th  - 1988 - 51st ENC Conference

RECENT EXTENSIONS OF NOESYSIM, A PROGRAM FOR RAPID COMPUTATION
112 [OF NOESY INTENSITY MATRICES FROM ATOMIC COORDINATES AND EXPERI-
MENTAL CONDITIONS. Hugh L. Eaton*, Niels H. Andersen, and Xiaonian Lai, University of Wash-
ington, Seattle, WA 98195.
The NOESYSIM program has been extended to allow calculation of NOESY matrices at any of
four levels of theory or approximation: 1) linear-limit isolated spin pairs; 2) isolated spin pairs with
leakage correction; 3) summation over all possible three spin systems; and 4) complete experiment
simulation by numeric integration (J. Magn. Reson. 74, 212). The latter implicitly includes
all spins and is equivalent to CORMA (Borgias and James, 28th ENC) for idealized experiments
with an effectively infinite repetition interval. Methods 3) and 4), which yield matrices that lack
diagonal symmetry for experiments with short repetition times, give superior agreement with
experimental data. We find that the three spin approximation, which requires considerably less
computation time than method 4, holds to 4- 10% for all systems examined throughout the range
wr c = 0.18 - 12. The three spin approximation thus appears to be suitable for conformational
refinement procedures based on experimental NOESY data. In contrast, we find that isolated spin
pair approximation frequently yields NOESY cross-peak intensities that differ from exact theory by
as much as a factor of 3, which corresponds.to as much as a 40% error in distance estimates.
For the purposes of NOESY analysis carried-out without computer assistance, the three spin
approximation is equivalent to
1 {S~j Sj~ rm
27" m k Sii "}" Sj3 } = Oi~ "}- -2- Z tTjk Oki
k
an equation that holds to better than 10% throughout the correlation time range that is found for
molecules in solution states amenable to high resolution NMR.
[--
{ 115 [alp MAGNETIC RESONANCE IMAGING OF SOLID CALCIUM PHOSPHATES:
POTENTIAL FOR CHEMICAL IMAGING OF BONE: Jerome L. Ackerman, a Daniel P. Raleigh, *b,c and
Melvin J. Glimcher; a aDepartment of Radiology, NMR Facility, Massachusetts General Hospital Boston, MA
02114; bDepartment of Chemistry, Massachusetts Institute of Technology Cambridge, MA 02139; CFrancis Bit-
ter National Magnet Laboratory, Massachusetts Institute of Technology Cambridge, MA 02139; aLaboratory
for the Study of Skeletal Disorders and Rehabilitation, Department of Orthopedic Surgery, Harvard Medical
School, Children's Hospital Medical Center, Boston, MA 02115
The use of NMR imaging of biological systems has been almost exclusively restricted to the fluid com-
ponents of tissues. An exciting possible application of NMR imaging is that of imaging of the phosphate
resonance in mineralized tissue. Previous spectroscopic experiments have demonstrated the potential of 31p
NMR in elucidating the chemical composition of the mineral phase of bone. With the eventual goal of ex-
tending these measurements to the imaging domain, we have been developing methods for the production of
phosphorus images in calcium hydroxyapatite (an accepted model for the major mineral phase of bone).
We have obtained one-dimensional projections of the alp resonance in synthetic hydroxyapatite for speci-
mens oil the order of 0.5 to 1.0 cm in linear extent at 7.4 T field strength. Because of the solid state nature of
these samples, short alp spin-spin relaxation times under 1 msec occur, necessitating echo times on the order:
of 1 msec and phase-encoding magnetic field gradient pulses under 500 /zsec. Although such T3 values are
easily managed by spectrometers, they are well below the range of minimum echo times (typically 15 msec or
greater) achievable by clinical imagers. Optimal projection quality and shortest total image acquisition times
result from pulsed gradient phase-encoding of the spatial dimension, using a compensating gradient pulse to
cancel the distorting effects of gradient waveform transients. The exceedingly long alp spin-lattice relaxation
times could lead to potentially intolerable image acquisition times; we have reduced these with a flipback
pulse teclmique. These methods should be of general utility in the multinuclear imaging of a wide variety of
solids of interest in biophysics .and the materials sciences.
155