The importance of forces in biology has been recognized for quite a while but only in the past decade have we acquired instrumentation and methodology to directly measure interactive forces at the level of single biological macromolecules and/or their complexes. This review focuses on force measurements performed with the atomic force microscope. A general introduction to the principle of action is followed by review of the types of interactions being studied, describing the main results and discussing the biological implications.

Unfixed chicken erythrocyte chromatin fibers in very low salt have been imaged with a nning force microscope operating in the tapping mode in air at ambient humidity. These images reveal a threedimensional organizaffon of the fibers. The planar "6zig-zag" conformation is rare, and extended "beads-on-a-string" fibers are seen only in chromatin depleted of histes Hi and H5. Glutaraldehyde fixation reveals very similar structures. Fibers fixed in 10 mM salt appear somewhat more compacted. These results, when compared with modeing stude, suggest that chromatin fibers may exist as frregular three-dimensional arrays of nucleosomes even at low ionic strength.The structure of the chromatin fiber in low salt concentrations remains controversial. Electron microscopy (EM) experiments, most of which utilized the Miller spreading technique (1), typically showed extended "beads-on-a-string" or "open zig-zag" structures (refs. 2 and 3; for reviews, see refs. 4-6). At slightly higher ionic strength (-10 mM NaCI), somewhat more compact, "closed" zig-zags of nucleosomes were observed (7-9). Only upon further addition of NaCl to about 100 mM did these extended structures condense to form the so-called 30-nm fiber (8), which resembles structures observed in situ (7, 9, 10). However, there has been concern that strong interactions of the fiber with the EM support surface, and the dehydration produced by the high vacuum conditions, could distort the structure, especially at low ionic strength.Attempts to circumvent these problems used solution studies. The first scattering experiments suggested that the nucleosomes were densely packed in a linear array (11,12 To address this controversy, a study was performed using the three-dimensional imagi capabilities ofa scanning force microscope, which makes it possible to image chromatin fibers under less damaging conditions (22,23). The samples are never vacuum dried and are scanned in air at about 50%o relative humidity. Under these conditions a film of liquid water resides on the support surface (24). The newly developed tapping operation mode (25, 26) was employed, in which a stiff cantilever is oscillated near its resonance frequency with amplitudes typically in the range of 10-20 nm as the sample is scanned laterally. The oscillation amplitude is kept constant via feedback control. This operation has several advantages over the contact mode, in which a tip is pulled across the sample. Tip-sample forces are lighter than in the contact mode. Moreover, since most of the force is perpendicular to the surface, the sample experiences minimal lateral deformation during scanning, thus improving spatial resolution (25, 26). MATERIAL AND METHODSPreparation and Fization of Chromatin. Chicken erythrocyte chromatin was prepared essentially as described (27), with a reduction in the amount of micrococcal nuclease to allow isolation of long fibers (28). Soluble chromatin was dialyzed versus 5 mM triethanolamine/HC1 (pH 7.0), with or without 10 mM NaCl and was stored on ice. In a few experiments...

Single chromatin fibers were assembled directly in the flow cell of an optical tweezers setup. A single lambda phage DNA molecule, suspended between two polystyrene beads, was exposed to a Xenopus laevis egg extract, leading to chromatin assembly with concomitant apparent shortening of the DNA molecule. Assembly was force-dependent and could not take place at forces exceeding 10 pN. The assembled single chromatin fiber was subjected to stretching by controlled movement of one of the beads with the force generated in the molecule continuously monitored with the second bead trapped in the optical trap. The force displayed discrete, sudden drops upon fiber stretching, reflecting discrete opening events in fiber structure. These opening events were quantized at increments in fiber length of approximately 65 nm and are attributed to unwrapping of the DNA from around individual histone octamers. Repeated stretching and relaxing of the fiber in the absence of egg extract showed that the loss of histone octamers was irreversible. The forces measured for individual nucleosome disruptions are in the range of 20-40 pN, comparable to forces reported for RNA- and DNA-polymerases.

The N-terminal tails of the core histones play important roles in transcriptional regulation, but their mechanism(s) of action are poorly understood. Here, pure chromatin templates assembled with varied combinations of recombinant wild-type and mutant core histones have been employed to ascertain the role of individual histone tails, both in overall acetylation patterns and in transcription. In vitro assays show an indispensable role for H3 and H4 tails, especially major lysine substrates, in p300-dependent transcriptional activation, as well as activator-targeted acetylation of promoter-proximal histone tails by p300. These results indicate, first, that constraints to transcription are imposed by nucleosomal histone components other than histone N-terminal tails and, second, that the histone N-terminal tails have selective roles, which can be modulated by targeted acetylation, in transcriptional activation by p300.