Phosphorus-31 two-dimensional solid-state exchange NMR. Application to model membrane and biological systems.

Two-dimensional solid-state 31P NMR has been used to investigate the orientational exchange of phospholipids in gel and liquid-crystalline aqueous multilamellar dispersions and oriented multibilayers, and in biological membranes. In liquid-crystalline L alpha multilamellar dispersions, orientational exchange originates from the lateral diffusion of phospholipid molecules over the curved surface of the liposomes and is manifest by an increase in off-diagonal intensity, which correlates the 90 and 0 degrees orientations of the membrane normal with respect to the magnetic field when the system is fully exchanged. Spectral simulations of the time evolution of exchange allowed determination of the correlation times tau d for lateral diffusion. For DMPC and DPPC at comparable reduced temperatures, tau d values of 44 and 8 ms were obtained, respectively. The nature and rate of exchange observed for POPE at 30 degrees C is similar to that of DMPC at the same temperature. The measured correlation times are consistent with diffusion rates obtained by FRAP for liposomes with radii in the 1 micron range. In the gel phase of DPPC (30 degrees C), little orientational exchange is observed at mixing times up to 200 ms, demonstrating that the lateral diffusion is very slow. The correlation time for orientational exchange obtained from spectral simulations was approximately 900 ms; thus, exchange in the gel state is at least two orders of magnitude slower than in the liquid-crystalline state. In the P beta (ripple) phase, at temperatures between 34 and 39 degrees C, significant exchange is observed for mixing times between 50 and 200 ms. Exchange is also observed in oriented samples of DPPC in the P beta phase for mixing times of 50 ms, but not for oriented liquid-crystalline samples for mixing times up to 100 ms. The exchange observed in the ripple phase could originate from rapid lateral diffusion of "fast" diffusing phospholipid within defect structures, and/or from "slow" lateral diffusion of ordered phospholipid over the ripples. 2D experiments were also performed on pig erythrocyte ghosts and on intact pig spinal cord. Significant orientational exchange was observed with the erythrocyte ghosts at a mixing time of 200 ms, but almost no exchange was observed with the spinal cord at the same mixing time. Spectral simulations suggest tau d values of approximately 400 ms and 1.3 s for the erythrocyte ghosts and spinal cord at 30 degrees C.(ABSTRACT TRUNCATED AT 400 WORDS)     

A theoretical study of rotational diffusion models for rod-shaped viruses. The influence of motion on 31P nuclear magnetic resonance lineshapes and transversal relaxation.

Information about the interaction between nucleic acids and coat proteins in intact virus particles may be obtained by studying the restricted backbone dynamics of the incapsulated nucleic acids using 31P nuclear magnetic resonance (NMR) spectroscopy. In this article, simulations are carried out to investigate how reorientation of a rod-shaped virus particle as a whole and isolated nucleic acid motions within the virion influence the 31P NMR lineshape and transversal relaxation dominated by the phosphorus chemical shift anisotropy. Two opposite cases are considered on a theoretical level. First, isotropic rotational diffusion is used as a model for mobile nucleic acids that are loosely or partially bound to the protein coat. The effect of this type of diffusion on lineshape and transversal relaxation is calculated by solving the stochastic Liouville equation by an expansion in spherical functions. Next, uniaxial rotational diffusion is assumed to represent the mobility of phosphorus in a virion that rotates as a rigid rod about its length axis. This type of diffusion is approximated by an exchange process among discrete sites. As turns out from these simulations, the amplitude and the frequency of the motion can only be unequivocally determined from experimental data by a combined analysis of the lineshape and the transversal relaxation. In the fast motional region both the isotropic and the uniaxial diffusion model predict the same transversal relaxation as the Redfield theory. For very slow motion, transversal relaxation resembles the nonexponential relaxation as observed for water molecules undergoing translational diffusion in a magnetic field gradient. In this frequency region T2e is inversely proportional to the cube root of the diffusion coefficient. In addition to the isotropic and uniaxial diffusion models, a third model is presented, in which fast restricted nucleic acid backbone motions dominating the lineshape are superimposed on a slow rotation of the virion about its length axis, dominating transversal relaxation. In an accompanying article the models are applied to the 31P NMR results obtained for bacteriophage M13 and tobacco mosaic virus.     

Analysis of 31P nuclear magnetic resonance lineshapes and transversal relaxation of bacteriophage M13 and tobacco mosaic virus.

The experimentally observed 31P lineshapes and transversal relaxation of 15% (wt/wt) M13, 30% M13, and 30% tobacco mosaic virus (TMV) are compared with lineshapes and relaxation curves that are simulated for various types of rotational diffusion using the models discussed previously (Magusin, P. C. M. M., and M. A. Hemminga. 1993. Biophys. J. 64:1851-1860). It is found that isotropic diffusion cannot explain the observed lineshape effects. A rigid rod diffusion model is only successful in describing the experimental data obtained for 15% M13. For 30% M13 the experimental lineshape and relaxation curve cannot be interpreted consistently and the TMV lineshape cannot even be simulated alone, indicating that the rigid rod diffusion model does not generally apply. A combined diffusion model with fast isolated motions of the encapsulated nucleic acid dominating the lineshape and a slow overall rotation of the virion as a whole, which mainly is reflected in the transversal relaxation, is able to provide a consistent picture for the 15 and 30% M13 samples, but not for TMV. Strongly improved lineshape fits for TMV are obtained assuming that there are three binding sites with different mobilities. The presence of three binding sites is consistent with previous models of TMV. The best lineshapes are simulated for a combination of one mobile and two static sites. Although less markedly, the assumption that two fractions of DNA with different mobilities exist within M13 also improves the simulated lineshapes. The possible existence of two 31P fractions in M13 sheds new light on the nonintegral ratio 2.4:1 between the number of nucleotides and protein coat subunits in the phage: 83% of the viral DNA is less mobile, suggesting that the binding of the DNA molecule to the protein coat actually occurs at the integral ratio of two nucleotides per protein subunit.     

Identifying inter-residue resonances in crowded 2D 13C–13C chemical shift correlation spectra of membrane proteins by solid-state MAS NMR difference spectroscopy

The feasibility of using difference spectroscopy, i.e. subtraction of two correlation spectra at different mixing times, for substantially enhanced resolution in crowded two-dimensional ¹³C–¹³C chemical shift correlation spectra is presented. With the analyses of ¹³C–¹³C spin diffusion in simple spin systems, difference spectroscopy is proposed to partially separate the spin diffusion resonances of relatively short intra-residue distances from the longer inter-residue distances, leading to a better identification of the inter-residue resonances. Here solid-state magic-angle-spinning NMR spectra of the full length M2 protein embedded in synthetic lipid bilayers have been used to illustrate the resolution enhancement in the difference spectra. The integral membrane M2 protein of Influenza A virus assembles as a tetrameric bundle to form a proton-conducting channel that is activated by low pH and is essential for the viral lifecycle. Based on known amino acid resonance assignments from amino acid specific labeled samples of truncated M2 sequences or from time-consuming 3D experiments of uniformly labeled samples, some inter-residue resonances of the full length M2 protein can be identified in the difference spectra of uniformly ¹³C labeled protein that are consistent with the high resolution structure of the M2 (22–62) protein (Sharma et al., Science 330(6003):509–512, 2010).     

Quasielastic light scattering from solutions of filamentous viruses. I. Experimental

Intensity fluctuations of laser light scattered from filamentous viruses Pf1 [length L (Å) × diameter d (Å) = 20,000 × 90], M13 (9000 × 90), potato virus X (5150 × 130), and tobacco mosaic virus (3000 × 180) in sucrose density gradients were measured with a photon correlation spectrometer over a range of scattering angles from 15° to 120°. The experimental data can be approximated by two exponential decays, “slow” and “fast.” The slow decay rate constant t corresponds to the translational diffusion D of the virus, i.e., t = K²D, where K is the magnitude of the scattering vector. The amplitude of the slow component, i.e., translational diffusion, remains greater than that of the fast component, even at high KL. The fast decay rate constant t is also proportional to K² for viruses such as Pf1, M13, and even potato virus X. In the companion paper, we shall attribute the amplitude enhancement of the translational diffusion to the coupling of its anisotropy to the rotational diffusion modes. In order to explain the excessive decay rates in the fast component, we need to consider the bending mode of rodlike viruses, especially in the longer viruses such as M13 and Pf1, in addition to the usually expected rotational diffusion modes.     

Magnetic resonance studies of the binding site interactions between phosphorylcholine and specific mouse myeloma immunoglobulin

The interaction of phosphorycholine-binding mouse myeloma protein M603 and the isotopically substituted hapten phosphoryl[methyl-13C] choline has been investigated using 13C and 31P nuclear magnetic resonance (NMR) spectroscopy. Upon binding to antibody, upfield shifts of 0.7 and 1.5 ppm are observed for the hapten 13C and 31P resonances, respectively, and both spectra are in the "slow" exchange limit. Linewidth analysis indicates some immobilization of the phosphate group but essentially unrestricted methyl group rotation for the bound hapten. Hapten-antibody dissociation rate constants of 10 and 38 s-1 are calculated from 13C and 31P NMR spectra, respectively, suggesting the possibility of differential dissociation rates for the two opposing ends of the phosphorylcholine molecule. The NMR data are entirely consistent with the known x-ray structure of the M603 Fab'-phosporylcholine complex (Segal,D.M., Padlan, E.A., Cohen G.H., Rudikoff S., Potter,M., and Davies, D.R. (1974), Proc. Natl. Acad. Sci. U.S.A. 71, 4298).     

Identification of trapped and boundary lipid binding sites in M13 coat protein/lipid complexes by deuterium NMR spectroscopy

The major coat protein of M13 bacteriophage has been incorporated into bilayers of 1,2-dimyristoyl-sn-glycero-3-phosphocholine, deuterated in the trimethyl segments of the choline headgroup (DMPC-d9). Two-component deuterium and phosphorus-31 NMR spectra have been observed from bilayer complexes containing the coat protein, indicating slow exchange (on the deuterium quadrupole anisotropy and phosphorus-31 chemical shift averaging time scales) of lipid molecules of less than 10(3) Hz between two motionally distinct environments in the complexes. The fraction of the isotropic spectral component increases with increasing M13 protein concentration, and this component is attributed to lipid headgroups, which are disordered relative to their order in protein-free bilayers. The activation energy of the fast local motions of the trimethyl groups of the choline residue in the headgroup decreases from 23 kJ mol-1 in the pure lipid bilayers to 20 kJ mol-1 for the protein-associated lipid headgroups. The chemical exchange rate of lipid molecules between the two motionally distinct environments has been estimated to be 20-50 Hz by steady-state line-shape simulations of the deuterium spectra of DMPC-d9/M13 coat protein complexes using exchange-coupled modified Bloch equations. The off-rate was, as expected from one-to-one exchange, independent of the L/P ratio; tau off -1 = 0.23 kHz. It is suggested that the protein-associated lipid may be trapped between closely packed parallel aggregates of M13 coat protein and that the high local concentration of protein in a one-dimensional arrangement in lipid bilayers may be required for the fast reassembly of phage particles before release from an infected cell.     

The conformation of gramicidin A in dimethylsulphoxide solution. A full analysis of the one- and two-dimensional 1H, 13C, and 15N nuclear-magnetic-resonance spectra

The combined application of one- and two-dimensional high-field NMR techniques has led to the first assignment of the 1H, 13C, and 15N spectra of the pentadecapeptide gramicidin A in dimethylsulphoxide solution. The 62.9-MHz and 100.6-MHz 13C spin-lattice relaxation times and 13C-[1H] NOE factors for the backbone alpha carbons have been analysed in the 'model-free' approach to give a single correlation time (tau m) for isotropic overall molecular motion and an order parameter and internal correlation time for each C alpha H group in the backbone. The relatively high and constant values for the order parameter along the backbone indicate a degree of ordering of the structure, while the internal correlation times show that internal motions are progressively more rapid towards the N terminus. The average values of the vicinal HNC alpha H couplings are 7.4 Hz and 8.4 Hz respectively for the alternate L- and D-amino acid residues. The values are not consistent with either a ribbon conformation for the backbone or a right-handed beta 6.3 helix; they are consistent with the model proposed by Glickson et al. [Glickson, J. D., Mayers, D. F., Settine, J. M. & Urry, D. W. (1972) Biochemistry 11, 477-486] in which there is a rapid conformational order in equilibrium disorder equilibrium, the ordered structure being the left-handed beta 6.3 helix and the disordered state having local random-coil character.     

The detection of three viruses of hop (Humulus lupulus) by enzyme-linked immunosorbent assay (ELISA)

The recently described technique of enzyme-linked immunosorbent assay (ELISA) was used throughout the 1976 growing season to detect hop mosaic, arabis mosaic and prunus necrotic ringspot viruses in hop plants. On each occasion virus was detected quickly, conveniently and with great sensitivity. The technique was particularly suitable for processing numerous samples collected from the field. Serious difficulties and limitations were encountered in testing comparable material by established techniques. The serology test for the hop strain of arabis mosaic virus by double diffusion in agar gels was very insensitive and only worked satisfactorily early in the growing season. Grafting sensitive Golding hop varieties to detect hop mosaic virus was inconvenient and time-consuming and symptom expression was so slow and erratic that glasshouse space was utilized for long periods. It became impossible to detect prunus necrotic ringspot virus by sap inoculations to cucumber during an exceptionally hot period in mid-summer. The possibilities are discussed of exploiting the ELISA technique for use in large scale surveys and epidemiological studies on viruses of hop and other crops. Changes in the current methods of handling and extracting leaves are considered for increasing the throughput of samples.     

Analysis of 31P MAS NMR spectra and transversal relaxation of bacteriophage M13 and tobacco mosaic virus.

Phosphorus magic angle spinning nuclear magnetic resonance (NMR) spectra and transversal relaxation of M13 and TMV are analyzed by use of a model, which includes both local backbone motions of the encapsulated nucleic acid molecules and overall rotational diffusion of the rod-shaped virions about their length axis. Backbone motions influence the sideband intensities by causing a fast restricted reorientation of the phosphodiesters. To evaluate their influence on the observed sideband patterns, we extend the model that we used previously to analyze nonspinning 31P NMR lineshapes (Magusin, P.C.M.M., and M. A. Hemminga. 1993a. Biophys. J. 64:1861-1868) to magic angle spinning NMR experiments. Backbone motions also influence the conformation of the phosphodiesters, causing conformational averaging of the isotropic chemical shift, which offers a possible explanation for the various linewidths of the centerband and the sidebands observed for M13 gels under various conditions. The change of the experimental lineshape of M13 as a function of temperature and hydration is interpreted in terms of fast restricted fluctuation of the dihedral angles between the POC and the OCH planes on both sides of the 31P nucleus in the nucleic acid backbone. Backbone motions also seem to be the main cause of transversal relaxation measured at spinning rates of 4 kHz or higher. At spinning rates less than 2 kHz, transversal relaxation is significantly faster. This effect is assigned to slow, overall rotation of the rod-shaped M13 phage about its length axis. Equations are derived to simulate the observed dependence of T2e on the spinning rate.     

Quasielastic light scattering from solutions of filamentous viruses. II. Comparison between theories and experiments

We have compared four theoretical effects of rodlike macromolecules with the fast components, i.e., components other than translational diffusion, of our experimental data, which are presented as amplitude autocorrelation functions of electric field scattered from dilute solutions of monodisperse rodlike viruses with lengths from 3300 Å for tobacco mosaic virus to 20,000 Å for Pf1. The four effects are (1) the optic anisotropy treated by Aragón and Pecora, (2) coupled translational–rotational diffusion due to anisotropy in translational mobility recently reformulated by Gierke, (3) anisotropic rotational diffusion with respect to the direction of translational displacement first discussed by Berne and Pecora, and (4) the bending mode of a rod by Fujime and Maruyama. We show that both the first and second effects are required to explain the enhancement of amplitude of the translational diffusion at the expense of fast components. The experimental decay rates of the fast component exceed that of the rotational diffusions. In order to explain the excessive decay rate in the fast component, we need to include a minute amount (～1%) of bending mode of rodlike viruses, especially in longer viruses such as M13 and Pf1.     

Environmental modulation of M13 coat protein tryptophan fluorescence dynamics

The effects of detergent [deoxycholate (DOC) and phospholipid [dimyristoylphosphatidylcholine (DMPC)] environments on the rotational dynamics of the single tryptophan residue 26 of bacteriophage M13 coat protein have been investigated by using time-resolved single photon counting measurements of the fluorescence intensity and anisotropy decay. The total fluorescence decay of tryptophan-26 is complex but rather similar in DOC as compared to DMPC when analyzed in terms of a lifetime distribution (exponential series method). This similarity, in conjunction with the almost identical steady-state fluorescence spectra, indicates only minor differences between the tryptophan environments in DOC and DMPC. The reorientational dynamics of tryptophan-26 are dominated by slow rotation of the entire protein in both detergent and phospholipid environments. The resolved anisotropy decay in DOC can be approximated by a simple hydrodynamic model of protein/detergent micelle rotational diffusion, although the data indicative slightly greater complexity in the rotational motion. The tryptophan fluorescence anisotropy is not sensitive to protein conformational changes in DOC detected by nuclear magnetic resonance on the basis of pH independence in the range 7.5-9.1. In DMPC bilayers, restricted tryptophan motion with a correlation time of approximately 2 ns is observed together with a second very slow reorientational component. Resolution of the time constant for this slow rotation is obscured by the tryptophan fluorescence time window being too short to clearly locate its anisotropic limit. The possible contribution made by axial rotational diffusion of the protein to this slow rotational process is discussed.(ABSTRACT TRUNCATED AT 250 WORDS)     

Assignment of amide proton signals by combined evaluation of HN,NN and HNCA MAS-NMR correlation spectra

In this paper, we present a strategy for the ¹H^N resonance assignment in solid-state magic-angle spinning (MAS) NMR, using the -spectrin SH3 domain as an example. A novel 3D triple resonance experiment is presented that yields intraresidue H^N-N-C correlations, which was essential for the proton assignment. For the observable residues, 52 out of the 54 amide proton resonances were assigned from 2D (¹H-¹⁵N) and 3D (¹H-¹⁵N-¹³C) heteronuclear correlation spectra. It is demonstrated that proton-driven spin diffusion (PDSD) experiments recorded with long mixing times (4 s) are helpful for confirming the assignment of the protein backbone ¹⁵N resonances and as an aid in the amide proton assignment.     

THE INFLUENCE OF LIGHT INTENSITY ON THE SUSCEPTIBILITY OF PLANTS TO CERTAIN VIRUSES

Reducing the light intensity under which plants were grown in summer to one-third increased their susceptibility to infection with tobacco necrosis, tomato bushy stunt, tobacco mosaic and tomato aucuba mosaic viruses. With the first two viruses shading increased the average number of local lesions per leaf by more than ten times and by more than five times with the second two.
Reducing the light intensity increased the virus content of sap from leaves inoculated with Rothamsted tobacco necrosis virus by as much as twenty times. As it also reduced the total solid content of sap by about one-half, purification was greatly facilitated; crystalline preparations of the virus were readily made from shaded plants but not from unshaded controls.
Reducing the light intensity also increased the virus content of systemically infected leaves; the greatest effect was with tomato bushy stunt virus with which increases of up to ten times were obtained, but with tobacco mosaic and aucuba mosaic viruses there were also significant increases.
The importance of controlled illumination in raising plants for virus work and the possible mechanisms responsible for the variations in susceptibility are discussed.     

Etiology of two virus diseases of lilac (Syringa vulgaris L.) occurring in Czechoslovakia

Virus origin of lilac ringspot and chlorotic ringspot of lilac was identified serologically by means of double gel diffusion method. The former of these diseases is due to Arabis mosaic virus, the latter to a mixture of Arabis mosaic virus and cherry leaf roll virus. The occurrence of these viruses has been detected for the first time in Czechoslovakia.     

Optical rotatory dispersion by 5 viruses of the tobacco mosaic virus group and their components

Optical rotatory dispersion (ORD) spectra in 250 to 350 nm region were measured for preparations of five TMV-like viruses (TMV vulgare, HR and U2 strains of TMV dolihosenation mosaic virus and cucumber virus 4) and also for RNA and protein preparations of these viruses. The data obtained testify against the possibility that the double peak with maxima at 286 and 293 nm observed in ORD of all the five viruses is due to interaction of tryptophan residues in virus coat protein with the RNA of the virul particle. The spectra of intravirus RNA of the five viruses, calculated as the difference between ORD of the intact virus and of its coat protein, were found to differ significantly from each other and from ORD of free RNA. ORD spectra of hybrid viruses, reconstituted from RNA of one virus and coat protein of another, proved to be identical to the ORD of the virus, whose protein was used in reconstitution. We suppose that the difference in ORD of the intravirus RNA of the five viruses reflect differences of RNA-protein interactions in them.     

Genetic analysis of helper virus-specific selective amplification of cucumber mosaic virus satellite RNAs

Satellite RNAs (sat-RNAs) are small molecular parasites associated with a number of plant RNA viruses. The cucumber mosaic virus (CMV) sat-RNAs are ca. 335 nucleotides and have evolved to produce a large number of closely related sat-RNAs. Different cucumoviruses can act as helper viruses in the amplification of CMV sat-RNAs. We have found that different helper viruses show a preference for a particular sat-RNA in a mixed infection. In this study the specificity of WL47 sat-RNA amplification by LS-CMV and of D4 sat-RNA amplification by tomato aspermy virus in mixed infections was examined. Recombinant cDNA clones of D4 sat-RNA and WL47 sat-RNA were used to map the sat-RNA sequences responsible for the helper virus selection of a specific sat-RNA for amplification.Correspondence to: M.J. Roossinck     

Slow motion in the CAA*TTG sequence of a DNA decamer duplex studied by NMR

In DNA duplexes, pyrimidine-purine steps are believed to be flexible or conformationally unstable. Indeed, several DNA crystal structures exhibit a multitude of conformations for CpA*TpG steps. The question arises of whether this structural flexibility is accompanied by dynamical flexibility, i.e., a question pertaining to the energy barrier between conformations. Except for TpA steps, slow motions on the microsecond-to-millisecond time scale have not been detected in duplexes until now. In the present study, such slow motion was investigated by 1H, 13C, and 15N NMR relaxation measurements on a DNA decamer d(CATTTGCATC)*d(GATGCAAATG). The DNA decamer was enriched with 15% 13C and 98% 15N isotopes for each adenosine and guanosine residue. Three lines of evidence support the notion of slow motion in the CAA*TTG moiety. Analysis of (15)N relaxation showed that the order parameter, S2, of guanosine imino NH groups was about 0.8, similar to that of CH groups for this oligomer. The strong temperature dependence of guanosine NH S2 in the CAA*TTG sequence indicated the presence of a large-amplitude motion. Signals of adenosine H8 protons in the CAA*TTG sequence were broadened in 2D 1H NOESY spectra, which also suggested the existence of slow motion. As well as being smaller than for other adenine residues, the 1H T2 values exhibited a magnetic field strength dependence for all adenosine H8 signals in the ATTTG*CAAAT region, suggesting slow motions more pronounced at the first adenosine in the CAA*TTG sequence but extending over the CAAAT*ATTTG region. This phenomenon was further examined by the pulse field strength dependence of the 1H, 13C, and 15N T1rho values. 1H and 13C T1rho values showed a pulse field strength dependence, but 15N T1rho did not. Assuming a two-site exchange process, an exchange time constant of 20-300 micros was estimated for the first adenosine in the CAA sequence. The exact nature of this motion remains unknown.     

Interaction of actinomycin D, ethidium, quinacrine, daunorubicin, and tetralysine with DNA: 31P NMR chemical shift and relaxation investigation.

The binding of actinomycin D, ethidium, quinacrine, daunorubicin, and tetralysine to DNA has been investigated using 31P NMR. Titration of DNA with actinomycin yields a new downfield peak or overlapping peaks as would be expected from the slow dissociation kinetics of this compound. The other intercalators shift the DNA 31P signal downfield as a single exchange averaged peak. Tetralysine causes a slight upfield shift. The chemical shift titration curves for the intercalators are sigmoid curves suggesting that cooperative processes or competing effects on the chemical shift are being observed. The magnitude of the chemical shift change at saturation of DNA with the compounds is found to vary significantly and to be linearly related to the DNA base pair unwinding angle for the compounds. Analysis of 31P spin lattice relaxation times (T1) and linewidths as a function of temperature (below Tm) and titration with the above compounds indicates that T1 does not change significantly while linewidth increases with decreasing temperature and increasing bound intercalator. One interpretation of these results is that in both cases the overall motion of DNA becomes slower while the internal motion is not greatly affected.     

Incidence of amantadine-resistant influenza A viruses in sentinel surveillance sites and nursing homes in Niigata, Japan

We surveyed the incidence of amantadine-resistant influenza A viruses both at sentinel surveillance sites and at nursing homes, and verified their types of change by partial nucleotide sequence analysis of the M2 protein. Fifty-five influenza A viruses from 27 sentinel surveillance sites during six influenza seasons from 1993 to 1999, and 26 influenza A viruses from 5 nursing homes from 1996 to 1999 were examined for susceptibility to the drug by virus titration in the presence or absence of amantadine. While amantadine-resistant viruses were not found in sentinel surveillance sites, a high frequency of resistance (8/26, 30.8%) in nursing homes was observed. Resistant viruses can occur quickly and be transmitted when used in an outbreak situation at nursing homes, where amantadine is used either for neurologic indications or for influenza treatment. Eight resistant viruses had a single amino acid change of the M2 protein at residue 30 or 31. In vitro, all 11 sensitive viruses turned resistant after 3 or 5 passages in the presence of 2 microg/ml amantadine, and they showed an amino acid change at residue 27, 30, or 31. The predominant amino acid substitution in the M2 protein of resistant viruses is Ser-31-Asp (a change at 31, serine to asparagine). The results indicate that a monitoring system for amantadine-resistant influenza viruses should be established without delay in Japan.