Molecular-level resolution microscopy method based on two-color single-molecule positioning

Abstract

The invention discloses a molecular resolution microscopy (MITI) (Molecular-level Imaging Based on Twoo-color Imaging) based on two-color single molecule positioning, and relates to the technical field of microscopy, in particular to the MITI (Molecular-level Imaging Based on Twoo-color Imaging) based on two-color single molecule positioning. According to the invention, by using a two-color imaging method, the positioning points of the fluorescent molecules which cannot be distinguished originally successfully pass through a two-color method, different positioning point sets are distributed to corresponding fluorescent channels for successful distinguishing, and then the positioning points of the fluorescent molecules are obtained by fitting the distribution of the positioning points of the fluorescent molecules again. And the center position of the fluorescent molecule positioning point distribution and the positioning precision improved again are obtained, so that the resolution is improved. According to the method, the spatial resolution of 2nm can be achieved at most, and an important means is provided for analysis of a molecular-scale biological structure and an interaction process.

Description

Technical Field

[0001] The invention belongs to the technical field of super-resolution microscopy, and in particular relates to a molecular-level resolution microscopy method based on dual-color single-molecule positioning. Background Art

[0002] In the past two decades, super-resolution microscopy has become a key means to reveal complex biological structures and processes, and has been widely developed and applied. However, although the resolution of traditional single-molecule localization technology can break the optical diffraction limit, it is still limited to resolving scales below 10nm. The resolution of SMLM (single-molecule localization microscopy) is difficult to resolve molecular-level scales, that is, the study of complex biological processes at a scale of several nanometers, such as interactions between biomacromolecules and interactions between proteins. In recent years, in order to meet the requirements of resolving biological processes at 10nm, several studies have proposed different methods to increase the resolution of optical microscopes to a few nanometers, entering the era of nanoscopes.

[0003] The method based on DNA-PIANT (DNA point accumulation in nanoscale topography) sequential imaging effectively distinguishes the distribution of two fluorescent molecules and obtains angstrom-level positioning accuracy. However, DNA-PAINT requires the use of encoded modified single-stranded DNA and access to the target point of interest. This labeling method is more challenging than using fluorescent molecules to directly label the target site of interest, and not all biological molecules are suitable for DNA oligonucleotide labeling methods. In addition, DNA oligonucleotides do not pass through the cell membrane, and the imaging chain solution needs to be constantly replaced during imaging to maintain the number of fluorescent molecules. Most importantly, using DNA-PAINT sequential imaging to distinguish the distribution of the positioning points of fluorescent molecules requires constantly replacing and washing the imaging solution, thereby increasing the imaging time, which will also cause more serious drift. Therefore, it is extremely challenging for imaging at this molecular scale.

[0004] Many methods have improved the resolution by combining with single-molecule localization. Many microscopy techniques have effectively improved the positioning accuracy of SMLM by combining the method of structured light illumination with single-molecule localization, so that their resolution can reach the range of 1-5nm. However, they all require more complex optical systems and control systems or more complex labeling methods to achieve higher resolution. Therefore, how to achieve true molecular-level resolution imaging based on the traditional single-molecule localization imaging system remains a challenging scientific problem that needs to be solved urgently. Summary of the invention

[0005] In view of the above problems of the prior art, the present invention proposes a molecular-level resolution microscopy based on dual-color single-molecule localization, which improves the microscopic capability to the molecular level through dual-color single-molecule localization technology without introducing extremely complex optical systems and special labeling methods.

[0006] To achieve the above object, the technical solution adopted by the present invention is:

[0007] A molecular-level resolution microscopy method based on dual-color single-molecule localization is based on the traditional single-molecule localization method SMLM, and uses dual-color imaging to achieve further improved resolution. The specific steps are as follows:

[0009] 1) Two-color marking target:

[0010] To achieve dual-color imaging, two different fluorescent molecules must be used to label the target. Fluorescent molecules that can be used for SMLM are needed, including typical fluorescent dyes Cy5, Cy3, Alexa 647, etc.; light-controlled fluorescent proteins PcStar, PAmCherry1, ffDronpa, etc. Use fluorescent dyes or fluorescent proteins to label the two ends of the target, and note that the two fluorescent molecules must have different excitation and emission bands.

[0011] 2) Dual-color single-molecule localization super-resolution imaging

[0012] After obtaining the dual-color labeled sample, the signals of the two fluorescent molecules are collected by super-resolution optical microscopy based on single-molecule localization. Single-molecule-based super-resolution optical microscopy is usually equipped with 4 wavelength lasers, namely 405nm, 488nm, 561nm, and 647nm, so as to activate and excite most types of fluorescent molecules. When performing dual-color imaging, there are two imaging modes, one is dual-channel parallel imaging, and the other is sequential fluorescence imaging. When dual-channel parallel imaging is performed, the corresponding two excitation lasers are turned on at the same time, and a lower intensity 405nm laser is used to make the molecular density of the two channels as close as possible. The signals of the two channels are collected at the same time for subsequent processing. When using sequential imaging, the fluorescence signal with a longer excitation wavelength is collected first, and then the fluorescence signal with a short wavelength is collected. At the same time, the density of the fluorescent molecules imaged is controlled by a low-intensity 405nm laser. Collect enough fluorescence signals for subsequent processing.

[0013] 3) MITI implementation method

[0014] After obtaining the signals of two fluorescent molecules, the super-resolution data of single-molecule localization is obtained through standard SMLM data analysis methods, including the coordinates of the localization points of the fluorescent molecules, the intensity of the fluorescent molecules, the localization accuracy and other information. The localization position information of adjacent fluorescent molecules that were originally indistinguishable is distinguished by the two-color imaging method. Then, by fitting the position information of the fluorescent molecules of the two fluorescent molecule channels respectively, the center position of the fluorescent molecule is obtained. The final localization accuracy of MITI is determined by the localization accuracy of the original SMLM fluorescent molecules and the number of localization points of the fluorescent molecules.

[0015] Since the position distribution of the fluorescent molecule's positioning points basically conforms to the Gaussian distribution, the Gaussian function is used to fit the position distribution of the fluorescent molecule's positioning points. The parameter information of the Gaussian function used comes from the coordinate information and positioning accuracy of the point set. The Gaussian function form is as follows:

[0016] Where (x0, y0) is the coordinate of the center point of the Gaussian distribution. In this function, A can be regarded as a constant, B is the background intensity, and considering that there is no background when fitting the center of the positioning point, B can be taken as 0, σ 2 is the variance of the Gaussian function in the x and y directions. We fit the Gaussian function using the maximum likelihood estimation method to obtain the center coordinates of the corresponding positioning point set of the MITI image. The positioning accuracy of the final MITI image can be expressed by the relationship: Determine, where σ SMLM is the localization accuracy of single-molecule localization imaging, and K is the total number of localization points. The images of the two channels are fitted by this method to improve the localization accuracy, and the final localization accuracy is proportional to the number of localization points K.

[0017] Finally, the resolution-enhanced images of the two channels are merged to obtain an image with molecular-level resolution, thus realizing MITI imaging.

[0018] After obtaining the fluorescent molecular signals of the two channels in step 2), the two fluorescent molecular channels need to be aligned first. Use 100nm fluorescent beads to image in the two channels respectively, and move the position of the fluorescent beads with an appropriate step size (determined by the imaging area) to cover the entire imaging area. Extract the center position of the fluorescent beads, establish a correction matrix between the two channels, and apply it to the subsequent channel alignment of the fluorescent channels.

[0019] The method for obtaining the single molecule point positioning accuracy in step 3) is usually obtained by fitting the Gaussian distribution of the diffraction limited image point spread function using the least squares method, where the root mean square error is the positioning accuracy of the single molecule point, and the relationship between the positioning accuracy of the SMLM and the point spread function of the diffraction limited function is (2):

[0020] Where s is the standard deviation of the diffraction-limited image point spread function, a is the pixel size in the image (taking into account the system magnification), N is the total number of photons measured from the fluorescent molecules, and b is the number of background photons collected from the fluorescent molecules. Therefore, in order to obtain higher localization accuracy of SMLM, when designing SMLM experiments, the highest possible diffraction-limited resolution and photon collection efficiency should be achieved, while minimizing background noise.

[0021] Compared with the prior art, the present invention has the following advantages:

[0022] The present invention utilizes a traditional single-molecule localization super-resolution microscope system, and through a dual-color imaging technique combined with a single-molecule localization imaging method, fluorescent molecules of different emission wavelengths are distinguished through two imaging channels, and the distribution of two fluorescent molecule localization points is fitted respectively, and the central position of the distribution of the fluorescent molecule localization points and the localization accuracy of the distribution of the fluorescent molecule localization points, i.e., the localization accuracy of MITI, the localization accuracy of the MITI obtained by fitting is further improved relative to the localization accuracy of the SMLM image, and the imaging resolution can be increased to a maximum of 2nm, achieving molecular-level resolution. It is used to analyze the structure and interaction process of biological macromolecules below 10nm. And this method does not need to introduce a complex optical system and a special labeling method, and can achieve molecular-level resolution on the original single-molecule localization super-resolution microscopy system. The spatial resolution capability of super-resolution microscopy based on single-molecule localization is significantly improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 Schematic diagram of the molecular-level resolution microscopy method MITI based on dual-color SMLM of the present invention.

[0024] Figure 2 It is a simulation schematic diagram of the relationship between the number of fluorescent molecule localizations and the resolution capability of MITI according to the present invention.

[0025] Figure 3 As an example of MITI microscopy according to the present invention, a MITI image of a single-stranded DNA labeled at both ends was used. DETAILED DESCRIPTION

[0026] The technical solution of the present invention is further described below in conjunction with the accompanying drawings and embodiments, but the protection scope of the present invention shall not be limited thereto.

[0027] like Figure 1 As shown, a molecular-level resolution microscopy method based on dual-color single-molecule localization includes first designing two fluorescent molecules to be labeled at both ends of the target, wherein the distance between the two different fluorescent molecules is 2≤2.35σ SMLM, which is the distance that cannot be distinguished by SMLM. In order to verify whether it can be distinguished by MITI, the targets with different fluorescent molecules at both ends are arranged horizontally, where distance 1 is the distance from fluorescent molecule 1 to the fluorescent molecule 1 of another target, and distance 1>2.35σ SMLM , at this time, the two targets with fluorescent molecules arranged in this way are imaged. Channel 1 collects the signal of fluorescent molecule 1, and channel 2 collects the signal of fluorescent molecule 2. Due to the resolution limit of single-molecule localization super-resolution imaging, when distance 1 is designed to be 40nm, the positioning points of two adjacent fluorescent molecules 1 can be distinguished, and will not be indistinguishable due to overlap. Similarly, when the distance between two fluorescent molecules 2 is 40nm, the positioning points of two adjacent fluorescent molecules can be distinguished. However, it can be imagined that if channel 1 and channel 2 are not imaged separately, the positioning points of fluorescent molecules 1 and fluorescent molecules 2 at distance 2 will overlap and cannot be distinguished. After the two channels generate the corresponding SMLM images by generating histogram distributions from these points respectively, the center positions of these positioning points can be fitted to obtain the center positions of the fluorescent molecules in the final MITI image. Since the center positions and positioning accuracy of the fluorescent molecule positioning points are obtained by distinguishing the positioning points of the fluorescent molecules, the original SMLM positioning accuracy is improved again, which is the positioning accuracy of MITI, which can significantly improve the positioning accuracy of the image. After obtaining MITI images in two channels, the MITI images in the two channels are finally merged to obtain a MITI image with improved resolution. In the MITI image, a distance of 2nm that SMLM cannot resolve at all can be resolved. This method, which distributes the signals of different fluorescent molecules into different channels, fits them separately to improve the positioning accuracy, and finally merges them to obtain a higher-precision image, successfully improves the imaging resolution based on dual-color single-molecule positioning to the molecular level.

[0028] In order to explain the relationship between the positioning number K and the resolution of the MITI image, according to Figure 2 The simulation results can be seen.

[0029] exist Figure 2 When K = 1, it is the super-resolution image of the traditional SMLM. Figure 1The principle of the MITI method described above is to assign different fluorescent molecular signals to different channels for analysis. First, the distance between two adjacent identical fluorescent molecules is set to 30nm, which is the limit distance of the resolution of an SMLM super-resolution image. Then the distance between two adjacent different fluorescent molecules is set to 2nm, which is the molecular-level resolution distance that a MITI image expects to achieve. The image at the bottom is the image after the two channels are merged. When K=1, the images of the two channels are almost completely overlapped and cannot be distinguished. When the localization number K is increased to 50, the resolution of the fluorescent molecular images of the two channels is significantly improved. The 30nm distance that can barely be separated at the resolution of traditional SMLM can now be completely separated, but the image merged from the two channels still overlaps and cannot be distinguished. When the localization number K is increased to 100 again, the fluorescent molecules merged from the two channels begin to have distinguishable outlines, but in fact there is still a lot of overlap and cannot be distinguished. When the localization number K is increased to 150, the fluorescent molecules merged from the two channels can be completely distinguished, proving that when the localization number is increased to more than 150, the imaging resolution of MITI can reach 2nm. The relationship between the localization number K and the MITI resolution satisfies the formula:

[0030]

[0031] Therefore, the simulation results provide a theoretical basis for the relationship between the positioning number K and the resolution of MITI, and also confirm that Figure 1 The proposed idea is correct.

[0032] according to Figure 1 and Figure 2 The proposed ideas and theoretical verification are Figure 3 In the process, a single-stranded DNA with fluorescent dyes Cy3 and Cy5 that can be distinguished and labeled at both ends was designed, and a single-stranded DNA with a length of 6 bases (bp) was designed. It is known that the length of a base is approximately 0.314nm, so a single-stranded DNA of 6bp is approximately 2nm in length. First, wide-field imaging of this sample was performed to obtain a diffraction-limited image. It can be clearly concluded that the wide-field microscopy image cannot resolve a single-stranded DNA sample with a length of 6bp. Further, a stochastic optical microscopy (STORM) super-resolution image was obtained through single-molecule localization reconstruction. At a resolution of 30nm, it is obviously impossible to resolve the information of the sample at all, and the two fluorescent molecules completely overlap. Finally, according to the molecular-level resolution microscopy based on dual-color single-molecule localization of the present invention, MITI. In the MITI image, the sample Cy3-6bp-Cy5 with a length of 6bp and approximately 2nm was successfully resolved.

[0032] When performing the imaging experiment of sample Cy3-6 bp-Cy5, the specific steps are as follows:

[0033] Since the STORM super-resolution image is ultimately reconstructed from many frames of wide-field images through single-molecule localization, MITI is an image enhancement image obtained by refitting the set of localization points of fluorescent molecules in the single-molecule localization super-resolution image. Therefore, the imaging process required by the MITI method is consistent with that of STORM.

[0034] First, prepare the STORM imaging buffer solution before each imaging.

[0035] After the sample is diluted to an appropriate concentration, it is mixed with STORM imaging buffer, and fluorescent beads diluted to a certain concentration are added to prepare for the subsequent drift correction process.

[0036] In dual-color imaging, time-series imaging is used, using a 647nm laser to excite the Cy5 fluorescent molecules, while a low-intensity 405nm laser is used to keep the Cy5 fluorescent molecules at an appropriate density.

[0037] After collecting the signal of the Cy5 fluorescent molecule, turn off the 647nm laser, use the 561nm laser to excite the Cy3 fluorescent molecule, and use a low concentration of 405nm laser to keep the Cy3 fluorescent molecule at an appropriate concentration.

[0038] After collecting enough fluorescent molecule signals from two channels, STORM image reconstruction is performed to obtain the coordinate information, fluorescence intensity information, and positioning accuracy information of the fluorescent molecules in the corresponding channels.

[0039] First, use fluorescent beads to correct the chromatic aberration between the two channels and perform channel alignment.

[0040] After completing the chromatic aberration correction, the position change of the fluorescent ball in each frame is analyzed and the fluorescent ball is used as a standard for drift correction.

[0041] Note that due to the longer imaging time required for sequential imaging, drift correction is more important. Therefore, when performing MITI imaging, dual-color parallel imaging is preferred.

[0042] After chromatic aberration correction and drift correction, the coordinate information of the fluorescent molecules is extracted and the center position of the fluorescent molecule positioning point set is fitted by the Gaussian function using the least squares method. The positioning accuracy of the MITI image is obtained through the relationship between the positioning number K and the positioning accuracy of MITI. Finally, the fluorescent molecule information of the two channels is reconstructed and merged to obtain the MITI image.

[0043] Finally, it should be noted that the purpose of publishing the embodiments is to help further understand the present invention, but those skilled in the art can understand that various substitutions and modifications are possible without departing from the spirit and scope of the present invention and the appended claims. Therefore, the present invention should not be limited to the contents disclosed in the embodiments, and the scope of protection claimed by the present invention shall be subject to the scope defined in the claims.

Claims

1. A molecular-resolution microscopy method based on dual-color single-molecule localization, characterized in that: The steps include: S1. Dual-color labeling of the target: using two different fluorescent molecules to label the target, the two fluorescent molecules have different excitation and emission bands and are suitable for single-molecule localization microscopy; S2. Dual-color single-molecule localization super-resolution imaging: Use a super-resolution optical microscope equipped with at least four wavelength lasers to image dual-color labeled samples and collect signals from two fluorescent molecules. Imaging methods include dual-channel parallel imaging or sequential imaging. S3.MITI implementation method: Through standard single-molecule localization microscopy data analysis methods, the localization information of fluorescent molecules is obtained, including the coordinates of the localization point, the intensity of the fluorescent molecules, and the localization accuracy. The localization position information of adjacent fluorescent molecules is distinguished by two-color imaging, and the fluorescent molecule position information of the two fluorescent molecule channels is fitted separately to obtain the center position of the fluorescent molecules, thereby realizing MITI imaging with molecular-level resolution.

2. The molecular-level resolution microscopy method based on dual-color single-molecule localization according to claim 1, characterized in that: It also includes channel alignment and positioning accuracy optimization: before MITI imaging, fluorescent beads of known size are used to image in two channels separately, a correction matrix is ​​established to achieve precise alignment between the two fluorescent molecule channels, and the positioning accuracy of single molecule points is optimized by fitting the diffraction limited image point spread function of the Gaussian distribution.

3. The molecular-level resolution microscopy method based on dual-color single-molecule localization according to claim 1, characterized in that: The two different fluorescent molecules described in step S1. are selected from fluorescent dyes Cy3, Cy5, Alexa 647, or light-controlled fluorescent proteins PcStar, PAmCherry1, ffDronpa.

4. The molecular-level resolution microscopy method based on dual-color single-molecule localization according to claim 1, characterized in that: The super-resolution optical microscope in step S2 is equipped with four wavelength lasers of 405 nm, 488 nm, 561 nm and 647 nm, respectively, which are used to activate and excite different types of fluorescent molecules.

5. The molecular-level resolution microscopy method based on dual-color single-molecule localization according to claim 1, characterized in that: The positioning information in step S3 also includes the distribution information of the positioning points of the fluorescent molecules. The position distribution of the fluorescent molecule positioning points is fitted by a Gaussian function, and the central coordinates of the corresponding positioning point set of the MITI image are obtained using the maximum likelihood estimation method.

6. The molecular-level resolution microscopy method based on dual-color single-molecule localization according to claim 2, characterized in that: The size of the fluorescent sphere is 100 nm. The position of the fluorescent sphere is moved to cover the entire imaging area, the center position of the fluorescent sphere is extracted, and a correction matrix between two channels is established.

7. The molecular-level resolution microscopy method based on dual-color single-molecule localization according to claim 2, characterized in that: The positioning accuracy optimization method includes using the least squares method to fit the Gaussian distribution of the diffraction-limited image point spread function, and the positioning accuracy meets , where K is the total number of positioning points of the same fluorescent molecule.

Images