Ultrastructure of chromatin. II. Three-dimensional reconstruction of isolated fibers

Methods for generating high-resolution structural models from electron microscope tomography data

Reconstructed volumes generated by tilt-image electron-microscope tomography offer the best spatial resolution currently available for studying cell structures in situ. Analysis is often accomplished by creating surface models that delineate grayscale contrast boundaries. Here, we introduce a specialized and convenient sequence of segmentation operations for making such models that greatly improves their reliability and spatial resolution as compared to current approaches, providing a basis for making accurate measurements. To assess the reliability of the surface models, we introduce a spatial uncertainty measurement based on grayscale gradient scale length. The model generation and measurement methods are validated by applying them to synthetic data, and their utility is demonstrated by using them to characterize macromolecular architecture of active zone material at the frog's neuromuscular junction.

Electron microscopy of DNA-protein complexes and chromatin

This article focused on a number of aspects of the preparation of chromatin and other DNA-protein complexes for conventional transmission EM that are critical for success but may not have been addressed in a single chapter before. These include the importance of optimizing fixation, the generation of active supporting supports, and the use of negative staining as a means of obtaining higher resolution detail than can be garnered from shadow casting methods.

Automated electron microscope tomography of frozen-hydrated chromatin: the irregular three-dimensional zigzag architecture persists in compact, isolated fibers

The potential of electron microscope tomography as a tool for obtaining three-dimensional (3D) information about large macromolecular assemblies is greatly extended by automation of data collection. With the implementation of automated control of tilting, focusing, and digital image recording described here, tilt series of frozen-hydrated specimens can be collected with the requisite low dose. Long chromatin fibers were prepared in 90 mM monovalent ions to maintain a fully compact conformation, and after vitrification were completely contained within the ice layer. Tilt series of this material were recorded at 5 degrees tilt increments between +60 degrees and -60 degrees, with a cumulative dose of approximately 35 e-/A2 for the series. This extremely low dose data was successfully aligned, then reconstructed by weighted backprojection. The underlying architecture of the fibers is an irregular 3D zigzag of interconnected nucleosomes, with the linker DNA between successive nucleosomes in a largely extended conformation. The visualization of this structural motif within long, frozen-hydrated chromatin fibers at relatively high salt extends our previous studies on small fragments at low ionic strength and is in agreement with the observation of this architecture in chromatin fibers in situ in sectioned nuclei.

Alignment of electron tomographic series by correlation without the use of gold particles

Electron tomography requires that a tilt series of micrographs be aligned and that the orientation of the tilt axis be known. This has been done conveniently with gold markers applied to the surface of a specimen to provide easily accessible information on the orientation of each tilt projection. Where gold markers are absent, another approach to alignment must be used. A method is presented here for aligning tilt projections without the use of markers, utilizing correlation methods. The technique is iterative, drawing principally on the work of Dengler [Ultramicroscopy 30 (1989) 337], and consists of computing a low resolution back projection image from which computed tomographic projections can be generated. These in turn serve as reference images for the next alignment of the tomographic series. An initial alignment must be made before the first back projection, and this is done following the method of Guckenberger [Ultramicroscopy 9 (1982) 167] for translational alignment and by common lines analysis [Liu et al., Ultramicroscopy 58 (1995) 393] for identification of the tilt axis. Four tomographic series of a biological nature were aligned and analyzed, and the method has proven to be both accurate and reproducible for the data presented here.

DNA changes involved in the formation of metaphase chromosomes, as observed in mouse duodenal crypt cells stained by osmium-ammine. I. New structures arise during the S phase and condense at prophase into "chromomeres," which fuse at prometaphase into mitotic chromosomes

BACKGROUND

In the hope of understanding how chromosomes condense at mitosis, we took advantage of a subdivision of the cell cycle into 11 stages to examine the changes in DNA taking place during the stages preceding the emergence of metaphase chromosomes.

METHODS

To identify DNA changes, pieces of mouse duodenum were fixed in formaldehyde, and sections of the rapidly dividing cells of the crypts were stained by the osmium-ammine method, which is specific for the detection of DNA in the electron microscope.

RESULTS

Throughout the cell cycle, DNA is present in nucleofilaments composed of rows of 11-nm-wide nucleosomes. At stage I, during which the DNA-synthesizing or S phase of the cell cycle begins, some of the nucleofilaments are compacted in the heterochromatin accumulations associated with nuclear envelope and nucleoli, while the others are scattered in the nucleoplasm where they appear either "free" or "attached" to the heterochromatin. This DNA distribution is similar to that observed in the noncycling cells examined. After the beginning of the S phase, "free" nucleofilaments are seen to assemble into structures composed of compacted nucleofilaments and referred to as "aggregates"; these make their appearance at stage II and increase in size through stage III up to the end of S during stage IV. Meanwhile, the heterochromatin associated with nuclear envelope and nucleoli expands toward the nucleoplasm in the form of protrusions referred to as "bulges," which gradually enlarge during stages III and IV, while the heterochromatin shrinks and eventually vanishes. On average, a total of 1,171 aggregates and bulges are formed in the nucleus during the S phase. At the apparition of stage V, which corresponds approximately to prophase, aggregates and bulges are rapidly gathered into an average of 288 spheroidal bodies referred to as "chromomeres." These are connected to one another by nucleofilamentous bridges in such a way as to be lined up in rows. The formation of rows of chromomeres represents in the electron microscope the prophasic condensation observed in the light microscope. Finally, during stage VIa, which corresponds to prometaphase, the chromomeres approach one another within each row, make contact, and coalesce to become the 40 chromosomes of the mouse, which during stage VIb are organized in the equatorial plate of metaphase.

CONCLUSIONS

The condensation of metaphase chromosomes occurs in three main steps. The first and longest takes place during the S phase, as nucleofilaments are assembled into aggregates, while the heterochromatin gives rise to bulges. The brief second step occurs toward the beginning of prophase, when the numerous aggregates and bulges are congregated into a limited number of chromomeres, which are lined up in rows. The third step takes place during the brief prometaphase, when the chromomeres of a row coalesce into a mitotic chromosome.

Restoration of high-tilt electron micrographs using a focus series

Electron microscopy tomography requires high-tilt projections. However, the variation in defocus over a highly tilted specimen results in a blurred micrograph, where the blur is spatially varying blur can cause the micrograph to deviate from an acceptable approximation of the projection of the specimen's density function. In practice, this has been one factor limiting reconstructions of large cellular areas. We derive an algorithm to restore, or deblur, high-tilt electron micrographs using multiple defocused versions of the micrograph. This algorithm consists of digitization and registration of the micrographs, identification of the defocus blur using either a theoretical model or a zero-tilt focus series, and spatially varying restoration using the multiple high-tilt defocused micrographs. Numerical results using simulated and real data are presented, demonstrating the effectiveness of this algorithm.

Chromosome structure inside the nucleus

Recent in situ three-dimensional structural studies have provided a new model for the 30 nm chromatin fiber. In addition, research during the past year has revealed some of the molecular complexity of non-histone chromosomal proteins. Still to come is the unification of molecular insights with chromosomal architecture.

Transitions between in situ and isolated chromatin

We show that the mechanism by which chromatin displaying higher-order structure is usually isolated from nuclei involves a transition to an extended nucleosomal arrangement. After being released from nuclei, chromatin must refold in order to produce the typical chromatin fibers observed in solution. For starfish sperm chromatin with a long nucleosome repeat (222 bp), isolated fibers are significantly wider than those in the nucleus, indicating that the refolding process does not regenerate the native higher-order structure. We also propose that for typical eukaryotic nuclei, the concept that the native state of the (inactive) bulk of the genome is a chromatin fiber with defined architecture be reconsidered.

A 360 degrees single-axis tilt stage for the high-voltage electron microscope

A new type of specimen stage that permits more than 180 degrees of tilting about the axis of a side-entry rod has been developed for a high-voltage electron microscope (HVEM). Roughly cylindrical specimens, with radial dimensions of less than a few micrometers, that can be mounted on the tip of a microneedle or micropipette are applicable. For glass micropipettes, the energy of the 1-MeV beam of the HVEM is sufficient to image specimens through both walls. The stage employs a spindle mechanism that holds these needles or micropipettes coaxial with the tilt axis, allowing the specimen to be rotated without restriction. This arrangement, along with the cylindrical form of the specimen, is an important development for single-axis tomography, because it permits a complete 180 degrees set of projections to be recorded. The angular accuracy of the stage was demonstrated to be within +/- 0.20 degrees, with a cumulative error of less than 1.0 degrees over a 180 degrees span. The new stage was tested using puffball spores mounted on a micropipette. A 180 degrees tilt series was recorded and processed to yield a tomographic three-dimensional reconstruction which was displayed both as a cross-sectional view perpendicular to the tilt axis, and as a shaded surface viewed from different directions. The same computations were repeated using subsets of the tilt series to assess the effect of various amounts of missing information. Visual inspection of a selected cross-section from these reconstructions indicated that limiting the angular range to 160 degrees produced results nearly as good as the full data set. Limiting the range to 140 degrees, however, produced a noticeable geometric distortion, which became increasingly severe with ranges of 120 degrees and 100 degrees.