The amino-terminal domain of influenza A virus matrix protein (residues 1-164) was crystallized at pH 7 into a new crystal form in space group P1. This packing of the protein implies that M1(1-164) was monomeric in solution when it crystallized. Otherwise, the structure of the M1 fragment in the pH 7 crystals was the same as the monomers in crystals formed at pH 4 where crystal packing resulted in dimer formation [B. Sha and M. Luo, 1997, Nature Struct. Biol. 4, 239-244]. Analysis of intact M1 protein, the N-terminal domain, and the remaining C-terminal fragment (residues 165-252) in solution also showed that the N-terminal domain was monomeric with the same dimensions as determined from the crystal structure. Intact M1 protein was also monomeric but with an elongated shape due to the presence of the C-terminal part. Circular dichroism showed that the C-terminal part of M1 contained helical structure. A model for soluble M1 is presented, based on the assumption that the C-terminal domain is spherical, in which the N- and C-terminal domains are connected by a linker sequence which is available for proteolytic attack.

The M1 protein of influenza virus is thought to make contact with the cytoplasmic tails of the glycoprotein spikes, lipid molecules in the viral membrane, and the internal ribonucleoprotein particles. Here we show electron micrographs of negatively stained virus particles in which M1 is visualized as a 60-A-long rod that touches the membrane but apparently is not membrane inserted. Photolabeling with a membrane restricted reagent resulted in labeling of the transmembrane region of haemagglutinin but not of M1, also suggesting that most of M1 is not embedded into the hydrophobic core of the viral membrane. Finally, in vitro reconstitution experiments using soluble M1 protein and synthetic liposomes or Madin-Darby canine kidney cell membranes suggest that M1 can bind to negatively charged liposomes and to the cellular membranes and that this binding can be prevented under high-salt conditions. Although none of these experiments prove that there does not exist a minor fraction of M1 that is membrane inserted, it appears that most of M1 in the virus is membrane associated through electrostatic interactions.

Rabies virus nucleoprotein (N) was produced in insect cells

The adenovirus fiber appears as a long, thin projection terminated by a knob (head). The fiber consists of a trimeric protein whose head domain is thought to interact with cell receptors. The head part (amino acids 388 to 582) of adenovirus type 2 fiber was produced in a baculovirus expression system. The purified protein was shown to cross-link into trimers. It was very resistant to proteolytic attack and seemed to attain a high degree of compactness. The head domain efficiently inhibited attachment of adenovirus to receptors on the surface of HeLa cells, thereby confirming the hypothesis that the head domain interacts with viral receptors.One of the major constituents of the adenovirus outer capsid is an elongated structure, the fiber, protruding from each of the twelve fivefold vertices of the icosahedral virion (for a review, see reference 10). These fibers play a crucial role in adenovirus infection by attaching the virus to specific receptors on the cell surface (2,13,18).The adenovirus type 2 (Ad2) fiber appears as a long, thin projection terminated by a knob (head) (15,16) and is a trimer of three identical subunits (17). Ad2 fiber polypeptide (582 amino acids) can be divided into three regions (4): a short amino-terminal tail region, a shaft consisting of repeating units each of approximately 15 amino acids whose relative hydrophobicities rather than strict identities tend to be conserved, and a carboxy-terminal part of about 180 residues.