode. 1-msec wide pulses were delivered at up to 500 Hz. The images were triggered by the trailing edge of the pulses and had an exposure time of 1 msec or longer, as specified. The square-wave pulse train was provided by the Global Specialties Instruments signal generator. The Leitz fluorescent microscope is equipped with 365-nm pulsed excitation from a Nichia 230mW LED (Leif et al., 2006a; Jin et al., 2007b). The Atik 314L+ camera was cooled (ΔT = −27°C).

Figure 2.22.7 — Rear view of time-gated microscope. Starting from the top: An Atik 314L+ camera is held by a C-mount in a slider assembly that holds the emission filters. The slider assembly is mounted in the camera light exit of the microscope head. An aluminum tube is held by a bayonet mount similar to the one shown at the bottom of the microscope. The LED position assembly has a 3-dimensional positioning system. The x and y positioning is based on the design of a typical 3-point condenser mount.

Laserlab LED power supply (http://www.laserlab.com/); the custom-built unit specified in Basic Protocol 1, or any other that will follow a function generator with at least a frequency range of 0 to 100 kHz and provide the peak power suggested by the manufacturer, should suffice.

Signal generator (Global Specialties Instruments model 4001; http://globalspecialties.com/): this unit is inexpensive but still adequate. One other alternative is a board in a PC that can serve as a function generator and possibly an oscilloscope. The caveat, as always, is to make sure that the software is both adequate and easy to use.

Oscilloscope: any dual trace should be suitable or as stated above a PC board should be considered

Camera: Artemis Atik 314L+ (http://www.artemisccd.com/artemis-ccd-fs-range.html). The Atik 314L+ can be cooled (ΔT = −27°C} and employs a 1392 × 1040–pixel Sony ICX-285AL chip (pixel size of 6.45 × 6.45 μm²). This camera has a readout noise of 4 electrons and is equipped with a 16-bit ADC and USB 2.0 interface. It should be purchased with a C-mount adapter or as needed to fit the emission filter assembly. The Atik 314L+ camera has been succeeded by the FS14.

Emission filter wheel unit. The use of a commercial one is recommended, such as the Atik Electronic Filter Wheel 2, which is software and mechanically compatible with the. Atik 314L and its successor.

Artemis Capture software (http://www.artemisccd.com/)

Condenser mount

1
Machine a cylindrical aluminum adapter tube (Fig. 2.22.6A) to have an outside diameter (O.D.) of 3.6 cm that slip fits into the excitation entrance of the epi-illumination microscope head (Fig. 2.22.6A), and an inside diameter (I.D.) of 2.5 cm that provides a slip fit for the condenser. Tap a hole for a set metal screw with a knob into the aluminum adapter tube. Tap a pair of threaded holes into the excitation entrance and corresponding attachment holes into the adapter tube.

This secures the aluminum adapter tube to the excitation entrance.
2
Remove and invert the epi-illumination microscope head and place it upon a flat surface. Slip the aluminum adapter tube into the excitation entrance and tighten the two attachment screws.
3
Slip the Linos condenser (Fig. 2.22.6B) into the adapter tube and tighten the knob on the adapter tube to secure the condenser.

LED mount

4
Purchase the UV LED with a heat sink that was attached by the manufacturer.
5
Glue the LED heat sink to the flat head of an approximately 0.96-cm bolt. Thread the inside of a brass pipe, length 30 cm, outside diameter (O.D.) 1.32 cm, and inside diameter (I.D.) 0.86 cm (Fig. 2.22.6B), to accept the bolt.

Three-dimensional LED positioner

4
Examine the mount that holds the arc lamp housing. Attempt to purchase a mount that will attach to the arc lamp housing. Only if this fails, either precisely measure the arc lamp housing mount and have a copy made in a precision instrument shop, or, as a last resort, remove the mount from the microscope and replace it with a new mount and its mating unit combination.
5
Acquaint yourself with the design of the X and Y position part of the standard illumination condenser.

This design is used to position the brass pipe.
6
Have an aluminum tube with length ~7.6 cm, O.D. 7.6 cm, and I.D 5.8 cm fitted to match the mount that will attach it to the back of the microscope.

After machining, the wall of the tube should be about 0.5 cm thick.
7
Drill and thread in the aluminum tube three holes at 120° relative to one another. Attach this aluminum tube to the mount and then attach the unit to the back of the microscope.

Two knob-headed screws that are located at 120° relative to one another are threaded in the aluminum outer tube.
8
Construct an inner black plastic sleeve with length 10.2 cm, O.D. 4.8 cm, and I.D ~1.4 cm to slip fit the brass pipe. Notch the sleeve at its center with a V cut.

When inserted in the aluminum tube, the black sleeve should have ~0.5 cm clearance on both sides.
9
Insert the black plastic sleeve into the aluminum tube and tighten the knobbed screws to pass their tips through the aluminum tube and then into the V cut. Tighten the spring-loaded long throw spring pin, which is also located at 120° relative to the knob-headed screws, to contact the V groove. Seat the two knob-headed screws in the V cut. Approximately center the plastic sleeve in the aluminum tube by adjusting the knob-headed screws.
10
Slide the brass pipe into the hole in the center of the plastic sleeve.

Depending on the state of the microscope, the bolt with the attached LED heat sink can be screwed into the brass pipe, and then the assembly can be slid through the body of the microscope from either side, or in the event that the microscope is intact, the threaded end of the pipe can be slid through the body of the microscope; then, the bolt that is attached to the LED heat sink can be threaded into the brass pipe and tightened by rotating the pipe.
11
Position the LED by gently pushing or pulling brass pipe.

This controls the light focus on the first lens of the condenser or possibly at another place.
12
Remove an objective and place a business card on the microscope stage.

Since the UV excitation results in the card fluorescing, the LED can be centered by adjusting the two knob-headed crews and focused by sliding the brass pipe. The depth of the spring-loaded pin should be adjusted to be about half compressed. The instructions in Basic Protocol 3 are a very good place to start the optical alignment.

Interfacing the Atik camera to the microscope

The manufacturer-modified Atik 314L+ can operate in multiple modes, which permits changes to the operation of the CCD chip. The Sony progressive scan CCD in the Atik 314L+ camera has two different types of pixels (http://www.sony.net/Products/SC-HP/datasheet/01/index.html; then go to ICX285AL). The light-sensitive pixels convert photons into electrons, which can be transferred to the storage pixels. The storage pixels are the ones that are part of the CCD transferring system that sequentially delivers the charge packets to the ADC. Normally, after each time the light-sensitive pixels have been illuminated, the charge from each light-sensitive pixel is transferred to the corresponding storage pixel. The storage pixels are then sequentially read out to the ADC. The two-dimensional array is then transferred to the computer. This is basically the way the Atik 314L+ normally operates. The essential difference for the modified Atik camera is that the sequential delivery of the charge packets to the ADC does not occur after each image has been acquired; it only occurs after the last image has been completely acquired. This permits the combination of analog integration in the storage pixel of the output of its associated light sensing pixel together with the use of an electronic shutter, which limits the acquisition of light by the light-sensing pixel to the period when the light is off. Since, as shown in Figure 2.22.8, the acquired image is read out only once, the digitized image includes the noise from one readout.

Figure 2.22.8 — Chart describing analog charge summation. (A) The UV light is controlled by the square wave produced by the signal generator. The break after the eighth pulse indicates that multiple pulses have occurred between the two sections of the pulse train. (B) Shows the state of a CCD photosensitive pixel. The amount of light absorbed is shown in red. (C) Shows the state of the corresponding light insensitive (integrating) pixel (blue filled), which is part of the storage array, which is read out. Only after the last pulse (100 msec in this figure) is the summation of the pulses in the storage pixels sequentially read out and digitized to form the image. The number of images (pulses) is shown as being set for 50. Reproduced from Leif et al. (2012). For the color version of this figure, go to http://www.currentprotocols.com/protocol/cy0222.

Imaging modes with this modified configuration include:

Mode 1: Normal mode, which takes normal snap shots at set exposures (modification wires not connected)
Mode 5: Time-gated mode, which takes multiple analog summed images of time-gated images. Both modification wires are connected in parallel to pulse generator and send current to the LED and CCD. The CCD electronic shutter opens when voltage drops from pulse generator. The switching time of the camera provides the time delay. The software can also set the time delay. However, the switching time is sufficient to eliminate background fluorescence.

As in normal exposure mode, the integration time is controlled by the combination of the electronic shutter and the transfer of charge from the photosensitive pixels to the photo-insensitive storage pixels. In time-gated mode (Mode 5), the difference is that charges from many short exposures are accumulated in the storage pixels before digitization. Figure 2.22.8A shows the modulation of the UV light, which is identical to that of the pulse generator. As shown in Figure 2.22.8B, the falling edge of the light pulse square wave (light to dark transition) triggers the electronic shutter, which clears the photosensitive pixels of any signal acquired during the LEDs on phase. The signal acquired by the photosensitive pixels during the light-off period is transferred to the photo-insensitive pixels when triggered by the positive going edge of the signal generator (dark to light transition). In time-gated mode (Mode 5), the difference is that the photo-insensitive pixels are neither read out nor digitized until the end of the acquisition of the image. The number of repetitions is controlled by the frequency of the light pulse and the software parameter defining the total time over which the image is collected. The readout and digitization phases are the same as the normal exposure mode. The advantages of analog integration in the light-insensitive pixels over addition of a series of digitized images is that the final image only contains the read noise contribution of a single read as opposed to multiple camera reads. Also, analog addition is much quicker, which allows the camera to spend more time collecting photons as opposed to sending images to a computer for storage or digital summation by a microprocessor system within the camera. This mode of image acquisition is called by the IIDC 1394 Specification 11 “multiple shutter pulse width mode” or Mode 5 (1394 Trade Association, 2004).

TIFF monochrome files were directly produced by the modified Atik 314L+ camera. Three types of images were obtained. The single images were obtained in conventional mode; the time-gated analog summed images of the beads and their background control images were obtained in the electronically time-gated mode. The emission light was blocked for the background control images. The number of single images that were analog summed was varied.

4
Attach the filter changer to the camera tube of the microscope head.

Note that this location was chosen only for simplicity. Location of the filter changer beneath the microscope head is preferred because the emission filters are then available in a place which permits visual observation.
5
Attach the C-mount of the camera to the C-mount of the filter changer. If necessary, insert the appropriate extension tubes or lenses.

Connecting the electronics

4
As shown in Figure 2.22.9, connect the signal generator to the inputs of the camera, LED power supply, and oscilloscope.

The wires attached to the camera shown in Figure 2.22.7 are the ones that are connected. The length of the wires to the camera should be minimized.
5
Connect the LED power supply to the LED.

Figure 2.22.9 — Electronic connections. The abbreviations T and V on the signal generator knobs stand for time and voltage, respectively. The signal generator is connected to the LED power supply, the oscilloscope, and the camera. The LED power supply is also connected to the oscilloscope. All connections include a ground.

Running the camera software

4
Turn on the computer and log in.
5
Plug in USB and AC cords to connect the CCD by the USB cable to the computer.

This runs the camera including its cooling fan. There should be air current from the fan; if not, the cord might not be plugged into the wall outlet.
6
Run Artemis Capture (by selecting Programs/ArtemisCCD/ArtemisCapture or the icon for the program, illustrated in Fig. 2.22.10).

The Artemis Capture screen is shown in Figure 2.22.11.
7
Software will open window (shown in Fig. 2.22.11) and start default cooling of the sensor.
8
If the two smaller windows are not shown, click on View/exposure settings and Cooler settings.
9
Allow CCD to cool down for a couple of minutes before taking long exposure images to increase signal-to-noise ratio (S/N).

Figure 2.22.10 — Artemis Capture software icon.

Figure 2.22.11 — Captured screen from Artemis Capture program.

Image collection

4
With visible light on and sample on stage, find and focus the object through eyepieces.
5
Switch to camera by pushing the pin in on the right of the eyepiece manifold.
6
Set the signal generator to 4 V.

This can be D.C. or a symmetrical square-wave with a 1-msec on and off cycle. For time-gated luminescence, the square wave must be used; for conventional (visible), images, either the time-gated or D.C. setting can be used. The exposure with the square-wave for a conventional image will take twice as long as the D.C. image. Depending on the lifetime of the lumiphore, it may be useful to shorten or lengthen the period of the square wave.
7
If new or modified settings are to be used, test the system with the combination of a lanthanide standard, such as lanthanide complex–labeled beads and a conventional fluorochrome labeled beads (an example of this is described in Anticipated Results).
8
Using the Artemis software, adjust the exposure settings (0.1 sec) and set a 5 × 5–in. subframe by clicking on Camera/subframe (or the corresponding icon, illustrated in Fig. 2.22.12).
9
To start the live view, activate the loop function by clicking on Camera/loop (or the corresponding icon, illustrated in Fig. 2.22.13).
10
This will allow much faster refresh rate to aid in focusing on the object on the slide.
11
After object is found, use mode 1 (Snap) or mode 5 (Gated; with two external wires connected) to take the images.

These modes operate the same way.
12
Set the binning desired (1 × 1 or 2 × 2, etc.).
13
Set the exposure time desired.
14
Deactivate the subframe and loop function to take full-view images by either clicking the icons or going to the menu to uncheck the functions Camera/Subframe and Camera/Loop.
15
Take the image by clicking on the Snap function, Camera/Snap or its corresponding icon, illustrated in Figure 2.22.14 (or use Control + N).

Software will take the image and, depending on the time setting, the image will appear on screen after the exposure period. There will be a count down on the lower right corner of the window (live update of what is going on).
16
Save the image by clicking on File/Save the Image As …, or its corresponding icon, illustrated in Figure 2.22.15.
17
Set the desired directory, type in the filename, and click the Save button.
18
Repeat, if desired, with different exposure times and lighting or filters or in a different mode.

ImageJ was used to process the analog summed image files. For the 0.5-μm bead images (see Anticipated Results), the image was thresholded (880–3758) to show virtually all of the beads. Some of the touching particles in the resulting binary image after conversion to 8 bits were then split by employing the binary image watershed method. This was followed by a binary open. A mask was then created by employing the Analyze.Analyze Particles method (size 6 to 60, d circularity 0.0 to 1.0, and Bare Outlines). The image calculator method was then used to multiply the 8-bit versions of the mask and the inverted image to produce a 32-bit real result, which was then saved as an 8-bit image.

Figure 2.22.12 — Subframe icon from Artemis Capture program.

Figure 2.22.13 — Loop icon from Artemis Capture program.

Figure 2.22.14 — Snap icon from Artemis Capture program.

Figure 2.22.15 — Save icon from Artemis Capture program.

SUPPORT PROTOCOL 1 BEAD SAMPLE PREPARATION

The capacity of the time-gated luminescence system developed in Basic Protocol 2 was tested with commercially available beads. FireRed beads (Newport Instruments) were created for this purpose.

Materials

DNA-Check beads (Beckman Coulter, part no. 6603488)

FireRed 5- and 0.5-μm beads that contain Eu(TTFA)₃ (Newport Instruments; http://www.newportinstruments.com/)

HERMLE Z-180-M microcentrifuge (http://www.hermle-labortechnik.de/english/circulate/circulate.html)

Branson Ultrasonifier model 450 (http://www.sonifier.com) with micro-tip

Centrifugal Cytology Buckets (Newport Instruments)

Plain frosted end glass slide (Fisher Scientific, cat. no. 12-550-15)

Beckman GPR centrifuge (http://www:beckmancoulter.com)

Glass coverslips

Transfer 0.2 ml of DNA-Check beads to a 1.5-ml microcentrifuge tube and pellet by microcentrifugation for 10 sec at 14,000 rpm.
Subsequently, aspirate the supernatant.
Resuspend the beads with 0.5 ml distilled water.
Add 30 μl of the FireRed 5-μm beads (Leif et al., 2009) and sonicate at 10% amplitude for 10 sec with the Branson Ultrasonifier model 450 equipped with a micro-tip.
Assemble Centrifugal Cytology Buckets (Leif, 2002) with 12-well inserts and add 20, 40, and 60 μl of the bead suspension to the respective buckets.
Sediment the beads on to a plain frosted end glass slide in a Beckman GPR centrifuge for 5 min at 300 × g, room temperature. Aspirate the supernatant and air dry the 6-mm circular bead spots. Then, apply a glass coverslip to the bead containing area and secure in place with nail polish at the four corners and let dry.

For some of the studies performed with the Atik 314L+ and the 0.5 FireRed beads, after sedimentation, the beads were coated with a transparent plastic solution consisting of Zeonor 330R, Zeon Chemicals. (http://www.zeonex.com) diluted with n-octane (Leif, 2013), which was optically coupled to a coverslip by glycerol. Glycerol was also used as the immersion fluid for the objective.

BASIC PROTOCOL 3 PREPARATION AND USE OF TIME-GATABLE INTENSIFIED CCD CAMERA

Miller (Galhlaut and Miller, 2010) adapted a conventional epi-fluorescence microscope for time-gated imaging by using only commercially available components including a collimated UV LED, a pulse generator, and an ICCD. The instrument is capable of rapidly (in 1 to 4 sec) acquiring images of dim lanthanide specimens including 40-nm nanospheres (containing ~400 europium complex molecules; Gahlaut and Miller, 2010), terbium complexes in the cytoplasm of living mammalian cells (cellular concentration, 1 to 10 μM; Rajapakse et al., 2010; Mohandessi et al., 2012; Rajapakse and Miller, 2012), and long-lifetime (20 to 200 μsec) non-metal luminescent probes (Vaasa et al., 2012). The ability to quickly acquire high-contrast images from minimal numbers of emitting molecules is critical for live-cell imaging, as overly long exposure times increase phototoxicity and limit temporal resolution. Under such imaging conditions, only a few tens of photons per pixel are acquired in a single camera frame (Gahlaut and Miller, 2010). Therefore, the gain function of the intensifier is critical for producing images with adequate contrast and signal-to-noise ratio. The benefits, then, of using a gated ICCD are commercial availability of all components, relative simplicity of setup and operation, and high sensitivity for low-light imaging; these benefits are offset by higher equipment costs.

Materials

The required components and their relevant capabilities for ICCD-based, time-gated luminescence imaging are listed in Table 2.22.2. The system reported by Miller and Gahlaut (2010) that serves as the example here was built around a Zeiss Axiovert 200 microscope, but any inverted or upright wide-field microscope will suffice (Gahlaut and Miller, 2010). As excitation illumination is delivered through the objective, a lens with adequate UV transmission is necessary. Here, a 63× objective with ~50% transmittance at 350 nm was used (EC Plan Neofluar, 63× 1.25 N.A., Carl Zeiss). A collimated, UV LED (e.g., Mic-LED-365, Prizmatix, Ltd., http://www.prizmatix.com) emitting at 365 nm is required for excitation. The LED must be capable of external, TTL-mediated switching with μsec-scale rise/fall times, must have adjustable output (up to ~50 mW at the exit window), and must be fittable to the epi-illumination port of the microscope. The Prizmatix LED is one of several commercially available units that meet these requirements. A pulse generator with at least two output channels (TTL) and programmable burst mode operation (e.g., DG645, Stanford Research Systems) is needed to switch the LED and synchronize its operation with the intensifier component of the camera.

Table 2.22.2.

Necessary Hardware Components and Relevant Specifications for Building a Time-Gated Microscope Using an Intensified CCD Camera with Pulsed UV LED Excitation

Component	Manufacturer/model no.^a	Specifications
Epi-fluorescence microscope	Carl Zeiss, Inc. Axiovert 200	High N.A. (≥1.0) objective lens with ≥50% transmittance at 350 nm (e.g., EC Plan Neofluar, 63×, 1.25 NA)
Collimated UV LED (365 nm emission)	Prizmatix, Ltd. MIC-LED-365	External on/off modulation (TTL) with ~ 1 μsec rise/fall time Adjustable output to at least ~50 mW at exit window
Digital pulse generator	Stanford Research Systems, Inc. DG645	At least 2 independent delay outputs Programmable burst mode capability Externally triggered start (TTL)
Intensified CCD camera	Stanford Photonics, Inc. MEGA-10EX	GenIII photocathode, peak Q.E. = ~0.4 External triggering (TTL) of intensifier On-chip integration Adjustable intensifier gain voltage

Open in a new tab

Components listed here are for system originally described in (Gahlaut and Miller, 2010).

With respect to the type of intensified CCD camera, there is some latitude in selecting components. Firstly, it is possible to separately purchase an intensifier and lens-couple it to a CCD. This option is more cost effective if one already has a suitable CCD camera; however, lower resolution, lower coupling efficiency, and greater complexity in hardware integration may be expected in comparison to a fiber-optically coupled, integrated ICCD platform. Whether lens coupling or fiber coupling is chosen, the type of photocathode incorporated into the intensifier has the greatest impact on system performance. GenIII photocathodes have peak quantum efficiencies of 40% to 50% with sensitivity ranging from near-UV to IR, depending on the particular material (GaAs or GaAsP; with or without ion barrier film). Miller employed an integrated ICCD system (Mega-10EX, Stanford Photonics) that consists of a filmless, GaAsP photocathode (quantum efficiency = ~0.4 from 450 to 600 nm) fiber optically coupled to a Sony XX285 CCD sensor. The CCD sensor is a 1380 × 1024 array of 6.47 μm pixels, and the effective pixel size of the intensifier/camera is 10.35 μm. The CCD can be read out at a full-frame rate of 15 frames-per-sec.

This protocol assumes computer control of the ICCD system, and proprietary software packages are available from most camera manufacturers (e.g., Piper, v2.4.05, Stanford Photonics). At minimum, a computer and software are required that allow the user to control CCD frame length, and that can generate an external TTL trigger pulse synchronous with the beginning of image acquisition. Additionally, the camera system must allow for external TTL gating of the intensifier and on-chip integration.

Hardware configuration

1
Mount the LED so that its collimated output enters the epi-illumination light path of the microscope.

In the system built by Miller, a dual-lamp housing adapter (Zeiss, cat. no. 447230) is installed that makes it possible to switch between the LED and a mercury lamp for conventional fluorescence excitation.
2
Once mounted, align the LED output for Köhler illumination so that it fills the back aperture of the objective lens.

In this way, a reasonably uniform field of illumination is achieved at the focus.
3
Adjust the LED output so that the excitation light intensity at the image plane equals ~0.2 to 1.0 W/cm².

Light intensity at the sample may be calculated by measuring the steady-state power at the back aperture of the objective using a light meter and by accounting for the objective magnification, field number, and transmittance at the excitation wavelength, according to the method of Grunwald et al. (2008). For example, with the EC Plan Neofluar, 63 ×, 1.25 NA objective, a measured power of 1.33 mW at the objective back aperture yields an estimated illumination intensity of ~0.5 W/cm² at the specimen plane.
4
Attach the ICCD to an appropriate camera port.

Typically, cameras use a C-mount coupling, but an adaptor may be required depending on the camera or microscope used.
5
Once the LED and camera are physically mounted on the microscope, electronically connect the LED, pulse generator, and camera.

The exact configuration of these connections will necessarily be hardware dependent; here, connection of the aforementioned Mega-10EX camera system, DG645 delay generator, and MIC-LED-365 light source is described (Fig. 2.22.16).
6
First, link the capture card of the PC to the External TTL input of the delay generator. Next, connect the TTL output of one of the delay generator channels to the auxiliary gate input of the Mega-10EX camera controller.

Note that the Mega-10EX system comprises three physical components: the camera itself, a separate camera controller that is used to select and modulate intensifier gain, and a PC with a Windows operating system running Piper Control software. Manufacturer-supplied connectors link these three components. The camera runs on its own internal clock set to a predetermined frame rate (15 frames/sec, in this case), and the camera sends the frame clock and digital image data to the frame-grabber card in the computer. A link between the camera controller and camera allows for external modulation of the gain voltage and timing of the intensifier. In order to control the intensifier operation via the delay generator, two additional connections must be made.

When properly configured, the delay generator mediates synchronization of the LED and the image intensifier. Piper Control software synchronizes the camera frame clock with an external TTL pulse routed to the delay generator to begin the image acquisition sequence.

The delay generator can be programmed to generate a burst of TTL output signals directed to the LED and the intensifier (routed through the camera controller). This configuration allows the user to define the LED pulse width (T) and pulse period (T’), the intensifier gate delay (Δt), the intensifier gate width (T₀), and number of excitation/emission cycles that occur in a single camera frame (Fig. 2.22.16). The emission signal from multiple cycles is integrated on the CCD and read out to the image capture card of the computer at the end of the frame. The frame length can be varied as multiples of the camera frame clock, from 1 clock (66.7 msec) up to a maximum of 30 clocks (2 sec). The camera control software allows for either summation or averaging of an arbitrary number of frames, and generates images in Tagged Image File (TIF) format.

Figure 2.22.16 — Electronic connections and signal-timing parameters for time-gated imaging with an intensified CCD camera and a pulsed LED light source. (A) Schematic showing electronic connections between an ICCD (Stanford Photonics, Inc., Mega-10EX), a PC running camera control software (Stanford Photonics, Inc., Piper, v2.4.05), a digital pulse generator (Stanford Research Systems, Inc., DG645), and a UV LED (not shown, Prizmatix, Ltd., Mic-LED-365). Thick lines represent manufacturer’s camera connections that mediate camera control and data flow. Thin arrows represent TTL connections to/from the pulse generator. (B) Sample pulse generator timing parameters and their effects on LED and intensifier operation.

Programming the pulse generator

7
Program the pulse generator to achieve the desired LED pulse width, gate delay, and intensifier gate width.

A good approximation is to set the LED pulse width and the intensifier-gate width equal to one another (T = T₀) at a value that is on the order of the emission lifetime of the sample. The gate delay must be at least 2 μsec to allow for complete turn-off of the LED before the intensifier is gated on. For example, the delay parameters used by Miller and co-workers to image luminescent terbium complexes in live cells (emission lifetime = ~2000 μsec; Gahlaut and Miller, 2010; Rajapakse et al., 2010; Mohandessi et al., 2012; Rajapakse and Miller, 2012) are shown in Figure 2.22.16. In this case, the delay generator is programmed to generate a burst of 22 output pulses (TTL, +4.0 V) 100 μsec after receiving an external TTL trigger pulse from the camera computer at the start of CCD integration. The AB output channel (rise, t = 0 μsec; fall, t = 1500 μsec) is sent to the LED, and the CD output channel (rise, t = 1510 μsec; fall, t = 2990 μsec) is sent to the external gate input of the intensifier, gating each component on for approximately 1500 μsec (T = T₀ = ~1500 μsec). The time between the end of the AB pulse and the beginning of the CD pulse determines the gate delay (here, Δt = 10 μsec).

See Critical Parameters and Troubleshooting for additional information.

Image acquisition

Here, a procedure is given for imaging bubbles of an aqueous solution of a terbium complex, Lumi4-Tb, suspended in mineral oil (see Support Protocol 2 for preparation). Lumi4-Tb is brightly luminescent (quantum yield = 55%), with appreciable absorbance at 365 nm (ε₃₆₅ = ~8400 M⁻¹cm⁻¹; Xu et al., 2011). The water-in-oil emulsion method can be used with any water-soluble lanthanide complex, and specimen brightness can be varied by changing complex concentration. Alternatively, commercially available europium microspheres (200 nm diameter, Life Technologies, cat. no. F20881) may be used as test specimens, as described in Gahlaut and Miller (2010).

7
Prepare a luminescent lanthanide sample (Support Protocol 2) for imaging and mount it on the microscope stage.
8
Once mounted on the stage, image in bright field (transmitted light) mode, bringing a bubble into focus in the center of the field of view.

The Mega-10EX camera allows for real-time display with automated control of intensifier gain; this detection mode should be used for bright-field imaging, as it protects the intensifier from inadvertent exposure to high light levels that may damage the photocathode.
9
Switch the microscope to fluorescence excitation mode with pulsed UV excitation. Select a filter set that is appropriate for the particular lanthanide specimen (terbium or europium). See Critical Parameters and Troubleshooting for more information
10
Dim all room lights and switch the ICCD to manual gain control mode using the camera controller. Ensure that the microscope optics are properly aligned so that only luminescence emission from the specimen is directed to the camera.

Direct exposure to excitation light or other bright light sources during operation of the ICCD in manual mode could result in irreversible damage to the photocathode.
11
Select an intensifier gain voltage using the camera controller.

For the Mega-10EX, a setting of 0.700 corresponds to ~780 V and a gain of ~5000. This is a modest gain level and is a good starting point for image optimization.
12
Using the camera control software, select a CCD integration time (frame length) and program the pulse generator to output the appropriate burst sequence (refer to step 2 and Fig. 2.22.16). Start with a short integration period (e.g., 1 clock or 66.7 msec) and increase as needed to achieve desired image brightness.
13
Lastly, choose a desired number of frames to be summed into a single image.

Frame summing increases image signal-to-noise ratio at the expense of longer acquisition times, and Piper control software can remove intensifier ion-feedback noise when 2 or more frames are summed. Frame summing generates composite images (.TIFF) with a pixel depth of 1024 (10-bit) per frame (e.g., with 4 frames, pixel depth = 4096, or 12-bit).

SUPPORT PROTOCOL 2 PREPARATION OF WATER/OIL EMULSION STANDARDS FOR MICROSCOPE CALIBRATION AND TESTING

This support protocol describes the preparation of water/oil emulsions that can be used to prepare luminescent standards for microscope optimization or calibration (see Basic Protocol 3). Aqueous droplets containing a luminescent species are formed in mineral oil, and an aliquot of the emulsion is sandwiched between a coverslip and a microscope slide. Such test specimens can be easily prepared in any laboratory, and sample brightness can be varied by changing the concentration of the luminescent solution. Moreover, the droplets range in size from 5 to 50 μm, and the lower end of this range corresponds to thickness of adherent, cultured mammalian cells (5 to 10 μm). Thus, the signal levels detected from water-in-oil bubbles reasonably approximate those that may be observed when imaging lanthanide luminescence diffusely distributed through the cytoplasm of cultured cells.

Materials

Tris base

NaCl

1 M HCl

Mineral oil (Sigma-Aldrich, cat. no. M5904)

Span 80 (Sigma-Aldrich, cat. no. 85548)

Tween 80 (Sigma-Aldrich, cat. no. P1754)

Lumi4 (Lumiphore, Inc., http://www.lumiphore.com/)

TbCl₃·6H₂O (Sigma-Aldrich, cat. no. 204560)

Valap (1:1:1 mixture of vaseline, lanolin, and paraffin)

Volumetric glassware (for buffer preparation)

pH meter (for buffer preparation)

Separatory funnel, 60 ml (e.g., Chemglass, cat no. CG-1743-07)

Glass reaction/storage vials, 20 ml (e.g., Chemglass, cat no. CG-4904-03)

Magnetic stir bar, 7 mm × 2 mm (e.g., Chemglass, cat no. CG-2003-14)

Bench-top magnetic stirrer

Double-sided adhesive tape

Microscope slides

Prepare water-in-oil emulsion

1
Prepare Tris-buffered saline (TBS, 50 mM Tris Cl, pH 7.6, 150 mM NaCl) by dissolving 6.05 g Tris base and 8.76 g NaCl in 800· ml of water. Adjust pH to 7.5 with 1 M HCl and adjust volume 1 liter with water.
2
Pre-extract mineral oil (25 ml) in a separatory funnel with an equal volume of TBS to remove acidic contaminants. Repeat three times.
3
Add Span 80 (450 μl, 447 mg) and Tween 80 (50 μl, 55 mg) to 9.5 ml (8 g) of pre-extracted mineral oil and vortex for ~30 sec.
4
Add 9 ml of pre-extracted mineral oil to 1 ml of above solution and vortex (~30 sec) to obtain a 0.45% (v/v) Span 80 and 0.05% (v/v) Tween 80 solution in mineral oil.
5
Prepare a stock solution (1 mM) of Lumi4 by dissolving the compound in an appropriate volume of TBS. Add ~1.1 equivalents of TbCl₃ (predissolved in TBS at a concentration of at least 10 mM). Vortex for ~30 sec and allow stock solution to rest at room temperature for ~30 min. Check for bright green luminescence by placing the vial containing stock solution under a hand-held UV lamp.

Once complexed with terbium, the Lumi4-Tb stock solution may be stored for months at 4°C.

Note that metal complexation procedures may differ for other types of terbium complexes or for europium complexes.
6
Dilute Lumi4-Tb to desired working concentration (e.g., 5 μM) in TBS.
7
Aliquot 1 ml of the 0.45% (v/v) Span 80 and 0.05% v/v Tween 80 mineral oil solution (from step 4) into a glass vial with a magnetic bar.
8
While stirring, add 5 μl of the working Lumi4-Tb solution (in TBS) to the oil solution. Continue stirring for 1 min.

Prepare slide mount

9
Place two pieces of double-sided tape perpendicular to the long axis of a standard microscope slide (1 in. × 3 in.) so that the edges of the tape are approximately 20 mm apart. Place two more pieces of tape on each edge of the slide to create a chamber.
10
Pipet 60 μl of the water-oil emulsion into the chamber formed by the tape on the slide.
11
Place a coverslip over the water/oil mixture and be sure to avoid forming large bubbles under the coverslip. Seal the edges of the coverslip with pre-warmed Valap.

COMMENTARY

Background Information

Comparison of instruments

The time-gated systems that we have described consist of two subsystems: (1) the light source and (2) the camera system. All three protocols use standard epifluorescence microscopes that are equipped with a UV LED light source. Basic Protocols 1 and 2 describe inexpensive systems consisting of components that are appropriate for commercial, clinical or environmental use. Basic Protocol 3 describes an expensive research system that should be sufficiently sensitive to make single photon measurements.

Comparison of LEDs:

Basic Protocols 1 and 2 use Nichia LEDs mounted on an integral Nichia heat sink with custom microscope mounts. Basic Protocol 3 describes an off-the-shelf item with a similar Nichia LED. Multi LED assemblies with a common power supply are becoming commercially available, such as one from Prizmatix (http://www.prizmatix.com/pdf/mic/Triple-LED.pdf).

Comparison of cameras:

Basic Protocol 1 allows naked-eye observation, as well as the use of any camera for image collection, and Basic Protocol 2 uses solid-state cameras, whereas Basic Protocol 3 uses an image intensifier attached to a camera. The red-sensitive photocathodes of image intensifiers tend to have increased noise compared to blue ones, and low quantum efficiencies in the far red and near infrared compared to solid-state cameras. Gated-intensified CCDs are relatively expensive, but depending on the wavelength of the light, often can be more sensitive, and presently are simpler in terms of synchronization than other means of time gating. All three protocols can use the other protocols’ light sources, and Basic Protocol 1 can use Basic Protocol 2’s camera, including a color version of the modified ATIK camera described in Basic Protocol 2. All three protocols could improve their signal-to-noise ratios by employing a camera that operates at lower temperatures. Basic Protocols 1 and 2 would also benefit from the use of an electron multiplier CCD camera and/or greater cooling; however, these changes would significantly increase their costs. Basic Protocol 3 obtains its signal amplification by use of an image intensifier. Many image intensifiers do not have major limitations on the types of cameras to which they can be coupled.

Camera sensitivity depends on: (1) the quantum efficiency, (2) the operating temperature, (3) the mode of operation, (4) the processing on the chip, and (5) the gain mechanism. The quantum efficiency at different wavelengths is affected by the chemistry of the photocathode of an image intensifier and of silicon chips. In the case of an image intensifier, many types of cathodes are available. Besides cost, there is a tradeoff in regard to using a red-sensitive noisy photocathode or a blue-sensitive low-noise photocathode. In the case of CCDs, besides changes in the chemistry, the intellectually simple means of reducing noise is cooling. There is an obvious decrease in noise even at a modest reduction of 25°C from ambient. This type of cooling and greater decreases in temperature are available for both regular CCD or CMOS chips and intensified cameras. The three sections of this protocol describe the effects of the mode of operation, particularly gating of the image detector. Presently, one of the most efficient means for amplifying the charge on each pixel is electron multiplication by the last group of light-insensitive CCD shift register elements.

The following significant benefits accrue from the use of multiple cameras: (1) the range of the emission wavelengths observed can be optimized for each camera; (2) the cameras can be independently focused and the images aligned by software, which eliminates the requirement for the use of complex optics that have been optimized to minimize chromatic aberration, which will decrease the cost of the optics, particularly the objectives, and improve the efficiency of light collection.

Monochrome or color detection:

Visually scanning the luminescent slides in conjunction with a synchronized inexpensive chopper is presently consistent with current research and clinical practice, which uses the human eye. The use of a chopper should also be appropriate for imaging with a time-delay integration (TDI) camera. Presently, there is a trade-off between the use of monochrome and color cameras. Commercial color cameras are less expensive than comparable research-grade monochrome cameras. However, monochrome research cameras are more sensitive than comparable commercial color cameras. Research cameras can have a longer integration period because of cooling and the use of chip designs that are optimized for lower noise. Monochrome cameras, because of their larger area per pixel, can collect more photons and have increased resolution compared to color cameras because all of the pixels are devoted to one emission. Simultaneous detection of multiple labels (colors) with monochrome cameras requires the use of multiple cameras and imaging paths. An inexpensive chopper can be combined with either a commercial color or a research cooled color camera, such as the Artemis FS14 (http://www.artemisccd.com/uploads/DOCS/fs14-jun-2013.pdf). The FS14t uses a Sony ICX285AL CCD as does the Atik 314L+ (http://www.sony.net/Products/SC-HP/datasheet/01/index.html), which should make it reasonably economical to detect background-free luminescence with a multicolor camera. The manufacturer has stated that the FS14 can also be modified for electronic gating. Image intensifiers presently can only use monochrome cameras. As of yet, an interline TDI camera for time-gated image acquisition is unavailable.

Cost and commercialization

Since the cost of a UV LED is considerably less than an arc lamp, and the Atik 314L+ costs about U.S. $2000, the instrumentation for time-delayed detection of lanthanide complexes is now less expensive than standard arc-based fluorescence microscopy and thus should be considered for use in less developed countries and as a point-of-care technology. Mass production of the rotor of Basic Protocol 1, which increases the number of available cameras, should also be cost effective; however, the rotor noise may be a problem unless a carefully engineered enclosure is used for acoustic protection. The combination of sensitive out-of-phase lanthanide luminescence, which is essentially noise-free, and in-phase DAPI DNA detection, should be of use in the detection of malaria and other pathogenic organisms. This combination, possibly together with a third marker, should permit the development of a rational molecular biology–based Papanicolaou monolayer dispersion test (Leif et al., 1976). It should be noted that automation of the Pap test was one of the major initial motivations for the development of cytometry. The availability of the practical image- intensified system (Basic Protocol 3) provides a powerful research tool for precise, low-noise measurements, with a very large range of fluorescence and luminescence lifetimes.

The present low-cost technical solutions (Basic Protocols 1 and 2), as well as the very powerful research solutions (Basic Protocol 3) to time-gated luminescent imaging, can be combined with multiple forms of lanthanide luminescent labels. These include organic and inorganic lanthanide–containing nanoparticles, densely labeled polymers (Leif et al., 2001; Leif and Vallarino, 2011), conventional functionalized single metal-organic complexes, and Resonance Energy Transfer Enhanced Luminescence (RETEL; Leif et al., 2006b,c). In short, the use of lanthanide, background-free labels is now both practical and desirable.

Critical Parameters and Troubleshooting

Basic Protocol 1

LED

The UV LEDs unfortunately also emit broadband visible emissions due to the semiconductor defects, a critical issue limiting their performance in TgLM, previously reported by us (Connally et al., 2006; Jin et al., 2006). What makes it even worse is that these visible emissions are long-lived emissions with lifetimes in the hundreds of microseconds range, adding substantial background noise to the TgLM. This makes the additional excitation UV filter essential to produce quality TgL images.

Chopper blade

To minimize the “sunrise-sunset” effect for a chopper blade blocking an optical beam, the beam size should be as small as possible. This is why our chopper should be used behind a condenser lens. The key to producing rapid time gating is to chop the focused beam on the far edge of a chopper blade. This is because the linear speed can be maximized on the chopper blade edge, since normally the maximum angular rotating speed for a mechanical chopper is around 150 to 170 cycles per sec. In our case, the chopper rotates at maximum rate of 167 cycles per sec, so that for a 4.2-cm-radius chopper blade, the fastest speed to fully switch a 1-mm beam on/off was tested as 11 μsec (Jin, 2011b; Jin and Piper, 2011), which is sufficient for TgL application.

Synchronization

Unlike the other two basic protocols, synchronization of pulsed LED excitation and gated chopper unit could be tricky, since the mechanical choppers never listen to an external trigger! Fortunately, commercial choppers have a build-in light detector to monitor the chopper blade in real time. The light detector can produce a synchronized TTL signal for triggering other devices. The trick here is to always use a synchronized signal from the chopper detector to trigger the LED to achieve highly stable synchronization.

Basic Protocol 3

Pulse generator

Note that the pulse period is determined by the “burst period” parameter (here, T′ = 3000 μsec), and this period must exceed the length of the longest delay event, in this case, the falling edge of the CD pulse at t = 2990 μsec. Further, note that the burst sequence must be completed within the duration of the CCD frame length, which is set using the camera control software. In this example, a burst of 22 excitation/emission events, each with a period T′ = 3000 (total time = 66 msec), occurs during a single frame of 66.7-msec duration (equivalent to 1 frame clock of the Mega-10EX camera). If the frame length is increased, then increase the burst count commensurately while holding other parameters constant (e.g., burst count = 44 for a frame length of 133.33 msec, or 2 frame clocks). For bright samples, it may be desirable to program a lower burst count to avoid saturating the ICCD. For example, with the above-described delay parameters, a burst count of 10 and a burst period of 6000 μsec would yield 10 events during a single, 66.7 msec frame.

Excitation filter

An excitation filter (e.g., 350 ± 25 nm) should be used to block long-lifetime, visible emission from the LED. Better signal-to-background ratios will be obtained by using a narrow-pass (20 nm bandwidth) emission filter centered on one of the lanthanide emission peaks (e.g., 495 nm or 545 nm for terbium, 615 nm for europium). Avoid the use of long-pass emission filters, as they do not adequately block nonspecific, long-lived background luminescence from microscope optics or coverslips, even though they will allow detection of a larger fraction of the lanthanide emission spectrum. In this example, a narrow-pass emission filter centered at 495 nm was used.

Anticipated Results

Basic Protocol 1

After the optimization of the mechanical/optical designs and mechanical engineering works, a detachable chopper unit can be produced as shown in Figure 2.22.17.

Figure 2.22.17 — Time-gated luminescence microscopes can be readily modified from an Olympus X71 inverted microscope of Olympus or the X51 upright microscope using a synchronized pulsed UV diode with a chopper. Reproduced from J. Lu et al. (2012).

Figure 2.22.18 shows the expected result using this configuration for true-color background-free imaging of europium complex–labeled Cryptosporidium cysts in environmental water concentrate.

Basic Protocol 2

Analog-to-digital conversion (ADC) noise

This test applies to any system where there could be any significant noise generated by one or more analog to digital conversions. As stated above, the previous digital integration, because of the 10 frames per second acquisition rate, was painfully slow. It also had the serious problem of a horizontal undulating background that was present in the experimental and control images. This is the test that led to the discovery of the source of the undulating background. In the data images, the intensity of this undulation could be reduced but not eliminated by background subtraction. As shown in Figure 2.22.19, this undulation was also present in images where the emitted light was blocked.

Figure 2.22.19 — Digitally summed background images obtained with the Retiga camera (http://www.qimaging.com/products/cameras/scientific/). The number of 1-msec exposures summed are shown at the top upper left. For all of these images, the number of readouts equals the number of exposures. All of the multi-readout images except for the 1000-exposure images were cropped to the area shown in the upper left of the 1000 exposure-readouts image. These background images show a periodic undulating horizontal pattern, the contrast of which increases with the number of images. The 1000 ×1 image shown at the lower right consists of 1000 light-pulse exposures, but only 1 readout; it only shows a few white defective pixels without any evidence of the periodic horizontal pattern. Reproduced from Leif and Yang (2010).

Fluorescence background test

This test checks the suppression of the fluorescence background. It applies to all lifetime gated systems.

Nano-bead test

This test permits the determination of the shortest permissible exposure time. A version of this test could be used to determine the optimum light and dark pulse periods. It requires a robust signal for noise measurement, such as that described in Figure 2.22.18D of Basic Protocol 1 and Table 2.22.3. It applies to all lifetime gated systems.

Table 2.22.3.

Exposure Statistics for TgLM Set Up by Basic Protocol 2^a

Identifier	Num. images summed (frames)	Mean	Std. dev.	CV in %	Mode	Signal/noise	Median	Min	Max
Beads	50	671	215.	32.2	601	2.17	624	389	3,969
	100	836	299.	35.8	745	2.62	769	465	5,441
	200	1,221	486.	39.8	1,065	3.64	1,109	679	9,049
	500	2,323	1,00	43.3	2,025	6.04	2,085	1,359	18,702
	1,000	1,583	541.	34.2	1,401	3.57	1,456	943	10,020
Black	50	287	23.3	8.1	289		287	187	395
	100	294	24.4	8.3	297		293	181	403
	200	307	27.3	8.9	305		305	197	469
	500	349	42.0	12.0	337		345	203	761
	1,000	420	72.3	17.2	401		408	223	1,276

Open in a new tab

Signal/noise is the ratio of the median of the bead images to that of the corresponding black background images.

The use of a cooled camera that has a 16-bit ADC permitted a longer exposure, which produced a very high contrast image (Fig. 2.22.20, D500), compared to the previous result (not shown) obtained with a room temperature camera.

Figure 2.22.20 — Plastic-embedded 0.5-μm FireRed beads with glycerol immersion medium sandwiched between the coverslip and the embedded beads. The time-gated images were acquired with Atik 314L+ camera with 2 × 2 binning. The total number of binned pixels was 360,705. The images of the beads (A and D)were obtained with a 60×, N.A. 1.25, oil-immersion objective, and the illumination was accomplished with a Nichia UV LED, which produced 500 pulses per sec, each of which had a duration of 1 msec for excitation, which was followed by 1 msec for image acquisition. The number of images summed is shown in the upper left-hand corner. For instance, for image A, the total period for the analog summation of 50 images was 0.1 sec. (B) Image A after processing with ImageJ, as described in the Image Analysis section with the Analyze Particles method set at size 6–60,d circularity 0.0–1.0, and Bare Outlines. Image D is the same as image A except that the total exposure was 1 sec (500 images summed). C, E, and F are black images that were obtained with the emission light blocked. The images have been optimized by ImageJ to have a range that is comparable to their content. Images C, E, and F when viewed with Photoshop are totally black (not shown). Reproduced from Leif et al. (2012).

The images of the 0.5-μm beads (A and D), which are shown with a black background, are virtually identical, as are the 100 and 200 summed images (not shown). The black images in Figure 2.22.21 do not show any internal structure that is the size of a bead. This is a very significant improvement from the periodic undulating horizontal pattern shown in the previous digitally summed background black images obtained with the cooled Retiga camera, shown in Figure 2.22.19.

Figure 2.22.21 — Coulter DNA-Check (Beckman Coulter) and 5-μm FireRed (Newport Instruments) beads imaged with the Atik 314L+ camera without binning at 4°C. (A) A conventional 1-sec exposure. (B) A gated 4-sec exposure (1 msec on and 1 msec off), 500 images/sec, total excitation period = 2 sec. The black, smaller (5-μm) luminescent long-lifetime microspheres contain Eu(TTFA)₃, and the larger fluorescent short-lifetime microspheres are DNA-Check microspheres. Time gating results in an obvious decrease in emission from the larger DNA-Check beads. Reproduced from Leif et al. (2012).

As shown in Table 2.22.3, in the case of the 50 summed images, the black image (Fig. 2.22.20C) has a maximum, 395, that is comparable with the minimum of the bead image (Fig. 2.22.20A), 389. For the 100, 200, and 500 summed images, the minimum of the bead image is greater than the maximum of comparable black image. Since the images were taken sequentially, the low values of the 1000 summed images of the beads are presumably due to photo-bleaching. The limiting step in maximizing the signal to noise is no longer the camera but photo-bleaching.

Surprisingly, in the case of both the bead and black images, increasing the number of summed frames from 50 to 500 raises the CV. As would be expected, the emissions from the beads increase the CV, which is a simple measure of the information present in the image. Admittedly, a considerable part of the CV is the result of uneven illumination. In the case of the black images that are not illuminated, the effects of uneven illumination and the presence of luminescence particles are irrelevant to the measurement of the CV. As expected, the CVs of the black images are much lower than those of the luminescence of the comparable bead containing image. However, the CVs of the black images are significant and increase with the number of frames acquired. The oversimplified but still useful and quick measurement of signal/noise, the ratio of the median of the bead images to that of the corresponding black background images, was at a maximum at 500 frames.

Basic Protocol 3

The quality of time-gated images obtained using a gated intensified camera, as indicated by perceived contrast and signal-to-noise ratio, is a function of the specimen itself (extinction coefficient, quantum yield, concentration) as well as the image-acquisition parameters including excitation light intensity, exposure time, intensifier gain level, and the number of frames summed. Altering the time-gating parameters (LED pulse width, intensifier gate width, number of pulse/detection cycles per frame) may also be expected to affect image quality. Figure 2.22.22 shows the results of imaging bubbles containing different amounts of Lumi4-Tb under varying conditions of intensifier gain, frames summed, and CCD frame length, which is proportional to the exposure time or total number of pulse/detection events. The images shown here were processed and analyzed using NIH ImageJ (v. 1.42).

Figure 2.22.22 — Effects of CCD frame length, intensifier gain, and frame summing on net signal and signal-to-noise ratio (SNR) of time-gated images of terbium luminescence. An aqueous solution containing a luminescent terbium complex, Lumi4-Tb, at indicated concentrations was emulsified into mineral oil, mounted on a microscope slide, and imaged with an ICCD-based, time-gated luminescence microscope (see Support Protocol 2 for specimen preparation). LED and camera timing parameters were set to values given in step 2 and Figure 2.22.16. Intensifier gain levels, CCD frame lengths, and number of frames summed are shown for each image in the figure. The net emission signal, S, is the mean pixel gray value in a region of interest (ROI) centered in the bubble image minus the mean pixel gray value in a nearby ROI of equivalent area (e.g., circles in rightmost image of A). The SNR was calculated as SNR = S/σ_s, where σ_s is the standard deviation of the signal in the bubble ROI. Micrographs (A-C): BF, bright field; TGL, time-gated luminescence (λ_em = 494 ± 10 nm) acquired under indicated conditions. Scale bars = 10 μm. (A) Both S and SNR increase with increasing CCD frame length. (B) Increasing intensifier gain level results in increased S, but no change in SNR. (C) Doubling the number of frames summed (from 4 to 8) increases SNR, but has no effect on net signal (as a percentage of image pixel depth).

Firstly, note that a longer exposure time, or frame length, has the greatest effect on image quality, increasing both mean, background-subtracted signal level (S), and SNR (Fig. 2.22.22A). If possible, adjust exposure time so that net signal levels are approximately equal to 50% to 70% of the maximum image pixel depth (e.g., S >2000 for a 12-bit image with a pixel depth of 4096). This will maximize SNR and image contrast, albeit at the expense of increased image acquisition time; it may not be possible to achieve such high signal levels with dim specimens. Secondly, note that an increase in intensifier gain level (from 780 V to 835 V) yields an increase in net signal level, and thus perceived image contrast, but has no effect on SNR (Fig. 2.22.22B). This is because the intensifier adds noise to the measurement in a gain-dependent fashion. Finally, image SNR may be increased by summing multiple frames (Fig. 2.22.22C). Under optimal operating conditions, ICCD images are photon noise-limited, and SNR ∝ S/√S, or √S. Thus, frame summing may be expected to increase S additively while increasing SNR as the square root of S. In actuality, SNR < √S because of intensifier noise and the detection of nonspecific background photons. Frame summing also increases image bit depth, and the overall magnitude of S increases with each additional frame, but its value as a percentage of image pixel depth remains constant.

Time Considerations

Setup time

For Basic Protocol 1, once all components are purchased/manufactured, it will take a single day to build the equipment. The assembly of the UV-LED in the dichroic cube and the alignment of imaging optics takes about 1 to 2 hr. The alignment of the chopper and the optimization of the delay configuration requires about 1 to 2 hr.

For Basic Protocol 2, after you have purchase all components, it will take one day to build the equipment.

For Basic Protocol 3 and Support Protocol 2, the assembly, initial testing, and optimization of image-acquisition parameters using water/oil emulsion standards may be completed in a single day once all hardware components, materials, and reagents are purchased. Mounting of the LED and ICCD to the microscope and connection of PC, ICCD, pulse generator, and LED require 1 to 2 hr. Measurement of LED illumination intensity at the objective back aperture takes approximately 1 hr. Preparation of water/oil emulsion test specimens requires about 2 hr. Once microscope components are assembled and test specimens prepared, image acquisition may begin, and a series of images can be acquired under different timing parameters and gain settings in as little as 1 to 2 hr.

Acknowledgements

D. Jin wishes to acknowledge Olympus Australia for generous technical supports as well as demo versions of microscopes. The research for Basic Protocol 1 was supported, in part, by the Australian Research Council (DP1095465; LP130100517).

FireRed is a trademark of Newport Instruments, San Diego, Ca. Newport Instruments supported the research for Basic Protocol 2.

Lumi4 is a registered trademark of Lumiphore, Inc., Richmond, Ca. The research for Basic Protocol 3 was supported, in part, by the National Institutes of Health (National Institute of General Medical Sciences Grant R01GM081030-01A1).

Literature Cited

1394 Trade Association, 2004. IIDC 1394-based Digital Camera Specification Ver.1.31. TA Document 2003017.
Bae E, Patsekin V, Rajwa B, Bhunia AK, Holdman C, Davisson VJ, Hirleman ED, and Robinson JP. 2012. Development of a microbial high-throughput screening instrument based on elastic light scatter patterns. Rev. Sci. Instrum. 83:044304–044310. [DOI] [PMC free article] [PubMed] [Google Scholar]
Bunzli JCG and Piguet C. 2005. Taking advantage of luminescent lanthanide ions. Chem. Soc. Rev. 34:1048–1077. [DOI] [PubMed] [Google Scholar]
Campbell R, Tour O, Palmer A, Steinbach P, Baird G, Zacharias D, and Tsien R. 2002. A monomeric red fluorescent protein. Proc. Natl. Acad. Sci. U.S.A. 99:7877–7882. [DOI] [PMC free article] [PubMed] [Google Scholar]
Chan WCW and Nie SM. 1998. Quantum dot bioconjugates for ultrasensitive nonisotopic detection. Science 281:2016–2018. [DOI] [PubMed] [Google Scholar]
Chattopadhyay PK, Price DA, Harper TF, Betts MR, Yu J, Gostick E, Perfetto SP, Goepfert P, Koup RA, and De Rosa SC. 2006. Quantum dot semiconductor nanocrystals for immunophenotyping by polychromatic flow cytometry. Nat. Med. 12:972–977. [DOI] [PubMed] [Google Scholar]
Connally R, Jin D, and Piper J. 2006. High intensity solid-state UV source for time-gated luminescence microscopy. Cytometry A 69:1020–1027. [DOI] [PubMed] [Google Scholar]
Deiters E, Song B, Chauvin AS, Vandevyver CDB, Gumy F, and Bunzli JCG. 2009. Luminescent bimetallic lanthanide bioprobes for cellular imaging with excitation in the visible-light range. Chemistry 15:885–900. [DOI] [PubMed] [Google Scholar]
Eliseeva SV and Bunzli J-CG. 2010. Lanthanide luminescence for functional materials and biosciences. Chem. Soc. Rev. 39:189–227. [DOI] [PubMed] [Google Scholar]
Enderlein J. 2012. Polymer dynamics, fluorescence correlation spectroscopy, and the limits of optical resolution. Phys. Rev. Lett. 108:108101. [DOI] [PubMed] [Google Scholar]
Gahlaut N and Miller LW. 2010. Time-resolved microscopy for imaging lanthanide luminescence in living cells. Cytometry A 77:1113–1125. [DOI] [PMC free article] [PubMed] [Google Scholar]
Grégori G, Patsekin V, Rajwa B, Jones J, Ragheb K, Holdman C, and Robinson JP. 2011. Hyperspectral cytometry at the single-cell level using a 32-channel photodetector. Cytometry A 81:35–44. [DOI] [PubMed] [Google Scholar]
Grunwald D, Shenoy SM, Burke S, and Singer RH. 2008. Calibrating excitation light fluxes for quantitative light microscopy in cell biology. Nat. Protoc. 3:1809–1814. [DOI] [PMC free article] [PubMed] [Google Scholar]
Harma H, Soukka T, and Lovgren T. 2001. Europium nanoparticles and time-resolved fluorescence for ultrasensitive detection of prostate-specific antigen. Clin. Chem. 47:561–568. [PubMed] [Google Scholar]
Hemmilá I and Mukkala V-M. 2001. Time-resolution in fluorometry technologies, labels, and applications in bioanalytical assays. Crit. Rev. Clin. Lab. Sci. 38:441–519. [Google Scholar]
Huang B, Wang W, Bates M, and Zhuang X. 2008. Three-dimensional super-resolution imaging by stochastic optical reconstruction microscopy. Science 319:810–813. [DOI] [PMC free article] [PubMed] [Google Scholar]
Jin D. 2011a. Background-free cytometry using rare earth complex bioprobes. Methods Cell Biol. 102:479–513. [DOI] [PubMed] [Google Scholar]
Jin D. 2011b. Demonstration of true-color high-contrast microorganism imaging for terbium bioprobes. Cytometry A 79:392–397. [DOI] [PubMed] [Google Scholar]
Jin D and Piper JA. 2011. Time-gated luminescence microscopy allowing direct visual inspection of lanthanide-stained microorganisms in background-free condition. Anal. Chem. 83:2294–2300. [DOI] [PubMed] [Google Scholar]
Jin D, Connally R, and Piper J. 2006. Long-lived visible luminescence of UV LEDs and impact on LED excited time-resolved fluorescence applications. J. Phys. D Appl. Phys. 39:461–465. [Google Scholar]
Jin D, Connally R, and Piper J. 2007a. Practical time-gated luminescence flow cytometry. I: Concepts. Cytometry A 71:783–796. [DOI] [PubMed] [Google Scholar]
Jin D, Connally R, and Piper J. 2007b. Practical time-gated luminescence flow cytometry. II: Experimental evaluation using UV LED excitation. Cytometry A 71:797–808. [DOI] [PubMed] [Google Scholar]
Jin D, Piper JA, Leif RC, Yang S, Ferrari BC, Yuan J, Wang G, Vallarino LM, and Williams JW. 2009. Time-gated flow cytometry: An ultra-high selectivity method to recover ultra-rare-event mu-targets in high-background biosamples. J. Biomed. Opt. 14:024023. [DOI] [PubMed] [Google Scholar]
Kamentsky L and Kamentsky L. 1991. Microscope-based multiparameter laser scanning cytometer yielding data comparable to flow cytometry data. Cytometry A 12:381–387. [DOI] [PubMed] [Google Scholar]
Leif RC. 2002. Methods for preparing sorted cells as monolayer specimens. 2002. In Living Color, Protocols in Flow Cytometry and Cell Sorting (Diamond RA and DeMaggio S, eds.) pp. 592–619. Springer, New York. [Google Scholar]
Leif RC. 2013. Centrifugal Cytology System, Chamber Block and Method for the preparation of Treated Monolayers of Sample Material. U.S. Patent 8,501.096.
Leif RC and Yang S. 2010. An analog method to produce time-gated images. In Imaging, Manipulation, and Analysis of Biomolecules, Cells, and Tissues VIII Proc. SPIE 7568 (Farkas DL, Nicolau DV, and Leif RC, eds.) pp. 756–801. 1A-75681A doi: 10.1117/12.858683. SPIE, Bellingham, Wash. [DOI] [Google Scholar]
Leif RC and Vallarino LM. 2011. Conjugated Polymer Tag Complexes. U.S. Patent 8,048,659.
Leif RC, Clay SP, Gratzner HG, Haines HG, Rao KV, and Vallarino LM. 1976. Markers for instrumental evaluation of cells of the female reproductive tract: Existing and new markers. In The Automation of Uterine Cancer Cytology (Wied GL, Bahr GF, and Bartels PH, eds.) pp. 313–344. Tutorials of Cytology, Chicago, Ill. [Google Scholar]
Leif RC, Becker MC, Bromm AJ Jr., Vallarino LM, Williams SA, and Yang S. 2001. Increasing the luminescence of lanthanide(III) macrocyclic complexes by the use of polymers and lanthanide enhanced luminescence. In Optical Diagnostics of Living Cells IV, Proc. SPIE 4260 (Farkas DL and Leif RC, eds.) pp. 184–197. doi: 10.1117/12.426770. SPIE, Bellingham, Wash. [DOI] [Google Scholar]
Leif RC, Becker MC, Vallarino LM, Williams JW, and Yang S. 2003. Progress in the use of Quantum Dye ^® Eu(III)-macrocycles. In Imaging, Manipulation, and Analysis of Biomolecules, Cells, and Tissues VIII, Proc. SPIE. 4962 (Farkas DL, Nicolau DV, and Leif RC, eds.) pp. 341–353. SPIE, Bellingham, Wash. [Google Scholar]
Leif RC, Becker MC, Bromm JA, Vallarino LM, and Yang S. 2006a. Fluorescence resonance energy transfer enhanced luminescence (FRETEL) of Quantum Dyes. In Ultrasensitive and Single-Molecule Detection Technologies, Proc. SPIE (Enderlein J and Gryczynski ZK, eds.) pp. 609206–1–609206–13. doi: 10.1117/12.645069. SPIE, Bellingham, Wash. [DOI] [Google Scholar]
Leif RC, Vallarino LM, Becker MC, and Yang S. 2006b. Increasing lanthanide luminescence by use of the RETEL effect. Cytometry A 69:940–946. [DOI] [PubMed] [Google Scholar]
Leif RC, Vallarino LM, Becker MC, and Yang S. 2006c. Increasing the luminescence of lanthanide complexes. Cytometry A 69:767–778. [DOI] [PubMed] [Google Scholar]
Leif RC, Jin D, Vallarino LM, Zucker RM, Williams JW, Piper J, and Yang S. 2009. Calibration beads containing luminescent lanthanide ion complexes. J. Biomed. Optics 14:024022–024022. [DOI] [PubMed] [Google Scholar]
Leif RC., Yang S., Lu Y., Jin D., and Chambers SA. 2012. A cost-effective analog method to produce time-gated luminescence images. In Imaging, Manipulation, and Analysis of Biomolecules, Cells, and Tissues X, Proc. SPIE 8225 (Farkas DL, Nicolau DV, and Leif RC, eds.) 8225–35 82250Z. SPIE, Bellingham, Wash. [Google Scholar]
Liu Y, Ding Y, Alonas E, Zhao W, Santangelo PJ, Jin D, Piper JA, Teng J, Ren Q, and Xi P. 2012. Achieving λ/10 resolution CW STED nanoscopy with a Ti:sapphire oscillator. PLoS One 7:e40003. [DOI] [PMC free article] [PubMed] [Google Scholar]
Lu J, Martin J, Lu Y, Zhao J, Yuan J, Ostrowski M, Paulsen I, Piper JA, and Jin D. 2012. Resolving low-expression cell surface antigens by time-gated orthogonal scanning automated microscopy. Anal. Chem. 84:9674–9678. [DOI] [PubMed] [Google Scholar]
Lu Y, Jin D, Leif RC, Deng W, Piper JA, Yuan J, Duan Y, and Huo Y. 2011. Automated detection of rare-event pathogens through time-gated luminescence scanning microscopy. Cytometry A 79:349–355. [DOI] [PubMed] [Google Scholar]
Lu Y, Xi P, Piper JA, Huo Y, and Jin D. 2012. Time-gated orthogonal scanning automated microscopy (OSAM) for high-speed cell detection and analysis. Sci. Rep. 2:837. [DOI] [PMC free article] [PubMed] [Google Scholar]
Makhluf SBD, Arnon R, Patra CR, Mukhopadhyay D, Gedanken A, Mukherjee P, and Breitbart H. 2008. Labeling of sperm cells via the spontaneous penetration of Eu3+ ions as nanoparticles complexed with PVA or PVP. J. Phys. Chem. C 112:12801–12807. [Google Scholar]
Mirabelli P, Scalia G, Pascariello C, D’Alessio F, Mariotti E, Noto RD, George TC, Kong R, Venkatachalam V, and Basiji D. 2012. ImageStream promyelocytic leukemia protein immunolocalization: In search of promyelocytic leukemia cells. Cytometry A 81:232–237. [DOI] [PubMed] [Google Scholar]
Mitchell AC, Wall JE, Murray JG, and Morgan CG. 2002a. Direct modulation of the effective sensitivity of a CCD detector: A new approach to time-resolved fluorescence imaging. J. Microsc. 206:225–232. [DOI] [PubMed] [Google Scholar]
Mitchell AC, Wall JE, Murray JG, and Morgan CG. 2002b. Measurement of nanosecond time-resolved fluorescence with a directly gated interline CCD camera. J. Microsc. 206:233–238. [DOI] [PubMed] [Google Scholar]
Mitchell AC, Wall JE, and Morgan CG. 2007. True-color nanosecond time-gated fluorescence imaging using a CCD detector. In Three-Dimensional and Multidimensional Microscopy: Image Acquisition and Processing XIV, SPIE 6443 (Conchello JA, Cogswell CJ, and Wilson T, eds.) pp. 64430B-1–64430B-8. SPIE, Bellingham, Wash. [Google Scholar]
Mohandessi S, Rajendran M, Magda D, and Miller LW. 2012. Cell-penetrating peptides as delivery vehicles for a protein-targeted terbium complex. Chemistry 18:10825–10829. [DOI] [PMC free article] [PubMed] [Google Scholar]
Peng T, Xie H, Ding Y, Wang W, Li Z, Jin D, Tang Y, Ren Q, and Xi P. 2012. CRAFT: Multimodality confocal skin imaging for early cancer diagnosis. J. Biophoton. 5:469–476. [DOI] [PubMed] [Google Scholar]
Perfetto SP, Chattopadhyay PK, and Roederer M. 2004. Seventeen-colour flow cytometry: Unravelling the immune system. Nat. Rev. Immunol. 4:648–655. [DOI] [PubMed] [Google Scholar]
Petoud S, Cohen SM, Bunzli JCG, and Raymond KN. 2003. Stable lanthanide luminescence agents highly emissive in aqueous solution: Multidentate 2-hydroxyisophthalamide complexes of SM3+, Eu3+, Tb3+, Dy3+. J. Am. Chem. Soc. 125:13324–13325 [DOI] [PubMed] [Google Scholar]
Rajapakse HE and Miller LW. 2012. Time-resolved luminescence resonance energy transfer imaging of protein-protein interactions in living cells. Methods Enzymol. 505:329–345. [DOI] [PubMed] [Google Scholar]
Rajapakse HE, Gahlaut N, Mohandessi S, Yu D, Turner JR, and Miller LW. 2010. Time-resolved luminescence resonance energy transfer imaging of protein-protein interactions in living cells. Proc. Natl. Acad. Sci. U.S.A. 107:13582–13587. [DOI] [PMC free article] [PubMed] [Google Scholar]
Robinson JP, Rajwa B, Patsekin V, and Davisson VJ. 2012. Computational analysis of high-throughput flow cytometry data. Exp. Opin. Drug Discov. 7:679–693. [DOI] [PMC free article] [PubMed] [Google Scholar]
Selvin PR. 2002. Principles and biophysical applications of lanthanide-based probes. Annu. Rev. Biophys. Biomol. Struct. 31:275–302. [DOI] [PubMed] [Google Scholar]
Shapiro H. 2003. Practical Flow Cytometry. Wiley-Liss, Hoboken, N.J. [Google Scholar]
Song B, Wang GL, Tan MQ, and Yuan JL. 2006. A europium(III) complex as an efficient singlet oxygen luminescence probe. J. Am. Chem. Soc. 128:13442–13450. [DOI] [PubMed] [Google Scholar]
Song CH, Ye ZQ, Wang GL, Jin DY, Yuan JL, Guan YF, and Piper J. 2009. Preparation and time-gated luminescence bioimaging application of ruthenium complex covalently bound silica nanoparticles. Talanta 79:103–108. [DOI] [PubMed] [Google Scholar]
Thibon A and Pierre VC. 2009a. A Highly selective luminescent sensor for the time-gated detection of potassium. J. Am. Chem. Soc. 131:434–335. [DOI] [PubMed] [Google Scholar]
Thibon A and Pierre VC. 2009b. Principles of responsive lanthanide-based luminescent probes for cellular imaging. Anal. Bioanal. Chem. 394:107–120. [DOI] [PubMed] [Google Scholar]
Vaasa A, Ligi K, Mohandessi S, Enkvist E, Uri A, and Miller LW. 2012. Time-gated luminescence microscopy with responsive nonmetal probes for mapping activity of protein kinases in living cells. Chem. Commun. 48:8595–8597. [DOI] [PMC free article] [PubMed] [Google Scholar]
Vereb G, Jares-Erijman E, Selvin PR, and Jovin TM. 1998. Temporally and spectrally resolved imaging microscopy of lanthanide chelates. Biophys. J. 74:2210–2222. [DOI] [PMC free article] [PubMed] [Google Scholar]
Weibel N, Charbonniere LJ, Guardigli M, Roda A, and Ziessel R. 2004. Engineering of highly luminescent lanthanide tags suitable for protein labeling and time-resolved luminescence imaging. J. Am. Chem. Soc. 126:4888–4896. [DOI] [PubMed] [Google Scholar]
Wu J, Wang GL, Jin DY, Yuan JL, Guan YF, and Piper J. 2008. Luminescent europium nanoparticles with a wide excitation range from UV to visible light for biolabeling and time-gated luminescence bioimaging. Chem. Commun. 365–367. [DOI] [PubMed] [Google Scholar]
Wu J, Ye ZQ, Wang GL, Jin DY, Yuan JL, Guan YF, and Piper J. 2009. Visible-light-sensitized highly luminescent europium nanoparticles: Preparation and application for time-gated luminescence bioimaging. J. Mater. Chem. 19:1258–1264. [Google Scholar]
Xu J, Corneillie TM, Moore EG, Law GL, Butlin NG, and Raymond KN. 2011. Octadentate cages of Tb(III) 2-hydroxyisophthalamides: A new standard for luminescent lanthanide labels. J. Am. Chem. Soc. 133:19900–19910. [DOI] [PMC free article] [PubMed] [Google Scholar]
Yuan J and Wang G. 2006. Lanthanide-based luminescence probes and time-resolved luminescence bioassays. TrAC Trends Anal. Chem. 25:490–500. [Google Scholar]
Yung PT and Ponce A. 2008. Fast sterility assessment by germinable-endospore biodosimetry. Appl. Environ. Microbiol. 74:7669–7674. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R1] 1394 Trade Association, 2004. IIDC 1394-based Digital Camera Specification Ver.1.31. TA Document 2003017.

[R2] Bae E, Patsekin V, Rajwa B, Bhunia AK, Holdman C, Davisson VJ, Hirleman ED, and Robinson JP. 2012. Development of a microbial high-throughput screening instrument based on elastic light scatter patterns. Rev. Sci. Instrum. 83:044304–044310. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R3] Bunzli JCG and Piguet C. 2005. Taking advantage of luminescent lanthanide ions. Chem. Soc. Rev. 34:1048–1077. [DOI] [PubMed] [Google Scholar]

[R4] Campbell R, Tour O, Palmer A, Steinbach P, Baird G, Zacharias D, and Tsien R. 2002. A monomeric red fluorescent protein. Proc. Natl. Acad. Sci. U.S.A. 99:7877–7882. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R5] Chan WCW and Nie SM. 1998. Quantum dot bioconjugates for ultrasensitive nonisotopic detection. Science 281:2016–2018. [DOI] [PubMed] [Google Scholar]

[R6] Chattopadhyay PK, Price DA, Harper TF, Betts MR, Yu J, Gostick E, Perfetto SP, Goepfert P, Koup RA, and De Rosa SC. 2006. Quantum dot semiconductor nanocrystals for immunophenotyping by polychromatic flow cytometry. Nat. Med. 12:972–977. [DOI] [PubMed] [Google Scholar]

[R7] Connally R, Jin D, and Piper J. 2006. High intensity solid-state UV source for time-gated luminescence microscopy. Cytometry A 69:1020–1027. [DOI] [PubMed] [Google Scholar]

[R8] Deiters E, Song B, Chauvin AS, Vandevyver CDB, Gumy F, and Bunzli JCG. 2009. Luminescent bimetallic lanthanide bioprobes for cellular imaging with excitation in the visible-light range. Chemistry 15:885–900. [DOI] [PubMed] [Google Scholar]

[R9] Eliseeva SV and Bunzli J-CG. 2010. Lanthanide luminescence for functional materials and biosciences. Chem. Soc. Rev. 39:189–227. [DOI] [PubMed] [Google Scholar]

[R10] Enderlein J. 2012. Polymer dynamics, fluorescence correlation spectroscopy, and the limits of optical resolution. Phys. Rev. Lett. 108:108101. [DOI] [PubMed] [Google Scholar]

[R11] Gahlaut N and Miller LW. 2010. Time-resolved microscopy for imaging lanthanide luminescence in living cells. Cytometry A 77:1113–1125. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R12] Grégori G, Patsekin V, Rajwa B, Jones J, Ragheb K, Holdman C, and Robinson JP. 2011. Hyperspectral cytometry at the single-cell level using a 32-channel photodetector. Cytometry A 81:35–44. [DOI] [PubMed] [Google Scholar]

[R13] Grunwald D, Shenoy SM, Burke S, and Singer RH. 2008. Calibrating excitation light fluxes for quantitative light microscopy in cell biology. Nat. Protoc. 3:1809–1814. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R14] Harma H, Soukka T, and Lovgren T. 2001. Europium nanoparticles and time-resolved fluorescence for ultrasensitive detection of prostate-specific antigen. Clin. Chem. 47:561–568. [PubMed] [Google Scholar]

[R15] Hemmilá I and Mukkala V-M. 2001. Time-resolution in fluorometry technologies, labels, and applications in bioanalytical assays. Crit. Rev. Clin. Lab. Sci. 38:441–519. [Google Scholar]

[R16] Huang B, Wang W, Bates M, and Zhuang X. 2008. Three-dimensional super-resolution imaging by stochastic optical reconstruction microscopy. Science 319:810–813. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R17] Jin D. 2011a. Background-free cytometry using rare earth complex bioprobes. Methods Cell Biol. 102:479–513. [DOI] [PubMed] [Google Scholar]

[R18] Jin D. 2011b. Demonstration of true-color high-contrast microorganism imaging for terbium bioprobes. Cytometry A 79:392–397. [DOI] [PubMed] [Google Scholar]

[R19] Jin D and Piper JA. 2011. Time-gated luminescence microscopy allowing direct visual inspection of lanthanide-stained microorganisms in background-free condition. Anal. Chem. 83:2294–2300. [DOI] [PubMed] [Google Scholar]

[R20] Jin D, Connally R, and Piper J. 2006. Long-lived visible luminescence of UV LEDs and impact on LED excited time-resolved fluorescence applications. J. Phys. D Appl. Phys. 39:461–465. [Google Scholar]

[R21] Jin D, Connally R, and Piper J. 2007a. Practical time-gated luminescence flow cytometry. I: Concepts. Cytometry A 71:783–796. [DOI] [PubMed] [Google Scholar]

[R22] Jin D, Connally R, and Piper J. 2007b. Practical time-gated luminescence flow cytometry. II: Experimental evaluation using UV LED excitation. Cytometry A 71:797–808. [DOI] [PubMed] [Google Scholar]

[R23] Jin D, Piper JA, Leif RC, Yang S, Ferrari BC, Yuan J, Wang G, Vallarino LM, and Williams JW. 2009. Time-gated flow cytometry: An ultra-high selectivity method to recover ultra-rare-event mu-targets in high-background biosamples. J. Biomed. Opt. 14:024023. [DOI] [PubMed] [Google Scholar]

[R24] Kamentsky L and Kamentsky L. 1991. Microscope-based multiparameter laser scanning cytometer yielding data comparable to flow cytometry data. Cytometry A 12:381–387. [DOI] [PubMed] [Google Scholar]

[R25] Leif RC. 2002. Methods for preparing sorted cells as monolayer specimens. 2002. In Living Color, Protocols in Flow Cytometry and Cell Sorting (Diamond RA and DeMaggio S, eds.) pp. 592–619. Springer, New York. [Google Scholar]

[R26] Leif RC. 2013. Centrifugal Cytology System, Chamber Block and Method for the preparation of Treated Monolayers of Sample Material. U.S. Patent 8,501.096.

[R27] Leif RC and Yang S. 2010. An analog method to produce time-gated images. In Imaging, Manipulation, and Analysis of Biomolecules, Cells, and Tissues VIII Proc. SPIE 7568 (Farkas DL, Nicolau DV, and Leif RC, eds.) pp. 756–801. 1A-75681A doi: 10.1117/12.858683. SPIE, Bellingham, Wash. [DOI] [Google Scholar]

[R28] Leif RC and Vallarino LM. 2011. Conjugated Polymer Tag Complexes. U.S. Patent 8,048,659.

[R29] Leif RC, Clay SP, Gratzner HG, Haines HG, Rao KV, and Vallarino LM. 1976. Markers for instrumental evaluation of cells of the female reproductive tract: Existing and new markers. In The Automation of Uterine Cancer Cytology (Wied GL, Bahr GF, and Bartels PH, eds.) pp. 313–344. Tutorials of Cytology, Chicago, Ill. [Google Scholar]

[R30] Leif RC, Becker MC, Bromm AJ Jr., Vallarino LM, Williams SA, and Yang S. 2001. Increasing the luminescence of lanthanide(III) macrocyclic complexes by the use of polymers and lanthanide enhanced luminescence. In Optical Diagnostics of Living Cells IV, Proc. SPIE 4260 (Farkas DL and Leif RC, eds.) pp. 184–197. doi: 10.1117/12.426770. SPIE, Bellingham, Wash. [DOI] [Google Scholar]

[R31] Leif RC, Becker MC, Vallarino LM, Williams JW, and Yang S. 2003. Progress in the use of Quantum Dye ^® Eu(III)-macrocycles. In Imaging, Manipulation, and Analysis of Biomolecules, Cells, and Tissues VIII, Proc. SPIE. 4962 (Farkas DL, Nicolau DV, and Leif RC, eds.) pp. 341–353. SPIE, Bellingham, Wash. [Google Scholar]

[R32] Leif RC, Becker MC, Bromm JA, Vallarino LM, and Yang S. 2006a. Fluorescence resonance energy transfer enhanced luminescence (FRETEL) of Quantum Dyes. In Ultrasensitive and Single-Molecule Detection Technologies, Proc. SPIE (Enderlein J and Gryczynski ZK, eds.) pp. 609206–1–609206–13. doi: 10.1117/12.645069. SPIE, Bellingham, Wash. [DOI] [Google Scholar]

[R33] Leif RC, Vallarino LM, Becker MC, and Yang S. 2006b. Increasing lanthanide luminescence by use of the RETEL effect. Cytometry A 69:940–946. [DOI] [PubMed] [Google Scholar]

[R34] Leif RC, Vallarino LM, Becker MC, and Yang S. 2006c. Increasing the luminescence of lanthanide complexes. Cytometry A 69:767–778. [DOI] [PubMed] [Google Scholar]

[R35] Leif RC, Jin D, Vallarino LM, Zucker RM, Williams JW, Piper J, and Yang S. 2009. Calibration beads containing luminescent lanthanide ion complexes. J. Biomed. Optics 14:024022–024022. [DOI] [PubMed] [Google Scholar]

[R36] Leif RC., Yang S., Lu Y., Jin D., and Chambers SA. 2012. A cost-effective analog method to produce time-gated luminescence images. In Imaging, Manipulation, and Analysis of Biomolecules, Cells, and Tissues X, Proc. SPIE 8225 (Farkas DL, Nicolau DV, and Leif RC, eds.) 8225–35 82250Z. SPIE, Bellingham, Wash. [Google Scholar]

[R37] Liu Y, Ding Y, Alonas E, Zhao W, Santangelo PJ, Jin D, Piper JA, Teng J, Ren Q, and Xi P. 2012. Achieving λ/10 resolution CW STED nanoscopy with a Ti:sapphire oscillator. PLoS One 7:e40003. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R38] Lu J, Martin J, Lu Y, Zhao J, Yuan J, Ostrowski M, Paulsen I, Piper JA, and Jin D. 2012. Resolving low-expression cell surface antigens by time-gated orthogonal scanning automated microscopy. Anal. Chem. 84:9674–9678. [DOI] [PubMed] [Google Scholar]

[R39] Lu Y, Jin D, Leif RC, Deng W, Piper JA, Yuan J, Duan Y, and Huo Y. 2011. Automated detection of rare-event pathogens through time-gated luminescence scanning microscopy. Cytometry A 79:349–355. [DOI] [PubMed] [Google Scholar]

[R40] Lu Y, Xi P, Piper JA, Huo Y, and Jin D. 2012. Time-gated orthogonal scanning automated microscopy (OSAM) for high-speed cell detection and analysis. Sci. Rep. 2:837. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R41] Makhluf SBD, Arnon R, Patra CR, Mukhopadhyay D, Gedanken A, Mukherjee P, and Breitbart H. 2008. Labeling of sperm cells via the spontaneous penetration of Eu3+ ions as nanoparticles complexed with PVA or PVP. J. Phys. Chem. C 112:12801–12807. [Google Scholar]

[R42] Mirabelli P, Scalia G, Pascariello C, D’Alessio F, Mariotti E, Noto RD, George TC, Kong R, Venkatachalam V, and Basiji D. 2012. ImageStream promyelocytic leukemia protein immunolocalization: In search of promyelocytic leukemia cells. Cytometry A 81:232–237. [DOI] [PubMed] [Google Scholar]

[R43] Mitchell AC, Wall JE, Murray JG, and Morgan CG. 2002a. Direct modulation of the effective sensitivity of a CCD detector: A new approach to time-resolved fluorescence imaging. J. Microsc. 206:225–232. [DOI] [PubMed] [Google Scholar]

[R44] Mitchell AC, Wall JE, Murray JG, and Morgan CG. 2002b. Measurement of nanosecond time-resolved fluorescence with a directly gated interline CCD camera. J. Microsc. 206:233–238. [DOI] [PubMed] [Google Scholar]

[R45] Mitchell AC, Wall JE, and Morgan CG. 2007. True-color nanosecond time-gated fluorescence imaging using a CCD detector. In Three-Dimensional and Multidimensional Microscopy: Image Acquisition and Processing XIV, SPIE 6443 (Conchello JA, Cogswell CJ, and Wilson T, eds.) pp. 64430B-1–64430B-8. SPIE, Bellingham, Wash. [Google Scholar]

[R46] Mohandessi S, Rajendran M, Magda D, and Miller LW. 2012. Cell-penetrating peptides as delivery vehicles for a protein-targeted terbium complex. Chemistry 18:10825–10829. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R47] Peng T, Xie H, Ding Y, Wang W, Li Z, Jin D, Tang Y, Ren Q, and Xi P. 2012. CRAFT: Multimodality confocal skin imaging for early cancer diagnosis. J. Biophoton. 5:469–476. [DOI] [PubMed] [Google Scholar]

[R48] Perfetto SP, Chattopadhyay PK, and Roederer M. 2004. Seventeen-colour flow cytometry: Unravelling the immune system. Nat. Rev. Immunol. 4:648–655. [DOI] [PubMed] [Google Scholar]

[R49] Petoud S, Cohen SM, Bunzli JCG, and Raymond KN. 2003. Stable lanthanide luminescence agents highly emissive in aqueous solution: Multidentate 2-hydroxyisophthalamide complexes of SM3+, Eu3+, Tb3+, Dy3+. J. Am. Chem. Soc. 125:13324–13325 [DOI] [PubMed] [Google Scholar]

[R50] Rajapakse HE and Miller LW. 2012. Time-resolved luminescence resonance energy transfer imaging of protein-protein interactions in living cells. Methods Enzymol. 505:329–345. [DOI] [PubMed] [Google Scholar]

[R51] Rajapakse HE, Gahlaut N, Mohandessi S, Yu D, Turner JR, and Miller LW. 2010. Time-resolved luminescence resonance energy transfer imaging of protein-protein interactions in living cells. Proc. Natl. Acad. Sci. U.S.A. 107:13582–13587. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R52] Robinson JP, Rajwa B, Patsekin V, and Davisson VJ. 2012. Computational analysis of high-throughput flow cytometry data. Exp. Opin. Drug Discov. 7:679–693. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R53] Selvin PR. 2002. Principles and biophysical applications of lanthanide-based probes. Annu. Rev. Biophys. Biomol. Struct. 31:275–302. [DOI] [PubMed] [Google Scholar]

[R54] Shapiro H. 2003. Practical Flow Cytometry. Wiley-Liss, Hoboken, N.J. [Google Scholar]

[R55] Song B, Wang GL, Tan MQ, and Yuan JL. 2006. A europium(III) complex as an efficient singlet oxygen luminescence probe. J. Am. Chem. Soc. 128:13442–13450. [DOI] [PubMed] [Google Scholar]

[R56] Song CH, Ye ZQ, Wang GL, Jin DY, Yuan JL, Guan YF, and Piper J. 2009. Preparation and time-gated luminescence bioimaging application of ruthenium complex covalently bound silica nanoparticles. Talanta 79:103–108. [DOI] [PubMed] [Google Scholar]

[R57] Thibon A and Pierre VC. 2009a. A Highly selective luminescent sensor for the time-gated detection of potassium. J. Am. Chem. Soc. 131:434–335. [DOI] [PubMed] [Google Scholar]

[R58] Thibon A and Pierre VC. 2009b. Principles of responsive lanthanide-based luminescent probes for cellular imaging. Anal. Bioanal. Chem. 394:107–120. [DOI] [PubMed] [Google Scholar]

[R59] Vaasa A, Ligi K, Mohandessi S, Enkvist E, Uri A, and Miller LW. 2012. Time-gated luminescence microscopy with responsive nonmetal probes for mapping activity of protein kinases in living cells. Chem. Commun. 48:8595–8597. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R60] Vereb G, Jares-Erijman E, Selvin PR, and Jovin TM. 1998. Temporally and spectrally resolved imaging microscopy of lanthanide chelates. Biophys. J. 74:2210–2222. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R61] Weibel N, Charbonniere LJ, Guardigli M, Roda A, and Ziessel R. 2004. Engineering of highly luminescent lanthanide tags suitable for protein labeling and time-resolved luminescence imaging. J. Am. Chem. Soc. 126:4888–4896. [DOI] [PubMed] [Google Scholar]

[R62] Wu J, Wang GL, Jin DY, Yuan JL, Guan YF, and Piper J. 2008. Luminescent europium nanoparticles with a wide excitation range from UV to visible light for biolabeling and time-gated luminescence bioimaging. Chem. Commun. 365–367. [DOI] [PubMed] [Google Scholar]

[R63] Wu J, Ye ZQ, Wang GL, Jin DY, Yuan JL, Guan YF, and Piper J. 2009. Visible-light-sensitized highly luminescent europium nanoparticles: Preparation and application for time-gated luminescence bioimaging. J. Mater. Chem. 19:1258–1264. [Google Scholar]

[R64] Xu J, Corneillie TM, Moore EG, Law GL, Butlin NG, and Raymond KN. 2011. Octadentate cages of Tb(III) 2-hydroxyisophthalamides: A new standard for luminescent lanthanide labels. J. Am. Chem. Soc. 133:19900–19910. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R65] Yuan J and Wang G. 2006. Lanthanide-based luminescence probes and time-resolved luminescence bioassays. TrAC Trends Anal. Chem. 25:490–500. [Google Scholar]

[R66] Yung PT and Ponce A. 2008. Fast sterility assessment by germinable-endospore biodosimetry. Appl. Environ. Microbiol. 74:7669–7674. [DOI] [PMC free article] [PubMed] [Google Scholar]

PERMALINK

How to Build a Time-Gated Luminescence Microscope

Dayong Jin

Yiqing Lu

Robert C Leif

Sean Yang

Megha Rajendran

Lawrence W Miller

Abstract

INTRODUCTION

Figure 2.22.1.

Table 2.22.1.

BASIC PROTOCOL 1 CONSTRUCTION AND USE OF A MECHANICAL CHOPPER TgL UNIT

Pulsing the UV LEDs for excitation

Optical chopper unit

Materials

Figure 2.22.2.

Figure 2.22.3.

Pulsed UV excitation

Figure 2.22.4.

Figure 2.22.5.

Fast-gating chopper unit

Synchronization

BASIC PROTOCOL 2 CONSTRUCTION AND USE OF CAMERA WITH ANALOG SIGNAL SUMMATION

Materials

Figure 2.22.6.

Figure 2.22.7.

Condenser mount

LED mount

Three-dimensional LED positioner

Interfacing the Atik camera to the microscope

Figure 2.22.8.

Connecting the electronics

Figure 2.22.9.

Running the camera software

Figure 2.22.10.

Figure 2.22.11.

Image collection

Figure 2.22.12.

Figure 2.22.13.

Figure 2.22.14.

Figure 2.22.15.

SUPPORT PROTOCOL 1 BEAD SAMPLE PREPARATION

Materials

BASIC PROTOCOL 3 PREPARATION AND USE OF TIME-GATABLE INTENSIFIED CCD CAMERA

Materials

Table 2.22.2.

Hardware configuration

Figure 2.22.16.

Programming the pulse generator

Image acquisition

SUPPORT PROTOCOL 2 PREPARATION OF WATER/OIL EMULSION STANDARDS FOR MICROSCOPE CALIBRATION AND TESTING

Materials

Prepare water-in-oil emulsion

Prepare slide mount

COMMENTARY

Background Information

Comparison of instruments

Comparison of LEDs:

Comparison of cameras:

Monochrome or color detection:

Cost and commercialization

Critical Parameters and Troubleshooting

Basic Protocol 1

LED

Chopper blade

Synchronization

Basic Protocol 3

Pulse generator

Excitation filter

Anticipated Results

Basic Protocol 1

Figure 2.22.17.

Figure 2.22.18.

Basic Protocol 2

Analog-to-digital conversion (ADC) noise

Figure 2.22.19.

Fluorescence background test

Nano-bead test

Table 2.22.3.