参见作者原研究论文

本实验方案简略版
Mar 2021

本文章节


 

A High-throughput Pipeline to Determine DNA and Nucleosome Conformations by AFM Imaging
通过 AFM 成像确定 DNA 和核小体构象的高通量管道   

引用 收藏 提问与回复 分享您的反馈 Cited by

Abstract

Atomic force microscopy (AFM) is a powerful tool to image macromolecular complexes with nanometer resolution and exquisite single-molecule sensitivity. While AFM imaging is well-established to investigate DNA and nucleoprotein complexes, AFM studies are often limited by small datasets and manual image analysis that is slow and prone to user bias. Recently, we have shown that a combination of large scale AFM imaging and automated image analysis of nucleosomes can overcome these previous limitations of AFM nucleoprotein studies. Using our high-throughput imaging and analysis pipeline, we have resolved nucleosome wrapping intermediates with five base pair resolution and revealed how distinct nucleosome variants and environmental conditions affect the unwrapping pathways of nucleosomal DNA. Here, we provide a detailed protocol of our workflow to analyze DNA and nucleosome conformations focusing on practical aspects and experimental parameters. We expect our protocol to drastically enhance AFM analyses of DNA and nucleosomes and to be readily adaptable to a wide variety of other protein and protein-nucleic acid complexes.

Keywords: Atomic force microscopy (原子力显微镜), AFM (原子力显微镜), DNA (脱氧核糖核酸), Nucleosome (核小体), Image analysis (图像分析)

Background

Nucleosomes are the basic units of compaction of eukaryotic DNA into chromatin and function as regulators of gene readout and activity (Bowman and Poirier, 2015; Sadakierska-Chudy and Filip, 2015; Baldi et al., 2020). Canonical nucleosome core particles consist of two copies each of the four histones H2A, H2B, H3, and H4 assembled into a histone octamer that is tightly wrapped by 147 bp of DNA (Luger et al., 1997; Richmond and Davey, 2003). Accessibility to the genetic code for readout and processing is facilitated by (partial) unwrapping of nucleosomal DNA and can be achieved either by active processes involving, e.g., RNA polymerase or nucleosome chaperones that exert forces and torques on the nucleosomes (Sirinakis et al., 2011; Mueller-Planitz et al., 2013; Narlikar et al., 2013; Ordu et al., 2016), or spontaneously by thermal fluctuations (Li and Widom, 2004). Using single-molecule micromanipulation techniques such as optical tweezers, the energetics of force-induced nucleosome unwrapping has been probed at high resolution (Mihardja et al., 2006; Hall et al., 2009; Schlingman et al., 2014; Ngo et al., 2015; Kaczmarczyk et al., 2020). However, the unwrapping landscape in the absence of force has been more difficult to access.


Atomic force microscopy (AFM) is a powerful tool to probe nucleosome structure and interactions due to its capability to image molecular complexes at the single-molecule level, label-free, and with sub-nanometer resolution, well suited to visualize the DNA and protein components of nucleosomes (Shlyakhtenko et al., 2009; Bintu et al., 2011; Miyagi et al., 2011; Katan et al., 2015; Ordu et al., 2016). Recent improvements in hardware make fast imaging of thousands of molecules possible, and combination with automated image analysis enables highly quantitative and reproducible studies of DNA and nucleoprotein complexes (Dufrêne et al., 2017; Ando, 2018; Brouns et al., 2018; Würtz et al., 2019; Bangalore et al., 2020).


By combining large field of view AFM imaging and automated image analysis of DNA and nucleosomes, we have recently elucidated the nucleosomal wrapping landscape for passive invasion of nucleosomes with linker DNA, in contrast to the previous force-induced unwrapping assays (Konrad et al., 2021). While we have demonstrated the strength of our methodology by quantitatively capturing the conformational ensemble of wild-type and CENP-A nucleosomes – centromeric nucleosomes where histone H3 is replaced with the CENP-A variant – the methodology can be easily adapted to study a wide range of open questions such as the effect of post-translational modifications on nucleosome wrapping or the impact of DNA sequence on nucleosome positioning on the single-molecule level.


The current protocol describes all steps necessary to study DNA and nucleosomes by AFM imaging, starting with the surface deposition of the molecules and ending with the quantitative image analysis after AFM imaging. The protocol describes AFM imaging of dry samples in air. However, it can be readily adapted for AFM measurements in liquid. In liquid, the deposition protocol and imaging parameters have to be adjusted. In particular, instead of drying the surface after depositing the sample, the sample buffer solution remains on the surface for imaging. Examples of how to perform liquid AFM measurements have been published previously (Bussiek et al., 2003; Brouns et al., 2018). Subsequent analysis of the AFM images might also require adjustment of the image analysis parameters (see AFM image analysis).

Materials and Reagents

  1. For surface deposition of the sample

    1. Mica Grade V-1 25 mm discs (SPI Supplies, catalog number: 01926-MB)

    2. Marking tape ROTI (Carl Roth, catalog number: 8000.1)

    3. 50 ml irrigation syringes (Braun, catalog number: 4617509F)

    4. Parafilm (Carl Roth, catalog number: H666.1)

    5. Protein LoBind Tubes 0.5 ml (Eppendorf, catalog number: 0030108094)

    6. Petri dishes (Carl Roth, catalog number: 0690.1)

    7. Kimwipes (Kimtech, catalog number: 5511)

    8. Milli-Q H2O (Merck, catalog number: Z00Q0V0WW)

    9. Poly-L-lysine (Sigma Aldrich, catalog number: P0879 – diluted to 0.01% in Milli-Q H2O)

    10. N2 gas (to blow dry the surface)

    11. DNA/nucleosome sample – prepared as described previously (Krietenstein et al., 2012)

    12. Ethanol (Carl Roth, catalog number: T171.4 – diluted to 80% with Milli-Q water)

    13. Deposition buffer (see Recipes)


  2. For AFM imaging

    1. Glass slides (Thermo Scientific Menzel, catalog number: 15998086)

    2. Double-sided adhesive discs (SPI Supplies, catalog number: 05095-AB)

Equipment

  1. Self-closing tweezers (SPI Supplies, catalog number: SN5AP-XD)

  2. Vortex mixer (Scientific Industries, catalog number: SU-0236)

  3. Centrifuge (to fit 0.5 ml Eppendorf tubes and spin down tube content; Carl Roth, catalog number: T464.1)

  4. Tweezers ESD-safe (SPI Supplies, catalog number: 0CFT07PE-XD)

  5. AFM cantilevers for high-speed imaging in air; we used FASTSCAN-A (resonance frequency 1400 kHz, spring constant 18 N/m; Bruker) or AC160TS (200-400 kHz, 26 N/m; Olympus) cantilevers

  6. Imaging AFM; we employed a Nanowizard Ultra Speed 2 (Bruker) and a MultiMode 8 (Bruker)

Software

  1. Software to plane-correct the raw AFM data (see section “AFM image analysis”)

  2. Microsoft Excel

  3. Python 3

  4. Software toolbox to analyze DNA and nucleosomes in AFM images (previously described in Konrad et al., 2021), and available for download at https://github.com/SKonrad-Science/AFM_nucleoprotein_readout)

Procedure

  1. Surface deposition of the sample

    Note: Contamination can affect the quality of the imaging surface and thus imaging quality in general. It is therefore important that all instruments are kept clean throughout the process.

    1. Clean the workbench with ethanol and kimwipes thoroughly. Flush the tip of the self-closing tweezers with ethanol and blow-dry the tweezers with N2 gas. Place the tweezers on a kimwipe such that the tip does not get contaminated and does not touch the bench (Figure 1A).



      Figure 1. Surface deposition steps. A. Overview of materials required for the surface deposition of the sample as described in Materials and Reagents. B. A mica disc is placed under marking tape. C. Tearing off the tape removes a layer of the mica plate and leaves behind a flat and clean surface for the subsequent sample deposition. D. Poly-L-lysine and sample solutions are pipetted on the center of the mica plate and incubated for 30 s each, with washing and drying steps E and F directly after each incubation. E. After 30 s of incubation, the surface is rinsed by gently dropping 50/20 ml Milli-Q H2O of the syringes on the surface and letting it flow off by rotating the mica plate. F. After rinsing, the surface is dried by perpendicularly pointing a nozzle with a gentle stream of N2 gas onto the surface.


    2. Place two stripes of the marking tape next to each other on the bench with a small overlap (Figure 1A) such that they are wide enough to completely cover the mica disc. Tear off part of the marking tape and put the mica disc underneath. Apply pressure such that the tape fully attaches to the surface of the mica (Figure 1B). Tear off the tape with a quick movement to cleave the mica (Figure 1C). It is important that a full layer of the mica is removed. If only part of a layer was removed or if there are small cracks on the remaining surface, repeat this step until a whole layer is removed. Store the cleaved mica disc in a Petri dish while preparing the next steps.

    3. Remove two sterile syringes from their packing and remove the plunger. Make sure that the front of the plunger and the syringe barrel do not make contact anywhere to avoid contamination. Seal the syringe barrels with the clean side of parafilm and fill with Milli-Q water. Place the plunger back in the barrel and press down the plunger such that the parafilm tears and water flows out of the syringes. Remove the parafilm and adjust the plunger such that one syringe holds 50 ml and the other syringe holds 20 ml of Milli-Q water. Filling the syringes directly from the storage bottle from the back, as opposed to drawing up fluid with the plunger, helps to avoid contaminations.

    4. Prepare a 20 µl aliquot of the 0.01% poly-L-lysine (see Materials and Reagents), shortly vortex it, and briefly spin down the content of the Eppendorf tube in a centrifuge (~2 s at 700 rcf). Keep the aliquot on ice.

    5. Pipette the 20 µl poly-L-lysine solution onto the center of the freshly cleaved mica (step 2) and incubate for 30 s (Figure 1D). Make sure not to touch the surface with the pipette. During the 30 s incubation, pick up the mica plate with the self-closing tweezers and pick up the 50 ml syringe and move to a sink or a waste container to be able to start flushing after exactly 30 s. It is important to keep the mica surface horizontal and as still as possible during the movements to ensure a high quality of the surface deposition. Flush the surface by dropping droplets from the syringe on the edge of the mica plate (not in the center where the poly-L-lysine was placed) and periodically tilting the mica plate such that the water flows off (Figure 1E). After flushing with 50 ml, make sure to leave some water on the surface to avoid unintentional drying.

    6. With the surface still covered in water, start drying the mica surface with a gentle stream of N2 gas by quickly tilting the mica surface to a vertical position and targeting the center of the mica with the stream perpendicularly, at about 2 cm distance (Figure 1F). Once the center of the surface is dry, move the stream to the edges until the mica is completely dry.

    7. Dilute the sample solution with buffer (in our case, 200 mM NaCl + 10 mM Tris) to achieve the desired concentration for surface deposition and incubate on ice for 60 s. Nucleosome samples and buffer are stored at 4°C and put on ice before starting the surface deposition. A total volume of at least 25 µl is required for surface deposition. We typically dilute 1 µl nucleosome/DNA sample solution (containing roughly 30 ng/µl 486 bp DNA and 10 ng/µl histones, corresponding to ~120 nM nucleosomes; prepared as described in Krietenstein et al., 2012) with 40 µl buffer solution resulted in a good surface density for AFM imaging (i.e., dense enough to have many molecules in one field of view but not too dense to have too many molecules overlap with each other).

    8. After 60 s of incubation on ice, pipette 25 µl of the buffered sample solution on the center of the poly-L-lysine coated mica plate and incubate for 30 s. Again, proceed as in steps 5 and 6, rinsing – this time with a 20 ml volume of Milli-Q water – and subsequently drying the surface.

    9. Store the mica disc in the Petri dish at room temperature until starting the AFM measurement. The AFM surface should always be prepared on the measurement day.


  2. AFM imaging of nucleosomes and DNA

    Note: The imaging is performed on a Nanowizard Ultraspeed 2 system, and the steps presented here will slightly vary for other instruments. The steps to start the measurement are only briefly described since they can be found in the user manual of the respective AFM system in detail. The focus of this section lies on tips and tricks on how to tune AFM imaging and what to look out for to achieve the highest image quality for large datasets of DNA and nucleosomes.

    1. To prepare the final imaging surface, place three double-sided adhesive discs on a glass slide while leaving an area in the center free (Figure 2A). Place the sample mica plate on the adhesive region such that the central area of the mica aligns with the adhesive-free region of the glass slide and apply gentle pressure on the mica above the three adhesive discs, i.e., not in the center where the sample was placed, with the tweezers to fixate the mica more strongly.



      Figure 2. Mounting the sample in the AFM. A. The mica plate with the deposited sample is fixated on a glass slide with three double-sided adhesive discs. The glass side is then placed on the AFM stage. B. The AFM cantilever is mounted onto the glass block that was delivered with the AFM system (Ultra-Speed Glassblock, JPK, catalog number: 22229-E-01). C. Subsequently, the glass block is placed in the designated spot at the bottom of the AFM scanner head. D. To finalize the setup, the scanner head is placed on the AFM stage.


    2. Install the glass slide with the mica plate on the sample holder of the AFM (Figure 2A).

    3. Place the cantilever in the cantilever holder glass block (Figure 2B) and mount the glass block in the AFM scanner head (Figure 2C). Afterwards, place the AFM head on the stage (Figure 2D).

      Start the JPK SPM Desktop software and select AC Mode Fast Imaging (Figure 3A).



      Figure 3. AFM software and settings for imaging. A. The JPK SPM Desktop (7.0.128) software has several imaging modes available. The desired mode for this experiment is AC Mode Fast Imaging. B. Via an optical microscope, a view of the AFM cantilever allows to place the cantilever centrally within the scanner head (green cross). C. Laser alignment onto the tip of the cantilever to maximize the amount of signal reflected towards the detector. D. Alignment of the detector such that the maximum of the signal is in the center of the quadrant detector. E. Cantilever calibration based on the thermal noise spectrum yields an estimated spring constant of 19.7 N/m. F. AC Feedback Mode Wizard to select the drive frequency and the drive amplitude of the AFM cantilever. G. Typical imaging settings used for our DNA and nucleosome images. H. System status after successful approach of the cantilever towards the surface. I. Scanning of the first lines of a 6 µm × 6 µm field of view displaying both image trace (left) and retrace (right).


    4. Go to the “Setup Experiment” tab and focus the green crosshair on the AFM cantilever tip as seen in the camera view (Figure 3B).

    5. Align the laser on the cantilever tip using the screws indicated by the green arrows in Figure 3C to maximize the signal that is collected by the detector. The sum of the collected signal is represented by the blue bar (Figure 3C). For FASTSCAN-A cantilevers, a sum signal between 1.3 and 1.5 is typical. Other cantilevers yield different sum signals based on the reflection achieved by the back of the cantilever tip.

    6. Align the detector such that the maximum of the signal is in the center of the quadrant detector using the screws marked by the green arrows in Figure 3D. In case that proper alignment of the laser is not possible, adjust the mirror that reflects the laser signal towards the detector first (red arrow Figure 3D), and then do alignment fine-tuning with the screws afterwards.

    7. To calibrate the cantilever, select the room temperature and press on “calibrate” (Figure 3E). The calibration uses the thermal noise spectrum (Mullin and Hobbs, 2014) of the cantilever and determines the spring constant and the inverse lever sensitivity, which is used to convert the measured cantilever deflection from V to nm. The spring constant should lie within the specifications of the cantilevers used (for FASTSCAN-A cantilevers typically between 17-19 N/m but may vary for different batches).

    8. Go to the “Acquire data” tab and scan the cantilever for its resonance frequency in the AC Feedback Mode Wizard (Figure 3F). For FASTSCAN-A cantilevers, it should be around 1.2-1.4 MHz. Select the driving frequency for the measurement by placing the horizontal dashed line (Figure 3F) slightly left of the peak. Place the horizontal dashed line accordingly to reach a setpoint of 85-90% at a target amplitude of 12 nm.

    9. Start approaching the surface by pressing the “Approach” button (Figure 3G). It is recommended to coarse approach the surface with the head internal Z scanner as it covers a larger Z range and switch to the smaller but faster “Fast HG” scanner before starting to scan (this can be done in the Z Scanner Selection menu). Our typical imaging parameters are shown under Force Control and Scan Control (Figure 3G). After the approach succeeded, the z-position and the Laser Align should be in green color (Figure 3H).

    10. Once approached to the surface, verify the resonance peak again in the AC Feedback Mode Wizard window, as the cantilever resonance can change under the influence of a nearby surface.

    11. Start imaging (Figure 3I).

      Notes: Choose image size, scan speed, and pixels once and keep the settings constant. In our experience, the best results were obtained at a resolution of 1.46 nm/pixel either scanning 2,048 × 2,048 pixels in a 3 µm × 3 µm field of view or scanning 4,096 × 4,096 pixels in a 6 µm × 6 µm field of view. Scanning is then performed at 3 or 1.5 lines per second, respectively. The image size represents a compromise between the number of molecules imaged and the imaging time required. The field of view should not be smaller than 3 µm × 3 µm since it is important to have enough molecules (>100 DNA and nucleosomes) for good statistics in the subsequent analysis. Conversely, when scanning even larger areas (e.g., 12 µm × 12 µm), the time required for recording one image at scanning speeds low enough to enable excellent resolution starts to exceed 1 h, such that cantilever wear and drift become problematic. In addition, very large images are computationally cumbersome to process.

      1. Scanning speed matters! When scanning too fast, the molecules will appear less “sharp” and thus image quality will overall be worse. However, scanning too slowly will increase the effect of drift on the image. The scanning speed chosen should thus be as high as possible while maintaining the sharp imaging of molecules. For this purpose, we usually measure the diameter of DNA on the surface as it appears in the AFM image. Due to tip convolution, the DNA does not have a visible diameter of 2 nm, as expected from its crystal structure. A 6-8 nm DNA apparent full width at half maximum is a good value to target for ongoing AFM imaging making sure that the molecular resolution is high enough for quantitative assessment of the structural parameters. Sometimes, achieving a stable 6-8 nm DNA diameter can be difficult and requires tuning of the imaging parameters (i.e., adapting the drive frequency, the drive amplitude, the setpoint, and feedback gain) or exchanging the cantilever.

      2. For the large images with hundreds of molecules imaged here, nonlinear behavior and hysteresis of the piezos can cause artifacts and distortions that affect the structural parameters of DNA and nucleosomes. As an example, when imaging 3 µm × 3 µm images on a MultiMode8 AFM (Bruker), DNA molecules appear shorter in the beginning (bottom) of the scan than at the end (top of the image, Figure 4A). This shortening effect is strongest for the first scans but still occurs at reduced intensity in subsequent scans. In contrast, scanning even larger 6 µm × 6 µm images on the Nanowizard Ultraspeed 2 does not show these nonlinear effects (Figure 4B). Typically, while continuously imaging, these nonlinear effects and drift tend to decrease over time once the system stabilizes after warming up.



        Figure 4. DNA length as a quality control parameter to detect AFM scanning artifacts. A. Example DNA strand at the bottom of an AFM image acquired on a MultiMode 8 AFM system. B. Example DNA strand at the top of an AFM image acquired on a MultiMode 8 AFM system. The traced DNA contour is indicated by the yellow line in A and B. C. AFM image with a field of view of 3 µm × 3 µm imaged at 2,048 × 2,048 pixels acquired on a MultiMode 8 AFM system. The DNA molecules shown in detail in panels A and B are indicated by the boxes in the image. D. Distribution of DNA lengths measured at different y-positions (bottom to top) from a total of 10 AFM images on a MultiMode 8 AFM equipped with a tube scanner. Nonlinear effects in the AFM system, likely piezo creep, cause the DNA strands in the bottom of the image to appear to have different lengths than the DNA strands elsewhere in the image. E. Distribution of DNA lengths measured at different y-positions (bottom to top) for the JPK Nanowizard Ultraspeed 2. For this instrument, the drift effects due to piezo creep are significantly reduced, likely due to its linearized scanner design.


      3. All AFM cantilevers are different, and important properties such as tip radius, resonance frequency, and spring constant can vary significantly between batches. Even cantilevers that are within the specifications of the respective model can show significant variations in obtainable image quality. On the Nanowizard Ultraspeed 2 AFM setup, we had the best imaging results with FASTSCAN-A cantilevers (Bruker). On a MultiMode 8 AFM system that was used in the past, we had the best imaging results using AC160TS cantilevers (Olympus). However, due to the lower resonance frequency of the AC160TS cantilevers compared to FASTSCAN-A cantilevers and due to the smaller maximum image size of the MultiMode 8 (20482 pixels), both the scanning speed (1 Hz) and the field of view (3 µm × 3 µm to keep the pixel size constant) are generally smaller and make the system more prone to drift and limited statistics when analyzing the image. Still, it was possible to take data sets of similar quality – despite the imaging taking longer on the MultiMode 8 – on both instruments.


  3. AFM image analysis

    Note: We have developed an analysis pipeline written in Python to analyze the AFM images, which has been described in detail previously (Konrad et al., 2021). A detailed guide on how to set up the Python analysis pipeline can be found via https://github.com/SKonrad-Science/AFM_nucleoprotein_readout. Therefore, the image analysis pipeline is described only briefly here, and the focus will lie on tips and tricks on how to test image quality from the structural parameters obtained from the image readout. In addition, we discuss possible further analyses to obtain additional parameters such as DNA persistence length and states of nucleosome wrapping.

    1. To preprocess the raw AFM images, apply either a plane fit or an average profile fit to the surface and subsequently apply a line-wise leveling to remove observable steps between subsequent scan lines due to noise in the scanner system. The plane correction can be performed using either the image analysis software supplied by the AFM manufacturer (JPK Data Processing from JPK in our case, Figure 5A-C) or using the commercially available software SPIP (Image Metrology, Figure 5D-E). Save the leveled AFM image as an ASCII file for further processing.



      Figure 5. Plane correction of the raw AFM image. A. Raw AFM image displayed in the JPK Data Processing software (version 7.0.128). B. AFM image after applying a plane fit of first order. C. AFM image after consecutively applying line leveling. D. Plane correction parameters best used when plane correcting the raw AFM image using SPIP (Parameters: Mode, Custom; Global Correction, Average Profile Fit; Estimation Volume, Entire Image; Line-wise Correction, Histogram Alignment; and Z Offset Method, Set Mean to Zero). E. The same raw image as shown in panel A (left) and after applying the plane correction in SPIP (right).


    2. Open the custom Python code as described in the installation guide (Konrad et al., 2021) and select the desired analysis parameters. See Figure 6A for example analysis parameters used for analyzing the DNA and nucleosomes in our images. Depending on the background noise level, parameter tuning might be needed to maximize molecule detection. If the value of the background threshold chosen is too small, the molecules cannot be separated properly from the background, and if the value chosen is too high, the molecules might become fractured. Still, thresholding does not affect the final nucleosome parameters such as volume, opening angle, or height.

    3. Run the code and select the image you want to analyze from the file dialogue (Figure 6B). DNA and nucleosomes are then detected, and their structural parameters are analyzed automatically (Figure 6C). The results are saved to an Excel worksheet. As described in the user manual, you can also choose to manually help to categorize molecules that cannot be categorized automatically (such as two slightly overlapping DNA strands) by setting the manual filtering parameter (Figure 6A) to True.



      Figure 6. Image analysis post-processing software. A. Example input parameter settings for automated readout of the AFM images. B. File dialogue to select the desired AFM image for automated readout. C. Output of the analysis software during processing. The number of molecules detected in one example AFM image.


    4. Repeat the image analysis for all images of the data set to have all DNA and nucleosomes of the imaging run analyzed.

    5. The structural parameters stored in the Excel worksheet can be used to gain a broad understanding of the DNA and nucleosomes in the images and serve as an input for further analysis, for example, by principal component analyses or clustering. For an example, data set of wild-type nucleosomes reconstituted on a DNA segment of 486 bp, possible further analysis and plots are shown in Figure 7A-7G.



      Figure 7. Example analysis of DNA and nucleosome conformations for a dataset obtained with the protocol presented here. A. Distribution of bare DNA lengths. The solid line is a Gaussian fit centered at 151 ± 3 nm (mean ± STD). B. The DNA length is determined by tracing its contour with segments of 5 nm length, and the relative orientation of consecutive segments yields DNA bend angles. The solid line is a fit with a folded Gaussian, as described previously (van Noort et al., 1999), to obtain the DNA persistence length (lp = 52.7 nm). C. Distribution of DNA length wrapped around nucleosomes. The solid line is a double Gaussian fit to the data. The peaks are centered at 120 ± 14 bp and 168 ± 12 bp (mean ± STD). D. Distribution of nucleosome opening angles. E. Example image and tracing of a fully wrapped nucleosome. F. Example of an analyzed partially unwrapped nucleosome. G. 2D kernel density profile (bandwidth = 2.5°, 2.5 bp) of nucleosome opening angles and wrapped lengths. The cartoons in the insets depict the qualitative shape of fully and partially wrapped nucleosomes.


    Notes:

    1. We typically deposit both bare DNA and nucleosomes on each surface for imaging since the average length of the bare DNA is used to estimate the amount of wrapped DNA of a nucleosome by subtracting the length of the two arms from the average bare DNA length. Using the DNA length determined from co-deposited molecules is more accurate than using an average DNA length from separate imaging runs since even if the tip geometry does not change while measuring, the bare DNA length might differ slightly between images due to changes in drift.

    2. Looking at the average values of structural parameters such as the average length of bare DNA and the average nucleosome volume provides insights into the change in data quality for a specific dataset. As a rule of thumb, for a well-imaged high-quality dataset, the average bare DNA length should not differ by more than 2 nm over multiple images. Similarly, constant average nucleosome volume (exhibiting no more than 5-10% difference for images of the same dataset) is a good measure for image quality, and an increase or, in general, variation of nucleosome volume during a measurement run indicates changes in the tip geometry and will affect the structural parameters in general (Figure 8).



      Figure 8. DNA length and nucleosome volume as quality control parameters for AFM imaging. A. Average DNA length in subsequent images (mean and standard deviation over ~200 DNA strands per image). B. Average nucleosome volumes in subsequent images (mean over ~400 nucleosomes per image). Shown are two data sets obtained on a Nanowizard Ultraspeed 2, each comprising multiple 6 × 6 µm2 images. All images in one data set were obtained using the same AFM tip. Data set 1 consists of 7 AFM images, and data set 2 consists of 9 AFM images that were analyzed with respect to DNA and nucleosome structural parameters such as DNA length and nucleosome volume. For data set 1, the system stabilized during the first three images as indicated by almost constant DNA lengths and nucleosome volumes after image 3. In contrast, in data set 2, the volume parameter still shows fluctuations after several hours of imaging (~45 min per image), indicating that the system is less stable. Overall, in data set 2, the DNA lengths and the volumes are larger than for data set 1, indicating that the AFM cantilever has a less sharp tip.


    3. A detailed description of how to extract 5 bp unwrapping populations from the data obtained using this protocol can be found in our previous publication (Konrad et al., 2021).

    4. Typically, datasets of ~1,000 nucleosomes or more are required to allow for a detailed analysis of the unwrapping landscape.

    5. Additional parameters that are stored in the output Excel sheet, such as radius of gyration, end-to-end distance, or length of the individual nucleosome arms, allow for many different analyses. As an example, the 2D distribution of arm lengths and opening angles can be used to test for anti-cooperative unwrapping, or the ratio of arm lengths can be used to assess nucleosome positioning along the DNA strand.

Notes

  1. While this protocol was developed for imaging and analysis of DNA and nucleosomes, it can be readily adapted to other nucleo-protein complexes.

  2. Short chained poly-L-lysine should be used to guarantee a monolayer on the surface.

  3. Our surface deposition and imaging protocol is compatible with a broad range of ionic conditions. For example, we obtained high-quality images using 10 mM NaCl, 200 mM NaCl or 2 mM MgCl2 (always with 10 mM TRIS, pH 7.6). Importantly, ionic conditions significantly affect DNA and nucleosome geometry (Lipfert et al., 2014; Gebala et al., 2019; Konrad et al., 2021).

Recipes

  1. Deposition buffer

    200 mM NaCl + 10 mM TRIS, pH 7.6 – filtered.

    The concentration of salt can be varied: for example, we have obtained high-quality images at 200/50/10 mM NaCl or 50 mM NaCl + 2 mM MgCl2, always in 10 mM TRIS, pH 7.6.

Acknowledgments

We thank Philipp Korber and Felix Muller-Planitz for help with initial nucleosome reconstitutions, Pauline Kolbeck, Tine Brouns, Wout Frederickx, Herlinde De Keersmaecker, Steven De Feyter, and Björn H. Menze for discussions and assistance with AFM imaging, and Thomas Nicolaus for help with sample preparation. This work was funded by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) through SFB863 – Project ID 111166240. This protocol is based on our previously published study (Konrad et al., 2021).

Competing interests

The authors declare no conflict of interest.

References

  1. Ando, T. (2018). High-speed atomic force microscopy and its future prospects. Biophys Rev 10(2): 285-292.
  2. Baldi, S., Korber, P. and Becker, P. B. (2020). Beads on a string-nucleosome array arrangements and folding of the chromatin fiber. Nat Struct Mol Biol 27(2): 109-118.
  3. Bangalore, D. M., Heil, H. S., Mehringer, C. F., Hirsch, L., Hemmen, K., Heinze, K. G. and Tessmer, I. (2020). Automated AFM analysis of DNA bending reveals initial lesion sensing strategies of DNA glycosylases. Sci Rep 10(1): 15484.
  4. Bintu, L., Kopaczynska, M., Hodges, C., Lubkowska, L., Kashlev, M. and Bustamante, C. (2011). The elongation rate of RNA polymerase determines the fate of transcribed nucleosomes. Nat Struct Mol Biol 18(12): 1394-1399.
  5. Bowman, G. D. and Poirier, M. G. (2015). Post-translational modifications of histones that influence nucleosome dynamics. Chem Rev 115(6): 2274-2295.
  6. Brouns, T., De Keersmaecker, H., Konrad, S. F., Kodera, N., Ando, T. and Lipfert, J. (2018). Free Energy Landscape and Dynamics of Supercoiled DNA by High-Speed Atomic Force Microscopy. ACS Nano12(12): 11907-11916.
  7. Bussiek, M., Mücke, N., Langowski, J. (2003). Polylysine-coated mica can be used to observe systematic changes in the supercoiled DNA conformation by scanning force microscopy in solution. Nucleic Acids Res 31(22): e137.
  8. Dufrêne, Y. F., Ando, T., Garcia, R., Alsteens, D., Martinez-Martin, D., Engel, A., Gerber, C. and Müller, D. J. (2017). Imaging modes of atomic force microscopy for application in molecular and cell biology. Nat Nanotechnol 12(4): 295-307.
  9. Gebala, M., Johnson, S. L. and Narlikar, G. J. (2019). Ion counting demonstrates a high electrostatic field generated by the nucleosome. eLife 8: e44993.
  10. Hall, M. A., Shundrovsky, A., Bai, L., Fulbright, R. M., Lis, J. T. and Wang, M. D. (2009). High-resolution dynamic mapping of histone-DNA interactions in a nucleosome. Nat Struct Mol Biol 16(2): 124-129.
  11. Kaczmarczyk, A., Meng, H., Ordu, O. and Noort, J. V. (2020). Chromatin fibers stabilize nucleosomes under torsional stress. Nat Commun 11(1): 126.
  12. Katan, A. J., Vlijm, R., Lusser, A. and Dekker, C. (2015). Dynamics of nucleosomal structures measured by high-speed atomic force microscopy. Small 11(8): 976-984.
  13. Konrad, S. F., Vanderlinden, W., Frederickx, W., Brouns, T., Menze, B. H. and De Feyter, S. (2021). High-throughput AFM analysis reveals unwrapping pathways of H3 and CENP-A nucleosomes. Nanoscale 13(10): 5435-5447.
  14. Krietenstein, N., Wippo, C. J., Lieleg, C. and Korber, P. (2012). Genome-wide in vitro reconstitution of yeast chromatin with in vivo-like nucleosome positioning. Methods Enzymol 513: 205-232.
  15. Li, G. and Widom, J. (2004). Nucleosomes facilitate their own invasion. Nat Struct Mol Biol 11(8): 763-769.
  16. Lipfert, J., Doniach, S., Das, R. and Herschlag, D. (2014). Understanding nucleic acid-ion interactions. Annu Rev Biochem 83: 813-841.
  17. Luger, K., Mäder, A. W., Richmond, R. K., Sargent, D. F. and Richmond, T. J. (1997). Crystal structure of the nucleosome core particle at 2.8 A resolution. Nature 389(6648): 251-260.
  18. Mihardja, S., Spakowitz, A. J., Zhang, Y. and Bustamante, C. (2006). Effect of force on mononucleosomal dynamics. Proc Natl Acad Sci U S A 103(43): 15871-15876.
  19. Miyagi, A., Ando, T. and Lyubchenko, Y. L. (2011). Dynamics of nucleosomes assessed with time-lapse high-speed atomic force microscopy. Biochemistry 50(37): 7901-7908.
  20. Mueller-Planitz, F., Klinker, H. and Becker, P. B. (2013). Nucleosome sliding mechanisms: new twists in a looped history. Nat Struct Mol Biol 20(9): 1026-1032.
  21. Mullin, N. and Hobbs, J. K. (2014). A non-contact, thermal noise based method for the calibration of lateral deflection sensitivity in atomic force microscopy. Rev Sci Instrum 85(11): 113703.
  22. Narlikar, G. J., Sundaramoorthy, R. and Owen-Hughes, T. (2013). Mechanisms and functions of ATP-dependent chromatin-remodeling enzymes. Cell 154(3): 490-503.
  23. Ngo, T. T., Zhang, Q., Zhou, R., Yodh, J. G. and Ha, T. (2015). Asymmetric unwrapping of nucleosomes under tension directed by DNA local flexibility. Cell 160(6): 1135-1144.
  24. Ordu, O., Lusser, A. and Dekker, N. H. (2016). Recent insights from in vitro single-molecule studies into nucleosome structure and dynamics. Biophys Rev 8(Suppl 1): 33-49.
  25. Richmond, T. J. and Davey, C. A. (2003). The structure of DNA in the nucleosome core. Nature 423(6936): 145-150.
  26. Sadakierska-Chudy, A. and Filip, M. (2015). A comprehensive view of the epigenetic landscape. Part II: Histone post-translational modification, nucleosome level, and chromatin regulation by ncRNAs. Neurotox Res 27(2): 172-197.
  27. Schlingman, D. J., Mack, A. H., Kamenetska, M., Mochrie, S. G. J. and Regan, L. (2014). Routes to DNA accessibility: alternative pathways for nucleosome unwinding. Biophys J 107(2): 384-392.
  28. Shlyakhtenko, L. S., Lushnikov, A. Y. and Lyubchenko, Y. L. (2009). Dynamics of nucleosomes revealed by time-lapse atomic force microscopy. Biochemistry 48(33): 7842-7848.
  29. Sirinakis, G., Clapier, C. R., Gao, Y., Viswanathan, R., Cairns, B. R. and Zhang, Y. (2011). The RSC chromatin remodelling ATPase translocates DNA with high force and small step size. Embo J 30(12): 2364-2372.
  30. van Noort, J., Orsini, F., Eker, A., Wyman, C., de Grooth, B. and Greve, J. (1999). DNA bending by photolyase in specific and non-specific complexes studied by atomic force microscopy. Nucleic Acids Res 27(19): 3875-3880.
  31. Würtz, M., Aumiller, D., Gundelwein, L., Jung, P., Schütz, C., Lehmann, K., Tóth, K. and Rohr, K. (2019). DNA accessibility of chromatosomes quantified by automated image analysis of AFM data. Sci Rep 9(1): 12788.

简介

[摘要]原子力显微镜(AFM)是一种强大的工具,以图像大分子复合物具有纳米分辨率的第二精美单-分子的灵敏度。虽然 AFM 成像已经很好地用于研究 DNA 和核蛋白复合物,但 AFM 研究通常受到小数据集和手动图像分析的限制,这些分析速度缓慢且容易出现用户偏见。最近,我们已经表明,大规模 AFM 成像和核小体自动图像分析的结合可以克服 AFM 核蛋白研究的这些先前局限性。使用我们的高通量成像和分析管道,我们已经解决了核小体包裹中间体的五个 碱基对分辨率,并揭示了不同的核小体变异和环境条件如何影响核小体 DNA 的解包途径。在这里,我们提供了我们的工作流程的详细协议,以分析 DNA 和核小体构象,重点是实际方面和实验参数。我们希望我们的协议能够大大增强对 DNA 和核小体的 AFM 分析,并且能够很容易地适应各种其他蛋白质和蛋白质-核酸复合物。


[背景]核小体是将真核 DNA 压缩成染色质的基本单位,并作为基因读出和活性的调节剂发挥作用(Bowman 和 Poirier,2015 年;Sadakierska-Chudy 和 Filip,2015 年;Baldi等人,2020 年)。典型核小体核心颗粒由四个组蛋白H2A,H2B,H3的两个拷贝的,和H4组装成的组蛋白八聚体是紧密由147 bp的DNA的包裹(卢格等人,1997;里士满和Davey,2003) 。可接近用于读出和处理遗传密码是通过促进(部分)核小体DNA的展开,并且可以通过涉及活动进程任一实现,例如,RNA聚合酶或核小体的分子伴侣该施加力和扭矩的核小体(Sirinakis等人。, 2011 年;Mueller-Planitz等人,2013 年;Narlikar等人,2013 年;Ordu等人,2016 年),或由热波动自发地(Li 和 Widom,2004 年)。使用光镊等单分子显微操作技术,力诱导核小体展开的能量学已被高分辨率探测(Mihardja等人,2006 年;Hall等人,2009 年;Schlingman等人,2014 年;Ngo等人)., 2015; Kaczmarczyk等人, 2020) 。然而,在没有武力的情况下展开的景观更加难以进入。
原子力显微镜(AFM)是探测核小体结构和相互作用的有力工具,由于其能力,以图像的分子复合物在所述单-分子水平,无标记的,并用亚纳米分辨率,非常适合以可视化的DNA和蛋白质核小体的成分(Shlyakhtenko等人,2009 年;Bintu等人,2011 年;Miyagi等人,2011 年;Katan等人,2015 年;Ordu等人,2016 年)。在硬件的最新改进使数千种可能分子的快速成像,并用自动图像分析组合使小号DNA和核蛋白复合物的高度的定量和可再现的的研究(DUFRENE等人,2017;安藤,2018; Brouns等人,2018; Würtz等人,2019 年;班加罗尔等人,2020 年)。
通过结合大视野 AFM 成像和 DNA 和核小体的自动图像分析,我们最近阐明了核小体被动侵入核小体与接头 DNA 的核小体包裹景观,这与之前的力诱导解包裹分析形成对比(Konrad等人, 2021) 。虽然我们已经证明了STREN通过定量地捕捉构象ENSEMBL我们的方法GTH野生电子-类型和CENP-A核小体-在组蛋白H3替换为CENP-A变种着丝粒核小体-的方法可以很容易地适应研究中范围广泛的开放性问题,例如翻译后修饰对核小体包裹的影响或 DNA 序列对单分子水平上核小体定位的影响。
当前的协议描述了通过 AFM 成像研究 DNA 和核小体所需的所有步骤, 从分子的表面沉积开始, 以AFM 成像后的定量图像分析结束。该协议描述了空气中干燥样品的 AFM 成像。然而,它可以很容易地适用于液体中的 AFM 测量。在液体中,必须调整沉积协议和成像参数。特别是,样品缓冲溶液不会在沉积样品后干燥表面,而是保留在表面上进行成像。之前已经发布了如何进行液体 AFM 测量的示例(Bussiek等人,2003 年;Br ouns等人,2018 年)。AFM 图像的后续分析可能还需要调整图像分析参数(请参阅AFM 图像分析)。

关键字:原子力显微镜, 原子力显微镜, 脱氧核糖核酸, 核小体, 图像分析


材料和试剂

 
用于样品表面沉积
1. Mica Grade V-1 25 mm 圆盘(SPI S供应,目录号:01926-MB)       
2.标记胶带ROTI(Carl Roth,目录号:8000.1)       
3. 50 ml 冲洗注射器(Braun,目录号:4617509F)       
4. P arafilm (卡尔罗斯,目录号:H666.1)       
5. Protein LoBind Tubes 0.5 ml(Eppendorf,目录号:0030108094)       
6.培养皿(Carl Roth,目录号:0690.1)       
7. Kimwipes(Kimtech,目录号:5511)       
8. Milli-Q H 2 O(默克,目录号:Z00Q0V0WW)       
9.聚-L-赖氨酸(Sigma Aldrich公司,目录号:P0879 -在稀释到0.01%中号ILLI-Q H 2 O)       
10. N 2气(吹干表面)   
11. DNA/核小体样品——如前所述制备(Krietenstein等,2012)   
12.乙醇(卡尔罗斯,目录号:T171.4 -稀释到80%的中号ILLI-Q水)   
13.沉积缓冲液(见配方)   
 
用于 AFM 成像
载玻片(Thermo Scientific Menzel,目录号:15998086)
双-双面胶盘(SPI小号upplies,目录号:05095-AB)
 
设备
 
自闭合镊子(SPI S 供应品,目录号:SN5AP-XD)
涡流混合器(科学工业,目录号:SU-0236)
离心机(适合 0.5 ml Eppendorf 管和离心管内容物;Carl Roth,目录号:T464.1 )
镊子 ESD 安全(SPI S电源,目录号:0CFT07PE-XD)
用于空气中高速成像的 AFM 悬臂;我们使用 FASTSCAN-A(共振频率 1400 kHz,弹簧常数 18 N/m;Bruker)或 AC160TS(200-400 kHz,26 N/m;Olympus)悬臂
成像原子力显微镜;我们采用了 Nanowizard Ultra Speed 2 (Bruker) 和 MultiMode 8 (Bruker)
 
软件
 
对原始 AFM 数据进行平面校正的软件(参见“AFM 图像分析”部分)
微软Excel
蟒蛇 3
用于分析AFM 图像中DNA 和核小体的软件工具箱(之前在Konrad等人,2021 年描述),可在https://github.com/SKonrad-Science/AFM_nucleoprotein_readout下载)
 
程序
 
样品表面沉积
注意:污染会影响成像表面的质量,从而影响成像质量。因此,在整个过程中保持所有仪器清洁非常重要。
用乙醇和 kimwipes 彻底清洁工作台。用乙醇冲洗自闭合镊子的尖端,并用 N 2气吹干镊子。将镊子放在 kimwipe 上,这样尖端就不会被污染,也不会接触到工作台(图 1A)。
 
 
图 1. 表面沉积步骤。一个。如材料和试剂中所述,样品表面沉积所需材料的概述。乙。云母盘放置在标记带下。Ç 。撕下胶带会去除一层云母板,并留下平坦干净的表面,用于后续的样品沉积。d 。将聚-L-赖氨酸和样品溶液移至云母板的中心,各孵育 30 秒,每次孵育后直接进行洗涤和干燥步骤E和F。乙。孵育 30 秒后,通过将注射器的 50/20 ml M illi-Q H 2 O轻轻滴在表面上并通过旋转云母板使其流出来冲洗表面。˚F 。冲洗后,通过将喷嘴与 N 2气体的温和流垂直指向表面来干燥表面。
 
将两条标记带并排放置在工作台上,并有少量重叠(图 1A),使它们足够宽以完全覆盖云母盘。撕下部分标记带,将云母盘放在下面。施加压力,使胶带完全附着在云母表面(图 1B)。快速撕下胶带以切割云母(图 1C)。去除一整层云母是很重要的。如果仅去除了一部分层或剩余表面上有小裂缝,请重复此步骤,直到去除整个层。存放在切割的云母盘P ETRI菜,在准备下一个步骤。
从包装中取出两个无菌注射器并取出柱塞。确保柱塞前部和注射器筒不接触任何地方以避免污染。密封注射器筒用石蜡膜和填充物的清洁侧与中号ILLI-Q水。将柱塞放回桶中并按下柱塞,使封口膜撕裂,水从注射器中流出。取下封口膜并调整柱塞,使一个注射器容纳 50 毫升,另一个注射器容纳 20 毫升M illi-Q 水。从后面直接从储存瓶填充注射器,而不是用柱塞吸取液体,有助于避免污染。
准备 20 µl 0.01% 聚-L-赖氨酸的等分试样(参见材料和试剂),短暂涡旋,并在离心机中短暂旋转 Eppendorf 管中的内容物(约 2 秒,700 r cf )。将等分试样置于冰上。
将 20 µl 聚 L-赖氨酸溶液移至新切割的云母中心(步骤 2)并孵育 30 秒(图 1D)。确保不要用移液器接触表面。在 30 秒的孵化过程中,用自闭合镊子拿起云母板,拿起 50 毫升注射器,移到水槽或废物容器中,以便能够在 30 秒后开始冲洗。重要的是要保持云母表面水平并且在移动过程中尽可能保持静止,以确保表面沉积的高质量。通过将注射器中的液滴滴在云母板边缘(而不是在放置聚 L-赖氨酸的中心)并定期倾斜云母板,使水流出(图 1E)来冲洗表面。用 50 毫升冲洗后,请确保在表面留下一些水,以避免意外干燥。
在表面仍被水覆盖的情况下,开始用温和的 N 2气流干燥云母表面,方法是将云母表面快速倾斜到垂直位置,并在大约 2 厘米距离处垂直对准云母的中心(图1F)。一旦表面中心干燥,将水流移至边缘,直到云母完全干燥。
用缓冲液(在我们的例子中为200 mM NaCl + 10 mM Tris)稀释样品溶液,以达到表面沉积所需的浓度,并在冰上孵育 60 秒。核小体样品和缓冲液在 4°C 下储存,并在开始表面沉积之前放在冰上。表面沉积需要至少 25 µl 的总体积。我们通常稀1μl的核小体/ DNA样品溶液(含有大约30毫微克/微升486 bp的DNA和10ng /μl的组蛋白,对应于〜120nm的核小体;如所述制备Krietenstein等人。(2012))用40μl缓冲溶液产生了用于 AFM 成像的良好表面密度(即,密度足以在一个视场中包含许多分子,但又不会太密集以致于有太多分子彼此重叠)。
在冰上孵育 60 秒后,将 25 µl 缓冲样品溶液移至聚 L-赖氨酸包被的云母板的中心并孵育 30 秒。再次,进行如在步骤5和6 ,漂洗-此时用20ml的体积的中号ILLI-Q水-和随后干燥的表面上。
存储在云母盘P在室温下ETRI培养皿直到开始AFM测定。AFM 表面应始终在测量当天准备好。
 
核小体和 DNA 的 AFM 成像
注意:成像是在 Nanowizard Ultraspeed 2 系统上执行的,此处介绍的步骤对于其他仪器会略有不同。开始测量的步骤只是简要描述,因为它们可以在相应 AFM 系统的用户手册中找到详细信息。本节的重点在于有关如何调整 AFM 成像的技巧和窍门,以及为实现大型 DNA 和核小体数据集的最高图像质量而需要注意的事项。
为了准备最终的成像表面,将三个双面胶盘放在载玻片上,同时在中心留出一个区域(图 2A)。将样品云母板上的粘合剂区域,使得与粘接云母对齐的中心区域-自由载玻片和区域应用上的三个粘合剂盘上面的云母,温和的压力即。,不要在放置样品的中心,用镊子更牢固地固定云母。
 
 
图2. 在 AFM 中安装样品。A.带有沉积样品的云母板固定在带有三个双面胶盘的载玻片上。然后将玻璃面放置在 AFM 载物台上。乙。AFM 悬臂安装在随 AFM 系统(Ultra-Speed Glassblock,JPK,目录号:22229-E-01)一起交付的玻璃块上。Ç 。随后,玻璃块被放置在原子力显微镜扫描头底部的指定位置。d 。为了完成设置,将扫描头放置在 AFM 载物台上。
 
在 AFM 的样品架上安装带有云母板的玻璃滑块 (图 2A)。
将悬臂放在悬臂支架玻璃块 (图 2B) 中, 并将玻璃块安装在 AFM 扫描头 (图 2C) 中。然后,将 AFM 头放在舞台上(图 2D)。
启动JPK SPM 桌面软件并选择AC 模式快速成像(图 3A)。
 
 
图 3. AFM 软件和成像设置。A.该JPK SPM桌面(7.0.128)软件有几个可用的成像模式。本实验所需的模式是交流模式快速成像。乙。通过光学显微镜,AFM 悬臂的视图允许将悬臂放置在扫描头(绿色十字)内的中央。Ç 。激光对准悬臂的尖端,以最大化反射到探测器的信号量。d 。对齐检测器,使信号的最大值位于象限检测器的中心。乙。基于热噪声谱的悬臂校准产生 19.7 N/m 的估计弹簧常数。˚F 。交流反馈模式向导,用于选择 AFM 悬臂的驱动频率和驱动幅度。格。用于我们的 DNA 和核小体图像的典型成像设置。H . 悬臂成功接近表面后的系统状态。我。扫描 6 µm × 6 µm 视野的第一行,显示图像轨迹(左)和回扫(右)。
 
转到“设置实验”选项卡,并将绿色十字准线聚焦在 AFM 悬臂尖端,如相机视图中所示(图 3B)。
使用图 3C 中绿色箭头指示的螺钉将激光对准悬臂尖端,以最大化检测器收集的信号。收集到的信号的总和由蓝色条表示(图 3C)。对于 FASTSCAN-A 悬臂,1.3 和 1.5 之间的总和信号是典型的。其他悬臂根据悬臂尖端背面实现的反射产生不同的总和信号。
使用图 3D 中绿色箭头标记的螺钉对齐检测器,使信号的最大值位于象限检测器的中心。如果激光的正确对准是不可能的,首先调整将激光信号反射到探测器的镜子(红色箭头图 3D),然后进行对准微调-之后用螺钉进行调整。
要校准悬臂,请选择室温并按“校准”(图 3E)。校准使用悬臂的热噪声谱(Mullin 和 Hobbs,2014 年)并确定弹簧常数和反向杠杆灵敏度,用于将测量的悬臂挠度从 V 转换为 nm。弹簧常数应在所用悬臂的规格范围内(对于 FASTSCAN-A 悬臂,通常在 17-19 N/m 之间,但可能因不同批次而异)。
转到“获取数据”选项卡,并在交流反馈模式向导(图 3F)中扫描悬臂的共振频率。对于 FASTSCAN-A 悬臂,它应该在 1.2-1.4 M Hz左右。通过将水平虚线(图 3F)放置在峰值的左侧,选择测量的驱动频率。相应地放置水平虚线,以在 12 nm 的目标幅度下达到 85-90% 的设定点。
按“接近”按钮开始接近表面(图 3G)。建议使用头部内部 Z 扫描仪粗略接近表面,因为它覆盖了更大的 Z 范围,并在开始扫描之前切换到更小但速度更快的“快速 HG”扫描仪(这可以在Z 扫描仪选择菜单中完成)。我们的典型成像参数显示在力控制和扫描控制下(图 3G)。方法成功后,z 位置和激光对齐应为绿色(图 3H)。
一旦接近表面,在AC 反馈模式向导窗口中再次验证共振峰,因为悬臂共振会在附近表面的影响下发生变化。
开始成像 (图 3I)。
注意: 一次选择图像大小、扫描速度和像素,并保持设置不变。根据我们的经验,最好的结果是在 1.46 nm/像素的分辨率下获得的,要么在 3 µm × 3 µm 视野中扫描 2,048 × 2,048 像素,要么在 6 µm × 6 µm 视野中扫描 4,096 × 4,096 像素。扫描然后以每秒3个或1.5行执行,分别。图像大小代表了成像的分子数量和所需的成像时间之间的折衷。视场不应小于 3 µm × 3 µm,因为在随后的分析中拥有足够的分子(>100 个 DNA 和核小体)对于获得良好的统计数据非常重要。相反地,扫描甚至更大的区域时(例如,12微米× 12微米),在扫描记录一个图像所需的时间速度足够低,以使极好的分辨率开始超过1个小时,以使得悬臂磨损和漂移成为问题。此外,处理非常大的图像在计算上很麻烦。
扫描速度很重要!扫描速度太快时,分子会显得不那么“锐利”,因此图像质量总体上会变差。但是,扫描速度太慢会增加漂移对图像的影响。因此,选择的扫描速度应尽可能高,同时保持分子的清晰成像。为此,我们通常测量表面上 DNA 的直径,因为它出现在 AFM 图像中。由于尖端卷积,DNA 没有 2 nm 的可见直径,正如其晶体结构所预期的那样。6-8 nm 的 DNA 表观半高全宽对于正在进行的 AFM 成像来说是一个很好的目标值,可确保分子分辨率足够高,以便对结构参数进行定量评估。有时,实现稳定的6-8纳米的直径的DNA可以是困难的,需要的摄像参数的调谐(即,调整驱动频率,驱动幅度,设定点,以及反馈增益)或交换悬臂。
对于此处成像的数百个分子的大图像,压电的非线性行为和滞后会导致影响 DNA 和核小体结构参数的伪影和失真。例如,当在 MultiMode8 AFM (Bruker) 上对3 µm × 3 µm 图像进行成像时,DNA 分子在扫描的开始(底部)比在结束时(图像顶部,图 4A)显得更短。这种缩短效应在第一次扫描时最强,但在随后的扫描中仍然以降低的强度发生。相比之下,在 Nanowizard Ultraspeed 2 上扫描更大的 6 µm × 6 µm 图像不会显示这些非线性效应(图 4B)。通常,在连续成像时,一旦系统在预热后稳定下来,这些非线性效应和漂移就会随着时间的推移而降低。
 
 
图 4. DNA 长度作为检测 AFM 扫描伪影的质量控制参数。一个。在 MultiMode 8 AFM 系统上获取的 AFM 图像底部的示例 DNA 链。乙。在 MultiMode 8 AFM 系统上获取的 AFM 图像顶部的示例 DNA 链。所追踪的DNA轮廓通过在A和B的黄线表示Ç 。具有 3 µm × 3 µm视场的 AFM 图像,在 MultiMode 8 AFM 系统上以 2 , 048 × 2 , 048 像素成像。图 A 和 B 中详细显示的 DNA 分子由图像中的框表示。D.从配备管式扫描仪的 MultiMode 8 AFM 上的总共 10 个 AFM 图像在不同 y 位置(从下到上)测量的 DNA 长度的分布。AFM 系统中的非线性效应(可能是压电蠕变)导致图像底部的 DNA 链看起来与图像中其他地方的 DNA 链具有不同的长度。乙。在 JPK Nanowizard Ultraspeed 2 的不同 y 位置(从下到上)测量的 DNA 长度分布。对于该仪器,由于压电蠕变引起的漂移效应显着降低,这可能是由于其线性化扫描仪设计。
 
所有 AFM 悬臂都不同,并且不同批次的尖端半径、共振频率和弹簧常数等重要特性可能会有很大差异。即使是在相应型号规格范围内的悬臂梁,在可获得的图像质量方面也会出现显着变化。在 Nanowizard Ultraspeed 2 AFM 设置中,我们使用 FASTSCAN-A 悬臂 (Bruker) 获得了最佳成像效果。在过去使用的 MultiMode 8 AFM 系统上,我们使用 AC160TS 悬臂(奥林巴斯)获得了最佳成像效果。然而,由于 AC160TS 悬臂与 FASTSCAN-A 悬臂相比共振频率较低,并且由于 MultiMode 8 的最大图像尺寸较小(2048 2像素),扫描速度 (1 Hz) 和视野 ( 3 µm × 3 µm 以保持像素大小不变)通常更小,使系统在分析图像时更容易出现漂移和有限的统计数据。尽管如此,尽管在 MultiMode 8 上成像需要更长的时间,但还是可以在两种仪器上采集质量相似的数据集。
 
AFM 图像分析
注意:我们开发了一个用 Python 编写的分析管道来分析 AFM 图像,这在之前已经详细描述过(Konrad et al., 2021) 。有关如何设置 Python 分析管道的详细指南可通过https://github.com/SKonrad-Science/AFM_nucleoprotein_readout找到。因此,这里仅简要描述图像分析管道,重点将放在如何根据从图像读出中获得的结构参数来测试图像质量的技巧和窍门。此外,我们讨论了可能的进一步分析,以获得额外的参数,如 DNA 持久性长度和核小体包裹状态。
要对原始 AFM 图像进行预处理,请对表面应用平面拟合或平均轮廓拟合,然后应用逐行调平以消除由于扫描仪系统中的噪声而导致的后续扫描线之间的可观察步骤。可以使用 AFM 制造商提供的图像分析软件(在我们的案例中来自 JPK 的 JPK 数据处理,图 5A-C)或使用市售软件 SPIP(图像计量学,图 5D-E)进行平面校正。保存平整AFM图像的进一步处理ASCII文件。
 
 
图 5.原始 AFM 图像的平面校正。一个。JPK 数据处理软件(7.0.128 版)中显示的原始 AFM 图像。乙。应用一阶平面拟合后的 AFM 图像。Ç 。连续应用线调平后的AFM图像。d 。使用 SPIP 对原始 AFM 图像进行平面校正时最适合使用的平面校正参数(参数:模式、自定义;全局校正、平均轮廓拟合;估计体积、整个图像;线方向校正、直方图对齐;以及Z 偏移方法,将均值设置为零)。乙。与面板 A(左)和在 SPIP 中应用平面校正后(右)相同的原始图像。
 
按照安装指南(Konrad et al ., 2021)中的说明打开自定义 Python 代码,然后选择所需的分析参数。有关用于分析图像中 DNA 和核小体的示例分析参数,请参见图 6A。根据背景噪音水平,可能需要调整参数以最大化分子检测。如果选择的背景阈值太小,分子不能从背景中正确分离,如果选择的值太高,分子可能会断裂。尽管如此,阈值设置不会影响最终的核小体参数,如体积、打开角度或高度。
运行代码并从文件对话框中选择要分析的图像(图 6B)。然后检测 DNA 和核小体,并自动分析它们的结构参数(图 6C)。结果将保存到 Excel 工作表中。如用户手册中所述,您还可以通过将手动过滤参数(图 6A)设置为True来选择手动帮助对无法自动分类的分子(例如两条略微重叠的 DNA 链)进行分类。
 
 
图 6. 图像分析后处理软件。一个。用于自动读取 AFM 图像的示例输入参数设置。B.文件对话框以选择所需的 AFM 图像以进行自动读数。C.处理过程中分析软件的输出。在一个示例 AFM 图像中检测到的分子数量。
 
对数据集的所有图像重复图像分析,以分析成像运行的所有 DNA 和核小体。
Excel 工作表中存储的结构参数可用于广泛了解图像中的 DNA 和核小体,并作为进一步分析的输入,例如,通过主成分分析或聚类。例如,在 486 bp 的 DNA 片段上重组的野生型核小体的数据集,可能的进一步分析和绘图如图7A-7 G 所示。
 
 
图 7. 使用此处介绍的协议获得的数据集的 DNA 和核小体构象分析示例。一个。裸 DNA 长度的分布。实线是以 151 ± 3 nm(平均值 ± STD)为中心的高斯拟合。乙。的DNA长度由与5nm的长度的段跟踪其轮廓确定,并连续节段的相对取向产生DNA弯曲角度。实线是具有折叠高斯拟合,如先前所描述(v一个Noort等人,1999) ,以获得持续长度的DNA(L p = 52.7纳米)。Ç 。包裹在核小体周围的 DNA 长度分布。实线是数据的双高斯拟合。峰值集中在 120 ± 14 bp 和 168 ± 12 bp(平均值 ± STD)。d 。核小体开口角的分布。乙。完全包裹的核小体的示例图像和追踪。˚F 。部分展开的核小体分析示例。格。核小体开口角度和包裹长度的二维核密度分布(带宽 = 2.5°,2.5 bp)。插图中的卡通描绘了完全和部分包裹的核小体的定性形状。
 
笔记:
我们通常将裸 DNA 和核小体都沉积在每个表面上进行成像,因为裸 DNA 的平均长度用于通过从平均裸 DNA 长度中减去两条臂的长度来估计核小体的包裹 DNA 量。使用由共沉积分子确定的 DNA 长度比使用来自单独成像运行的平均 DNA 长度更准确,因为即使测量时尖端几何形状没有变化,由于漂移的变化,裸露的 DNA 长度可能会在图像之间略有不同。
综观结构参数的平均值,例如裸DNA的平均长度和平均核小体体积提供见解成用于特定的数据集数据质量的变化。根据经验,对于成像良好的高质量数据集,多个图像中平均裸 DNA 长度的差异不应超过 2 nm。类似地,恒定的平均核小体体积(对于同一数据集的图像显示不超过 5-10% 的差异)是图像质量的良好衡量标准,并且在测量运行期间核小体体积的增加或变化通常表明尖端几何形状,并会影响一般的结构参数(图 8)。
 
 
图 8. DNA 长度和核小体体积作为 AFM 成像的质量控制参数。一个。后续图像中的平均 DNA 长度(每个图像约 200 条 DNA 链的平均值和标准偏差)。B.后续图像中的平均核小体体积(平均每张图像超过 400 个核小体)。显示的是在 Nanowizard Ultraspeed 2 上获得的两个数据集,每个数据集包含多个 6 × 6 µm 2图像。一个数据集中的所有图像都是使用相同的 AFM 尖端获得的。数据集 1 由 7 个 AFM 图像组成,数据集 2 由 9 个 AFM 图像组成,这些图像针对 DNA 和核小体结构参数(如 DNA 长度和核小体体积)进行了分析。对于数据集 1,系统在前三个图像期间稳定,如图像 3 之后几乎恒定的 DNA 长度和核小体体积所示。相比之下,在数据集 2 中,体积参数在成像几个小时后仍然显示波动(~ 45 min per image) ,表明系统不太稳定。总体而言,在数据集 2 中,DNA 长度和体积大于数据集 1,表明 AFM 悬臂的尖端不那么锋利。
 
有关如何从使用该协议获得的数据中提取 5 bp 展开种群的详细说明,请参见我们之前的出版物(Konrad 等人,2021)。
通常,数据集 ~ 1 , 000 或更多核小体需要允许展开景观的详细分析。
存储在输出的Excel工作表的附加参数,诸如回转半径,端至端的距离,或单独的核小体臂的长度,允许许多不同的分析。例如,臂长和张角的 2D 分布可用于测试反合作展开,或者臂长的比率可用于评估核小体沿 DNA 链的定位。
 
笔记
 
虽然该协议是为 DNA 和核小体的成像和分析而开发的,但它可以很容易地适应其他核蛋白复合物。
应使用短链聚-L-赖氨酸以保证表面的单层。
我们的表面沉积和成像协议与广泛的离子条件兼容。例如,我们使用 10 mM NaCl、200 mM NaCl 或 2 mM MgCl 2 (始终使用 10 mM TRIS,pH 7.6)获得了高质量图像。重要的是,离子条件会显着影响 DNA 和核小体几何形状(Lipfert等人,2014 年;Gebala等人,2019 年;Konrad等人,2021 年)。
 
食谱
 
沉积缓冲液
200 mM NaCl + 10 mM TRIS,pH 7.6 – 过滤。
Ť盐的他浓度可以变化:例如,我们获得高-在200/50/10毫NaCl或50mM NaCl中+ 2毫摩尔MgCl质量的图像2在10mM TRIS,pH 7.6的,始终。
 
致谢
 
我们感谢 Philipp Korber 和 Felix Müller-Planitz 在初始核小体重建方面的帮助,感谢 Pauline Kolbeck、Tine Brouns、Wout Frederickx、Herlinde De Keersmaecker、Steven De Feyter 和 Björn H. Menze 对 AFM 成像的讨论和帮助,以及 Thomas Nicolaus帮助样品制备。这项工作由 Deutsche Forschungsgemeinschaft(DFG,德国研究基金会)通过 SFB863 – 项目 ID 111166240 资助。该协议基于我们之前发表的研究(Konrad等人,2021 年)。
 
利益争夺
 
作者声明没有利益冲突。
 
参考
1. Ando, T. (2018)。高速原子力显微镜及其未来前景。生物物理学修订版10(2):285-292。                    
2. Baldi, S.、Korber, P. 和 Becker, PB (2020)。串核小体阵列排列和染色质纤维折叠上的珠子。27(2):109-118。      
3.班加罗尔, DM, Heil, HS, Mehringer, CF, Hirsch, L., Hemmen, K., Heinze, KG 和 Tessmer, I. (2020)。DNA 弯曲的自动 AFM 分析揭示了 DNA 糖基化酶的初始病变传感策略。科学报告10(1): 15484。                    
4. Bintu, L.、Kopaczynska, M.、Hodges, C.、Lubkowska, L.、Kashlev, M. 和 Bustamante, C. (2011)。RNA聚合酶的延伸率决定了转录核小体的命运。Nat Struct Mol Biol 18(12): 1394-1399。      
5. Bowman, GD 和 Poirier, MG (2015)。影响核小体动力学的组蛋白的翻译后修饰。化学修订版115(6):2274-2295。                    
6. Brouns, T., De Keersmaecker, H., Konrad, SF, Kodera, N., Ando, T. 和 Lipfert, J. (2018)。通过高速原子力显微镜观察超螺旋 DNA 的自由能景观和动力学。ACS 纳米12(12):11907-11916。                    
7. Bussiek, M., Mücke, N., Langowski, J. (2003)。聚赖氨酸包被的云母可用于通过扫描力显微镜在溶液中观察超螺旋DNA 构象的系统变化。核酸研究31(22):e137。      
8. Dufrêne, YF, Ando, T., Garcia, R., Alsteens, D., Martinez-Martin, D., Engel, A., Gerber, C. 和 Müller, DJ (2017)。用于分子和细胞生物学的原子力显微镜成像模式。Nat Nanotechnol 12(4): 295-307。      
9. Gebala, M.、Johnson, SL 和 Narlikar, GJ(2019 年)。离子计数表明核小体产生的高静电场。网上生活8:ë 44993 。      
10. Hall, MA, Shundrovsky, A., Bai, L., Fulbright, RM, Lis, JT 和 Wang, MD (2009)。核小体中组蛋白-DNA 相互作用的高分辨率动态映射。Nat Struct Mol Biol 16(2): 124-129。                 
11. Kaczmarczyk, A.、Meng, H.、Ordu, O. 和 Noort, JV (2020)。染色质纤维在扭转应力下稳定核小体。国家通讯社11(1):126。   
12. Katan, AJ, Vlijm, R., Lusser, A. 和 Dekker, C. (2015)。通过高速原子力显微镜测量的核小体结构动力学。小11(8):976-984。   
13. Konrad, SF, Vanderlinden, W., Frederickx, W., Brouns, T., Menze, BH 和 De Feyter, S. (2021)。高通量 AFM 分析揭示了 H3 和 CENP-A 核小体的解包途径。纳米级13(10):5435-5447。   
14. Krietenstein, N.、Wippo, CJ、Lieleg, C. 和 Korber, P. (2012)。具有体内样核小体定位的酵母染色质的全基因组体外重组。方法 Enzymol 513:205-232。   
15. Li, G. 和 Widom, J. (2004)。核小体促进了它们自身的入侵。Nat Struct Mol Biol 11(8): 763-769。                 
16. Lipfert, J.、Doniach, S.、Das, R. 和 Herschlag, D.(2014 年)。了解核酸-离子相互作用。Annu Rev Biochem 83:813-841。   
17. Luger, K., Mäder, AW, Richmond, RK, Sargent, DF 和 Richmond, TJ (1997)。2.8 A 分辨率下核小体核心颗粒的晶体结构。自然389(6648):251-260。                 
18. Mihardja, S., Spakowitz, AJ, Zhang, Y. 和 Bustamante, C. (2006)。力对单核小体动力学的影响。Proc Natl Acad Sci USA 103(43): 15871-15876。   
19. Miyagi, A.、Ando, T. 和 Lyubchenko, YL (2011)。用延时高速原子力显微镜评估核小体的动力学。生物化学50(37):7901-7908。   
20. Mueller-Planitz, F.、Klinker, H. 和 Becker, PB (2013)。核小体滑动机制:循环历史中的新曲折。Nat Struct Mol Biol 20(9): 1026-1032。                 
21. Mullin, N. 和 Hobbs, JK (2014)。一种用于校准原子力显微镜横向偏转灵敏度的非接触式热噪声方法。Rev Sci Instrum 85(11): 113703。                 
22. Narlikar, GJ, Sundaramoorthy, R. 和 Owen-Hughes, T. (2013)。ATP 依赖性染色质重塑酶的机制和功能。单元格154(3):490-503。                 
23. Ngo, TT, Zhang, Q., Zhou, R., Yodh, JG 和 Ha, T. (2015)。在 DNA 局部柔性指导下的张力下核小体的不对称展开。单元格160(6):1135-1144。   
24. Ordu, O.、Lusser, A. 和 Dekker, NH (2016)。体外单分子研究对核小体结构和动力学的最新见解。生物物理学修订版8(增刊 1):33-49。                 
25. Richmond, TJ 和 Davey, CA (2003)。核小体核心中 DNA 的结构。自然423(6936):145-150。   
26. Sadakierska-Chudy, A. 和 Filip, M.(2015 年)。表观遗传景观的综合视图。第二部分:组蛋白翻译后修饰、核小体水平和 ncRNA 对染色质的调控。Neurotox Res 27(2): 172-197。                 
27. Schlingman, DJ, Mack, AH, Kamenetska, M., Mochrie, SGJ 和 Regan, L. (2014)。DNA 可及性途径:核小体解旋的替代途径。生物物理学杂志107(2): 384-392。   
28. Shlyakhtenko, LS, Lushnikov, AY 和 Lyubchenko, YL (2009)。延时原子力显微镜揭示的核小体动力学。生物化学48(33):7842-7848。   
29. Sirinakis, G., Clapier, CR, Gao, Y., Viswanathan, R., Cairns, BR 和 Zhang, Y. (2011)。RSC 染色质重塑 ATP 酶以高力和小步长转移 DNA。Embo J 30(12): 2364-2372。   
30. van Noort, J.、Orsini, F.、Eker, A.、Wyman, C.、de Grooth, B. 和 Greve, J. (1999)。通过原子力显微镜研究的特定和非特定复合物中光解酶的 DNA 弯曲。核酸研究27(19):3875-3880。   
31.维尔茨,M 。,奥米勒,D 。,Gundelwein ,大号。,荣格,P 。,茨,Ç 。,莱曼,K 。,托特,K 。和Rohr , K. (2019)。通过 AFM 数据的自动图像分析量化染色体的 DNA 可及性。科学报告9(1): 12788。                 
版权所有 © 20 2 1作者;独家被许可人 Bio-protocol LLC。1                                                                                                                             
登录/注册账号可免费阅读全文
  • English
  • 中文翻译
免责声明 × 为了向广大用户提供经翻译的内容,www.bio-protocol.org 采用人工翻译与计算机翻译结合的技术翻译了本文章。基于计算机的翻译质量再高,也不及 100% 的人工翻译的质量。为此,我们始终建议用户参考原始英文版本。 Bio-protocol., LLC对翻译版本的准确性不承担任何责任。
Copyright: © 2021 The Authors; exclusive licensee Bio-protocol LLC.
引用:Konrad, S. F., Vanderlinden, W. and Lipfert, J. (2021). A High-throughput Pipeline to Determine DNA and Nucleosome Conformations by AFM Imaging. Bio-protocol 11(19): e4180. DOI: 10.21769/BioProtoc.4180.
提问与回复
提交问题/评论即表示您同意遵守我们的服务条款。如果您发现恶意或不符合我们的条款的言论,请联系我们:eb@bio-protocol.org。

如果您对本实验方案有任何疑问/意见, 强烈建议您发布在此处。我们将邀请本文作者以及部分用户回答您的问题/意见。为了作者与用户间沟通流畅(作者能准确理解您所遇到的问题并给与正确的建议),我们鼓励用户用图片的形式来说明遇到的问题。

如果您对本实验方案有任何疑问/意见, 强烈建议您发布在此处。我们将邀请本文作者以及部分用户回答您的问题/意见。为了作者与用户间沟通流畅(作者能准确理解您所遇到的问题并给与正确的建议),我们鼓励用户用图片的形式来说明遇到的问题。