Water elimination by T2 relaxation in 1H spin-echo FT NMR studies of intact human erythrocytes and protein solutions

Water elimination by T2 relaxation in 1H spin-echo FT NMR studies of intact human erythrocytes and protein solutions

JOURNAL OF MAGNETIC RESONANCE 36, 281-286 (1919) Elimination by TZ Relaxation in ‘H Spin-Echo FT NM ies of Intact Human Erythrocytes and Protei...

353KB Sizes 0 Downloads 29 Views

JOURNAL

OF

MAGNETIC

RESONANCE

36,

281-286

(1919)

Elimination by TZ Relaxation in ‘H Spin-Echo FT NM ies of Intact Human Erythrocytes and Protein Solutio It recently has been shown that ‘H NMR spectra can be obtained for some of the small molecules naturally present in intact human erythrocytes by spin-echo Fourier transform NMR (1,Z). With this technique, ‘H NMR spectra can also be obtaine for small molecules in protein solutions (2). D20 is generally used as the solvent in these and other ‘H NMR studies of biological molecules in aqueous solution because of the problems associated with detecting weak signals in the presence of the strong E&Q signal. A variety of techniques have been described for reducing the solvent signal (3-11), some of which can also be used in the spin-echo J?T NMR experiment. For example, the Hz0 resonance in the spin-echo spectrum for a suspension of erythrocytes in isotonic Hz0 solution can be reduced in intensity by application of a selective saturation pulse (1). In this communication, we wish to report that, with a high field spectrometer, tbe or HzQ resonance from intact erythrocytes or from bovine serum albumin ns can be selectively reduced in intensity or completely eliminated by simply using a sufficiently long delay between the 90” pulse and acquisition of the free induction decay in the spin-echo experiment. This selective elimination of the HDO or Hz0 resonance is possible because of the anomolous frequency dependence of its spin-spin relaxation rate in these samples, specifically its relaxation rate increases as the frequency increases whereas the relaxation rates of other small molecules show little or no frequency dependence. Figure 1 shows a series of 400-MHz spectra measured by the spin-echo sequence (T-90”-r~-180°-~z-acquisition) for erythrocytes which had been washed with isotonic DZO-saline solution. Considerable spectrum simplification results as r2 is increased due to the shortness of the Tz values of resonances from hemoglobin as compared to those from the small molecules (1,Z). For example, at r2 = 0.015 set, the spectrum consists of sharp and broad resonances, due mainly to small molecules and the hemoglobin, respectively. At longer rz values, the resonances in the O- to S-ppm region are due mainly to the small molecules; the hemoglobin resonances are either much less intense or completely absent. The residual HDO signal (4.77 ppm) also decreases in intensity as r2 is increased, however its rate of decrease is somewhat faster than that for signals from other small molecules in the erythrocytes. For example, with r2 = 0.015 set, the HDO signal is considerably more intense than any other signal in the spectrum, whereas at 72 = 0.120 set it has almost disappeared. In fact, by increasing 72 further, the signal will completely disappear. The spectra in Fig. 2 illustrate this for erythrocytes which had been washed with isotonic I&O-saline solution. With Q = 0.060 set, other signals are of negligib intensity relative to the Hz0 resonance. When 72 is increased to 0.165 set, the HZ 281

OOZZ-2364/79/110281-06$02.00/O Copyright @ 1979 by Academic Press, Inc. All rights of reproduction in any form reserved. Printed in Great Britain

282

COMMUNICATIONS

.ObOsec

.120sec

FIG. 1. ‘H spin-echo FI NMR spectra (400 MHz) of packed intact human erythrocytes which bad been washed four times with isotonic Da0 solution. Spectra were measured on a Bruker WH-4OQ/DS spectrometer using quadrature detection and a SOOO-Hz spectral width. Spectra were measured at 25°C on approximately 0.4 to 0.5 ml of packed cells contained in 5-mm-o.d. NMR tubes. Four-hundred transients were accumulated using 8K of data points. Chemical shifts are reported relative to the methyl resonance of DSS, based on the resonance for the o-CHa protons of the Gly residue of glutathione having a chemical shift of 3.760 ppm (I).

resonance and the other resonances are of comparable intensity, and by r2 = 0.280 set, the Hz0 resonance has been completely eliminated by T2 relaxation. The selective elimination of the solvent resonance in the spectra in Figs. 1 and 2 results from its relatively rapid TZ relaxation. The spectra in Fig. 3 show its TZ relaxation rate to be frequency dependent, increasing as the frequency increases. Because there is much less frequency dependence for the T2 relaxation rates for resonances from the other small molecules, the selective elimination of the water resonance by T2 relaxation is much more efficient at the higher field.

COMMUNICATIONS

283

FIG. 2. ‘H spin-echo IT NMR spectra (400 MHz) of packed intact human erythrocytes which had been washed with isotonic Hz0 solution. See the legend for Fig. 1 for experimental details. The rs = 0.060-set spectrum is plotted to show the Hz0 signal; numerous other signals are observable when the display gain is increased. The FIDs for the ~a = 0.165 and 0.280-set spectra were normalized relative to the 7s = 0.060set spectrum; the increased display gains used in plotting the longer rs spectra are given in the figure.

The frequency dependence of the Hz0 TZ relaxation is opposite that expected for the dipolar mechanism (12). At the frequencies used in this research, the dipolar contribution to T2 relaxation is predicted to be frequency dependent for correlation times in the range of 6 x lo-r1 to 1 x lo-* set, however the dependence is such that T, is predicted to be longer at 400 MHz. Because the spectra shown in Figs. 1-3 were measured by the Hahn spin-echo sequence, diffusion might contribute to the observed relaxation rates if it is sufficiently rapid and if there are magnetic field gradients (13). Brindle et al. (14) have shown that there is a difference in the magnetic susceptibility of the media inside and outside the erythrocyte, and also that the magnetic susceptibility is different for different regions within the erythrocyte. Because the magnitude of gradients caused by differences in magnetic susceptibility would be dependent on the strength of the externally applied field, the contribution of diffusion to the observed Tz relaxation rates would be expected to be field dependent. However, we have made T2 measurements on the HDO resonance for a variety of samples, including packed erythrocytes which had been washed with isotonic saline in DzO, packed erythrocytes washed with hypotonic saline in hemolyzed erythrocytes and bovine serum albumin solutions at 200 and 400

284

COMMUNICATIONS 400 MHz

200 MHz

iI

0.060 set

0090 set 4x JU

10

8

6

2 PPk

II

32x --.-.-JddLJ

)\

0

10

8

i

111111 d’& 6

4 PPm

2

0

FIG. 3. ‘H spin-echo FT NMR spectra (200 and 40G MHz) of packed intact human erythrocytes which had been washed three times with isotonic DsO solution. The 200-MHz spectra were measured on a Bruker WH-200/DS spectrometer using quadrature detection, a 2500-Hz spectral width, and 4K of data points. Experimental details for the 400-MHz measurements are similar to those given in Fig. 1. The FIDs in each series were normalized relative to the 7s =0.030-set FID. The spectra are plotted so the HD8 signals are of comparable intensity; the relative display gains within each series are given in the figure.

both by the Hahn spin-echo and the Carr-Purcell-Meiboom-Gill pulse sequences, which suggest that diffusion is not the cause of the frequency dependence. epresentative results from several of these experiments are presented in Table 1. Even though the origin of the anomolous frequency dependence of the Hz relaxation rate has yet to be established, it is clear from the spectra in Figs. l-3 the rapid rate at high field makes possible the easy elimination of the I-ID8 or resonance in ‘H spin echo FT NMR measurements on small molecules in intact erythrocytes and other protein solutions. This should be of particular importance in the study of metabolic processes directly in intact cells by ‘H NMR. In these studies, it generally is desirable that Hz0 be the intracellular solvent because many metabolic processes involve the addition of solvent hydrogen to a carbon framework. Also, fewer manipulations of the cells are necessary prior to the NMR measurements.

COMMUNICATIONS TABLE

285 1

'& VALUES FOR THE HDO RESONANCE OF HUMAN ER~ROCYTES SERUMALBUMINSOLUTIONS~

AND BOVINE

T2(set) Pulse sequence

Sampleb A

T (4

Hahn CPMGd Hahn Hahn

A B C

72

(se4

10 10 2 2

variable’ 0.001 variable’ variable’

200 MHz

400 MHz

0.064 0.066 0.083 0.116

0.031 0.036 0.038 0.085

a Value of Tr for the HDO resonance for intact erythrocytes is -1.7 set at 200 MHz and -1.9 set at 400 MHz. b Sample A-packed erythrocytes, washed three times with DsO solution containing 0.154 A4 NaCl and 0.005 M glucose. Sample B-hemolyzed erythrocytes which had been washed three times with a DsO solution containing 0.154 M NaCl and 0.005 M giucose. SampIe B was hemolyzed in the NMR tube by several freeze-thaw cycles. Sample C-O.003 M bovine serum albumin in DaO; the BSA was purified by dialysis. ’ At 200 MHz, rs values ranged up to 0.180 set in O.OlO-set increments. At 400 MHz, ra values ranged up to 0.105-set in 0.00%set increments. d T-90°-(rs-1800-r2)n-acquisition.

We presently are investigating the frequency dependence of the Hz0 T2 relaxation rate in other cellular systems and solutions of macromolecules, and we have experiments in progress to identify the origin of the effect in such samples. ACKNOWLEDGMENTS This Council related

research was supported in part by a grant from the Natural Sciences and Engineering of Canada and by the University of Alberta. It is a pleasure to acknowledge numerous to this research with Dr. Tom Nakashima and Glen Bigam.

Research discussions

REFERENCES

1. F.F.BROWN,I.D.CAMPBELL,P.W.KUCHEL,ANDD.L.RABENSTEIN, 2. 3. 4. 3: 6. 7.

Left. D. L.

Fed.Eur.Biochem.Soc.

82, 12 (1977).

RABENSTEIN, Anal. Chem. 50, 1265A (1978). S. L.PA'~~AND B. D. SYKES,~ Che~.Phys. S&3182(1972). F. W.BENZ,J.FEENEY, AND G.C.K. ROBERTS,.~. Magn.Reson.8,114(1972). T.R.KRuGH AND W.C. SCHAEFER, J. Magn.Reson. 19,99 (1975). E.S.MOOBERRYANDT.R.KRUGH, J. Magn.Reson. 17,128 (1975). I.D.CAMPBELL,C.M.DOBSON,G.JEMINET,ANDR.J.P.WILLEAMS, Fed.Eur. Left.

49,115

Biochem.§oc.

(1974).

8. 9. IO. 11.

J.P.JESSON,P.MEAKIN, AND G.KNEISSEL,J. Ame: Chem. Soc.95,618(1973). B.L.ToMLINsoNANDH.D.W.HILL,J. Chem.Phys.59,1775 (1973). A.G.REDFIELD,S.D.KUNZ,AND E.K.RALPH,J. Mugn.Reson. 19,114(1975). D. ~.LOWMANAND G.E. MACIEL, Anal. Chem. 51,85 (1979).

12.

R. A. DWEK, 1975.

“Nuclear

Magnetic

Resonance

in Biochemistry,”

p. 26, Oxford

Univ.

Press, London,

COMMUNICATIONS

286

13. R.L.VOLD,R.R.VOLD,AND H.E.SIMON, J. Magn. Reson. 11,283(1973). 24. K.M.BRINDLE,F.F.BROWN,I.D.CAMPBELL,C.GRATHWOHL,ANDP.W.KUCIIEL,

Biochem.

J. 180,37 (1979). DALLAS L. RABENSTBIN" ANVARHUSEIN A. ISA_B

Department of Chemistry University of Alberta Edmonton, Alberta, Canada

T6G 2G2

Received July 2, 1979 *To

whom

correspondence

should

be addressed.