A model is presented for the mechanism of packaging double-stranded DNA into phage heads. The model is based on, and rationalizes, the mismatch in symmetry between the heads and tails of large bacteriophages. DNA movement is postulated to be mediated by a rotating protein structure at the tail-proximal vertex of the head. A long-standing problem of virus structure concerns the fact that the symmetries of the heads and tails of large bacteriophages are different. All bacteriophage tails for which adequate data are available are now known to have 6-fold rotational symmetry, and they attach to the icosahedral head at one of the axes of 5-fold rotational symmetry (1-5). I propose here a general model for the structure of the large double-stranded DNA-containing phages that accounts for the symmetry mismatch and argue that the symmetry mismatch makes available an interesting new class of models for how double-stranded DNA is packaged into phage heads. The phages to which this discussion applies include X, T4, P22, P2/P4, T7, T3, T5, 029, and others. Casjens and King (6) have recently reviewed the general features of the structure and assembly of these viruses.Phage structure A structural feature that is probably common to all the phages under consideration is a small protein knob that lies just inside the head membrane at the unique corner of the icosahedral head where the tail attaches. This structure, which I will call the connector, is assembled as part of the head and, in at least several phages, has been implicated in the earliest steps of head assembly. The phage tail attaches directly to the connector after DNA packaging. Connectors appear to have a roughly cylindrical shape, and in some cases (X, T7, T4, P22) an axial hole has been visualized oriented along the head-tail axis (4,7,8; W. C. Earnshaw, personal communication). During injection, the DNA passes out of the head and into the tail through the corner where the connector lies, and it is generally assumed that the DNA passes out through the connector. It is also likely that the DNA passes in through the connector when it is being packaged into the head.Although it is clear that tails have a different symmetry from the icosahedral head membrane, it has never been established whether the interface between the 5-and 6-fold symmetries is at the junction between the tail and the connector or at the junction between the connector and the head membrane. [A third possibility (1)-that the connector is an adaptor with both 5-and 6-fold symmetry (i.e., with 30-fold symmetry)-seems ruled out for at least many of the phages because no proteins are present in the required 30 copies.] The symmetry of the connector has not been determined directly for any of the phages. Table 1 lists the available measurements of the numbers of copies of the proteins in connectors for several phages. Most of the numbers can plausibly be interpreted as multiples of either five or six copies, and none of the data rule out either hypothesis conclusively.I propose that the connect...