Tobacco Documents Online

v THE EFFECTS OF TOBACCO SMOKE ON LUNG TISSUE AS MEASURED BY ELECTRON SPIN RESONANCE John R. Rowlands, David G. Cad'ena, Jr. and Arthur L. Gross* Southwest Research Institute . .. 8500 Culebra Road San Antonio, Texas This work was supported by the National Cancer Institute, National In:,titutes'of Health under Contract No. PI-I43-65-100. *Present Address Pico- Laboratories, Incorporated 8222 Broadway San Antonio, Texas 0

I. INTRODUCTION The presence of free radicals in tobacco smoke has been previously reported. During the course of a program concerned with free radicals ; and alkylating agents in tobacco smoke it was observed that the concen- tration oi free radicals in tobacco smoke condensate varied considerably as a function of methods used for collecting and subsequent treatment with solvents. In order to determine the biological effects of free radicals it became apparent that a method was needed whereby smoke could be directly applied to lunb tissue without first colltcting and processing the condensate. The method reported herein enabled lung tissue to be directly exposed to cigarettc smoke in a manner similar to normal smoking and resulted in the observation of distinct changes in the lung tissue as measured by electron spin resonance. •

II. EXPERIMEN'TAL Large adult rabbits of unkown strain and obtained locally were used. Each animal was sacrificed by cervical dislocation and the lungs and the trachea were excised intact as rapidly as possible. This was accomplished within ten minutes after the animals were sacrificed. The trachea was then attached to the apparatus shown in Figure 1 using thin twine to hold it in place. The smoking apparatus (Figure 1) consists of a bell jar, valving, and a Phipps and Bird smoking machine that was modified to partially exhaust the bell jar, hold it in the exhausted state, and then admit atmospheric air into the jar. The timing of the cycle is controlled by the cam and motor arrangement contained in the smoking machine. The smoking parameters used were: puff frequency ---------------------30 seconds puff duration----------------------- 2 seconds inhalation time------------------- - 4 seconds butt length-- ---- ----------------- 15 mm The puff volume varied with the capacity of the lungs during each experiment. The volume was adjusted so that the lungs would fully inflate without rupturing. During the initial phases of this work several lungs did rupture thereby allowing smoke to enter the bell jar. This was later prevented by - first using the mininial puff volume that would inflate the lungs and step-wise increasing the voiur.-±e until the lungs would inflate with a turgid appearance. King size, non-filtered cigarettes, manu actured in the United States were 10Q3546928

~ ~ CIGARET,T:E: ;~ b h1ORMAL-~ ::: I I ::) NIO~-M?.LLY ~ LY i CLOSE.. ~ ..._.__~t=V E CLOSED c i VALUE _11 ~ i I L , .~~ _ FIGURE 1. SMOKING APPA tiATUS

used. The cigarettes were not preconditioned prior to use.. They were inserted into a holder composed of rubber tubing and were ignited by means of a lit cigarette of the same brand. A total of six cigarettes were smoked into the lungs for each experiment. As soon as the smoking was completed the lungs were removed from the apparatus and homogenized at room temperature for thirty seconds in an ©mnimixer. A cylindrical sample of the homogenate was then prepared and placed into a Varian liquid nitrogen dewar. Electron spin resonance spectra . of the sample were then recorded at 77 0 K using a Varian V 4500 X band spectrometer using 100 mc/sec modulation. A typical spectrum is shown in Figure 2 together with the spectrum obtained from an unsmoked lung using the same instrument parameters, and sample size. The observed spectrum was reasonably reproducible although small changes in the relative magnitudes of the three line pattern superimposed~ on the broad resonance P line were noted from sample to sample indicating that the observed pattern is composite. A considerable amount of additional data is required before • we can hope to arrive at any definite conclusions about the species giving rise to the signal. However, the similar pattern that Swartzob-erved in irradiated blood suggests that the observed spectrum may be produced by the interaction of the free radicals in tobacco smoke with the red blood Q . O cells. This is being further investigated on further lung experiments and ~ on a series of model systerns including metal porphyrin complexes. In ~ Fi;ure 3 we have included the spectrum obtained from freeze dried smoke ~ exposed lung tissue. Figure 3A represents the spectrum obtained at 77 °i{,

B FIGURE 2. (A) ESR'OF LUNG WHICH HAS SMO1'11D SIX CIGARETTES (B) EPR OF CONTROL

and Figure 3B the same sample run at room temperature. The room - temperature signal consists of a single narrow line which is typical of the electron spin resonance signal obtained from tobacco sinoke condensate. On refreezinb to T?°K the electron spin resonance signal observed reverts measurement at low temperature is consistent with it being due to a paramagnetic metal complex resonance, which becomes saturated and to that shown in Figure 3A. The reappearance of the broad signal upon hence not detected at room temperature. Conclusions ~ It is. apparent from this work that the smoking system that is reported has proven to be a reliable method for exposing intact lung tissue to fresh cigarette smoke comparab2e to the manner of' human exposure. The results are not biased by changes in the smoke that occur as a result of trapping and processing. We cannot at this time be certain of the origin of our observed electron spin resonance signal. However, due to the similarity of it to that ob ! served by Swartz upon irradiation of red blood cells we anticipate that both his and our observations may be explained by free radical attack of the haernoglobin molecule at the sixth coordination position leading to a covalent hexacoordinated complex. N ~ ~ W L1"1 ~ ~ • ~ ~ _:.~ ,.:.....: .._ _.~.::.::.,.,4.:..,

FIGURE 3A. EPR OF VACUIUVM L~TF %~-UO_~=JL L"u NG AT 770X . . .. .~.a=.....j . .. . . ..:.. _ . _. . . .. f.S:! .~ .._ . ~ _

[j~Viy~ .,:\ o FIGURF, 3B. EPR OF VACUUM L`RTED LUZG AT ROOM ' Ai'URE .

APPENDIX A. 0

PROPOSED PROGRAM 1. To establish that the origin of our observed electron spin resonance signal' is in fact as we have outlined. selves to the investigation of representative examples of these common contain many constituents in common, at this s.tage we will restrict our- study we will restrict ourselves to compounds which are normally found in tobacco smoke, smog, and automobile exhaust. Since all three systems served from the reaction of tobacco smoke with haemoglobin. In this . To determine the types of molecules and free radicals that will react with haemoglobin to produce the type of complex we have ob- 3. Having established the necessary requirements for the pro- duction of such a complex, to investigate the stability of the complex, i. e. , is the complex once formed stable or is the decomposition of the complex to reform undamaged haemoglobin a probable reaction? 4. To determine what effect the formation of such a complex has on the general reactivity of the haemoglobin molecule. Since the bonding of the iron atom to both the porphyrin ring and the globin molecule changes upon formatio.n of such a complex this is accompanied by a change in electron distribution throughout the whole molecule. Such a change in electron O 0 distribution, will necessarily cause changes in the chemical reactivities of W ~ different areas of this complex molecule. ~ ~ W ~

lung into which had been smoked six popular bra.r,d cigarettes. A short pre- paramagnetic resonance signal from anhomogenate of a freshly excised rabbit • print of a publication,.Appendix A, which outlines these results is included with this proposal supplement. In our proposal #5-4727, I indicated that I had observcd an electron In a telephone conversation I was asked if I could give some additional information to strengthen~ my grant application. I believe the foilowing dis- cussion will supply the necessary information. As indicated in Appendix A, I believe the observed electron para- magnetic resonance pattern is-produced by the interaction of free radicals and/or reactive diamagnetic molecules with a metal complex present in the rabbit lungs. To be more specific I believe the bulk of the observed signal is due to the reaction of these free radicals and/or reactive molecules with the haemoglobin molecules present in the red blood cells. Such an, interaction is represented in Figure 1. In the haemoglobin molecule the iron is present as a ferric ionic 3+ complex i. e. Fe . The paramag.netic resonance spectra of Fe3+ ionic complexes have been investigated by many workers. In particular however, detailed electron spin resonance studies have been performed on four derivatives of the haemoglobin molecule, in order to discover the detailed form of bonding between the central iron atom and the particular group on the sixth coordination point, the point at which oxygen attachment occurs. The tentative explanation of the reported ESR signal from smoked rabbit lung is the attachment of an active molecule (s) from tobacco smoke to the selfsame sixth coordination point. On the basis of this hypothesis several soo3s4ss37

• I %

, extremely important questions arise: 2. Why does the lung membrane pass these molecules so readily? 3. What molecules or radicals are thus passed and picked up 4. Is there a connection between these modified ha emoglobin . molecules and the observed vessel constriction in smokers? Since none of these questions lie within the scope of our present NCI program (primarily because of the unexpected nature of this discovery), TRC support is essential to allow the investigation of any or all of these. Before we attempt an exploration of the more complex biological questions raised above, there are several rather direct questions which must first be dealt with; for example: 1. What types of mol'ecules and/or radicals will react to form haemoglobin complexes such as we have observed'? 2. What is the stability of these complexes at physiological conditions? 3. Are such complexes formed by other atmospheric contaminents;i. e. , smog, automobile exhaust fumes, etc. ? P

dicular to the axis joining the iron atom to the globin molecule and the ligand is represented in Fig. 1 where the haem-porphyrin plane is shown perpen- at the sixth coordinatio.n point. The main feature of the haemoglobin molecule is the relative ease with which the ligand of the sixth~ coordination point can be changed which makes it feasible to study a whole s ries of derivatives formed by substituting different ligands into this position. THEORETICAL CONSIDERATIONS The structure of the central portion of the haemoglobin molecule Single crystal studies-have been made on the acid methaemoglobin, acid metmyoglobin, metmyoglobin fluoride and metmyoglobin azide. The extreme g factors for each derivative are listed.in Table I. TABLE I • Derivative Group attached . gCl µ . in Bohr magnetons Acid methaemoglobin . H2O 2.0 6. 0 5.84 - _ ~ Acid metmyoglobin H20 2.0 6. 0 5.85 O Metmyoglobin fluoride F 2.0 6. 0 5.92 CJ ~ Metmyoglobin azide N 2.8 1.70 2.84 ~ 3 ~ ~ The parallel suffix refers to directions parallel to the axis through the iron atorn~ normal to the haem plane and the perpendicular suffix to directions at right angles to this axis (i. e. , in the haem plane). It can be seen from Table I that the four derivatives so far studied are divided into two classes by both

the resonance a.nd susceptibility,measureme.nts. As a result of the suscepti- . bility data these two groups were classified as "essentially ionic" and "essentially covalent" corresponding to the cases of a spin quantum number = 5/2 or 1/2. Thus, _if ionic binding is assumed, the central positively charged ferric ion will be left with five 3d electrons. These will line up with their spins parallel according*to Hund's rule and so half fill each of the five vacant 3d orbitals. A total electronic spi.n moment of 5/2 is thus formed which will itself be quantized in any applied field. On the other hand, if covalent bonding takes place six pairs of electrons are required to form with the surrounding ligands. These electrons donated by the surrounding atoms will occupy certain orbitals of the ferric iron, and it can be shown that the most stable configuration for octahedral bonding is the one in which two 3d orbitals and the one 4s orbital and three 4p orbitals are employed. The five non bonding 3d electrons of the ferric iron are therfore, left with only three 3d orbitals to occupy and hence four are forced to pair leavi.ng one unpaired and a resultant S= 1/2. If this simple interpretation were correct the resonance spectrum obtained from the essentially ionic compounds woul& be expected to have the same general features as that of the other ionic ferric salts. Thus, the total spin and associated magnetic moment corresponding to the configuration with S = 5/2 would~ take up six different orientations varying from +5/2 to -5/2. Transitions can then be induced between each of these levels according to the selection rule NS=±1 and thus five electronic absorption lines would be expected centered on a g factor of 2. 0. However, the work on single crystals of 1003546941

acid methaemoglobin, metmyoglobin and metrnyoglobin fluoride showed that the g factor was in fact, anisotropic varying from a gf/ of 2. 0 to a - g j. of 6. 0, and that there was only one electronic transition observable. A theoretical explanation of the observed anisotropic values has been : provided by assuming that very strong asymmetrical crystalline fields different orientations of the total spin vector of 5/2 are much larger than are active on the iron atoms so that the energy difference between the six the microwave quanta. This highly anisotropic g factor of the "essentially ionic'"' haemo- 0 globin derivatives xnake s the detection of the paramagnetic resonance signal of amorphous samples difficult, since the resonance line extends from approximately 1300 gauss to 3250 gauss. This of course, accounts for the each of any apparent signal from the sample of untreated rabbit lu.ngs. In contrast to the large variation from g = 2 to g = 6 observed for the ionic derivatives the azide derivative which is "essentially covalent'T' has a g factor spread across the free spin value, indicating an S of 1/2 with considerable orbital interactions. The signal we observe from rabbit lungs which have been treated~ with tobacco smoke is also located~ at approximately the free spin value. Consequently, we have tentatively attributed at least part of our observed signal to "essentially covalent"' derivatives of haemo- globin being formed by reactio.n of the normal haemoglobin with constituents of the tobacco smoke. So far the above discussion of the nature of the resonance is hypothesis based upon what is known about iron in other complexes and has 1003546942

to be verified experimentally. However, assuming the hypothesis is correct, such a complex provides a vehicle for carrying the chemicals present in heW ~ crn~OJ'n3s. cigarette smoke to the liver for the production of fo-r_e-ce:ntpl , le. •

SUBMISSION! OF REPORTS A bi-annual report will be submitted' to the Tobacco Research Council throughout the course of this program, with a final report summarizing the entire program. It is anticipated that much of this work will be appropriate for publication in the open Scientific Literature, but prior to publication the Sponsor requested to review the manuscript and give his consent to the publication. 41

REFERENCES Ingram, Free Radicals as Studied by Electron Spin Resonance, Academic Press, (1958). Swartz, et a1 , Biochem. and Biophysical Res. Commun. 21:61 (1965).

rank C. Whitmore Deputy Director Department of Physical and Biological Sciences e Approved: Approved: A. C. Hulen, Treasurer ._ J6hn R. Rowla.nds Senior Research Scientist