Patent Application
20100283835
The present invention relates generally to microscopic imaging and, more specifically, to three-dimensional (“3D”) photoactivation localization microscope imaging.
A conventional light microscope provides a resolution on the order of about 250 nanometers in a lateral two-dimensional (“2D”) plane (i.e., x-y plane) and on the order of about 600 nanometers, or worse, in an axial direction (i.e., z axis). To obtain higher resolution images, other microscopy methods must be utilized.
A number of techniques have been able to provide a higher resolution in the axial direction to achieve an improved 3D image of structures. For example, 4Pi confocal microscopy and I5M microscopy can provide a resolution on the order of about 250 nanometers in the lateral plane (same as conventional microscopy) and on the order of about 100 nanometers in the axial direction (approximately 6-fold higher resolution in the axial direction than conventional microscopy).
4Pi-microscopy is a fluorescence far-field technique that enhances the diffraction-limited optical resolution of confocal laser scanning microscopy in the axial direction by utilizing two opposing objectives. For example, 4Pi-microscopy is described in an article titled “Fundamental improvement of resolution with a 4Pi-confocal fluorescence microscope using two-photon exciation” by S. Hell and E. H. K. Stelzer (1992, Opt. Comm. 93: 277-282), which is incorporated herein by reference in its entirety. A coherent superposition of two counter-propagating wavefronts at a common focus point modulates an axially elongated focus of a single objective resulting in a sharp central maximum and two axially shifted side-maxima. By applying 2-photon excitation (“2PE”), the side-maxima height can be lowered and the artifacts produced by the side-maxima can be removed by post-processing, resulting in 3D imaging with an axial resolution of 100 to 150 nanometers. A more detailed description of the 2PE is described in an article titled “Two-photon laser scanning fluorescence microscopy,” by W. Denk, J. H. Strickler, and W. W. Webb (1990, Science 248, 73-76), which is incorporated herein by reference in its entirety.
I5M is a “wide-field version” of 4Pi that can utilize a spatially incoherent light source, e.g., a mercury lamp, and a CCD camera instead of a confocal scanner in the same kind of interference geometry. While the high I5M side-maxima are often difficult to remove, localization measurements of single particles can be performed more quickly. For example, I5M microscopy is described in an article titled “I5M: 3D widefield light microscopy with better than 100 nm axial resolution” by M. G. L. Gustafsson, D. A. Agard, and J. W. Sedat (1999, J. Microsc. 195: 10-16), which is incorporated herein by reference in its entirety.
A “Type C” mode of 4Pi microscopy, which increases 4Pi-resolution by 30% down to about 100 nanometers, is also the base for implementation of I5M into current 4Pi microscopes. A more detailed description of this 4Pi microscopy is described in an article titled “Cooperative 4Pi excitation and detection yields 7-fold sharper optical sections in live cell microscopy” by H. Gugel, J. Bewersdorf, S. Jakobs, J. Engelhardt, R. Storz, and S. W. Hell (2004, Biophys. J. 87, 4146-4152), which is incorporated herein by reference in its entirety. 4Pi Type C imaging recently identified histone H2AX structures, with improvements in imaging resolution that enabled the definition of boundaries of γ-H2AX spreading based on their size and distribution. A more detailed description of these improvements is included in an article titled “Novel H2AX Chromatin Structures Revealed by 4Pi Microscopy,” by J. Bewersdorf, B. T. Bennett, and K. L. Knight (2006, Proc. Nat. Acad. Sci., doi:10.1073/pnas.0608709103), which is incorporated herein by reference in its entirety.
A further improvement in the lateral resolution has been achieved by utilizing Fluorescence Photo Activation Localization Microscopy (“FPALM”). Specifically, FPALM imaging techniques can be utilized to facilitate visualization of multiple structures down to about a demonstrated 30 nanometer length scale (and smaller in principle) in the cell nucleus. Thus, FPALM provides nearly a one order of magnitude higher resolution in the lateral plane than conventional, 4Pi, or I5M microscopy. However, a problem with current FPALM imaging techniques is that it does not provide improved 3D resolution in the axial direction. Although FPALM can provide a 3D image of a particle by combining images of multiple 2D planes of a sample, this type of image has only a very poor resolution in the axial direction comparable to conventional microscopy because it fails to provide any axial resolution within any of the combined images. In other words, FPALM can only image a structure that can be observed in a particular slice of a sample, i.e., within a particular 2D plane of the sample, with improved resolution. FPALM cannot identify any fine structure along the axial direction of the particular slice.
Thus, FPALM has not yet been extended to 3D imaging. This is due to the currently used standard widefield or total internal reflection (TIR) microscopy geometries not being suitable to 3D FPALM imaging.
FPALM measures the positions (by localization) of large numbers (e.g., 104 to 106) of single fluorescent molecules to generate images with nanometer resolution (demonstrated resolution 20-30 nanometers). FPALM is predicated on single molecule detection methods that allow position localization of spatially isolated fluorophores with much higher precision than the theoretical resolution limit of a light microscope.
The key to high resolution FPALM is the activation of only a sparse subset of available fluorescent probes at any one time, such that nearly all fluorophores are spatially separated by at least the optical resolution of the system and can be localized individually. The molecules are generally localized using single molecule detection methods. Fluorescent probes, such as photoactivatable green fluorescent protein (“PA-GFP”), are initially in a non-fluorescent or weakly fluorescent state. A more detailed description of some fluorescent probes is provided in an article titled “A photoactiavatable GFP for selective photolabeling of proteins and cells,” by G. H. Patterson and J. Lippincott-Schwarz (2002, Science 297, 1873-1877), which is incorporated herein by reference in its entirety. Wide-field irradiation with a short wavelength (˜405 nanometers) source activates a stochastic, sparse subset of the fluorophores (irreversibly in the case of PA-GFP) in the field of view. The activated molecules are then excited as for normal single molecule fluorescence using a longer wavelength (˜480 to 500 nanometers) source, and then imaged with a sensitive CCD camera until photobleached. Photobleaching may be actively induced, or may occur spontaneously after a certain number of fluorophore excitation-emission cycles. This process is repeated until the entire probe population is exhausted, or until sufficient information about the sample has been obtained. The positions of observed and localized molecules are then plotted to construct a 2D map.
The resolution in FPALM is referred to as a localization-based resolution, in contrast to the diffraction-limited optical resolution of conventional microscopes. While the localization-based resolution is approximately proportional to the optical resolution of the microscope, it can exceed the optical resolution by a large factor (e.g., more than a factor of 10) if a sufficient number of photons is collected per fluorescent molecule. A more detailed description of the localization-based resolution is provided in an article titled “Precise Nanometer Localization Analysis for Individual Fluorescent Probes,” by R. E. Thompson, D. R. Larson, and W. W. Webb (2002, Biophys. J. 82: 2775-2783), which is incorporated herein by reference in its entirety.
Thus, there is a need for improved ultra high 3D resolution in fluorescence microscopy. None of the current fluorescence microscopy techniques can enhance resolutions of light microscopy to levels of electron microscopy. A need exists for a fluorescence microscopy technique that can resolve particles in 3D with a resolution higher than the current resolution of 100 nanometers in 4Pi fluorescence microscopy. The present invention is directed to satisfying one or more of these needs and solving other problems.
According to one embodiment, a method of performing 3D photoactivation microscope imaging includes providing a sample having a plurality of probes, each of the plurality of probes including a photo-activatable material. Probes from the plurality of probes are activated to form a sparse subset of probes, the sparse subset of probes having probes that are spatially separated by at least a microscope resolution. The sample is illuminated with a readout light source, and light emitted from activated probes is detected. Based on the light emission detected from the activated probes, localized three-dimensional positions of the activated probes are obtained.
According to another embodiment, a photoactivation microscope system is configured to generate 3D images of a sample. The microscope system includes an activation light source, a readout light source, a spatially resolving detector, and a controller. The activation light source is configured to activate probes of at least one probe subset, the probes being selected from a plurality of photo-activatable probes within a sample. The readout light source is configured to illuminate the sample with a readout light. The spatially resolving detector is configured to detect light emission from activated probes. The controller is configured to construct a 3D image of the sample in accordance with localized three-dimensional positions of the activated probes. The localized three-dimensional positions are based on the detected light emission from the activated probes.
According to yet another embodiment, a method of performing 3D photoactivation microscope imaging includes providing a sample having a plurality of probes, each of the plurality of probes including a photo-activatable material. A first subset of probes, from the plurality of probes, is activated, and the sample is illuminated with a readout light source. Light emitted by the activated probes of the first subset of probes is detected. A localized three-dimensional position is obtained for at least one of the activated probes of the first subset of probes to form a first detected image. The activated probes of the first subset of probes are deactivated and a second subset of probes, from the plurality of probes, is activated. Light emitted by the activated probes of the second subset of probes is detected. A localized three-dimensional position is obtained for at least one of the activated probes of the second subset of probes to form a second detected image. A 3D image is constructed based on the first detected image and the second detected image.
Additional aspects of the invention will be apparent to those of ordinary skill in the art in view of the detailed description of various embodiments, which is made with reference to the drawings, a brief description of which is provided below.
While this invention is susceptible of embodiment in many different forms, there is shown in the drawings and will herein be described in detail preferred embodiments of the invention with the understanding that the present disclosure is to be considered as an exemplification of the principles of the invention and is not intended to limit the broad aspect of the invention to the embodiments illustrated.
According to one embodiment, Three-Dimensional Fluorescence Photo Activation Localization Microscopy (“3D FPALM”) methods provide a localization-based resolution of at least 30 nanometers in three-dimensions (“3D”) that may be achieved on a single molecule level. In fact, the 3D FPALM methods can resolve particles on the order of about 20-30 nanometers in the lateral plane (more than 10 times greater than conventional, 4Pi, and I5M microscopy) and on the order of about 10 nanometers in the axial direction (about 60 times greater than conventional microscopy and about 10 times greater than 4Pi and I5M microscopy). The axial resolution of 3D FPALM is something that has simply not been achieved by previous FPALM methods, which are unable to resolve particles in the axial direction, or any other light microscopy technique.
It has been found that 2-photon excitation (“2PE”) can activate photoactivatable (“PA”) green fluorescent protein (“GFP”) in a single focal plane that has a thickness of only a few hundred nanometers. Accordingly, combining features of 2PE, 4Pi, I5M, and FPALM into a single instrument can facilitate the imaging of nuclear structures, such as GFP-tagged histones, with unprecedented resolution.
According to an alternative embodiment, a photoactivation process that depends non-linearly on the light source intensity can be used to activate probes (e.g., PA fluorescent molecules) within a sample. A sparse subset of the probes is activated within a volume constrained within a depth of the sample, wherein the probes are predominantly spatially separated by at least a lateral resolution of the microscope, e.g., 250 nanometers.
The sample is illuminated with a readout light source and light emission from the activated probes is detected. The activated probes are deactivated and localized positions of the activated probes, for which light emission was detected, are obtained (e.g., calculated) based on the detected light emission. The activation, readout, and detection steps can be performed for different volumes constrained within different depths of the sample, wherein the different volumes include different sparse subsets of probes. A 3D image of the sample is constructed based on the obtained localized positions of the probes from different volumes constrained within different depths of the sample.
For example, according to one embodiment at least one of an illumination light pattern and a detection efficiency pattern is modulated in an axial direction. At least two different axial patterns are applied for the sparse subset of probes, and images are collected via a detector for each applied pattern. Based on the collected images, probes of the sparse subset of probes are located in three-dimensions.
In alternative embodiments, an I5M setup can be used with FPALM to enable a readout technique that can provide a localization-based axial resolution equivalent or even superior to the lateral direction. For example, a Leica TCS 4PI microscope featuring 2PE, 4Pi, and I5M imaging modes is an ideal base for realizing 3D FPALM.
Referring to
The CCD camera 102 is used as a spatially resolving detector and is configured to detect light emission from activated probes of a sample (e.g., sample 142). The activated probes are generally selected from a plurality of photo-activatable probes, such as fluorescent molecules. The argon-ion laser 108 is a laser source that is used as a readout light source and is configured to excite probes for emission from the sample 142.
To use the camera port 103 for imaging and the laser scanner 110 for 2PE activation from a 2PE activation laser source (e.g., a pulsed Ti:sapphire laser) 112, without moving parts mechanically in between, the IR beam splitter 106 is inserted into the camera transmission path. A computer 114 with Leica-compatible software code is added to read out the EM-CCD camera 102, to coordinate with other microscope components including a z-stage 116 and laser shutters, and to control the interference phase by moving a piezo-mounted mirror 118. Accordingly, the computer 114 is communicatively coupled to the EM-CCD camera 102 and to a position sensitive detector 122.
A feedback control including a laser diode 120 and the position sensitive detector 122 that detects the laser light transmitted through two objectives (i.e., a first objective 146 and a second objective 144) can be included to compensate for drift with increased acquisition times. The argon-ion laser 108 (e.g., operating at 488, 477, or 496 nanometers) can be coupled into an Hg-lamp port for readout of the switched-on PA-GFP. One or more filters 124 can be added to the microscope at different locations to filter out laser light. Optionally, the interference phase can be shifted by moving the piezo mirror 118 over several periods to filter out other signal-influencing effects, such as blinking and bleaching.
The 3D FPALM microscopy system 100 can further include other components, such as a 4Pi control box 126, a Leica TCS 4Pi computer 128, a plurality of lenses 130-134 (including conventional first lens 130, second lens 131, and third lens 132), a first mirror 136, a second mirror 137, a plurality of glass wedges 138, a piezo objective stage 140, the first objective 144, the second objective 146, a glass window 148, and a beam splitter 150. Other components can include shutters and optics for coupling the laser 108. The 4Pi control box 126 is communicatively coupled to the computer 114, the Leica TCS 4Pi computer 128, the piezo z-stage 116, and the piezo mirror 118. The 4Pi control box 126 controls the moving of a sample on the piezo z-stage 116. In addition to the 4Pi control box 126, the Leica TCS 4Pi computer 128 is communicatively coupled to the computer 114. One or more of the computer 114 and the Leica TCS 4Pi computer 128 can be configured (e.g., programmed) to construct a 3D image of the sample 142 based on detected light emission from the activated probes.
An exit pupil 152 of the confocal microscope is shown in the 4Pi module above the first lens 130. An intermediate image plane 154 is shown between the first mirror 136 and the beam splitter 150. The sample 142 is shown on the z-stage 116, between the first objective 144 and the second objective 146. The z-stage 116 remains stationary during detection of probes from a sparse subset of probes.
The 3D FPALM microscopy system 100 is operated to perform volumetric imaging (e.g., 3D imaging). PA-GFP molecules in a lateral optical section of the 3D volume can be switched on by scanning the field with the 2PE activation 112 in 4Pi mode in this plane. The fluorescence signal can be read out in I5M mode (analogous to the widefield mode of regular FPALM) by illuminating the sample 142 with the beam of the argon-ion laser 108 (made spatially incoherent) in Koehler mode and detecting the light with the EM-CCD camera 102. The lateral position of the molecules can be determined analogously to conventional FPALM.
To obtain the axial position of the molecules, images at different I5M interference phases are recorded. To avoid axial localization artifacts due to blinking and bleaching of the molecules, approximately 5 runs of the phase scan (which are not scans of the z-stage 116) can be performed, resulting in 10 to 100 images per activation cycle. For lateral localization, these images can be added up, while for axial localization the interference modulation function can be fitted to the profiles. The axial localization accuracies are about 10 nanometers, based on the higher axial resolution of I5M compared to the lateral detection of point-spread functions (“PSF”) profiles.
The axial position of the molecules can be obtained at least in part by measuring a phase shift of interference phase modulation. Analogous to lock-in amplifiers in electronics, the periodic modulation of the signal by varying the interference phase additionally allows differentiating between diffuse background and localized signal. Molecules outside the 2PE plane are not activated and will not be bleached by the read-out light. The activation-recording cycle can be repeated analogous to conventional FPALM (but in planes at different axial positions in the sample 142) after most of the molecules have been bleached. Combining the information of these cycles can result in the construction of a final 3D position map of PA-GFP molecules.
Thus, the disclosed methods of 3D FPALM, such as the exemplary 4Pi FPALM setup, can provide 3D images with at least 30 nanometers resolution in all directions of PA-GFP labeled nuclear structures (e.g., histones). A second color that labels, for example, a certain small gene domain with quantum dots can be imaged conventionally and allows its localization relative to the highly resolved chromatin structure.
Optionally, if background fluorescence is not too strong, imaging can be performed with whole cells. Alternatively, cells can be cut in sections having a thickness of about 1 micrometer. Background fluorescence can be suppressed by pre-bleaching the sample 142 before photoactivation. The signal can also be filtered out from diffuse background by its phase-variation dependency.
Imaging of a chromosome territory (“CT”) having a thickness of about 1 micrometer may take approximately 1 to 2 hours (assuming 10 activation cycles/second). Drift, which may occur during acquisition times, can be counteracted by a feedback control that locks the relative objective positions. The feedback control, for example, is controlled by the computer 114 and includes the laser diode 120, two lenses 133-134, one filter 124, and the position sensitive detector 122. Sample and phase drift can be corrected by imaging non-bleaching (fiduciary) reference structures at defined time intervals and recording their positions. Ambiguity in the axial position due to activation of molecules by the side maxima of the 2PE 4Pi-PSF can be countered by carefully analyzing the variation of the signal fluctuations with the phase variation. Molecules that cannot be localized with sufficiently high certainty can be analyzed separately or excluded from the final combined position map. This does not change the overall recorded structure since this process is position independent.
Referring to
In general, the only moving component during a phase scan is the piezo mirror 216, which moves from a first position (shown in dashed lines) to a second position (shown in solid lines). The sample 222 (and corresponding stage) does not move during the phase scan. As described in more detail below, the movement of the piezo mirror 216 results in a variance in the detected fluorescence signal for a number of particles. Based on the variance in the detected fluorescence signal, the particles are imaged in 3D. More specifically, the movement of the piezo mirror 216 shifts the interference pattern in the sample created by coherent superposition of two counter-propagating light wavefronts via the two opposing objective lenses, the first objective lens 218 and the second objective lens 220.
Referring to
Referring to
As shown, in the first position of the piezo mirror 216 (
The middle particle is shown near a peak in the first position of the piezo mirror 216 (
Referring to
For example, the fluorescence signal for the top particle is a line on which intensity points 300a, 300b are located. Consistent with the description provided above, the top particle has a generally intermediate intensity 300a in the first position of the piezo mirror 216 and a somewhat similar intensity 300b in the second position of the piezo mirror 216. As shown, in the first position the intermediate intensity 300a is on a right side of a peak, between the adjacent peak and the adjacent valley of the respective fluorescence signal. In the second position, the intermediate intensity 300b is on a left side of a peak, between the adjacent valley and the adjacent peak of the respective fluorescence signal.
Also consistent with the description provided above are the chart illustrations of the middle particle and the bottom particle. As shown, the middle particle is shown having the high intensity 302a, which is illustrated in a peak of the respective fluorescence signal, and the low intensity 302b, which is illustrated in a valley of the respective fluorescence signal. Similarly, the bottom particle is shown having the low intensity 304a, which is illustrated in a valley of the respective fluorescence signal, and the high intensity 304b, which is illustrated in a peak of the respective fluorescence signal.
Various aspects of the present invention may be used alone, in combination, or in a variety of arrangements not specifically discussed in the embodiments described in the foregoing, and the aspects of the present invention described herein are not limited in their application to the details and arrangements of components set forth in the foregoing description or illustrated in the drawings. The aspects of the invention are capable of other embodiments and of being practiced or of being carried out in various ways.
Each of these embodiments and obvious variations thereof is contemplated as falling within the spirit and scope of the claimed invention, which is set forth in the following claims.
This application is related to and claims priority to U.S. Provisional Patent Application Ser. No. 60/879,833 filed Jan. 11, 2007, titled “Microscopic Imaging Techniques,” which is incorporated herein in its entirety.
| Number | Date | Country | |
|---|---|---|---|
| 60879833 | Jan 2007 | US |
