Faults in the square lattice of mammalian skeletal muscle z-disks

In recent years there have been a number of reports on Z-disk ultrastructure in vertebrate striated muscle (Knappies & Carlsen, 1962; Franzini-Armstrong & Porter, 1964; Reedy, 1964; Kelly, 1967; Landon, 1970). Although there is still some debate as to the exact nature and origin of the filaments of the Z-disk, all of these reports provide evidence in support of a square-lattice arrangement into which the actin filaments of adjacent sarcomeres insert. Recently Landon’s (1970) findings demonstra ted that the method of fixation has a great influence on the spacing and orientation of the planes of symmetry of the Z-disk lattice; it does, however, remain as a square lattice. From this work a number of models of Z-disk structure have been put forward, all of which envisage a well-ordered array of filaments.

There are a number of published electron micrographs of the square lattices found in the Z-disk and in the I-bands immediately adjacent to the Z-disk in which fault lines can be seen (Reedy, 1964; Page, 1965; Edge & Walker, 1970; Landon, 1970). The purpose of this paper is to draw attention to the presence of these faults and to discuss the effect they may have on myofibril structure and in particular the role they may play in imposing an upper limit on individual myofibril size.

Bovine and rat muscle, including material from both ‘red’ and ‘white’ muscle, was taken from the animal either by biopsy or immediately after death. Bundles of fibres, or in the case of rat material whole muscles, were fastened to wooden splints to prevent shortening during fixation. The restrained material was immediately fixed at 0 °C for 2 h in a mixture of 1 % osmium tetroxide and 2·5 % glutaraldehyde (2:1 by volume) buffered to pH 7·3 with 0 1 M cacodylate buffer (Hirsch & Fedorko, 1968). After this initial fixation the muscle was cut from the splint and small bundles of 2 or 3 fibres were teased from the surface of the material in 0·1 M cacodylate buffer (pH 7·3). Only the black surface fibres which showed evidence of osmium fixation were processed further. Small pieces (1 or 2 mm long) were cut from these and transferred to fresh fixative for a further 2 h at 0 °C. Subsequently the tissue was washed overnight in changes of 0·1 M cacodylate buffer at 4 °C, dehydrated in graded ethanol solutions, embedded in Araldite and sectioned on an LKB ultratome using glass knives. Sections were stained in uranyl acetate (saturated solution in 50% ethanol) and in Reynolds’ (1963) lead citrate. The sections were examined in a Philips EM 300 electron microscope.

The Z-disk of striated muscle is made up of protein filaments which form the lattice and give it the appearance of a well-ordered structure. In close association with this lattice is additional Z-disk material which has an amorphous appearance and tends to mask the structure of the lattice. Information on the Z-disk lattice can be gained indirectly from the arrangement of the I-filaments in the I-band immediately adjacent to the Z-disk. In this region the I-filaments are in a square array (Fig. 2). The I filaments from the sarcomeres on either side of the Z-disk are displaced relative to one another so that when viewed in transverse section they superimpose to give the arrangement found in the Z-disk lattice. Fig. 1 is a diagram of the Z-disk and adjacent I-band filaments viewed enface. The exact orientation of the Z-filaments joining the I-filaments in the Z-disk is debatable (see Kelly, 1967; Landon, 1970; Rowe, in preparation). These Z-filaments have been omitted as they do not contribute directly to the argument presented here. From Fig. 1 it should be noted that only at the line of transition from I-band lattice to Z-disk lattice is there any possibility of a change in the angle of orientation of the sides of the square lattices (view Fig. 1 at an acute angle and rotate the diagram). This change in angle is 45°.

Figs. 3–5 are all transverse sections of myofibrils through the Z-disk and/or adjacent I-bands. They all show fault lines in the lattices. The different angles of orientation in the lattices cannot be explained by different levels of section through the myofibril. A slightly oblique section moving from one I-band through the Z-disk to the next I band would produce angle changes of 45° only at the transition line as in Fig. 1. The faults shown here are not all at 45° and are between filaments which can be identified as belonging to the same I-band or in the Z-disk itself.

The Z-filaments joining the I-filaments in the Z-disk presumably do so by chemical bonding. The model system in which there is uniformity of spatial configuration enables all similar bonding to be carried out at similar angles. When the spatial relationship breaks down, as in the fault lines, the bonds joining the adjacent I filaments are either different or existing under different conditions of strain. In either case it would endow that particular part of the Z-disk with altered mechanical strength.

Brandt, Lopez, Reuben & Grundfest (1967), Carlsen, Knappeis & Buchthal (1961) and Rome (1968) have reported that the myofilament lattice (A- and I-bands) expands laterally when the sarcomere shortens. Whether the lateral expansion of the Z-disk lattice is as great as the expansion of the hexagonal lattice of the A-band is not yet clear. Landon (1970) postulated that activation of the muscle, with the release of Ca²⁺ from the sarcoplasmic reticulum, actually initiated a change in the lattice structure in a manner similar to the influence of divalent cations on the crystalline structure of tropomyosin as reported by Cohen & Longley (1966). Clearly the exact nature of the changes in the Z-disk lattice during active tension development requires further investigation. However, it would appear that there will be an increase in the transverse mechanical strain, which would tend to pull the Z-disk apart (Goldspink, 1970). The fault lines occurring in the lattices would have a different mechanical strength from the rest of the lattice predisposing these as the sites of rupture. Figs. 4 and 5 show myofibrils which appear to have started to split along the lines of such faults. The resulting fissures are occupied by sarcoplasmic reticulum elements and are therefore unlikely to have resulted from internal stress arising during fixation.

Hypertrophy of muscle fibres has been shown to involve an increase in total myo-fibril cross-sectional area (Rowe, 1969). Morkin (1970) has shown that newly syn-thesized myofilaments are laid down on the periphery of existing myofibrils. The close correlation of fibre diameter and total myofibril number (Rowe, 1967) shows that there must be a mechanism for increasing myofibril number. Goldspink (1970) has suggested a mechanism in which the increase in myofibril number is achieved by longitudinal splitting.

Faults in the Z-disk lattice are commonly found in most bovine and rat muscle samples including foetal material. We regard them as being mistakes in the spatial arrangement of the filaments made when the new filaments are added on to the existing myofibrils. Such mistakes would appear to occur both during embryonic and post-natal growth of the muscle fibres. The consequence of the existence of such faults is seen as being a lowering of the ability of the Z-disk to withstand internal stresses, which increase with increased myofibril area (growth) and with sarcomere shortening (contraction).

These faults are regarded as imposing an upper limit on individual myofibril size. This influence would be greatest at a time of myofibril protein anabolism, which is the time of myofibril number increase, stimulated by either normal growth or by the stimulus of increased high work-load activity.

We wish to thank Mr J. F. Weidemann and Mrs Sue Tucek for their skilled assistance in preparing the tissue for electron microscopy, and Mr F. D. Shaw for performing the muscle biopsies.