'hysiol.59,9-14.

tems (eds S. N. TIMASHEFF

mmun. 66,505-513-8,89-92.77) Biochim- biaphys. Aaa

243, M. Piccin, Padova.

J BoNl.\'ENrun4 J. (1977):.

H. BenNrs) pp. 673-681

7-472.

bins since this manuscript r'is of Marphysa sanguinea'nd C. Bonaventura (1978):he blood from Arenicoli;160). The self-diffirsion of'978) 22,453-468) and aseleech has been made byli

' :ri

'restris erythrocruorin by rta for the erythrocruorin':;.ius 5.4nm. tn" -od"t of 'li

ent studies. Two of tleselest for Hb-3 (C. Van den)A, 185-187; J. D'hondt,[b-2 was tle first haemo-iveiy (J. Heip, L. Moens,;en on the synthesis of ther and M. Kondo (1978)'i

from the clam77-85. lt was concluded;

ing domains. Neither thefunctional properties on

ssq. Biopltys. molec. Biol.. Vol. 35, pp. 103_134.

A i"rguron Press Ltd.. 1979. Printed in Crear Brirai!00'7 9-4 107 l'7 9 103014103 $05.00/0

106106108110110

1 1 1r721121 1 4

1 1 51211 ) 4126127127128

129

130

130

THE NMR STUDIES OF WATE'R IN BIOLOGICAL SYSTEMS

R. M.q,rHun-Ds Vn6uniuersitb Libre de Bruxelles, Laboratoire de Radiobiologie Molbctiaire, rue des Chet:aux 67,

rrstitut d,Hygiine et tEpid6riffrYj;r*;#:::';,u,:]ly##!^",t,rue ruriette wvtsmanT4.10 50 Bruxelle s. B el a ium

CONTENTSI. INTRODUCTION

IJ, HYDRATIoN WATER

ru. NMR eNo rrs Appr,rcATroN ro W,c,TEn SruorEs1. Basic Concepts2. The Relaxation Rates3. Difusion Constant4. A Comparison of Diferent Water Nuclei

lV. Tgn NMR Srunrrs on Werrn rN DrrnnnNr Brorocrc.lr Sysrpus1.. Randomly Oriented Systems

(a) Biopolymer solutions and hyfuated proteins@) Efects of y-irradiation on hydratioi water(c) Muscles and tissues(d\ Membranes and ceII water(e) Tumours and cancerous cells(f) NMR imaging by zeugmatography

2. Oriented Fibres(a) Collagen

. (b\ DN A Fibres

V. SUMMARY AND CoNcLUsroNs

ACKNoWLEDGEMENTS

RmrnrNcns

103

104

I . INTRODUCTION

. The state of water in the body constituents of living organisms and in the vicinity ofbiological macromolecules differs signifi-cantly from t]re s-tate of water in solutions ofsimple molecules and in pure water. Nuclear magpetic resonance (NMR) spectroscopyis a powerful techniqle for studying in detail tie slructure, mobility, ani extent ofordering of water molecules in various biological systlms. The NMR

-spectra of waterpresent in a sample can be obtained either fr6m the total water conten! or specifically

from the hydration water or intracellular water depending on the nature of the sample.In-order to fully appreciate the unique contribution of Nun spectroscopy in itredomain of hydration studies, it is important to note that the behavio*; ;;'" ;;;

/ n \ ^

molecule (H/"\t) can be investigated by looking at the nuclear magnetjc resonanceof four different nucleij these are the three isotopes of hydrogen: proton (rH),deuterium (2H), tritium (3H),,and-oxygen 17o._of these, the mosi wioery uod "^t"r,riu"rystudied nucleus is 1H; the'applications of 2H and iro NMR of water have beenli$"1 to the study of relativeiy few systems. The NMR studies of tritiated water inolologlcal systems apparently have been neglected so far, but in the course of time theyare likely to become a centre of intensive and growing research as a result of ther€cent progress in the development of tritium NMR .p"itror"opy. The importance ofthe NMR studies of tritiated water lies in the fact that such invesiigations aie expectedLo,o*"t..u

ne.w a.nd promising method flor exploring the biological hazards of tritiumtttre radroactlve lsotope of hydrogen) present in the environment. The widespread risks

r04 R. Mlrsun-Dp Vnr

arise because tritium released from nuclear reactors inevitably pervades the environmentas tritiated water: a form in which tritium can be readily absorbed by plants andanimals.

The NMR parameters most useful in the study of water are the relaxation times.Tt Tz and T1o. The diffusion constant and the distribution of correlation times forwater can generally be evaluated from relaxation times measured under specificconditions. It is well known that the relaxation processes for protons, deuterons, and17O, are influenced by different factors. Therefoie, for a comprehensive study of thestate of water in any system, using NMR, it is important to compare the relaxationtime data for more than one nucleus.

The present article comprises: a brief discussion of the characteristics of hydrationwater, an outline of the theory of NMR relaxation times with emphasis on the factorsrelevant to the study of water, and final1y an extensive review covering the NMRstudies of the different water nuclei in various biological systems and in the damageinduced by y-irradiation in the aqueous medium. Different theories related to eitherthe structure of water in general, or to cell water in particular, have not been describedat any fength in this review. Several monographs have covered these subjects inconsiderable detail (Ling, 1.962, 1969; cope, 1970; Kavanau, 1964; Eisenberg andKauzmann, 1969; Franks,1972-1975). A variety of physical techniques other than NMRspectroscopy are also highly suitable for the study of the hydration of biologicalmacromolecules. A number of methods which have been applied to nucleic acidJ are ,discussed explicitly in a recent article by Texter (1978).

I I . HYDRATION WATERBiological macromolecules induce a characteristic water structure in their close

vicinity due to weak macromolecular-water interactions. The solvent water moleculesinteract with the solute species by electrostatic forces (dispersion and induction forces)because of the high dipole moment of water, as well as through extensive hydrogenbonding by virtue of the potentially available proton donor and proton accepting siies.Consequently, macromolecules form a well-defined hydration layer in solution, inhydrated fibres (DNA, collagen) and in all biological samples such as intracellular water,tissues, muscle, and membrane. The water molecules contributing to the hvdration laverare dynamically oriented and exhibit restricted motion due to a significant decrease in ,the translational and rotational modes of motion caused by macromolecular-water :interactions. As a result, the mgbility and the extent of ordering of hydration water imolecules are distinctly different from those characterizing the fast and random motionof the bulk water. Furthermore, the overall behaviour of water molecules in the hydrationlayer is influenced by factors other than simple molecular interactions. Berendsen (1975)has distinguished three aspects of hydration water: thermodynamic aspects, dynamicaspects, and structural aspects. A complete description of hydration should necessarilycomprise all three aspects.

The NMR studies of water illustrate that certain coherent and cumulative orocessescome into effect when biological macromolecules interact with water, either in solutionor in biological systems. These processes (outlined below) are favoured by the extended,ordered, and complex structure of macromolecules (such as proteins and nucleic acids)and they govern different dynamic states of the hydration water molecules. (1) Therestricted motion of water in different fractions of the hydration layer is not unifornr, ,and all the dynamic modes of these water molecules cannot be described completely bya single correlation time but require a distribution of correlation times. (4Th; diffusionof water molecules in the hydration layer occurs in an anisotropic manner. (3) Thehydrogen bonded interactions between water molecules and ,"u"rul proton donor andproton accepting groups of macromolecuies give rise to a continuous hydrogen bondedpath in the hydration layer. This process promotes enhanced proton transfer alone the

rIS1

tLtlI tinb(lta

prdernisrth

reldilofN]ofchibert'i o^ ' b

llflr

1in,etarirhy.stui s e

Idianeilmoprotha

ls the environment)ed bY Plants and

: relaxation times,rrelation times for

:ed under sPecificrns, deuterons, andrnsive studY of the

rare the relaxation

ristics of hYdrationrasis on the factors;overing the NMRand in the damage:s related to €ithernot been describedI these subjects in

64; Eisenberg andes other than NMRation of biological:o nucleic acids are

:ture in their close

ent water moleculesnd induction forces)extensive hYdrogen

oton accePting sites'Lyer in solution, in

.s intracellular water,r the hYdration laYer,gnificant decrease rn

acromolecular-water; of hydration waterand random motlon:ules in the hYdrationrns. Berendsen (1975)

nic aspects, dYnamicrn should necessarilY

cumulative Processester. either in solutionrred by the extended,ns and nucleic acids),:r molecules. (1) Thelaver is not uniform,scr-ibed comPletelY bY

imes. (2) The diffusion)Dic manner. (3) The

ral proton donor andous hydrogen bondedton transfer along the

I

{

The NMR studies of water in bioloeical svstems

nacromolecular chain in a continuous and an extensive path having a well-defined

structure.From the temperature-dependent changes in the relaxation times of water protons,

the activation energy values of the order of 10.0 and 5.0kcal/mole were calculated for

the processes of proton transfer and diffusion, respectively (Berendsen and Migchelsen,

1966; Migchelsen and Berendsen, 1973; Mathur-De Vr6 et al., 1976). These processes

involve the breaking of two hydrogen bonds for the proton transfer and one hydrogen

bond for the diffusion process. In general, the hydrogen bonds are very sensitive to the

nature and orientation of groups interacting with water, and to small changes in the

temperature. Therefore, the relaxation processes that are influenced predominantly byproton transfer in the hydrogen bonded patb are also expected to exhibit a characteristic

dependence on the above mentioned factors. When macromolecules are present in a

medium containing 2H2O and tHrO, the exchangeable or labile protons undergo

isotopic substitution by exchange. This leads to incorporation of 2H and 3H throughout

the system in preferential sites on the macromolecular chain. Furthermore, there is

evidence suggesting that the isotopic distribution in the hydration layer is not random.At temperatures well below the freezing point of solvent, the hydration water molecuies

remain unfrozen or mobile on the NMR time scale. This phenomenon is distinctlydifferent from the physical processes such as freezing point depression and formationof eutectic mixtures. For the macromoiecular solutions and all biological samples, theNMR spectra of water observed between - 5 and - - 60'C arise from only a fractionof the total water content, i.e. the hydration "water. From the temperature-dependentchanges in these spectra, activation energy values for the relaxation processes havebeen calculated. In frozen samples, the extra-hydration layer water freezes to form arigid ice-like structure (as in free water) giving rise to a very broad signal. The areaunder the water proton signals obtained from frozen solutions was shown to varylinearly with the concentration of macromolecules (Kuntz et al., 1969; Mathur-De Vr6et al., 1976). This indicates that the water NMR signals observed in frozen samplesarise from water associated with macromolecules. Kuntz and Kauzmann (1974) definedhydration water as the unfrozen water. The poJential of this phenomenon for detailedstudies of the properties of water in biological systems was soon recognized. Thisis evident from tle numerous examples cited in Section IV.

Franks (1971) has described water-protein interactions in solutions as illustrateddiagrammatically in Fig. 1. The A shell includes solvent molecules which are in theneighbourhood of the protein side chains or the backbone. The motion of watermolecules in the primary hydration sphere (A shell) is determined by the motion ofprotein molecules or their Tocalized groups. The iegion C represents the water moleculesthat are unperturbed by macromolecules. Finally, "incompatibility of the hydrogen

Frc. 1- (Franks,1977)Diagrammatic representation of the protein environment in solution. Theprotein can be regarded as a hydrodynamic sphere with a primary hydration shell A in which

the molecular motions are largely determined by those of the polar protein sites. C is the

unperturbed water "structute" and region B arises from the spatial and orientational mismatchbetween recions A and C.

106 R. Marnun-Dn Vnr

Reg ion d

H-A(':l\ i77,

H\

x

\

Ftc. 2. (Packer, 1'977) A schematic iliustration of small-scale heterogeneity and various dynamicplocesses which may be experienced by water molecules in such a system. The shaded regionsrepresent macromolecular structures characterized by dimensions d, x, y, etc., orientations d,, 0petc. with respect to an external fixed axis, and correlation times for tumblin& c.. wateimolecules free of the influence of the macromoiecules diffuse and rotate and exchange proronswith characteristic times ra, x, a:fld 2,, respectively. water molecules interacting with themacromolecule tumble anisotropically, this process being represented by a collective correlationtime zf and have a lifetime in this state designated by c'.. water molecules may diffuse fromone region to another, their lifetime in a given region being rd-d2l2D") whilst they exchange

protons with macromolecules with a lifetime zl.

, r , , ::':;' ''.:,1,1,,,,::,.r i l

;i,

iraiii . ,

1''i''' i ,

i': i

bonding in regions A and C" gives rise to the region B. Franks indicated that waterin regions A and B markedly influences the NMR spectra and contributes to theunfrozen fraction. In other words, regions A and B represent the hydration layer, andregion C represents the extra-hydration water or the frozen fraction.

Berendsen (1975) defined the hydration of macromolecules in terms of speciflc andnonspecific hydration. The interactions between water and specific binding sites on themacromolecular chains result in specific hydration. Nonspecific hydration is the amountof water affected by the macromolecules in such a manner that it exhibits slightlylower rate of rotation than molecules in the bulk liquid state, and it also contributesto the unfrozen fraction of water. The regions of specific and nonspecific hydrationmay be compared with regions A and B in Fig. 1.

Packer (1977) has represented diagrammatically (see Fig. 2) different modes of motionof dynamically oriented water near a macromolecular surface. Distribution of correlationtimes for water protons arise because water molecules are subjected to largely diversedynamic processes as a result of their interactions with a variety of sites and grouplconstrained in different environments on the macromolecular chain. In Fig. 2, a and frepresent two components of a heterogeneous macromolecular surface, oriented byand 0 p with respect to an external reference axis. Packer proposed that water moleculesnear the macromolecular surface move in an anisotropic potential whose spatiproperties remain unchanged during the reorientation of a water molecule; as a resulithe anisotropic potential experienced by each water molecule is not averaged oduring its reorientation time. On the other hand, in liquid and bulk water the potentiexperienced by a molecule.at any instant is also anisotropic but its axis undergoesrapid change during the reorientation of water molecules, consequently the eflects dueanisotropic potential are averaged out

I I I . NMR AND ITS APPL ICATION TO WATER STUDI .ES . iSeveral authors have discussed in detail the theory and techniques oflNMR'spectro:

scopy @msley et a1., 1965; canington and Mclachlan, 1969; Ferrar and Becker, 1971Shaw, 1976). Therefore, in this section only the basic principles of the NMRwill be described briefly and concisely, with special emphasis on the relaxation processes;

1. Basic ConceptsA nucleus with spin 1 is characterued by a spin angular momentum and possesses

magnetic moment (p). In addition. a nucleus also has an electric quadrupole moment:iiir l+3 l l

if I > 1,12deuteron ,rnagneticspins betvmanner tlno net uafield ̂ F11(of energyprovidedhv: g*F*phenomenprocess ofrelaxationIfowever,relaxationspin and(ii) molecuultimate eprocessesHerein lietool to mo

The p.tcknown asmagnetic ran additioOnly the rwith increznucleus rerepresentsalso definefrom the fis removedto the equtwith the tirrelaxationH11o1at rig)motions prlattice molaveraging crates.

. In liquidnteractionrstates (e.9.dipolar inteconstitute irates and rquadrupolaand orderirshowed thaoccurring aponding tofrequency c.@z: yHrc".occurring v(Frey et al.,

rmlcjons

", ol'ater

tonsthe

.tionromlnge

;: to the.equilibrium value with a time constant 71, whereas .n"^;;;;;;;;:;;;#

- *}l1jf..tiT:::l.lyt.Tz- A third reraxation time ""rri""t-?,, known as spin-rattice

The NMR studies of water in biologicai systems rcj

i f I>1 '12 'For theNMRstud ieso f water , thenuc ie io f in te res tare : p ro ton ( I : l l 2 ) ,deuteron (1 : 1), tritium (I

,: 112), and oxygen ,rO e : 512).In the pr.r"o". of u furg.

rnagnetic field IIe, the nuclear e,nergy levels split into 2I * i states. The distribution ofspins between these energy levels attains thermal equilibrium with the lattice in such amanner that the lower energy level has a slight "r".r, of spins. at .quiflurlo*-ifr"*'i,no net transfer of spins.-However, in the presence of a ,."ond oscillating .uaiof..l,r.n"yfield Ht(w1(RF field) yh.ose mag:retic component is perpendicular to r{, the absorptionof energy occurs and induces spin transfer from the lower to the upper energy ievel,provided -the frequency

.(v) of oscillating field satisfles the resonance condition;hv : gn7xHo; where gry is_ the nuclear g factor and 811 is nuclear magneton. Thisphenomenon.gives rise to the nuclear resonance absorpiion signals. It is clear that theprocess of spin transfer would eventually lead to u ,tui. of saiuration if there were noteTaxation processes to re-establish the excess spin population in the lower level,However, spins in the higher energy level can relax to ihe lo*., level by means ofrelaxation processes that are induced by: (i) coupling interactions between the nuclearspin and the local fluctuating fields arising fto- thermal motion of the lattice.(ii) molecular motion, (iii) proton exchange processes and (iv) puru-ug;.t; ";;H;ultimate effects of the interactions (i)-(iii) on different

'nrr"l"u, magnetic relaxation

processes are a function of the time scale and the nature of fluctuations in motion..Heyin

lies.the basis of the application of the NMR relaxation times as a sensitivetool Io monltor dynamlc processes.

The process of energy exchange between the magnetic nuclei and the lattice isknown as spin-lattice relaxation, designated by the tiire constant 11.In addition, themagnetic nuclei also interact with each other, therefore each nuclear magnet experiencesan additional small local field ,Fi'o" produced by the neighbouring nirclear -ugo"t*Only the nearest neighbours exert an important influence 6""uor" flr* falls off rapidlywith increasing distance. The spread of the steady magnetic field experienced by eachnucleus results in dipolar broadening. The spin-spin- interaction time constant T2represents the lifetime (or phase memory time) of a nuclear spin state. Tr and T2 arcalso defined as longitudinal and kansverse relaxation times, respectively. This followsfrom the fact that when the RF field applied initiary to a syitem of nuclear spins

llliT'-ll*,1f.:T,::::il'.?1-"11e*d along Ho_(M" "o*fo"."if returns exponentialty

relaxation in rotating frame corresponds to _the decay of ^u,[kt ration aligned alongHso at right angles to Hs,13th9r than along 116. Due to fast roiational and translationalrnotions persisting in the liquid state, thJlocal fluctuating fielJs of the surroundinglattice molecules are averaged thereby reducing the T1 "."lu"utio1

rates, while thea^veragin'8 of the dipolar coupling interactions between ,r.r"Li decreases the T2relaxationrates.

at was t o.yer, and

:ific and:s on theamountslightly

ntriydratiori

rrelationr diversI groua anded byrolecu) spaa

igedpotenergoes.s due

.- In liquid samples very sharp water NMR signals are observed because magneticrnteractions are averaged to zero. Howevel, due to restricted motion in more rigidstates (e'g' frozen tuotpl"s, macromolecular-water interface) the fluctuating fields anddipolar interactions are not totally averaged,. their residual effects on the NMR spectra"" constitute a valuable source of informaion concerning the mechanisms of relaxationrates and molecular interactions. At low ternperatures, the averaging of dipolar andquadrupolar interactions is less effective because of reduced thermal motion, orientation,and orderins of water molecules. Ferrar and Becker (1971), and Frey et at. (1972)showed thaithe relaxation time ?1, T2 and T1o are sensitive to dynamic processesoccurring at different frequencies. For instance, T, is most sensitive to motions corres-fl::3::9,::.the

Larmor frequency @s: rH9, Tr, detects motions corresponding to RF;;:",1;1rr. l1::1^To *

exhibits sensitivity to.morions characterizei by frlquency' ,:^t-:1111".. ror' protons, Tr,Tro, and T2 were shown to be sensitive to motionioccurring-with frequencies of the order of 30MHz, 50kHz and 10H2, respectively(frey et al., 7972). Generalry, proton exchange processes contribute to T2 and r1o. High

I iiii'ri'r

' t ] , ' .

'iii.i i ;

:ll ;:,]],i;'ilij

. [ 'i r t i I

i , l:l::'

r , li i

'

t:: r ;

i ti

R. Matrun-DnVm

frequency processes affect both T1 and 72, but the low frequency processes (such aschemical exchange and slow diffusion) influence mainly the T2 relaxation, causing T2 hbe much shorter than T1 provided aox") l.

The line broadening is an approximate measure of spin-spin relaxation time; ?,may be obtained fiom the half-linewidth ( ll2) of the signals observed in the steadystate, LIf2:IlnTz. The relaxation times are measured accurately by the pulsetechniques; for a detailed discussion of the theory and experimental pulse techniquesone may refer to Ferrar and Becker (1971) and Shaw (1976). In this method a strong RFfield is applied for short durations, i.e. in pulses. The effects of pulses depend on theirmagnitude and duration. The principle of the pulse method is that by applying eitheran initial90'RF pulse M, is reduced to zero, or it is reversed by a 180'pulse. A secondpulse is used to tip M, into the xy plane where the exponential regrowth of magnetizationcan be detected, and the relaxation time constants ?1 and T2 measured. T2 is generallyobtained by the "spin echo" experiment (Carr-Purcell method), i.e. by applying a 90"pulse followed by a succession of 180' pulses. T1 can be measured by applying a seriesof pulse sequences of the type 180'-r-90" or by monitoring the amplitude of theinduction decay signal M, following a 90" pulse. The advent of the pulsed Fouriertransform technique has greatly facilitated precise measurements of the relaxation times,particularly of nuclei present in low quantities.

2. The Relaxation Rates

The relaxation rates of the water molecules are governed by two important factors:(i) the strength of local magnetic interactions between water nuclei; (ii) the molecularmotion and proton exchange rates.

(i) The important interactions between water nuclei are: nuclear magnetic dipole-dipole coupling (inter- and intramolecular), and nuclear quadrupole coupling fordeuterium and 17O.

Quadrupole interactions result because the electric moment Qinteracts with the neighbouring electric field. gradient. Within the hydration layer, theimagnetic interactions are partially averaged by specific processes depending on theinteractions of water with macromolecules such as proton transfer, dynamic orientation:and diffusion of water molecules through regions of different orientations. Whereas, infree or bulk water, motional averaging of magnetic interactions dominate the relaxationbehaviour.

(ii) The effects of molecular motion are generally incorporated into the theory ofnuclear magnetic relaxation in terms of the correlation time, (r").r" is considered as thetime taken by a molecule to translate through a molecular distance, or the averagetime between molecular collisions for a molecule in its actual state of motion. Forfast motion, i.e. when llu*> @o, Tt and T2 are equal. For slow and restricted motion,both Tr and T2 decrease and may not necessarily be identical as already mentionedin the previous section. The nuclear spin relaxation times are sensitive to molecularmotions of 10-8-10-12sec. Water molecules tumble in liquid solutions at a rate ofabout 10-12sec; this motion is considerably slowed down when water molecules.interact with biologifal macromolecules in solution, in muscle, or in cells, but still fallswithin the limits of NMR sensitivity. For example, the rotational motion of watermolecules associated by hydrogen bonds with polar groups on the macromoleculalchains are reduced so that their correlation time is of the order of 10-6sec insteadof -10-12sec fcr the remaining water (Fung, I977a). Under the conditions of rapidexchange between the hydration and bulk water in liquid solutions of macromolecules,the observed relaxation rcte (1,17),," is given by:

(1 /T tLo" :Xr ( l lT ) r+ Xh( I lT lh (1 , }

Even though the fraction of free water Xy is much greater than the fraction ofhydration water X6, the term Xh(IlT)h can still make an important contribution to.

(1/?r),o,, since (1/71)y, > $lTt)t because of the restricted motion of water molecules it

the hvdration laver. If the rates of exchanse of water between different reEions in ar

ii.rlj j

sa

ex,sig

weH(

wh

frerhargivtthe^2^w q

ancI t ccanverJvah

Bt o tmru

wheresofieldiengo f adiffuat t.

\ r lIn

smgJlayermokinfluexprbelo,samlmotinorn

ocesses (such as'n, causing T2 to

ration time; T2:d in the steadyy by the Pulsepulse techniqueshod a strong RFdepend on their' applying either'pulse. A secondof magnetizationl. ?2 is generallYy applying a 90"applying a seriesmplitude of ther pulsed Fourierrelaxation times,

rportant factors:ii) the molecular

nagnetic diPole-rle coupling forctric moment Plration laYer, the:pending on thenmic orientationons. Whereas, inLte the relaxation

to the theorY of;onsidered as ther, or the averager of motion. For:estricted motion,.ready mentioned;ive to molecularons at a rate ofwater molecules

ells, but still fallsmotion of watermacromolecular10- 6 sec instead

nditions of raPidI macromolecules,

(1 )

t the fraction of

Lt contribution to

rater molecules in

:rent regions in a

The NMR studies of water in bioloeical systems

sample is fast compared with the regional relaxation rates, then an average waterresonance sigral is observed (fast-exchange condition); on the other hand, if theexchange rate is slow compared with relaxation rates in each site, then a resonancesignal cttaracteristic of water in each site may be observed (siow-exchange condition).

Assuming the motions involved in intramolecular relaxation, the following expressionsv/ere repoftedfor lfTy llT2 and lfTloin terms of the correlation time z (Finch andHomer, 1974; Knispel et al.,1974;Belton et al., 1973; Fung and McGaughy,1974).

where cr;6 is the Larmor angular frequency in the constant magnetic field .I{o and isrelated to the resonance frequency v6 by the relation aq :2nrs, crtl is the Larmorfrequency in the rotating frame in the presence of a RF field (H1), and C is a constanthaving the form k'y4hl(I + l)111,6 for protons. A similar expression for deuteron 1/?r wasgiven where the constant C is equal to k"n2(1. + 7llZ)1e2qglh)2 (Fung et al.,I975a); y isthe gyromagnetic ratio, h:PTanck constant hf2n,rj is the internuclear distance,etqQlh is the nuclear quadrupolar coupling constant, 4 is the asymmetry parameter,and finally k' and k" are the numerical values. Both are nd cr.r1 are variable parameters.It can be seen from eqns (2)-(4) that when @sx ) 1, information about r (and motion)can be obtained by studying the c.rs dependence of relaxation times. In pure water, r isvery small (<10-tt sec) and @6? ( 1 even at very high rrs values, therefore T1 and T2values are independent of the coo values.

Berendsen has shown that in the hydration layer of biopolymers, T2 can be relatedto the diffrision of water molecules and the exchange lifetime of protons in the followingmanner (Berendsen, 1975; Mathur-De Vr6 et al., 1976):

7lT2: (y2120)( H)2r, , (5)

where y is the proton gyromagnetic ratio; A.FI is the maximum splitting of the protonresonance in the event that the biopolymer is completely aiigned with the magneticfield, and x,:(6Dla2) * r*"lr,i D is the diffusion constant; c is the measure for thelength over which the macromolecules are oriented; and r"*"6 is the exchange lifetimeof a proton on the water molecule. It was shown that in frozen DNA solutions thediffusion process determines the behaviour of T2 or (L112) at low temperatures, whileat higher temperatures the exchange rate dominates the relaxation process: where( I l2 ) :1 lnTz .

In free or bulk water. the rotational and translational motions of water molecuiesare strongly coupled, as a result the motion of water molecules can be defined by asingle correlation time (Eisenberg and Kauzmann, 1969). Whereas, in the hydrationlayer the translational and rotational motions are no longer appreciably coupled due tomolecular interactions, restricted motion and ordering of water molecules. Under theinfluence of the decoupling effect, the motion of the entire mass of water cannot beexpressed by a single correlation time but requires a distribution function. At temperaturesbelow the freezing point of the solvent, the molecules of unfrozen water in biologicalsamples have been treated as spherical molecules undergoing translational and rotationalmotion, governed by a distribution of correlation times (Fung and McGaughy, 1974).Thenormalized log-Gaussian distribution function is given by the relation:

/ ) " R r \L l T r : C l , , L r

" - |1r - cofrr2 1. + 4ulr2 )

/ \ - ) , \1 l T 2 : c l 3 t + - - - - + . : - l

\ 1 l a f i r " l - 1 4 a f i " )/ ^

^ / 3 1 5 t 2 r \, I I Lo: L i -

-T :

- - . - n r ; - - - . - - - ; - '

\ r - f +c { r1 r - - 1 * r7 - g4 r6V1 '

(2)

r1)

(4)

, .p.8.35/2_D,

d . -g(t) : -+ x exp [-(a In t I tg)21.

,/ ftr(6)

R. Mernun-De Vns

The parameters 16 and a are temperature-dependent, rq is the median of distributi6nand a is a parameter that determines the width of the distribution. The distribution sicorrelation times leads to different frequency dependence of 71, T1o and T, frornthat given by eqns (2)-(4).

At temperatures above the freezing point of water, the following distribution functionwas suggested:

s'(r) : X x s(t1)+ (1 - X)a(r),

within the system of spins than between the surrounding lattice and the spin sys.tem ,,(Berendsen, 1975).

3. Dffision ConstantThe self-diffusion coefficient is a measure of the interchange of identical molecules by ,

way of thermal movements. The difftrsion constant (D) of water can be evaluatedfrom the spin-echo decay of the NMR signals by applying a known magnetic fieldgradient. Several authors have discussed in detail the NMR methods of ialculatinsthe self-diffirsion constant of water in biological systems (Andrasko, 1976; packer,1973; ,chang et al., 1972; Hazlewood et aI., Di[b; cleveland et al., 1976). Theoretically, thecontribution of diffusion to the spin-echo decay is given by the expression: ,

A(r, G) : exp | - 2 I 3y2 G2 D "t31,

where ,4(t, G) is the echo amplitude for a 90"-180' pulse separation r in the presence ,of an applied field gradient G, y is the gyromagnetic ratio (cleveland et at-, 1976).

The interest in the theory and measurements of the diffusion constant of water in .muscle and cel1 water has centred mainly on the following two objectives: (i) to study 1the extent of-ordering of water molecules by comparing the self-diffusion constant of ,water (H2O,'HzO) in various biological samples and in free water; (ii) to investigateiany possible effects of diflusion of water molecules through the internal field gradients(generated by the heterogeneity of the magnetic susceptibility in biological samples) .o_n the relaxation processes of water nuclei (packer, 1973; Hazlewood et al., lgTl;Chang et al-, 1972). Diffirsion of water in the hydration layer is an activation process in ,which a mcilecule must attain sufficient energy to cross over a potential birrier; the 'values of (D) depend on molecular interactions in the system under consideration. ,i.l

4. A Comparison of Dffirent Water Nucleicertain characteristic nuclear properties of proton, deuteron, tritium and, 17 o afi::

given in Table 1 (Emsley et al., 7965, p. 589). The relaxation of protons is influenced'by inter- and intramolecular dipolar interactions; whereas, t-he ielaxation behavioul ,of deuterium and 17O is dominated by the quadrupolar interactions. Glasel (1967) hast::reported the following equations for relaxation rates of 1H,2H and 17O. 1

a)where X : fraction of molecules at any instant having a log-Gaussian distribution ofthe correlation times. Equation (7) signifies that above the freezing temperatures, eachwater molecule spends part of its time behaving as isotropic water with a single correlationtimp.

The correlation between the motion of protons from the water molecules and frommacromolecules gives rise to the process of cross-relaxation or spin diffusion. Theobserved relaxation rates of water protons can be influenced significantly by thecross-relaxation rates. spin difrrsion occurs by way of mutual exchar-rge of spinmagnetization between water protons of the. hydration layer and protons oo ih,macromolecular chain. The contribution of cross-relaxation becomes very importantunder the conditions of slow molecular diffusion and when Tz4Tt lsuch asil;;;;biological systems). Under these conditions, the energy exchanges much more rapidly

(e)

of distributiondistribution of

The NMR studies of water in biological systems

T$,l.ern 1. Nuculn pnopeRrrrs or 1H,2H, 3H, 17O (Eusr,nv et aI., 1965)median

ion. TheTr Trp and T2 frorn

ng distribution function

NMRfrequencY

in Mc/sec for -Natural

a 10kG abundanceIsotoPe field (%)

Relative sensitivity forequal number ofnuclei

Atconstant Atconstantfield frequency

Magneticmomentp,

in multiples of the Spin I, innuciearmagneton multiples

(ehl4nMc) of hl2n

Electricquadrupoie

moment Q, inmultiples of

e x 10-24 cm2

iaussian distributionzrng temperatures,

1H-rL3H

170

42.s7'7 99.98446.536 7.56 x l0-2

45.4145.772 3.7 x l0-2

1.0009.64 x 10-3

1.212.91. x I0-2

1121 2.77 x 1,0-3L12512 -4 x 10-3

1.0000.4091.071.58

2.79270.857382.9788

- 1.8930

with a single correiati

ler molecules and fior spin diffusion. T,d significantly by.ual exchar-rge of' and protons oncomes very( ?r (such as in man;es much more rapide and the spin

identical molecules brter can be evalknown magneticrethods of calculatio, I976; Packer, 197,)76). Theoretically, tression:

tton r in thed et al.,1976).constant of water

rbjectives: (i) to stu:-diffusion constantter; (ii) to investigarternal field eradienn biological samzlewood et al.. I97activation procesi

potential barrier;consideration. ,i

tritium and 17O

protons is influerelaxation behavins. Glasel (1967)t70 .

? " ,4h2(I lT)o;or,^: ' j f t ,

(llT)o,or", : "yf:

Protons:

Deuterons:

1 7 0 :

? /o2^n\2(1 l r ' )a : i l # | , "o \ t t

/ 2 H

*(+):"'",

(10)

(1 1)

(r2)(7171)o :

wherez"iscorrelationtime;a: radiusofamolecule;D: microscopediffusioncoefficient;and N : number density of molecules

- In walel,.deuterons ielax about lO-times faster than protons, and 17O nuclei relaxabout 100-times faster than deuterons. Such fast relaxation rates of 17O rutirfv tLslow-exchange condition in the studies of cell rut.r. wh;;;;;, ii g"rr"rut the relaxationbehaviour of protons and deuterons is governed by the fast-excf,ange conditi;;. i;;spinrlattice relaxation time T1 for deuterons in iHro is governJd "";;i"i;ly byrotational time of the individual 2Hro molecules; u. u ,"rirlt the 2H NtilR ;"ril;detects the anisotropic motion of water. The dipolar interactions of water protons aresensitive to rotational and translational motion, proton exchange, and the presence ofparamagnetic centres in a given sample. The quadrupolar intJractions are higher inmagnitudethanthemagnetic dipolar interactions; therefore, for small changes in theiuclearenvironment the deuteron relaxation times exhibit much greater sensitivity than the protonrelaxation times. However, the "NMR" sensitivity for thJ detection of deuteror, ,"rooun."signal is much lower than for the proton signal (see Table 1). The NMR studies oftritiated water in biological systems should be favoured by high sensitivity to detectlrltium

resonance _(Table 1); eventually such measurements face a great drawbackoecause low quantities of 3H are required to be present in samples due to its radio-

1Lt_t:lt;,th. NMR spectroscopy of tritium has been developed and applied successfullyour{t8 the past few years mainly by the research group of Frofessor iirag" @loxsidglet a1.,1971; Al-Rawi et a1.,1974;Al-Rawi et al.,l9i5).

'

IV . THE NMR STUDIES oF wATER IN DIFFERENT BIoLoGICAL SYSTEMS

^,.{.Hflt variety of biological systems whose hydration properties have been investigateduv nvlK spectroscopy may be classified broadly into two groups on the basis oi the NIrAnspectra of water.

iliiriij:

r

R. Mernur-Dr Vnr

Randomly oriented systems such as macromolecular solutions, cell suspensions,muscle, tissues and membranes give rise to water proton and deuteron resonancesignals consisting of a single peak even at low temperatures in frozen samples.oriented samples: the water spectra from oriented fibres of DNA ani collagenappear either as a single broad resonance line or a signal split into fwo rii.r.The characteristic feature of these spectra, which distinguishes them from tt e groui ,(1) spectra, is that the linewidth or the splitting of proion and deurcron resonance ;

solutions by measuring the relaxation times, using the spin-echo technique. They also rinterpreted the results in terms of firmly bound hydiation ,h"ll, lnoo-.otational binding) ,:and rotationally bound water (freedom of movement as in free water). Lower io.rlc strengihwas found to favour an increase in the hydration of DNA (Lubas and wilczok , r97a).

1 .

L .

signals are a function of the orientation of ,u-pl". with respect to the magnetic field,

l. Randomly Oriented Systems(a) Biopolymer solutions and hydrated proteins

The initial efforts to study water by NMR had shown that the water signal frory1DNA solutions was much broader than the signal from pure water. In order Io explainthese results, Jacobsen et at. (1954) proposed that line-broadening was due to;rrcreaseJordering in the water structure and the formation of hydrati'on shells near DNA .This early concept of hydration water in terms of a static model describing shells oiwater molecules near macromolecules proved inadequate. The current theories and theproposed modeis for water in macromolecular solutions or in different biological systems rconsider hydration water in terms of dynamic processes. Several examples of dynamic ,models of water in hyd_rated biological systems are discussed in different ,""tLrs oi,this article. Lubas and wilczok (1966, tiel, olts studied the hydration of DNA in,,

Further studies revealed that the relaxation behaviour of water was characterizeO Uy,,l*^":::?r,-:Ji"i ,1d" strucrure of different biopolvmers in sorution. Giasel osiol'.'"i-l1"r"eiil'#;lT:919: 9:.l.Ott"g.

the deuteron.magnetic relaxarion-of water at 31jC. r'n" ioffo*t"g :polymers were studied: poly(r-glutamic acid), poly(r-lysine), poly(adenylic acidipoly(uridylic acid), poly(metlaciyric acid), porlvinyrox atoridinone methyr), andpoly(vinylpyrrolidone). He gave the following eqtation for the observed ;;hr";;;rate under the fast-exchange condition:

(1/Tt),u, : {l I T)r,* * CaKto, (1 3)where: (1/TrLo, is the measured relaxation rate for solutions; (rlTr)r*"is the relaxation,rate for pure solvent; C is the concentration of the polymer j , ir tii" ti-r-i"a"p""Jr"tweight of water associated per gram of polymer; r ir tn" quadrupole coupling term;,and ro is the rotational reorientation time of water molecules u..o"iutii witrr iil. p-"rvrr.r.Straight line plots of (1/71).6" - (IlTr)n*vs C were observed. ,:

In this work, the importance of polymer-water interactions by hydrogen bonding waspointed out; for example, poly(U) did not show any interaction, wherlas strong Inter-action was recorded for^ poly(A) and poiy(r,-glutamic acid). It was observed that"s;ngwater interactions were favoured by the stable topology of polymers.

The magnetic field dependence of the relaxation rates oi p-tonr, deuterons, and 17onuclei of water (termed as relaxation dispersion) was investigated in detail for a variety lo-f proteins in liquid solutions (Hallenga and Koenig 1976;{oenig and Schillinge r, l96i;Koenig et a1.,1975: Koenig et a1.,197g; Grcisch and Noack, 1972). Koenig et-at.'(1975):1,"T9

that for.lyso zyme and, haemocyanin solutions, the relaxation dispJrsion roi tg, ,-rr and ''u nuclel of water were essentially the same. In general, these authors observed.that the plots of 1lT1 -vs,Fr6 for tF',,H and 17o showln inflexion at a value of Ho :that corresponds to the Larmor frequency v" given by the ,"tutiorr-u": firilia",where ta is the rotational relaxation tims of the protein molecule. It was" pioporrithat the correlation time for orientation of water molecules in the neighbouihood of

protem ma

rnacromolec. to the srdisPersionsolutions itcross-relax:to the Prolrate of croibehaviourtitrable grccross-relaxiprotein-so.proteins isspin-magntrelaxation ,

Finally, tproton reliwater in Pr

The orignucleic acihydrationmeasure ofthe relaxallinewidth <immediatelthe hydralprocedure,demonstrasensitive t<coils at rothelical conestimatedsignal coulYft et al.poly(A), pcexert a ma( 112)_s..polynucleoobservatiorpolynucleoThe increarof complexin the liquthat the diimportant

Thewidtin the hy<- 180"C (Iof water rmoment oribosomalsteadily asfragmentswater sorlThey expli

The NMR studies of water in biological systems

protein macromolecules was determined by the Brownian motion (tumbling rate) ofmacromolecules. A hydrodynamic mechanism describing a small transfer of the protein,1, to the solvent molecules was considered to be the common cause for the observeddispersion in T1 for all three nuclei. By studying the dispersion behaviour of proteinsolutions in partially deuterated water, Koenig et aI. (1.978) were led to conclude thatcross-relaxation between solvent and solute protons makes an important contributionto the proton relaxation rates, but not to the deuteron relaxation rates of water. Therate of cross-relaxation between protein and solvent protons shows a similar dispersionbehaviour as the T1 relaxation rates. Proton transfer between water protons and thetitrable groups on the protein surface was suggested as a possible mechanism for the

oss-relaxation, and the cross-relaxation term was shown to be proportional to theprotein-solvent interface. A detailed description of the cross:relaxation effects withinproteins is given by Kalk and Berendsen (1976), Sykes er al. (1978). The exchange ofspin-magnetization was considered to occur at a rate faster than the rate of spin-latticerelaxation of protons.

Finally, Grcjsch and Noack (I976) interpreted the frequency-dependent changes of theproton relaxation rates (71, T) of BSA solutions in terms of a three-state model forwater in protein solutions.

The original work of Kuntz et al. (1969) showed that when solutions of proteins ornucleic acids were frozen, a relatively sharp and distinct signal was observed fromhydration water at temperatures as low as - 35"C. The area under this signal gave ameasure of the unfrozen water. The activation energy values of the processes influencingthe relaxation tates were calculated from the temperature-dependent changes in thelinewidth of water signal. These interesting observations reported by Kuntz et al. wereimmediately elaborated and applied by several groups of workers for studying extensivelythe hydration water characteristics in different biological systems. Using the sameprocedure, Kuntz (197ra,b) investigated the hydration of several polypeptides. Hedemonstrated that the linewidth of water signals observed in frozen solutions was verysensitive to the polypeptide conformation. All those systems known to be in randomcoils at room temperature exhibited sharper lines than the corresponding systems in thehelical conformafion. Kuntz indicated that the hydration of globular proteins could beestimated from the hydration of appropriate polypeptides, but the linewidth of watersignal could not be calculated from the amino acid composition. The results of Mathur-DeYft et aI. (1975) have shown clearly that the nature and structure of polynucleotides:poly(A), poly(U), poly(C) and their complexes: poly(A * U), poly(A + 2U) and poly(AH+)exert a marked influence on the linewidth of water proton signals in frozen solutions( Il2)-s'. The formation of double-stranded helical complexes from single-strandedpolynucleotides is accompanied by a large increase in the (Lrl2)-s" values. Thisobservation was explained by considering that proton transfer in the hydration layer ofpolynucleotides decreases due to the formation of inter-chain links in the complexes.The increase in the rigidity of the macromolecular structure accompanying the formationof complexes is expected to influence the water spectra when measuiements are performedin the liquid state. The NMR studies of frozen macromolecular solutions show clearlythat the diffusion motion, proton trarisfer, and macromolecular-water interactions areimportant factors influencing the relaxation behaviour of hydration warer prorons.. The wide-line proton magnetic resonance spectra of ribonuclease and BSA were studiedm the hydrated and vacuum dried states over the temperature range of - 140 to*^180"C (Blears and Danyluk, 1968). Considerabie translational and rotational motiohoI water was shown to persist at such low temperatures by comparing the secondmoment of ice with that of water in proteins. The hydration of ribonuclease and totalribosomal RNA, as studied from the water spectr a at -35oc, was found to increasesteadily as a 70S particle was successively broken into smaller and more expandedIragments (white et al., 1972). Fuller and Brey (1968) reported the NMR spectra ofwater sorbed on serum albumin as a function of temperature and water content.They explained that sorbed water could exist in different siates depending on the water

1 i 3

>lutions, cell suspensior and deuteron:s in frozen samoles.es of DNA andgnal split into two lishes them from the eror and deuteron resonarpect to the magnetic

at the water signalrater. In order toing was due toation shells near DNodel describing shellscurrent theories and

fferent biological systeral examples of dI in different sectionse hydration of DNAho technique. They(non-rotational binditer). Lower ionic stren; and Wilczok, D7q.ter was characterizedsolution. Glasel (1

:veral biolosicalr at 31'C. The followir), poly(adenylic acidlidinone methyl), a-he observed relaxati

lTt)t,". is the relaxatiis the time-indedrupole coupling termciated with the ool

'hydrogen bondingr, whereas strong inas observed that stron3rs.rns, deuterons. and 17

I in detail for a varieLg and Schillingel Lg6g'6). Koenig et aI. (197Ltion dispersion for 1

these authors obse:xion at a value of Illation v": r[l11Zm^)rcule. ft was propothe neighbourhood

114 R. M.lruun-Dn Vnr

content: water up to T5mglgprotein represented_water strongly bonded to polar groupsof the protein, the "primary water". The NMR signar fro-m the pnmary laler wnqstrongly influenced by the protein-water interactiins, further addition resulted t^strongly hydrogen bonded water to the.primary layer which ir -.;;;ilil:il ;1'

lf;1:i,"jl;*::"i:::11,i,^.T?1":"_f; *i,r, 1ffi,"f *o B in Fig 1 rrr,,amount of water strongry bound to a sorid protein is-ress rh;^;; ilr'o?"i,"'T

:l1;*:*:,,yT^*1,:111,.."1u11,io",.:y*, .or tn" piotJ erastin (rigamentunnuchae of mature beef) soaked in excess 'Hro-rh;wJ ;;;;" ffi;";J'?"i1,'ili,XT 'iJLr;1"1r,_Hl,t;:"T:.:"1,-,1"*,li:1" theieby ,"ee",,*J-,hi' "",,t",,". of distinct;regions of water in the sample. These are: water contai"io *ittin ;;;":,'jrlol.

l:"Ii::i#f::l111i,9: Tll ,l::lt:rhe hrdratio*;i;i;.* is particurarry importanriin rendering it rubberlike properties (Ellis and packer, 1976). H'i;ffi ilil?(r#istudied the relaxation times.of hydrated lysozyme powder as a function of temperaturjand water content. A model based on cross_relaxation was found trthe behaviour of proton relaxation rates. tvr'd'\'auon was lound to account satisfactorily

(b) Effects of y-irradiation on hydration waterWhen biologicai systems are subjected to 7-irradiation, the induced radiation damaseis known to localize o1tl" DNA moiecule (Brok and Loman, lg.z; r^rii;";"I ;:;a full understanding of the radiation effects.on DNA in an aqueous medium bnd the

::l:.::,::l::_:,^fl:]:::lT l"l"r .in mediatins th. ;";;dr-iJiug., it is essentiar tostudy the effects of irradiation on the characteiistics of hydrati"#;; ;;:;,[A;:,molecular-water interactions, on water mobility and proton trunu.. do;g fi;I;;.ffi;lir;l :j"::',y.:::ylqi,b".:lh:_:tr:o19i y_rr,uaiu,io,, on iie r,yoration water of ,

I l ,, :r

l i lii,,:i r . :

:: ]l ,i: It l .i

i

lri l

i :i:.

DNA and poiynucleotides in H2O, ,HrO, X%Hroly%2Hz1:s;;e;i;ffi;Hl;by Mathur-De vr6 et ar' (1976), Maihur-De vr6 ani Bertinchamps (r977a,b). Thesolutions were irradiated at 0, - g0, and - rg6"c, and the linewidth of the waterproton NMR signals was measured from -5 to -45.c. It is known rtu, *t.r, uqr;;;;solutions are subjected to 7-irradiation, free radicals are generated that are stable at lowtemperatures but decay rapidly at about 0"c. The soiutions irradiated at _ g0 and- 196"c were thawed at 0 to + 5'c and refrozen before the NMR measurements.This process curtailed the possible line-broadening effects of paramagnetic free radicalstrapped in the frozen irradiared sorutions, on the rinewidth "f *;;;;"il;i,

";;r#

lrom the irradiated solutions below _ 5.C.

.^ll_,:::^r.n:yl,Or j:Tpu.'n.s tinewidths at _5oC, that the irradiation of DNAsolutions at 0 and - 80'c resurted in a decrease in rinewidth "f *;;;;;;;; ;r-#;compared with that of the corresponding non-irradiated solution; furthermorq theobserved decrease was greater when irradiaiion was performed at - g0.c. No significantchange in linewidth was recorded after irradiating the solutions i, - ltla"c,by irradiatingthe dry solid (at - 195'c) before dissolution, or by sonication oi the DNA solutions. Aninteresting observation was that irradiation at 0.c of pory(A t.u)""* ;;i;A;;;jcomplexes resulted in a large.broadening whereas -o"h'.h'urier signals were observed

lffiifi il:#:,ffi :.xll?.T*.;ils:i,".':::4,-*:i*'l'ruii''i*g--;urgfly moDlle' and the segments of macromolecular chains possess considerable flexibility.Il*::-,:::tles"at

-80o9, jh" hvdration water remains unfrozen and mobile butlthe bulk water is frozen a',d the segmental moblity of the -"";;;;;;"1'""ili"r'lr,rrestricted' consequently, when the soiutions of macromolecules ;;" ;;*"0 "ri11li.""i.temperatures, the process of either radiation.induced cross-r*nru^", l"p"#;;';f ,h;'

::if,f::::::-*-t':yj1'l-:l r"'": p'otg,'. signars t' l;;;;;,, o,';";;;|i,,',dominates depending on the dynamic states of the hydratio" und;;;;#'fjiJ"#:;and on the flexibility of the segments of the macromolecotu, "trulor, favoured underthe conditions of irradiation. It was suggested that the striting liff"r"o"", observed inthe hydration water proton spectra after y-irradiutiott r"r" Gtg"ry due to important*:^t^.: ]: :f T-]-r:n

transfer atongthe hydration layer resultirrg ?rorn the modifications:iilii'.!

induced in the structure of macromilecules.

P r r l

Ice

MuBra

Frofr.

Frot,

PlasRed

Ranol

CellCell

thM

PartmlH,

NorrcellsSicklcells

Livel

Lung

Mel

Brear

Thtmethrmediepolyn

(c) M

Corin mumolecwell-dall tht

HealrC3mi,

MicelariM(Tu(8-

The NMR studies of water in biological systems

Test-n 2. RsrnxA.rroN Trurs ron a rrw Snrucrro Svsrnus

1 1 5

ed to polar groupsrrimary layer wasdition resulted inless influenced by

Temp Frequency('C) (MHz) Nucleus

Relaxation times(msec,

Tt Tto TzSamPle Rel'erence

B in Fig. 1. Thehat for protein inastin (ligamentueuteron relaxationistence of distir the bulk elastinticularly importanand Bryant (1on of tempera:ount satisfactorillr

radiation,atarjet, 1972). F; medium and tit is essential

ler, e.g. on macrorlong the hydratiLydration water ots was unclertaps (1977a,b).dth of the wateihat when aqt are stable at Iated at -80

[R measuremenpetic free radi

rdiation ofr proton signalfurthermore, t

''C. No significa5"C,by irradiati)NA solutions.rnd poly(A + 2als were orter is liquid

Pure water

MuscleBrain

Frog skeletalmuscle

Frog gastrocnemiusmuscle

PlasmaRed cell inertia

Membrane bound

water Phase

Random PoPulationof HeLa cells

Cell pelletsCeil susPension in

the presence ofMn2*

Partially hydratedmuscle fibres 1 gH2O/g protein

Normal Oxycells DeoxySickie Oxyce1ls Deoxy

22; l

- f l

17r )

- n

- f l

1 7 o

-irl

1 7 o

1 E

1H

170

' H

1 H

23002830

470

2300 Cope (1969)2830 James and Gillen (1972)450 Cope (1969)

6.5 Swift and Barr (1973)6(psec) Fosteret al.(1'976)

ozi Cope(1969)

1 ) )

ffi Swift and Barr (1973)

44 Finch and Homer (\974)

Fabry and Eisenstadt (1975)

Finch and Schneider (1975)

Beal1 et al. (1976)

Shporer and Civan (1975)

Cooke and Wien(1971)

Zipp et al.(1,976)

Damadian et aI.(1974)

Frey et al. (1972)

25

25

3010

3.7

921 3 1

700 180

3.91.725

z) . ) 91.9 22.2

667

1.0-1.10

1.0-1.05

200

'25

25

22

8.13

A a

s60550530550

44.4

88845626

Liver

Lung

Breast

570832788

1 1 1 0367

1080

NormalTumotousNormal )6Tumorous

-"

NormalTumorous

1 H

Healthvc3ll

Llver

mice Lung

Mice with , .large Llver

yC-1 LunsI UmOUr(8-37 cm'�)

Tumour

1 H

386 47641 11,7

561 1 4

2547

J J

49

47

461

665

853 208

derable flexibilitand mobile

clecular chainsdiated at diseparation of tor decrease)water mo

;, favoured:nces observedCue to im

The NMR studies of water in frozen samples proved to be a very informative

method in revealing that hydration water molecuies make a distinct contribution in

mediating the overall radiation-induced damage to the structure - of DNA and

Polynucleotides in solution

(c) Muscles and tissues

Considerable evidence has accumulated leading to the general conclusion that waterin muscles and tissues exists in more than one fraction, and that fast exchange of watermolecules occurs between different regions. One of the fractions has been assigned to awell-defined and ordered water phasJ. In addition, it was affirmecl that the mc'tion ofall the water molecules cannot be described satisfactorily by a single correlation time.

the modificatto

I 1 6 R. Meruun-Dr Vnr

In general, the relaxation times and the diflusion constant for muscleto be reduced with respect to the vaiues for pure water. The valuessystems are included inTable2.

Bratton et al. (1965) postulated a two-state model for water in muscle. Hazlewos6et al' (1969) concluded that at least two ordered phases of water (major and minortexist in muscle; water molecules in the major phas- were considered to exhibit d;Jmotional freedom than in the minor phase and exchange rapidly with free rvat"r.

-Coot,

and wien (1971) fitted their data of T1 and ?2 measurements 1::"c; on live rrro*i,fibres with a modei in which 4-5\ of the total water was associated wittr prot.insand represented the fraction with fast relaxation rate, while the bulk of the waterinside a muscle was considered to be free. It was suggested that fast exchange "f ;;;;protons could occur between these two phases. Three-state models for watei i" *ur.f,were proposed by Hazlewood er al. (1974b) and by Belton et al. (1972).

'

Deuteron magnetic resonance studies of Cope (1969) also indicaied the existence 61two distinct fractions of tissue water in muscle and brain of adult rats. He rugg.r,jthat each fraction may be composed of multiple subfractions. The 17o sf;; ;ifi;;i;in frog skeletal muscle has further provided evidence in favour of the ristricteO motionin muscle water as compared with pure water (swift and Barr, 1973). ln b"rh ;h;;studies, the reported T2 and T, values were found to be much smaller than the valueifor pure water (71 : T), see Table 2.

In an attempt to describe the relaxation behaviour of water in muscles, tissues andcells, most authors explained the observed shortening of relaxation times by "onriO.ringthe existence of an ordered phase of water. In this phase, the motionai freedom olindividual water molecules is restricted by their interactions with cellular macro-molecules reducing the relaxation time (Hazlewood et a1., 1969; Hazlewood et aI., 1,.971;Hazlewood et aI., !914b). Hansen and Lawson (1970) and Hansen (1971) pointed out thaithe line-broadening was induced, at least partially, by diffusion of water moleculesthrough microscopic magnetic field furhomogeneiiies present in the heterogeneoussamples. cooke and wien (1971) measured Tr and, T2 for solutions of F-aciin andG-actin. These authors reported a decrease in the T2 uul.,.. when actin solutions werepolymerized, and they emphasized that diffusion through increased magnetic field,heterogeneity contributes significantly to the relaxation behaviour of water. Cbang et al.'

(Edzes and Samulskl, l97B).

(1972)have contested the discussion of Hansen and Lawson; Chang et at. (1972) arguedthat unusually large values of the local field inhomogeneity must be assumed in order ,that the proposed mechanism be effective. On the tasis of the detailed calculations,Packer (1973) concluded that the effects of diffusion of water through inhomogeneousinternal field gradients in striated muscle were negligibie. He point"d out aialogiesbetween the effects of restricted diffusion, the motionit nu.rolviig;i;";;;;'fi;r,and diffusion through periodically heterogeneous and structured Jystems. yet anothei,mechanism describing the water relaxation rates was proposed based on the effects ofcross-relaxation between the water protons and macrornolecular protons of muscle

water were fouq4for a few selectq

Ratkovi6 and Sinadin ovi6 (1977) investigated the relaxation times for water protonsin, tissues of the thyroid glands of rats. they obtained evidence indicating drat therelaxation times decrease (as compared with free water) by long-range inter-actions ofwater with macromolecules and the effects of compartmentalizatiln, rather than due todiffusion of water through the microscopic magnetic field inhomogeneity inside thesample' It may be important to differentiate between the physical compartmen talizalronof water in macroscopic regions of complex biological rropt"r and the distinguishablefractions of water structured at the molecular level. Both these phenomena may producesimilar effects on the relaxation rates of water nuclei.

Hazlewood et al- (1974b) demonstrated, by decomposing the spin-echo decay curvesat24C, that at least three distinguishable fractions of walr-protons were required tofit the data for skeletal muscle: water associated with macromolecules representsapproximately 8% of the total tissue water and it does not exchange rapidiy with

r1.1The NMR studies of water in biological systems

the remaining intracelluiar water (T2 less than 5 msec), myoplasm fraction 821

17. :45msec), .*,ru""iloiu, space -io% Q': 196msec)' T2 ̂ of pure water or Ringer's

l^i,.;^,.l - 'l .6 sec. These authors showed that water in different fractions did not exchange5 U I u " ' " - -

rapidly and consloer"J th" possibility that each fraction may be composed of fast-

exchanging sub-Iractrons. Hazlewoo d et al- (lg'74b), and Belton et al' (19-12) attributed

the fastest relaxing liaction to the closely to,,ttd-water: three fractions detected by

B e l t o n e t a l . c o r r e s p o n d t o ? 2 v a l u e s o f 2 5 0 , 4 0 a n d g m s e g r e s p e c t i v e l y . T h i sinterpretation has since been contested by Foster et aI' (1976) and Fung (1977b); both

these groups of workers have postulated that the fastest relaxing fraction (4-9 msec) of

Drotons arises from ttre nonrigla protons of macromolecules rather than from "bound"

vater were foundrr a few selected. ,

rscle. Hazlewoodrajor and minor)to exhibit greateriree water. Cooke)) on live muscleed with Proteinsrlk of the water:xchange of waterr water in muscle

I the existence ofats. He suggestedspectra of Hz17Orestricted motionI3). In both these:r than the values

nscles, tissues andres by consideringtional freedom ofr cellular macro-rwood el al., 1971';t) pointed out thatI water molecules:he heterogeneousns of F-actin and:tin solutions wereed magnetic fieldwater. Chang et al.zt al. (1972) arguedassumed in order

tailed calculations,gh inhomogeneousrted out analogiesof resonance lines,ltems. Yet anotherd on the effects ofprotons of muscle

for water Protonsindicating that thenge interactions ofrather than due torgeneity inside thempartmentalizationthe distinguishable|mena may Produce

L-echo decaY curvesrs were required toLolecules representshange raPidlY with

t o 3

Frc. 3. (KnisPel er al.,

-fit; relaxation studies perforrned at a.single frequency furnished ample evidence to

show that water in *orrL, and tissues ls n"ot homogen"o"s but exists in more than

one fraction. Further ,r.* and revealing details about the dynamic states of water, e'g'

distribution of correlation times and dispersion of proton relaxation rates, were brought

into evidence by careful measurements of the refuxation times over a wide range of

frequencyandatvari"dtemperatures.outhredandGeorge(|973a,b)described"T^",1"^1for analysing the frequffi-dependent behaviour of reiaxation rates' They analysed

the distribution of corielation tirnes for toad muscle water from measurements at three

f r e q u e n c i e s : 2 . 3 , 8 . g a n d 3 0 M H z ( 1 9 7 3 b ) . A v e r y c l e a r a n d d e t a i l e d t r e a t m e n t o f t h edispersion of water pr;;; spin-tattice relaxation-times T1 and r1, at 25"c for selected

mouse tissues was given by Knispel et at. (1g14).tfhe dispersion (frequency dependence

of relaxation times) of ?ro was attributedto proton exchange between water moiecules'

whereas the ma.ior contribirtion to Tr came from processes such as.molecular rotational and

translational diffusion. The correlaiion times for "x"hange, mo-lecular rotation and fast

diffusion processes;;"';;;;^;;'1'* ntu,2 x 10-t and -10-10sec' Figure 3 shows

the dispersion of the totairelaxation for muscle water' In a subsequent paper' Diegel and

Pin tar (1975b) recogn izedtheexchangeprocessasar is ing . f romthes low-exchangedif fusionofwatermoleculesbetweenthehydrat ionlaye.r?lg.I i=waterarrdnotfromthe exchange of protons as discussed earlier irnispet et al., r974;.Thompson et al', 1973)'

Figure 4 illustrates different relaxation pto""t* di."u"td by Diegel and Pintar (1975b)'

one may compare;;6;;c states of Hzo molecules_participating in the slow-

exchange diffusion und ondergoing slow r"oriiotutioo as deflned by Diegel and Pintar'

with the water molecules designated by the lifetime ln and r", respectively' in Packer's

diagram (see Fig.2).

_ 1 6

o

t 2

0a

I Heo l thY 'Musc le I Tumor o

*

t 6 o I O 7l O 4 l o s l o b

F r e q u e n c Y , H z

1974) Dispersion of T1 and T1o in samples of mou'se muscle

t o B

R : relaxation rate.

dssue

1 1 8 R. M.crHun-DE VnE

T, experiment

Fos t d i f fus ion :

r . oo >I , l - '( 20MHz)s t s - l

oEo

o

F-

Very s low

exchonge diffusion:

r "o ,= l , r ooo . i o3

\or : to s-r , ot t , =te

l--t [-lS low

reor ieniof ion:

f . oo>l , Tr@t < I

T - l e A . - l

Fos i d i f fus ion :

rduo<l,fp r I s-l

McGaughy (1974) and Fung et al. (1975a) supported a two-phasd model: one lhase,exhibiting a distribution of correlation times (the unfrozen fraction) and the other riiith a ,single correlation time (the frozen fraction). It may be nored that a ,ilii;, ;;l"l ;r,also proposed by Finch and Homer (1974) from the resuhs obtained "uo". o;c tJit.ur"ain the previous section). Of the various models suggested to describe the state of watet ,in muscle (and other biological systems), this modet accounts in thl -"ti *,irr".-ylmanner the behaviour of water. contrary to cope (1969), Fdng er c/. conclude d thatall the 2Hro was ':NMR- visible. oun and Derbyshire (igl+) it o reported a complex,

ro oo<<lrT,-rrl s

t r lL o f t i c e

Ftc' 4. (Diegei and Pintar, 1975b) The three relaxation processes and their contribution to thehigh field and the rotating field Zeeman relaxation.

-The numerical values shown for the

relaxation rates are approximative.

Finch and Homer (1974) reported the values of 4, T2 and, 7.1, for frog muscls,water protons at different temperatures above 0'C and over a wide range of ft"qu.n.y. ,They obtained a distribution of correlation times for muscle water, ranging from - 10-'i .to -10-11sec- The results were interpreted in terms of exchange of"wa"ter -1f"Jo.between two fractions: one with a distribution of different degreei of restricted ;"6; ,and the other with unrestricted motion like ordinary water eTn.Fung (1977a;; ,n*ruirl ,T1 of water protons in mouse muscle in the frequency range from ton-to 108H2, "rJiirr,,deuteron T1 from 2.0 x 103 to 1.54 x 101H2. He proposed relaxation -*tr"rir.r iir..hydration water protons and deuterons based on the observed frequency "rJ;;;;;;"i"rr. .dependent behaviour of relaxation times above 0oC, and the isotope substitution eflects'l(discussed in a subsequent section). ,,

The NMR studies of muscle water performed on frozen samples have revealedithat the unfrozen fraction of water is characterized by the frequency-dlpendent variations iof relaxation times. From the pulsed NMR measurements of the transverse relaxationtimes of water protons in striated frog muscle, Belton et at. (1972) rfr"r.O tfr"t ii,bound water did not freeze; as a result, below -i to' -td.c iu"", zoy]"i ,rrr.signal was observed. In another paper, Belton et al. (1973) investigated in detail thespin-lattice relaxation times and the dynamics of the unfrozen friction of *ut., inm1sc1e at two frequencies (30 and 60 MHz) and over a wide range of temperature ( irO to .-75'C). A distribution of correlation times was indicated for ihe unf.ozen *ut"r. fung,and McGaughy (197q measured the relaxation times of water in rat gastrocnemiuimuscle at frequencies ranging from 4.5 to 60MHz at -t31to -70"C. Th; i;""rr., orH2o and 2Hro for muscle and liver were also reported at different f;;q;*;;;;--;in the temperature range *37 to -70'c (Fung er ur., r975a). I" ,u;;i;; ",-*u-,keenng temperatures, the unfrozen fraction of water was shown to exhibit a distribution ::of correlation times; whilg above -8'C a single correlation time r"r "Ur"r""J',"frl*iwas short enough to render T1 independent offrequency. Based on these results, nung and .

I

(((

I

I

t

b

II\(

t

!

htt.1bf l

( ra

The NMR studies of water in biologicai systems

behaviour of relaxation times (Tr, Tr, T1o) of the bound or unfrozen fraction ofwater in ftozen Porcine muscle.

119

ir contribution to thealues shown for the

I 71, for frogide range of fiequer, ranging from - 10=ge of waters of restricted motiFung (1977a) mea104 to 108 Hz, andration mechanismsLency and temperaturpe substitution

lmples have reveay-dependent variati) transverse relaxa972) showed thatC about 201 of:stigated in detailfraction of water

ftemperature(+10unfrozen water. Fuln rat gastrocneml)"C. The Q valuesrrent frequencies

In samples atexhibit a distributiwas observed whi

rese results, Fungse model: one phand the other witha similar modelabove 0'C (di

ibe the state of wthe most satizt al. concluded

several authors have demonstrated that the water content of muscles and tissuesexerts a determining role on the relaxation behaviour of water nuclei observed in avaiety of samples. cooke and wien (1971) measured ?1 and T2 of partially hylr;i";various muscle proteins by the pulsed spin-echo technique. The rela-xation times decreasedas the rutio of water to protein decreasid, and. r I Tl was found to fe atectty proportionalto the protein concentration. Fung (1977c) me-asured T1 of water protons for dehydratedmouse muscle at three frequencies (5, 30, 100 MHz) down to u"iy to* water contents.41.i11.11'::,,t::.:^1"T1:l,1-9"**:"1r 11 values *iih a""r.urinf ,iut", content (x) wasoh.l-ul9:r:1::"llranincreasein Tl ar very low contents (x i?di ffi;ilil'rrl#"may have artsen because of a change in the structure of hydration tayer it low water,,'.t 'contents.

':r' Belton and Packer (1974) undertook a detailed study of the elfects of water content,, on the water proton relaxation times T1 and, T.2. The stepwise dehydration of a _"r.r"'t was found to correlate with changes in the transverse ielaxation times in a manner,,' ln:::li*r:j"3..1*Ti=,:rth".i;.scre rouo*"Juyi;hrd,;;" was arso investigated.

A very important contribution of this experiment *u, io rh"; ;;;;*;;;T;;';r1unfrozen" water for fresh and rehydrated muscles is the same, but the relaxation. 9*^11]:r:"i1-t:I__r:*p:ratures.

is..quite differenr in these two cases. The fact that theit 11"^11 ̂ ":^Tt::,:r

water is similar in the two cases reflects that it depends ", tr,", :lT.^"-,i:,:".J:T1 111,, ?",

the state .of macromolecules in muscle. However, the'difference in the relaxation behaviour arises from marked changes in the distribution.,of ,water in different fractions. - ----'

An interesting study making use of the correlation between the water content andT?]."1?r

times- was conducted to investigate the action of cholera toxin (rJdan et: ar.,1975)' These authors measured the water content and T1 and T2 values from the controland cholera-infected small intestinal tissues of rats. Ii was found that the relaxationtimes qf water in cholera-infected tissues were longer and the tissue hydration wasgreater than in control tissue samples:

Control Cholera

Percentage of tissue water?1 (msec)

:l2 (msec)

79.49% 84.s2%521.22 667.96

+69 .5 + rq ?s62.34 80.35

+9.59 +21.46

The abovementioned results suggest that cholera-treated tissues exhibit greater motionalfreedom of water than the- "ontiil ru;;i";. This observation was considered to supportthe general view that cholera enterotoxin acts by influencing the intracellular protein-water interactions. qiving rise to increased hydration. As a result, the permeability of' cells to water is incrJased, leading to enhanced secretory activityof small intestines... The importance of the relatio-nship between water content of muscles and differenttissues, and the relaxation times for "rptuining the increase in T1 when tumours developull be discussed in a subsequent section.

"^o,u"ty important contribution of the NMR studies of water in tissues and musclesnas been to reveal that the relaxation rates of water nuclei can be correlated withthe actual state of muscle caused by strain and death. Bratton et at. (1965) reportedthat ?z of muscle water protons increased with contraction and exhaustion, whereas?r ^remained insensitive io. "nuog". in ?e state. They exprained that T2 increased

,T::::: l-h: rl,"F.. in tension re'ieased 20r or water;'pari oi trr" bound water was

I'r.dseo reversrblv during isometric contraction and irreversibly in death. Chan g et ;.(1976) studied the relaxa"ti"; ;il;;i;;; protons in skeretai muscle (gastrocnemius)at different time intervals after taking iit"-ru-pr" from the animal. They obtained two

r reported a co

r20 R. MlrHun-Dr VnE

t _- l -

> t >

I

t ' r

t a r t o

\ o * * .| + '

n l ^ \ ^ + ' '

! i r \ - * ; ; :

- l ' r $ \i \ 9 !

i b o ' tr q r

+ .+ a

* * * _ _ ! _ .

I \ \\a\.. -------l:-ili;;-*.---_

i b I - - - - _ ] _ t + * - - - _

L K '\ ----- ]--..-..' l Y:

' i.^ '-*-!t'+_o --rr-r-r_

i . ^ \ - . ' + * + - _ - C r _

i :ro \.^!' *5. XI i. A'rt \iI

o . o l

o 50 t oo 150 200 250 300 350Time , m sec

Ftc. 5. (Belton and Packer, 1974) The variation of the transverse relaxation behaviour of thewater protons in frog gastrocnemius muscle as a function of water content at -300K. Themeasurements wete made using a Carr*Purcell/Gil1-Meiboon pulse sequence and the symbolscorrespond to different water contents which, relative to that of fresh muscie taken as 700/", are:

a 100%, + 54%, A 18%, O 10%, f 6%. Not all the measured points are shown for claiity.

relaxation times: Trn and Tte. Ttn (characterized by the slow relaxation rate) wasrinfluenced by the early post-mortem changes, and its value increased with time after'ithe removal of tissues from the muscle. Trz (representing the weighted average of allwater protons) remained practically unchanged with the lapse of time. These authorsstated that cellular water molecules "recognize" a change of environment as the physio- :logical state of cells undergoes a change. Furthermore, post-mortem chang"i were,iobserved to be relatively slow, taking about 4hr for completion. Hazlewoo a et at. (tglt)classified T1 and T2 of water (28'C) as a function of age. The muscles from animalsless than 10 days old were defined as immature muscles, and those from animals greaterthan 40 days old were considered as mature muscles. The followinq relaxation timeswere reported:

T1 (sec) T2 (sec)

Immature muscleMature muscle

1.206 + 0.055 0.t27 + 0.690.723 + 0.049 0.047 + 0.004

r i

: ;: l j

. irll-i

It was proposed that the fraction of ordered water increased in the post-natal developmenlof muscle. They also suggested that the extent of ordering of muscle water tends toincrease with maturation of muscles. l

By plotting the amplitude of the spin-echo train of water protons of mouse muscle,i(at 37"C) as a function of time, Fung (1977b) observed that the decay curve wasjexponential soon after the dissection, but with time it changed into a non-exponential ,curve during the first 40 min as illustrated in Fig. 6. On the contrary, for brain tissues .very little change in the spin-echo decay curves at 37"C was observed durine thejfirst hour after death. He concluded that changes in the relaxation times of hydiatiot'.water protons observed after death were caused by changes in the conformation of,muscle proteins, rather than by a rapid redistribution of water in different parts of theJtissue. shporer et al. (1976) showed that the relaxation of r7o from H217o in ratilymphocytes was non-exponential in the fresh state but became exponential after celldeath. Contrary to the opinion of Fung, Shporer et al. suggested that necrosis could

I(sI(st

IT

p

Ibe4

Jpb,si1 (

AI

SIfr,

A .ut

sirwi

wlsn

The NMR studies of water in biological systems

o a o o

o t t

aoooo

121

to !

o C

o Co C

Ioott.:

c - '

o

o

i behaviour of thet at -300K Thee and the symbolsrken as 100)(, are:hown for clarity.

:elaxation rate)rsed with time;hted average oftime. These autment as the ph,rtem changeszlewood et al. (197rscles from animirom anlmalsng relaxation

t-natal develoscle water tends

rs of mouse: decay curvea non-exponent

y, for brain tirserved duringtimes of hydrati.e conformationfferent parts of:om H217O in 'nnential after

T i m e , s e c: Frc'6' (Fung, ,1'977b)

rH spin-echo data for mouse water at 3\"c.The initial intensities arenot included because the contribution from organic protons is not negligibie. (A) 5min,', , I

(B) lomin, (C) 3omin.

''' lead to mixing of water in different compartments of water in tissues, or between nuclear

: , lnd cytoplasmic water., . The

r:ry:ti9. moment of deuteron is much smailer than the magnetic moment ofprotons (Table 1) this :uu:..t reduced dipolar coupling int"ru"tion, between protons anddeuterons as compared with proton utd p.oton. Thirefore, it is expecteoirr"i-p"rt*r

substitution of H2o by 2Hro in a system should result in ,h*p", warer proton signars., such a behaviour is observed in free water. For water in striated muscle "ii.;;civan and shporer (tlzs)"19n9rted that T1 of water protons was unallected after partial', 'substitution of Hzo. by.'\ro.Resing et at. (1977-) ana rung (r977a) also reportedthat deuterium substitution had very little effect on the relaxatioi, ti-". of muscle water.''

Fung explained this isotope substitution effect by "o.rsiaering that the major relaxationrnechanism is the intermolecular dipole-dipole coupling literaction, between waterprotons in the hydration layer and piotons in the relativiy immobile macromolecules,assisted by the slow water diffusion of the-type defined by rnispet et ar. (1974), and,' Pt:g"l and Pintar (lg7 sb)'The importance of cross-relaxation in explaining the relaxationbehaviour.of water protons in irotein solutions and in "oitug"n has been discussedelsewhere (see sections II rtO and rY 2(a). rn addition, it was observed that for DNAand pollrrucleotide solutions :rr-jF.zol2[2b solvents of various compositions, unlike at+5oc, in each case the linewidth at -s"c was indepen;";t;i the H2olzH2o com-f:i::l

but dependent on the nature of macromoleculei. Furthermore, for DNA solutionsDetween -5 and -35oc, the temperature-dependence of the tinewidth of water protonsignal decreased with increasing'2Ezo content (Mathur-De vre and Bertinchamps,1977a'b)' These results could be 6etteiexplained by taking into account the influence ofcross-relaxation and indicate that dipolai interactions betiveen macromolecular protonsand water protons dominate the T2 pro""r, of unfrozen water. Furthermore, civan andlloj*'

(tsts), and Civan "t or. i6iet;;;; r;;."d',i,"'i"n"*,ng importanr resurtsIrom a-comparative study of the three nuclei^(1}i, tH, rlo) i" -"r"r" water. (i) The ?1relaxation rates of t'a, tH and 1H nuclei of muscle-water ,*hiut identical frequencydependence. (ii) The raj!o.(7.1)"ul(T,t; ;f muscle. water and pure water are closelystmilar while the *'i:lT')'rlai;j';;;r;; water is 2.l-times sreater than for muscrewater' (iii) The ratio (Tr)rtfri f;r;i u-"olrr was found to b"'in the range of 9_11,whereas this ratio for 17o was uppro*r,,'ut"ryl.s-ib. it"o]a"l""rr-*e motion between asmall immobilized fraction u"a u iurg. i;;;;;" of free water was proposed.ld) Membranes and cell water

The erythrocvte membrane-- is highly permeable to water, this results in a fastexchange between water in "elts aoipiariu 1.*"tru.rge time of the order of 10msec).r.PJ. 35/2_E

hat necrosis

r22

, j

' I

)

iF

R. M,rrnun-Dr Vnr

The NMR technique is particularly fitting for studying such fast-exchange processes.Water molecules present in these two compartments are constrained in widely differentenvironments. Consequently, each type of water exhibits a characteristic and distinctrelaxation behaviour. The use of Mn2 + in these systerns has proved to be of particulqvalue for the following two reasons: (i) in the presence of Mn2+, the relaxation ratesiof water nuclei can be enhanced by the paramagnetic contribution to such an extent ,that their NMR signals become unobservable; (ii) the cell membrane is known to fgeffectively impermeable to Mn2+. Consequently, the addition of Mn2+ to a cell suspensionresults in an enhancement of the relaxation rates of water nuclei in the extracellularregion. Therefore, it is possible to selectively resolve the contribution of the intracellularwater to the observed spectra. A similar effect can equally well be obtained by freezhssamples below 0"C; in this case, only the intracellular water remains unfrozen nn4therefore contributes to the resulting spedtra. It is surprising to note that applicationsof the freezing technique in this domain is very limited. The only example is that ofwater in haemoglobin solutions reported by Zipp et al. (1976). Certain groups 0l:workers have used the pellets of cells to eliminate the contribution of extracellular,watef.

Andrasko (1976) measured the water diffusion permeability of human erythrocytes'by using pulsed magnetic field gradient techniques. The following values were reportedfor the diftision constant (Dr) and the lifetime (re) of water within red blood cells.

Dl rB(cm2lsec) (sec)

bloodbiood + xp-Cl. HgBzO-

1.16 x 10-5 0.0177.'15 x L0-6 0.048

*p-Chloromercuribenzoate: known to drasticallyreduce the osmotic water permeability of human redce11s.

It was shown that deoxygenation of normal and sickle erythrocyte results in aconsiderable decrease in the T, values but causes no chanse in the ?, values of water'(Cottam et aI., 1974; Thompson et aI., 1975; Zipp et al., 1976\. A three-state modelwas proposed to explain the relaxation data, each state exhibiting a characteristiccorrelation time: bulk water= 10-11sec, water hydrated to hracromolecule (10-?1:z"> 10-11sec), finally the third region of water which is tightly bound exhibits a'correlation time similar to that of protein (t.210-7sec) (Thompson et al., 1975;Zipp et al., 1976). Zipp et a/. reported that upon deoxygenation of sickle cells andhaemoglobin S solutions the T2 values at room temperature decreased by a factor ofwhereas after deoxygenation of normal cells and haemoglobin A solutions, no changein T2 was observed. The low temperature studies of linewidth at (-15 to -36'C), andiTt at -20 to -80oC, for oxy- and deoxy-haemoglobin A and haemoglobin S solutisuggested that the characteristics of bound water were similar for all four species.the basis of the three-state model, Zipp et a/. proposed that the sickling processthe irrotationally bound water. Lindstrom and Koenig (1974) and Lindstrom et al. (Iinvestigated the effects of oxygenation, and aggregation of haemogiobin (HbA)sickle haemoglobin (HbS) solutiot'r" by studying the frequency dependence ofproton relaxation rate (llT) (dispersion curves). They calculated 2", the correlatton,times for the rotational motiorl of haemoglobin molecule from the inflexionv" (see Section IV.1(a), p. ll2). It was shown that under the conditions of compler;oxygenation, HbS molecules interact with each other more strongly than do themolecules. The orientation time of oxy-HbS molecules was shown to be larger than tnatof HbA molecules. The T1 values of water obtained at low frequency gave much rnofeinformation about the state of aggregation and rotational motion of the haemoglobtltmacromolecule as compared with the ?1 values obtained at a single high frequency. l

A simple NMR technique for measuring the water exchange between erii, , i i

ii.

and.o f Pawate)Thes'polarandnerve

shrin ta1In thtthe Pwatetto cY( T t :authoslowewere ICivansuperlT1 forfate (

activaimporwater.red blrMn2+The hrtime eresults

Fincand th'Fromby NNthree.in defibeen uon thepostulainteracweak ir

Withwater, ,of watrfrom tlconstar

The NMR studies of water in biological systemsL / J

rst-exchange processes.rted in widely differentracteristic and distinctved to be of particular*, the relaxation ratesion to such an extentbrane is known tot2* lo a cell susoensiei in the extracellularion of the intracellulaie obtained by freezin:emains unfrozen anote that applicatiLly example is that5). Certain groupsution of extracellula

I human eryyalues were

ed blood cells.

hrocyte results ine T1 values of waA three-stateting a characteri;tomolecule (10- 7-,

ly bound exhibitsmpson et al., I97L of sickle cellssed by a factor ofsolutions, no-15 to -36'C),noglobin S solall four species.kling process alndstrom et aI. (1roglobin (HbA)ependence of wa

Tr, the

inflexion fiditions of coy than do ther be larger thancy gave muchof the haemogigh frequency. ,',

Eggaibumin(101 solution

lgg sealed in Distilledwhite vacuum) HzO

,il ,:d {TTll:9^e,ed with Mn2* was described bv conlon and outhre d (rg,2).The use

il ::^l:'::,:'l: ror as a toor for distinguirrr,.rg irt" p."i"" ilun signal 0f intracenurar

il water liom that of extracellular water wii init[ily dl.",il by Fritzand Swift (1967).il These authors successfully applied this_methoJ i" i;;ffiate the state of water inil polarized and depolarized frog ""tu"r.-ri.oliu;r"r""'a5porurized.bythe chemical

Id#ffff'#*lI water was ascribed to nucleus izt : J.i ms99] and the more rapidry reraxing popuration| ,o cytoplasm (4: r:l -,:*) ^ii. i '"u ',o-;;;;;;.was

appreciabty longerI trt : 7'5msec)' The results of temferature effects on the lehxation times led theseJ authors to conclude that the exchanee rate of water between these two phases isJ slower than the t"lTu'jT-lute of 170 (ilow-exchange condition). Two different merhods[ *ete used to studv the NvrR of iio ir iirtto enriclied nu*uo "rytirrocytes (Shporer andI

ar:",, 1975):,(,rr d'.rt "o-furir"'r'"ii"r"ration rates "i;o in isorated pelrets andI supernatant, (2) relaxation rates measured in th" pr;;;;; of ilnr*. rt was noted that

1,,. l,loi^t1111!:tt"H water ws 4-5-times shorter trru" i". ii""supernatant. The values off, rate constants (ft") at 25 and 37.c were found to be 60 and 107sec_1, and the:l ' activation energy f9r ft' as equal to g-7 t t.o kcal/mole. The authors emphasized theimportance of the interaction between water and membrane during the transport of;111,::r3,::i"p,t"":lii:!,:i:llnvestigated tl;;;;;"r water between humanred brood cets and tn:-rtffxi pil" ilt ,;trt"***;i"d;.iilJirffiL ,"ffi::r*#llMd'*, and by measuring 170 relaxatioo tl*"i oiHrrro'i1 irr. uu..n"" of added Mn2+The halfJife for cell water at 25.C was founa to U" is _."" ;;;;"", and the exchangetime equar to 0'046 msec' The relaxation-iime values are reported in Table 2. Theresults were anarysed in terms of the classical two-compartment exchange model.,"51?J;::"::*:i1'J # T o*":: :f:i -l', ":*" I; r?; ffi ii", "", rro m 0 to 30 " c,': and the a,1 dependence ir r;in "qr""", ir.p;;;i{##ifi fi:ff#::?&l+il3;i:: From the avairable.data th6se uurrro..-.ourd not specfy *i"rn.. the water detectedby NMR was associar.d ,"ith ";;;'r;t. -"-b.uo", ripia, o, polysaccharides, or a'three. Nevertheress, they pointed out ihe importance ;; ti" "r"*brane_bound waterin defining the structure and functions of

,membranes. Simple lecithin systems havebeen used as models for the ui"r"gi*i^Lembranes. Klose ani stelzner (1974) repofied,- on the NMR studv of specific uiori['"r water in r"u,rrlr_u,"rzene systems. Theypostulated that water-membrane interactions a* fimri.J-io ,i,r". regions: (1) theinteraction of water ,with the pd;;; groups, e) water interacting by additionalweak interactions, and (3) water -"f"irfg U.yond both these resions.with a view to studlng th. .*;;;;i*ffb;;,";;;;;'i"u,"r,u,,

on the srate or- water, James and Gillen ttglZl neirir"i Tr, T2 and r",,lif,rJoo consranr (D) valuesof water from the unfert'izei "rriJ"r-Jgg. onry oo" .".orur"e signal was observed::1il1f:|]::ilkT-"uil"-*"i"'''fi," uut'., or,"ru"ution times and diffusion

Foo

yolk

Difusion constantwith respect topure waterQ (msec)72 (msec)

0.80 0.881180 1270

340

0.2567.027.0 2830

2830,'tween erythroc

T

;o

E

o

Fq

F

o-

t o

a

7

Ia

:6=

;=

o-

8 . O

7 . O

4 . O

124 R. M.qrHuR-Dr Vn6

---1 i

l il 9-1 roo ts

t :J s o ; � .

It -t --l 60 ;-l oI

i 4 0 o

l ol . =

1 , o L

+ ?t ;l o- o <

Hou rsFIc' 7. (Beail et al., 1976) T1 and water content as a function of HeLa ceil cycle. (O-O) fiduring the cell cycle (mean of eight-ten experiments). Bars denote standard irror of the mean.(o-o) water content during the ceil cycle. (---) Actinomycin-D binding ability of thechromatin. Data for the dashed line are ftom pederson and Robbins.

copyright 19'18 (25 May) by the American Association for the Advancement of Science.

Significant portions of cellular water are known to exhibit reduced mobility, wcan be attributed to the obstruction or hydration effects. On the basis of Tr rle4surernrthese authors proposed that probably the hydration effects are more important. Fung ef(1975a) also reported that the T1 values for egg yolk were considerably shortei thfor egg white.

Some authors have given evidence showing that the changes in ?1 of cell water ardue to alterations in the configuration of macromolecules. ft in mammalian cellstissues was measured as a function of the external ion concentration and total srwater content (Raaphorst et al., i975). The changes in the fraction of boundunbound water were shown to be associated with changes induced in the macromolecul,configuration by varying the salt concentration and the amount of water. BeaTl et r(1976) investigated ?1 of water protons (at 25"C) and the total water conient as a functiof HeLa cell cycle (Fig. 7). They demonstrated the effects of biological and physiologi,alterations during the growth and division of cells on the ?, of water. It was p.oportthat the relation between T1 and the cyclic pattern of cell growth and divisioninfluenced probably by the conformational changes of macromolecules (associated withe morphological changes during cell division), and by the water content.

(e) Tumours and cancerous cellsAn important development in the NMR studies of water in biological systems I

been to show that the measurements of 7�1, T2 and T1o of water in tiisues and orgahave a potential use in thd cancer research. The relaxation times are closely relatedthe structure, mobility and the content of water in normal and cancerous tissues acells. As a result, the effects of the presence and growth of tumours on the behavrof water can be investigated by the NMR technique. Nevertheless, the eventual usethis method for the diagnostic purpose rqmains a highly controversial issue. The basiproblem arises because, even though it is now fairly well established that there ismarked difference between the relaxation behaviour of water protons in the normal ancancerous states, there is no certitude that such a behaviour is strictly specific forcancerous state.

In his pioneer paper, Damadian (1971) reported that the water proton relaxation tiT1 and T2fot malignant and normal tissues were distinctly different. Watqr in malitissues exhibited longer relaxation times or a higher motional freedom. These observatiwere interpreted in terms of the postulate of Szent.Gycirgyi (1957): it states thatdegree of otganization of water in the cancerous tissues is much lower than in not

tissues. Thrworkers an

FreY et theart, mus<significantlltissues fronin the T2 viInch et al.1tissues fronfrom animafrom healthi.e. ft wasauthors emHollis et al.gave shorte(1974) concvaluable tortissue in cas

Hazlewotand the reltumour. Itincreased mtumour cellrthat macroronum in t}t,hydration. Idiseased tisinot strictly c1974a). Thevalues of th,as fibrocystili values vstates. Hazltspectroscop)cancer reseatin NMR rewater conten

Block andtumorous tishad lesser ancell populatierythrocytes)frequency, tlfor either oIeffect to themoving macrwith the tota

It has beerconsiderablypointed outand cancerorreduced at lo

I

I

E

z

o

c6

E

F

The NMR studies of water in biolosical svstems 1)\

tissues. The results reported by Damadian were soon confirmed by several groups ofworkers and initiated extensivg research in this field.

Frey et aL (1972) showed that many nonmalignant tissues from spleen, kidney, liver,heart, muscle, intestines, stomach, skin and lung for mice with a tumour on hindieg hadsignificantly ]ong9r relaxation times (T1 and T1) as compared with the "ort"rpo.rdiogtissues from healthy mice (see Table 2). They did not observe any systematic variationsn the Tz values. It was proposed that water was more ordered in tiisues with tumours.Incb et al. (I97$ measured the water content and ?1 for neoplastic and non-neoplastictissues from mice and humans. The I values for tissues of liver, spleen and kidney

' from animals with large, rapidly growing tumours were longer than Tl for similar tissuesfrom healthy animals. Whereas, fol slowly growing tumours this difference was negligible,i.e. Tl was found to be related to the rate of growth rather than to malignancy. iheseauthors emphasized that the ?1 values were related directly with the water content.Hollis et al. (1975) also observed that the slowly growing and well-differentiated tumoursgave shorter ?r values as compared with more rapidly growing tumours. Schara et al.' (1974) concluded that proton spin-lattice relaxation (at room temperature) can be avaluable tool for the characterization of the pathological changes in tire thyroid gland

,,, tissue in case of thyroid gland cancer.Hazlewood et al. (I974d studied in great detail the relationship between hydration

and. the refaxation times of water prJtons in tissues from mice with and withouttumour. -It was shown that T1 and T2 of water protons, and diffusion constant (D),increased monotonically and distinguishably from the normal, to nodule and finally totumour cells in the development of marnmary tumours in mice. These results indieatedthat macromolecular-water interactions were altered by the presence of a tumour oronum in the host. The changes in T1 and T2 were iniepenient of changes in organhydration. A comparison of the ?r and T2 of neoplasms from the breast of normal anddiseased tissues showed that the values of relaxation times were correlated with, butnot strictly dependent on the hydration of tissues (Medina et aI., I975; Hazlewood et aL,1974a). The tissues could be classified as fibrocystic or neoplastic depending on thevalues of the pair of Ty T2.If T1 < 792 and f, S SS.i, then the tissui was classifieOas fibrocystic; if T1>792, Tz> 58.1, then the tissue was classified as neoplastic. The?2 values were considered to be more discriminating than T1 in certain diseasedstates. Hazlewood er aI. (L974a) and Medina et at. (tns) pointed out that NMRspectroscopy could be employed as a useful tool for the detiction of cancer, and in:ul:!t research. Saryan et al. (1974), and Bov6e et al. (1974) indicated that the increasein NMR relaxation time ?r (at -25"c) is determined, in paft, by the increasedwater content of cancerous tissues.

Block and Maxwell (1974) studied the behaviour of water proton T1 for normal andtumorous tissues of rats. These authors considered a model in which tumorous tissueshad lesser amount of water with restricted mobility than the normal tissues. Three mousecqll

.populations (EL 4 ascitis tumour cells, normal spleen leukocytes, and normalerythrocytes) were studied at 13.56 and 100MHz by Block et al. (1977). At eachfrequency, the T1 values for tumour cells were found to be greater than the valuesfor either of the normal cell type (see Table 2). Qualitativjy, they attributed thisefflbct to the binding of a fraction oi water (exhibiting restricted *oUltity; to slowlymoving macromolecules. The 1/7, values were found-to vary approximatety linearlywith the total water content over the range investigated.

It has been demonstrated by several iuthors that T1 of water protons in tumours isconsiderably longer than in healthy tissues. Furthermore, Damadian et al. (1973) havepomted out that the discrimination between the relaxation times of water in normaland cancerous tissues appears to improve at lower frequency, i.e. the overlapping isreduced at low frequenciis. Diegel uod pirrtu, (r975a) defined the resolutio n ,,r,' as:

cycle. (O-O) T1d error of the mean.rnding ability of the

nt of Science.

luced mobility, whicfiis of 11 measuremenimportant. Fung e/

riderablv shorter

r T1 of cell watermammalian cellstration and totalaction of boundin the macromolecuof water. Beall et

r content as a funct;ical and physiologirater. It waswth and divisioncules (associated wiontent.

,iological systemsin tissues andare closely relatedcancerous tissues aurs on the behas, the eventual userrsial issue. Thelished that there isrns in the normallrictlv soecific for

'oton relaxation tt. Water in malim. These observatii7): it states that.ower tnan ln no

,.r,, _ T,r (tumorous) - Ti (healthy) : (!!:_ \\a,

T1 (tumorous) \ t/f, ) b

128 R. Mernun-Dn Vnn

3 x 10-asec with an activation energy of 4.8kca1/mo1e, and the proton exchange tins'was calculated to be 1.3 x LO-asec with an activation energy of 1O.0kcalimole. J6,addition of NH+CI enhanced the proton exchange 13rg. It may be remarked that thevalue of correlation time noted above for collagen is much longer than the value of2 x 10-8sec that was reported earlier for the correlation time of molecular rotation s1water in muscles (Knispel et aI., 1974). The rate of proton exchange that greaflyinfluences the water proton signal from collagen was found to depend on the temperature.pH, and buffer salts (Migchelsen and Berendsen,1973; Biefkiewicz et al., 1977).

In order to explain the non-averaging of dipolar and quadrupolar interactions which,are responsible for the splitting of lH and 2H signals, respectively, Dehl and Hoeve (1909;assumed a model in which certain preferential hydrogen bonded sttuctures of water were :formed. The water molecules could difuse rapidly between the highly oriented strandsof collagen fibres, but their motion was anisotropic. Chapman and Mclauchlan (1969) ,

also proposed a continuous chain model for water in collagen. Fung and Trautmann:(1971) proposed that the observed dipolar or quadrupolar splittings for water in collagen.,were the ayerage of two types of water: (1) water molecules adsorbed or bound to '

the collagen triple helix (the oriented water), and Q) the remaining free water moleculeq ,that undergo rapid reorientation. These authors reported the effects of ions on collagen .,hydration by studying the ion effects on the deuteron quadrupole splitting. Migchelsen andBerendsen ft97T were led to the conclusion that the chain model was not sufficient ,to account completely for the hydration of collagen; their results also favoured the,specific binding model. Field-dependent splitting of the water signal was also reported,for sciatic nerves of rabbits (Chapman and Mclauchlan, 1967). The maximum splitting ,

was observed when the nerve axis was parallel to the applied field. lDehl (1970) described a method for estimating the amount of unfrozen water in'

frozen fibres of collagen. The method is based on the fact that line separation is given ,by K (cos2 0 - I), where the splitting constant K decreases with increasing 2H20 rcontent, and 0 is the angle between the fibre axis and the magnetic fie1d axis. Fung and:,Wei (1973) applied this method to study in detail the effects of water content and:salts on quadrupolar splitting of 2H2O in hydrated collagen for the maximum splitting r(0:0'). The amount of "unfrozen water" decreased in the presence of salts. They.pointed out tlat hydrated ions block the binding sites for water in collagen. Only the iwater molecules bound directly to collagen were oriented and resulted in the dipolar or ,quadrupolar splitting for H2O and 2H2O, respectively. rjLadrupolar splitting for H2O and'H2O, respectively. ,ti

The NMR studies of collaeen water discussed so far were performed bv the continuous iwave technique at a single frequency. Fung e/ al. (1974) applied the pulse techniqueito measure the T1 .values for water in hydrated collagen at different frequencies and:over a wide temperature range Q5 to - 80'C). They observed that T1 was strongly':dependent on the temperature and frequency. The correlation times could be describedby a distribution function in a manner similar to that for water in muscles, discussedearlier in Section IV.1(c). Edzes and Samulski (1977) used the FT pulse technique to'study the proton spin-lattice relaxation decay of hydrated collagen under the conditionswhen the dipolar splitting is zero (fibre axis perpendicular to magnetic field). Byjstudying the effects of partial substitution of HzO by 2H2O, these authors proposedthat dipolar coupling between water protons and collagen protons, i.e. cross-relaxation,and spin diffusion make an important contribution to the water proton relaxation Imechanism. Edzes and Samulski (1978) further confirmed, by using the method ol-selective inversion of water proton magnetization with longer 180' pulses, that cross:trelaxation contributes to the relaxation of water protons in hydrated collagen. .

(b) DNAfibres

A study of water from the hydrated DNA (salmon sperm) in the form of orienteo,fibres was initially reported by Berendsen and Migchelsen (1965} They observed that thesecond moment of the water proton signal varied qualitatively with the angle betwee4ithe fibre direction and magnetic field. In the case of oriented DNA, the anisotropl 01,

re proton exchange time' of 10.0kcal/mole. They be remarked that thernger than the valueof molecular rotation

exchange that greatly'rend on the temperature,cz et al., 1977).colar interactions whi; Dehl and Hoeve (1structures of waterhighly oriented strandSand Mclauchlan (1. Fung and Trautmanngs for water inadsorbed or boundrg free water molecu,'cts of ions on collagenplitting. Migchelsen a.odel was not sufficienl;ults also favoured thergnal was also.reporThe maximum splitti

: of unfrozen waterline separation is giwith increasing 2H

tic field axis. Fungof water contentthe maximum soli

rresence of salts.r in collagen. Only;ulted in the dipolar

:med by the cont-'d the pulsefferent frequenciesthat T1 was strongl

rres could be descrir in muscles. diFT pulse technique,n under the conditi.o magnetic field).hese authors

Iter proton relaxatiusing the method

80" pulses, thated collagen.

L the form of oriefhey observed thatrith the angle

The NMR studies of water in biological systems I2g

water molecules in a direction perpendicular to the fibre axis was proposed in contrastto the model which was put forth for collagen water. Rupprecht (1966) prepared sampleso,f caTl thymus DNA (NaDNA) by the wet spinning method. He plotted peak-to-peakamplitude of the derivative signal recorded as a function of the angle between thedirection of molecular orientation and the magnetic field. From these results, Rupprechtwas led to conclude that the hydration structure in DNA is similar to the structurepresent in hYdrated collagen.

Finalln Migchelsen et al. (1968) investigated the proton NMR spectra of water inoriented NaDNA in the A form, and LiDNA in B and C forms. Similar to the resultsreported by Rupprecht (1966), Migchelsen et al. (1968) also observed a single protonsignal at room temperature for NaDNA whose linewidth depended on the anglebetween the fibre direction and the field. However, for LiDNA in the B and C forms,angular dependent splitting was recorded at room temperatures. The proton exchangeprcc€ss was considered to be an important factor influencing the water spectra of NaDNA.

V. SUMMARY AND CONCLUSIONS' ' Thi, article presents a general perspective of the multifold NMR studies of waterperformed in various biological systems. A considerable effort has been devoted toinvestigate the relaxation times of water nuclei (1H, tEl, t'O) for a variety of biologicalsamples as a function of temperature and frequency. One common and striking featureobserved in nearly all cases studied is that the relaxation times of water nuclei andthe diffirsion constants of water molecules are much lower than t}re values observedfor free water. Generally, these results can be interpreted in terms of: (1) the restrictedand anisotropic motion of water molecules and enhanced proton transfer in the hydrationlayer; (2) the preferential and dynamic orientations of water molecules in the vicinityof biological macromolecules. The characteristics (1) and (2) arise because a fraction ofthe total water content (in solution or in biological samples) is associated with proteinsand nucleic acids, forming a hydration layer in their close vicinity. In other words, watermolecules in the hydration layer exhibit distinctly different properties from thoseobserved for free or extra-hydration layer water. Furthermore, the behaviour of thehydration water molecules was found to depend strongly on the nature of the hydratedspecies. Two-state and three-state models were proposed to account for the relaxationbehaviour of water nuclei.

An important characteristic of hydration water is that it remains unfrozen or mobile(on the NMR time scale) at temperatures much lower than the freezing point of freesolvent. This phenomenon proved to be very valuable for investigating the state ofwater in systems such as muscle, collagen, tissues, membranes; as well as for studyingthe changes it -aoo-olecular-water interactions induced by external factors (y-irradiation) and the changes in water structure which result during natural, biologicaland physical processes such as growth and division of cells, muscle strain, cancerousgrowth. The proton NMR spectra obtained from hydration water in frozen samplesfurnish unprecedented information concerning the macromolecular-water interactionsand the state of water in biological systems.

There is a vast scope for the application of tle NMR studies of hydration water toexplore and study in detail the effects produced by certain toxins, drugs, carcinogensand radiations: to investigate the sensitivity and specificity of different organs to'tleseand other related perturbing factors.

There is increaiing evidince showing that cross-relaxation between the protons ofwater molecules and of the macromolecular chain contributes to the relaxation ratesof water protons. Eventually, it may become necessary to revise and reeonsider themterpretation of certain previously published results in the light of cross-relaxation.

Water constitutes the major component of all living systems, for example it representsabout 70-801 of the total cell constituent. There is conclusive evidenCe showing thatwater does not simply serve as an inert medium, but it participates at the molecularNA, the anisotropy

i

r l l 130 R. Marsun-Dr VnE

level in basic biological interactions and in fundamental biological processes. 1n fac!the hydration water molecules constitute an integral part of any macromolecular q1cellular system under consideration. The importance of witer in maintaining the structurnlintegrity of proteins is well-established. Nevertheless, investigators in different domainshave not fully recognized the important and crucial iole thaihydration *ut., -or""irr,may play in various biophysical and radiobiological processes. wirile posturating lng"nilu,theories and mechanisms to explain such processes, many authors have either totallvneglected the participation of water or considered it simply rn terms

-"i ,h, ;;;;;imedium effects. Hopefully, the NMR studies of water .u.ii"a out very extensively indifferent laboratories would largely contribute to unravel the vital

-functional anistructural roles played by water at the molecular level in many biologi.d i";;;d;

and biophysical processes.

ACKNOWLEDGEMENTSThe author expresses her gratitude to Dr A. J. Bertinchamps (Euratom-UlB) for hernand encouragement that made it possible to accomplish this article.- il;k, ;:acknowledged to Dr p. Lejeune (Institut

. d'Hygidne et d,Epid6miologie) ro, .*uuseful discussions and his keen interest in this r"ort. Einutty, g*t.f,rl tfr"itl ; "if;;;!o Dr w' Stone (universit6 de Louvain) and Dr r. iiJtenga (vrije UniversiteitBrussel) for reading the manuscript with interest, and giving several inuui,ruut" ,ogg"rtioo*

. REFERENCESAr-Rewr, J. M. A., Broxsroor, J- p., o'BnrrN, c., caool D. E., Er,vroop, J. A., JoNEs, J. R. and EvANq E. A.(1974) Tritium nuclear magnetic-re,sonance spectroscopy. II. chemical shifts,;eferencrng, and an application.J- Chem. Soc. Perkin Trans 2,1635-163g.Ar-Rnwr, J' M. A., ELVIDGE, J. A., JoNEs, J. R.. and EvlrNs, !. A- (ig75) Tritium nuclear magnetic resonancespectrosco.py' Ill._Coupiing constants and isotope effects. J. Chim. Sic. perki Trans 2, ui-qsz-ANDRASKq J' (1976) water diffusion permeability of humar, erythrocytes studied uy a pulsed gradient NMRtechnique. Biochim.biophys. Acta42g,3o4_3i1.

E rsrwv or4qwut rrrvr^

B*!? l- T.,.Eezr twoon, C: _l_ul^d^ \4o, I N (1976) NMR patterns of intraceilutar water as a functioa ofHeLa ceil cycle. Science, N.y. 192" 904-gO7 .BtrroN, P' S', Jn crsoN, R. R.. and .Pncxrn, K. J. (1972) Pulsed NMR studies of water in striated muscle. LTraasverse nuclear spin relaxation timei and freezin! effects. Br";i;.ir;;hy"r. ,tcta286, 16_25.B""t9\, P' s', Pecrnn, K. J..and Srrl-vT6ep-, T. c..lth) Pulsed NMR studies of warer m striated muscle: II.Spin-lattice relaxation times and the dynamics of ihe non-frel;; i.;;ti"r ot *ut"r. ntiiii*.-;i;;:;;Acta304.56-64 ,BELroN, P' S. and PACKBR, K. J. (1974) Pulsed NMR studies of water in stdated muscle. III. The efiects ofwater content. Biochim. biophy s. Act a 354, 305_314.BEnnNosnN, H' J. c. (1962) Nuilear -ugo"ii. resonance study of collagen hydration. J. chem. phys. 36,3297_3305.BnnnNnstN, H' J' c' andllrccnnrsru, c. (1965) Hydration structure of fibrous macromolecules. Ann. N. y.

Acad. S ci. l2S, 365-37 9.BERENDSEN, H' J' c. and^MrocuersrN, c. (1966) Hydration structure of collagen and influence of saitsF edn Pr o c. 25, 998 -1002.BERENDSEN, H' J' c' (1975) Specific interacfions of water with biopolymers. rn water: A comprehensiue_ .Treatise. (ed. F, Fneurs), Vo1. 5, pp.293_349. plenum press, New iork.BrEirrrrwr-cz, K. J., BsnrNoseN, H. l. c. ana A";-rr, p.'i.'usr;';."p-ries of warer in native andmodified collagen. Ann. Soc. Chim. polonorum.5l, 149_15g.5LEARS' D' J' and DANYLUK, s. s. (196s) Proton wide-line nuclear magnetic resonance spectra of hydrated_ proteins. Biochim. biophys. Acta 154, 17_27.BLocK' R- E' and MaxwsLr'-g.!:^(127-!) Proton magnetic resonance studies of water in normal and tumor rattissues. "I. magn. Reson. 14,329_334..Brocr, R' E., Maxwrr,r, G. p, pnuoHoMMg D. L. and. T]rDso,N, L L. (1911) High resolution protormagnetic resonance spectral characteristic of water, lipid and p*t.i" !ig"a. from three mouse cellpopulations. J. natn. Cancer 1nsr. 5g. 151-156.Br'or, J' and LoulN, H. (1973) The effects of 7-irradiation in DNA. curr. Top. Radiat. Res. e. g, 165-245.Bovi-q w, IIuwnN, P. and irtaor, r. (1974) romo. aetection and nuclear *ug;Ji" ,.ron ance. J . natn. cancer.Inst- 52,595-597.Broxsrncq J., ELVDGE, J. A., JoNrg J. R. and EvrNs, E. A. (rglL) Tritium NMR spectroscopy. I. Techniqug_ intemal referencing and.some preliminary resuits. Org. *ogn. i"ror. S, til_l'il.BnarroN,,c' B', HoprINs'.4: !.�uid wnNannc, r. w. juos; Nuclear -ulo"til'r.rooance studies of livrngmuscle. Sclence, N. y. I47.73g_739.carnrNcroN, A' and Mcl.lcltlY:.

{. D. (1967) Introd.uction to Magnetic Resonance. Harper and Row,New York, and John Weatherhill. Tokvo.cnaNc, D' C', Hezrrwoo-D, g. ̂ {r 1lrgJor.s, B. L. and Ronscur.cu, H. E. (1972) Spin echo studies oncellular water. N ature, Lond. 235. L70_1i 1.

rcal processes. In facqry macromolecular orintaining the structurals in different domains'ation water molecules: postulating ingenious'rs have either totallyterms of the overall

rt very extensively invital functional and

riological interactions:

ratom-UlB) for heloarticle. Thanks are

3miologie) for manyul thanks are offer"i,a (Vrije Universiteiirvaluable suggestions.

tng J. R. aad EvaNs, E. A.:ncing, and an application.

rclear magnetic resonance.'ans 2,449452.r a pulsed gradient NMR

' t

ar water as a function o[

[ter in striated muscle. I;,ta286,76-25ter in striated muscle:.Il:water. Biochim.

ruscle. III The effects

t1on. "I. chem. Phys.

:romolecules. Ann. N. y,

a:rd influence of

Vater: A Comorehensioi

)f water in native and

nce spectra of hydra

cneuc, D' C', Ronscn'qcH, H. E.' NIcuors, B. L. and HAzLEwooD, c. F. (1973) Implications of diffusioncoefrcient measuremerts for the structure of cenurar water. Ann. i. i. ;;rr. sci. 204, 434_443.CHANG, D' c' Hazrnwoop, c. F. and wopssNra,o.n. gslal i-n" rpii-r"r,t"e relaxation times of waterassociated with eariy post mortem changes in skeletal muscle.' Biochii. uopnyr. Acta 437, 253-25g.cnarr'rei'r, G' and Mcleucaui'r, K. e. a"qozioii.ra;"r., ;;;'r.tif o".u" of rabbit. Nature, Lond.215,391-392-,"^r#i;r?: and MclaudflaN, K. A. (1969) The hydration sftucure of collagen. proc. R. Soc. B. 173,

OVAN, M' M' and Suronrn, M. (1975) Puised nuclear magnetic resonance study of ,rO, ,D and 1H ofwater in frog striated muscle. Biophys. J. 1.5,299_306.' crvAN, M' M', Acrfieto, A' y. uita stoo*un, rra. filzsl The relationship between the transverse andlongitudinal nuclear magnetic resonance relaxation rates ofmuscle water. Biophys. J.21,127-136.cr.rvrr.rNo, G. G., cn.lNc, D. c., Hlzrswooo,

.c. F. "_rd R;il;r*;", ri. n (1976) Nuclear magneticresonance measurement _of skeletal muscle. Biophys . J. 16, 1043_1053.

,"*T:;f*ltl"r?ryTlr*;$. ltii_rlfMater

diffusion permeabitity of erythrocytes using an NMR technique.' coor4 R' and wIrN, \ (1971) The state of water in muscle tissue as determined by proton nuclear: magnetic resonance. Biophys. J. ll, 1002_1617.

cor-a {' w' (1969)-Nuclear magnetic resonance evidence using D2o for structured water in muscle andbrain. Biophys. J. 9, 303-319.to"1_1r* (1970) The solid state physics ol electron and ion transport in biology. Adu. biol. meil. phys. 13,

corralr4 G' L', YlrsNrrNr, K. M., Yaueore, K. and MrBnu,rry M. R- (1g74) The gelation of deoxyhemoglobin1 s in erythrocytes as detected by transverse water proton relaxation measurements. Archs Biochem.Biophys. 162, 487 -492.DeveoIeN, R' (1971) Tumor detection by nuclear magnetic reson ance. Science, N. y.171,1151-1153.DAMADTAN, R', ZnNnn, K, Hon, D., Dunro, T., MiNronr, r. u"a coinsrrnnn, M.-ifi?x irr.lJii -ugn"ti"

i;r_ll?r?r.* a new tool in cancer research: human tumors by NM* Ann. N. y. Acad. Siil-nZ,

D.a'uaoreN, R., ZlNnn, r". H9R: D. and Druaro, T. (r9|/4) Human tumors detected by nuclear magneticresonance. Proc. natn. Acad. Sci. U.5.A.71,1471_1473.Daueor'lv, R', MrNrorn, L., Gorosmrn, M., Sraxrono, M. and.Kourcurn, J. (1g76) Field focussing nuclearmagnetic resonance (FoNAR) : visualization of a tumor in a tive anlmai. s ,in"", N. y. lg4,1430*1432.DmaaoraN, R., Goroswrn. M lld MTNKoFF, L. (19j7) NMR in ;;, ;ti FoNAR image of the rive' human body. physinl. Chem. phys. g, Sl _tOO.DeHr, R. E. and Hoevr,-c,-4. J (1969) Broad-rine NMR study of H2o and D2o in collagen fibers.J. chem. Phys. 50, 3245-3251.DEHL, R. E. (1920) collagen : mobile water content of frozen fibers. science, N. y. r7o,73g_,r3g.Dncrr' J' G' and PrNr.ln, ̂ M. M. (1975a) A possible improvement'in tne resotution of proton spinrelaxation for the study olc-an_cgr "i lo* fr.q,r*" y. J. natn.'canci l^t.is,-iis-lza.DroEL J' G' and pNrAR, M' M. (1975b) o.G; of the ;;;pon.",lirity of the water proton spinrelaxation in tissues. Biophys. "r. 15, g55_360.

-'- '"rvqvu!rs'!r

o*:ij'_fiu-d Dmrvsmnq w- (1974) NMR investigations of frozen porcine muscle. J. magn. Reson. 15,EozEs' H' T' and Su'rursrr, E-.-T.(1977) Cross relaxation and spin diffusion in the proton NMR of hydratedcoilagen. N atwe, Lonil. 265. 5il,_523.EozE!' H' T' and sa'uursr<1 E-.T (197s) The measurement of cross-relaxation effects in the proton NMR spin-iattice relaxation ofwater_in biologicat systems rhydrated "ollug* ;J;rrJ". ,,. magn. Reson.3l,2o7-22g.EccrmroN, J' c', senvfit,.r. A 1ol Hoi"n, n]. (1975) Nuilear *ugo"ti"'."rooance investigations ofhuman neoplastic and abnormar non neoplastic tissues. carce, Res. 35, 1326-1332.Etserinrnc, D' and K'cuzlrnNN, w. (1969) ini it*"i" *d properties of water.clarendon presg oxford.tttlrf-rtt.*d Pacrn4 r. i. trsiei Nriir*'rpi**r"rutioo'rtoalrJ oi nyor:r"a e,usttn. Bioporymers ls,

_""t'fli.'iY;ilililll;J;X1'fr$ilfl?a..(1e6s) High Resotution Nuclear Masnetic Resonance spectroscopy.

reanv' M' E' and ErspNsmor, M. (1975) water exchange between red cells and plasm.ai Measurements by,-- ^1l"lgut^-ugnetic relaxatio n. Biophy s. J. 15, 1 101 _1 1 10.T'KKAR' l ' c' and tsEcKER' E'-D.11971) Pulse and Fourie,r Transform NMR. Academic press, New york.rwcn' E' D' and Holrnn'-! ^?a (1i7q Pr;ta; "o"t"u. magnetic t".rou*"-i"ruxation measurements in_ togmuscle.Biopftys.J.l4,goj-921.'

,"s6uvrrv rvDu!4uu' rl

rrNcn' E' D' and ScnNnroT*, * s: (1975) Mobility of water bound to biological membranes. A proton NMR

,^^_."lu13ri9 study- Biochim. bioph)s. eio q06, ii'e:b+.rosrEn' K' F, RESING, H. A. and.-Gnnno*^",'A- N. (1976) Bounds on "bound water,': transverse nucleara_^Jtgl.f."^:"^royr4_ce.relaxation in barnacle muscle. ,science, N. y. lg4, lZi_lZi.' \^'�\N' r ' \r9 / 2-19 /5) Water: A Comprehensive Treatise, vors 1-5. plenum press, New york.r RANKS' F . (19'7 j) solvation interactions or p*t"i". io-rolution. p&i/. Trans. R. soi. B. zlg, g9_96..REY, H. E., KNrsrnr, R. R., KRr,,uv, 1., S"o*", a. n" THoMpsoN, R. T.;JFrNrA& M. M. (1972) proton

_ ;ili|}[l"" relaxation studies of non rnalignant tissues of ;r*";-;i"r J. natn. cancer Inst. 49,

r nrtz' o' G' Jn and swrFr, T. J. (1967) The state of water in polarized and depolarized frog nerves. A proton*,,.

rn?Bnetic resonance study. Biophys. J. 7,675_6g6.r uiLERr M' E' and Bnnv' w' S'-Jn (1968) Nuclear magnetlc resonance study of water sorbed on serumalbumin. J. biot. Chem. 243, 27 4_�2g0.

n normal and tumor 1a1

{igh resolution protonlfrom three mouie cell

es. Q.9,165-245. .-onance. "I. natn, Cancer

;troscopy. L T

nance studies of living

rce. Hatpet and Ro{i;

Spin echo studies oi

r32 R. MernLR-Dn Vxs

t ,

I

FuNc, B. M. and Tnetmra.{NN, P. (1971) Deuterium NMR and EPR of hydrated collagen fibers in 6,presence of salts. Biopolymers 10, 391-397.FuNg -B' M. and Wn, S. c. (1973) The effects of alkaii and aikaline earth salts on the structure ̂ r_ hydrated collagen fibers as studied by deuterium NMR. Biopolymers 12, 1053-1062.

--- ----'w olFuNg B. M. (1974) Non-freezabie water and spin-lattice relixaion time in muscle containing a grow.rng

tumor. Biochim. biophys. Acta 362,209-21,4.FuNo, B. M., Wrrscnrr, J. Jn and McAurs, L. L. (1974) The stare of water on hydrared collagen 4studied by pulsed NMR. Biopolymers. 13,lj67-1776.FuNc, B. M. and McGauoHv, T. W. (1974) The state of water in muscle as.studied by pulsed NMR. Bioc6ia

biophys. Acta 343, 663-673.Ftxc, B. M., Dunn.Av, D L. and {.LssIr", D. A. (1975a) The state of water ir bioiogical systems as studiedby proton and deuterium reiaxation. Biochim. biophys. Acta 399, 191_202.FuNc, B. M., wAssrl, D. A.. Dunr'or"r. D. L.. cnesNur, R. w., Dunneu, N. N. and BenrrN, K. D. (t97sr.i

water in normal muscle and muscle with tumor. Biochim. biophys. Acta3g5,1g0-1g7. I -' -' \r' 'JU,

FtrNc, B- M. (1977a) Proton and deuteron relaxation of musile water over wide ranges of resonancefrequencies. Biophys. J. 18, 235-239.

FuNc, B. M. (1977b) Carbon-l_3 and- proton magnetic resonance of mouse muscle. Bioplrys . J . lg,3l5-3Ig.FuNc, B. M. (1977c) Correlation of relaxation time with water content in muscle and"brain tirsrne.. nio.irir,,

biophy s. Acta 497, 317 -322.Grelnrr J. A. (1967) A study of water in biological systems by ttO magnetic resonance spectroscopy. IL^ Relaxation phenomena in pure water. proc. natn. eiaa. Sci., U.S.A. Sg,h43.Gusrr, J. A. (1970) Participation of water in conformationai changes of

'biopoiymers as studied by deuteron

magnetic resonance. J. Am. Chem. Soc. 92. 375_3gl-Gorosunrn' M.' KourcHr& J. and D.{MAorAN, R. (1978) NMR in cancer XI. Application of the NMR- ..malignancy index to human gastro-intestinal tumors. Cancer 41, lg3-Ig1.Gntiscrr, L- and Noncr, F. (1976) NMR relaxation investigation oi water mobility ia aqueous bovine serun

albumin solutions. Biochim. biophys. Acta 453,2L9-232.Harr,rNcn, K. and Kori'"rc, S. H. (1976) Protein rotational relaxation as studied by solvent 1H and 2H ,

magnetic relaxation. Bio chernistrv 15. 425 5-4263.HeNseN. J. R. and L.r,wsoN, f. f5. 1tezol Magnetic relaxarion in ordered systems. Nature, Lond. 211-,542-543.HaNsnN, J. R. (1971) Pulsed NMR study of water mobility in muscle and brain tissue. Narare, Lond. 230,482-486.H,lzrnwooo, C. F., NrcHors, B. L. and CuelrnrnrarN, N. F. (1969) Evidence for the existence of a minimum- of two phases of ordered water in skeletal muscle. Nature , Lond.222,747-750.H,qzlrwooo, C. F.. NrcHorg B. L., CuaNc. D. C. and Bnowru, B. (197i) On the stare of water in developins

muscle: a study of the major phase of ordered water in skeletal muscle and its relationship t" ;"d1";concentration. J ohns Hopkins Med. J. l2f. Il7-131.Hrzrnwooo, C. F., CrBvEL,tro, G. and MnorN,c,, D. (1974a) Relationship between hydration and proton

nuclear magnetic resonance relaxation times in tissues of tumor-bearing and non i"*o.-U"-iog'roi..,implications for cancer detection. J . natn. C ancer I nst. 52, 1 g49_ 1 g53.

Hazr,rwooo, C' F', CneNo, D. C., NrcHor,s, B. L. and WorssNR, D. E. (1g74b) Nuclear magnetic resonancetransverse relaxation times of water protons in skeletal muscle. Bioph.jt. "1. 14 ss:_ooo.

HILroN, B' D, Hsl E. and BnvlNr, R. G. (1977) IH nuclear magneti" .".orrun.. rejaxation of water or-lysozyme powders. J. Am. Chem. Soc.99, g4g3-g490.HoLLIs, D. P., Sanv.c.N, L. A" EccrrsroN, D. and Monrus, H. P. (1975) Nuclear magnetic resonanc€studies of cancer. vI- Relationship among spin-lattice relaxation times, growtn rate and water content ol

Morris hepatomas. J. natn. Cancer Inst. 54.7469_1472.Hourr, D' I. (1977) Zeagmatogtaphy: a criticism of the concepts of a selective pulse in the presence of a fielclgradient. J. rnagn. Reson. 26, 165-167.Hurcr+rvsoN, J. M. s., surnrnr.aNo, R., Merleno, J. \.3nd FosrER, M. A. (r97j) In NMR in Biobgy (eds

R. A. DWEK; I. D. Ceurnnr,r,, R. E. Rrcnnnos and R. J. P. WLuerrls), fp. 36g-369. Academic press,London.

Imcrr, w. R., Mccnsprn, J. A., KNlspnr, R. R., TnouesoN, R. T. and hrren, M. M. (1974) water contentard proton spin relaxation time for neoplastic and non-neoplastic tissues from mrce and humans. J. rctr.C ancer Inst. 82, 353-356.

JAcoBsEN, B., ANornsoN, W. A and Ar.Nor-o, T. (1954) A proton magnetic resonance study of the hydrationof deoxyribonucleic acid. N ature, Lond. 173. 7i2_77 3.JrvBs'-T. L. and GLLEN,K.T. \1972) NMR relaxation time and seif-diffusion constant of water in hen egg

white aad yolk Biochim. biophys. Acta2g6,10_15.Kerr, A' and BpnrNosp*, -4.' J. C. (1976) Proton magnetic relaxation and sprn diffirsion in proteins.

J . magn. Reson- 24, 343-366.Krv.rN.r,u, J-L. (1964) water anr solute-water Interactions. Horden-Day, London.Krosg,G' and SterzNrn, F. (1974) NMR investigations of the interaction of water with lecithin in benzenesolutions. Biochim. biophys. Acta 363, l_g.KNrsrer, R R., TsonrpsoN, ft. f. and hNra& M. M. (1974) Dispersion of proton spinJattice relaxation intissues. J. magn. Reson.14,44-5I.KorNrc, S' H. and Scnrr,rru-d-n-r, W. E. (1969) Nuclear magnetic relaxation dispersion in protein solutions.J. biol. Chem. 2U, 3283-3289.KoeNro- SJI., Ear-reNrcl, K. and SHronrn, M. (1975) Protein-water interaction studied by soivent lH, 2H

. and''O magnetic relaxation. proc. natn. Acad. Sci. U.S.A. 72, 2667_267I.KorNro, S' H., BxvlNr' R. G., HaI-LrNo.t, K. and Jacon, G. S. (1978) Magnetic cross-relaxation amongprotons in protein solutions. Biochemistrv 17. 434g4359.li

;i:F r -

li

'drated collagen fibers in the

'th salts on tle structure oft-1062. :

muscle containing a growing

ter on hydrated collagen n5

.ied by pulsed NMR. Bjoclrjrrr.

biological systems as studied

. and^Brnr.n, K. D. (1975bj80-187.r wide ranges of resonance

Biophys. J. lg, 315-l:lg.e and brain tissues. Biocftin.

resonance spectroscopy. Il

mers as studied by deuteron

.. Application of the NMR

ty in aqueous bovine serum'

ied by solvent 1H and 2H

tstems. Nature, Lond. 225,

tissue. N,ature, Lond. 230,

.he existence of a mi.i6u;

tate of water- in developinglls retauonshtp to sodium

een hydration and protoiinon tumor-bearing mice:

luclear magnetic resonanc683-606..ce relaxation of water ol

clear magnetic r"sonao"erate and water content oi

: in the presence of a field

The NMR studies of water in biologicai systems n3

KourcnsR,.J' A., Gorosrvru", M. and .D,c,MAoreN, R. (197g) NMR in cancer X. A malignancy index todiscriminate normai-and cancerous ttssus. Cancer 41. 174_'lg}.

Knruen, +"yt"l' D. and EnNsr, R. R. (1975) NMR Fourier zeugmatography. J. magn. Reson.1g,69-g3.KuNrz, I. D: lllll"o_l?, T. S., Lrw, G. D. and puncwr, -c.

v.irdodl nya.utroo of macromorecures.Science, N. X 163, 1329-1331.

KuNrz, I. D' (1971a) Hydration of macromolecules. III. Hydration of polypeptides. J. Am. Chem. Soc. 93,514-51 6.KuNrz, I. D.(1971b) Hydration of macromolecules. IV. Polypeptide conformation in frozen solution. J. Am.Chem. Soc.93, 516-518.KuNrz, I. D' and KeuzuaNN, W. (1974) Hydration of proteins and polypeptides. Adu. protein Chem. Zg,239-345.LArARrEr, R. (1972) Iteraction of radiation.energy with_nucleic acids. curr. Top. Radiat.Res. Q. g, 1_3g.LAUTERBUR, P' c' (1973) Image formation by induced local interactio"r, "*uiipt.r employirfi NMR. Norrr",Lond. 242, 190-19L.Laurrruun, \ c: (1.9]-4) vtugT".tl" resonance zeugmatograph y. pure Appr. chem. 40, 149-157.LAUTERTUR, P' c. (1977) spatially-resolved studies of whoie iirro", o.luos *J orj*ir*. by NMR zeugmato-graphy. In NMR in Biology ("-dr. R. A. Dwnx, I. D. c,u,rarrr,, fr. B. n cr*'*"s and it. r. e. wrr?rius),pp. 323-335. Academic Press, London.LrNDsrRolr4

-T. R. and Konuc, S. H. (1974).Magneljs-6.16 dependent water proton spir-lattice relaxation' rates of hemogiobin solutions and whole blo oi. J . *ogr. Resin. 15,344_�3s3.Lmosrnol4 T. R., Konrrrc,-S-.H., Boussros, T. and BiBnrr,ng J. F. (1976) Intermoleculax interactions ofoxlceaFqggkle hemoglobin molecules in cells and cell-free sotutioit. ni'piyr.7.lil#-A;:'**^

-

LrNc,,G' ]\r' (1962) ln A Physical rheory of the Liuing stctte: The associationitiiuaion Hypothesis.Blaisdeil,New York.

-,,.!t""j..,9:-*,9?92..1_"*-mo{el fgr the living cell: a summary of the theory and recent experimentalevrctence m lts support. Int. Rers. Cytol.26, l-61.' Luses, B' and W[czor,- T (1966) spin-echo technique study of the non-rotational hydration of deoxy-ribonucleic acid. Biochim biophys. Actal20, A7-4*.L*{1 B-,and Wrrczor"'t. (196V} Thermal transition of DNA measured by NMR spin-echo technigue.Bi opo lymer s 5, 9 67 -97 4.

: Lunrs, B' and Wrczor, -T. ,(1971) NMR study on molecular mobility of DNA moiecules in solution.Biopo lymer s 10, 1267 - 127 6.

' MrNsrrnr,o, P., Meuosr,rv, A. A. and BerNns, T. (1976) Fast scan proton density rmaging by NMR. J. phys. E.9,27L-278.: Mlrlu1-D-r V.ni, R., BsnrINcHeMPs, A. J. and BEnrNosnN, H. J. c. (1g76) The effect of }-irradiation onthe hydration characteristics ol DNA and polynucleotides. I. An Ntran study of aor""'rrro-uoJ-orosolutions. Radiat. Res. 68. I9'7 -214.

MArrrun-Dr Vnn, R' a-nd bmrwcnelups, A. J. (1977a)_-The effects of y-irradiation on the hydrationcharacteristics of DNA- and polynucleotides. IL An NMR rtody "r -*a H2olD2o r.or.o ,6tu-tioor.Radia. Res. 72, 181-189..'MArEEUn-Dn vrg,-R. and Bmmscnaups, A. J. (1g'77b) Theeffects of y-irradiation on the hydration char-acteristics of DNA and polynucleotides. III. A comparative NMR study of f.o".o *o riq"iJ ,orutioor._ , Rad.iat. Enuiron. Biophys.14, 31 1-315.MEDINA, D., HazrrwooD, c. F., cr,wrr,,rNo, G. G, craNc, D..c., Srrul H. J. and Movrns, R. (1975)l-{y9leat magnetic resonance studies on human breast dysplasias'uJ?optur*, . J. natn. cancer Inst.54 .813_818.Mrc'crrnrsrN, C. and Bsnnwnsnx, H J c. (1967) Deuteron magnetic resonance on hydrated collagen. InMagnetic Resonance and Relaxation (ed. Ii. BLiNc), pp.76l-7{6. North-Holland, Amsterdam.Mrccurr.srN, c., BrnnNosrN, H. J. c. and Ruppnrcni A. lrsos; Hydrutio;-;ibNA. comparison of nuclearmagnetic resonance results for oriented DNA in fte e, s ua C ioti- i . ,.tic. Biol. 37 , 235-237.MrccnnrsEN, c. and BrnnNosuf:-B- J. C. (1973) Proton exchange and molecular orientation of water rnhvdrated collasen fibers. An_NMR stuoy of H2'o and D2o. l. "Ei.inii.-ii,296-205.ourrnro, R' r. aid Grono4 E..P-(19T;)a ou.t"". magnetic o"roouni" study of hydrated systems usingthe frequencv deoendence ofthe reraxation processes. Biphys. J.13,g3-96.o*ii?rl tra-and GeoncE E. P. (i973bid"t"r-""a ions in musclSs anJmodel systems. Biophys. J.13,PACKER,. K' J. (1973) The effects of dift-rsion through locally inhomogeneous magnetic fields on transversenuclear spin relaxation.rn heterogeneous systeis. Proton transverie relaxation in striated muscle tissue.* J. magn. Reson.9,43g-443.rAcKE& K. J. fi97T The dynamics.of water in heterogereous systems. phit. Trans.R. Soc. B. 27g,5g-&7.RleEronsr, G. P.,'Knuuv,-J. and.P-rNr.4R, rra. rra. irszsl Nuclear magnetic resonance study of mammalian* cell water. Biophys. J.15,391_402.KANADE, S. s., cruucnur,r,.s_._R., KAsruRr, S. R, N,cDrARNr, J. s., TAL*ALKAR, G. v-, w.^cH, U. v'- KoncloNren, K. s. and vrrev.l*oo*,ouo*, i. 1rsz9 Pulsed nuclear magnetic resonance studies onhuman malisnant tissues and ceiis in uitro. Indian J. Bio'chem. Biophys. D,t2g-232.Rrrrcovr, s. anisIrq,q,orNourc, I trszTj N;"i.- *"g""ri" ,"roo*".^.todies on the thl,roid giand. II. on theo""j,lr?:f ifu.le water_in normal thyroid gland,. Stidia Biophys.63,25_39.^NI{G'-tl' A', FosrE& K. R. and Gannomv, A. N. (1977) :'Bound water" in barnacle muscle as. indicated ino,.__ou.l"ur mag.netic resonance studies. Science, N. y. 19g; 1lg0_11g2.^"t

t::T;.o (1966) Hvdration of DNA, a wide line NMR study of oriented DNA. /ac Chem. scand.20,

satvaN, L. A., r{or,i,rs, D. P., EcoNouou, J. S. and EGGLESToN, J. c. (1974)Nuclear magnetic resonance studiesof cancer' rv. Correlation of water content with tissue relaxation times. J . natn. cancer Inst. 52. 599-602.

) In NMR in Bioloav Gds't68-369. Academic- press. I

M. (1974) Water content.nice and bumans. J. naii

ce study of the hydration

rtant of water in hen egg

in diffi-rsion in proteins;

with lecithin in benzene

spin-lattice relaxation in

ion in protein solutions.

rdied by solvent 1II, 2H

cross-relaxation among

134 R. M.qrHun-Dr VnE

Scu.qnl, M., SnNrrunc, M., Aurnsrrnc, M. and Goroun, R. (1974) Characterizatiol of malignant thyr6i6gland tissue by magnetic resonance methods. Br' J . Cancer 29, 483-486.

Snrw, D. (1976\ Fourier Trarrform NMR Spectroscopy. Elsevier, Amsterdam.Snpornn, M. and CrveN, M. M. (1975) NMR study of 17O form H2170 in human erythrocytes. Biochim. biophys.

Acta 385,8l-87.Suponrr., M., Ha.qs, M. and CrveN, M. M, (1976) Pulsed nuclear magnetic resonancs study of 17O froq

H2170 in rat lymphocytes. Biophys. "I. 16' 601-610'SrNeorNovrc, J., R,lrrovri, S. KnArNd,q.Nrc. M. and Joverr'ri, M. (L977) Relationship of biochemical a14

morphological changes in rat-thyroid and proton spin-relaxation of the tissue watet. Endokrinologis69, 55-66.

Swrrr, T. J. and Bnnn, E. M. (1973) An oxygen magnetic-resonance study of wator in frog skeletal muscle.Ann. N. Y. Acad. Sci. 204, 19f-196.

Svrts, B. D., Hurq W. E. and Sxvorn, G. H. (1978) Experimental evidence for the role of cross-relaxation i1proton nuclear magnetic resonance spin lattice relaxation time measurements in proteins. Biophys. J.21,13'7-146.

SzBur-GvdncvI, A. (1957) Bioenergetics. Academic Press, New York.TxrER, J. (1973) Nucleic acid-water interactions. Prog. Biophys. molec. Biol. 33' 83-97 -T11orr,rpsoN, R. T., KNrspg, R. R. and PrNrln, M. M. (1973) A study of the proton exchange in tissue water by

spin relaxation in the rotating frame. Chem. Phys. Lett.22,335-337.TgornrpsoN, B. C., W.q.TERMAN, M. R. and Correu, G. L. (1975) Evaluation of the water environments in

deoxygenated sickle ceils by longitudinal and transverse water proton relaxation rates. Archs Biochea.

biophys.166' 193-200.Uorrr, J. N., Arv,4nrz, L. A., NIcnom, B. L. and Hezrrwooo, C. F. (1975) The effects of cholera

enterotoxin on intestinal tissue water as measured by nuclear magnetic resonance spectroscopy. Physiol.Chem. Phys.7. 533-539.

We6rsrsy, R. H. and Snronrn, M. (1978) Surface induced NMR li:re splittings and augmented relaxationrates in water. J . chem. Phys. 68,2584-2590.

Wnrrr, J. P., KuNrz, I. D. and CaNron, C. R. (1972) Studies on the hydration of Escherichia coli ribosomesby nuclear magnetic resonance. J . molec. Biol. 64' 51 1-514. ,

Zrpp, A., J.cMrs, T. L., KuNrz, I. D. and SuoHnr, S. B. (1976) Water proton magnetic resonance studies ofnormalandsickleerythrocytes-Temperatureandvolumedependence. Biochim.biophys.Acta473,29l-303.

i:1,ll l; i1 j

: l, i

Ir i

infel

oflabrti(ceti(rh