Tobacco Documents Online

4P02 4P02 5 minutes FtG. 5. Effects of increasing pCO at constant pO. afler detectable oxygen uptake was completely inhibited riih 1.5 mif NaCN- t4ote a similar retaiaiion of isometric tension as occurred in experimenis shown in Fig. 4 in the absence of NaCN. Oxygen tension-dependeni mechanical tension persisls following NaCN plus CO. mechanism of CO toxicity, at least in vascular tissue. Second, the finding that CO relaxed aorta smooth muscle under conditions where CN inhibited electron chain transport in the mitochondria suggests a reaction of CO with an intracellutar enzyme other than cytochrome oxidase. This compound is probably not myoglo- bin since it is absent, or present only in very small quantities, in this tissue. The above data do not fit very well with previous estimates of relative affinities of Oi and CO for binding to cytochrome oxidase [discussed above (3, 28, 31)]. We do not know the mean CO/O; ratio in our tissue where there was probably a largepO. gradient between bath and tissue core. At organ bathpO, values less than 100 mm Hg, it is likely that core pOz is nearly zero and, therefore, mean tissue pO_ is less than 50 mm Hg. CO/Op is at least 10, whereas V0z is depressed only 30 to 30:E of control. MYOGLOBIN It has been demonstrated that there is significant binding of CO to myoglobin in nine skeletal muscle as well as in myocardium, even at HbCO as low as 0.5 to .1.•_ I f rif1C~~"~If

RONALD F. COBURN SUBJECT' B.JC. 1.5 MIN EXERCISE - I DAY 2- I DAY 3 I I 1 SUBJECT" D.G 7 MIN EXERCISE .5 DAY I. ~ C - 1 0 I 0 I ~ tu (3) t YS -51 tz) ~ 8 I ,t -1o f DAY 2 (2) (n -IS Fi0= 21% i B FiD: 13-14 % I . -20 ( ) Sepuence I - _ -- FIO. 9. CO shifts during maximal exercise in a normal human subject. CO shifts out of blood are given in terms of perceniage chance of blood CO. Exercise ~'.as performed on a bicycle ergometnr unlil ma.eimal oxygen uptake w'as achieved. After at least 1.5 min at maximal oxygen uptake venous blood was sampled and analyzed for HbCO percentage saturation. This figure shows that in this subject ihere were large shifts of CO out of blood during maximal ox}gcn uptake when the subject uas brealhing 13 a Or and only small shifts when the subject exercised to maximal oxygen uptake breathing ?1.^r oxygen (7). Reprinted with permission of Ihe publisher, from Ref. (7). I decreased in blood perfusing peripheral tissues. The lethal effect of inhaled CO is explained by the presence-of CO dissolved in plasma .vhich then has an adverse effect on intracellular enzymes. rI L. ~Cop~r~~'nr . - l0

I > i°i'F _. -- WORKSHOP: CARBON MONOXIDE :\ND CVD 100°!° CO Hb_~ Equilibrium 100°/° CO Hb E quilibrium 100°l° 02Hb Photodissociation 1 - 2•5 sec ~ 100 msec Fto. 10. Rate of reaction of CO with human red blood cells. Study was performed in a stop flow apparatus (19). Tracing is a spectropho9ometric reading showing rate of formation of carbost hemo¢lo bin uhen CO is mixed with red blood cellscontaining ox)'hemoglobin. Equilibrium occurs uiihin a few - hundred milliseconds. Reprinted with permission of the author and the publisher. from RcL (19). The data published by Goldbaum et a7. can be criticized as follo«s: (al The concentration of inspired CO, i.e., 139c, is expected to saturate blood leaving the lung, nearly completely, for a few seconds after inspiration of this high concentra- tion. Abboud el al. (I) had anesthetized dogs inspire 5 cl'c CO which resulted in peak values in aortic blood of 85-95 0. Thus death in Goldbaum's dogs could have been a result of the sequelae of severe anoxia prior to mixing inspired CO in the entire body CO stores. (b) There are data in the literature which report sur- vival of animals given CO very slowly at low concentrations in amounts sufficient to cause increases in HbCO as high as reported by Goldbaum er al. (24). (c) Death is a poor index and many desirable measurements were absent in Goldbaum's study which might provide insight into the cause of death, i.e., blood gas tensions. blood pressure. In our opinion the concept of rapid equilibrium bet~.een blood pCO and HbCO seems to be on solid ground. Figure 10 shows the rapidity of equilibrium when CO is suddenly mixed with red blood cells. There is evidence of near equilibrium in pulmonary and placental capillary blood under conditions where there is transport of CO into alveolar gas or fetal circulation (8, 22, 23). The finding that equations which assume "equilibrium" in pulmonary capillary blood can accura,ely-predict rates of CO excretion via the lungs supports equilibrium (8). The carbon monoxide diffusing capacity has been shown to be the same whether measuring CO uptake or CO excretion (during breath holding) supporting equilibrium (22). Longo,.Power, and Forster (23) have considered theoretical aspects of determinants ofpCO in capillaries of the gas-exchange tissues, lung and placenta. _ --- ----- -- ~o('J-- -- ----- - .. , - QD N N N - ~ r7

WORKSHOP. CARBON MONOXIDE AND CVD 4DOr [G:RRQ 300 V02 l / __~ pcc I -PtJ(9 r N)-200 / IOC~ %~~ /1020 S'1 60 BO i00 200 300 anaa v*w v,~ n Flaa 6. Effect of CO on relationship of organ bath pO, and ox)gcn uptake 11'O_I of rabbit aorta. Control data were obtained in Ihe absence of CO. CO caused a 30-35.^. reduction in VO- over the entire range ororgan bathpO, studied. Data from a typical experiment. Uean values from six experi. mems have been previously published (12). - 19c saturation (10, 11). Figure 7 shows measured carboxymyoglobin (61bCO) as a function of arterial p0: at the time of muscle biopsy. Total "CO or "CO was measured in biopsies and corrected for CO bound to hemoglobin in the specimen. In these tissues there is 20 to 100 times more myoglobin than cytochrome oxidase making it unlikely that CO binding to other hemoproteins is influencing the con- clusion that CO is binding to myoglobin. Furthermore, the same approaches using liver failed to provide evidence of intracellular CO binding at normal HbCO per- centage saturations. This figure shows a ratio of %IbCO/HbCO of approximately unity in both of these tissues at normal p,O:. This ratio remained constant until paO, fell below 40 mm Hg. It is also known that this ratio remains constant as HbCO was increased in a ramp manner over 60 min to levels slightly in excess of 509'c, but at higher HbCO more CO bound in tissues than to hemoglobin ('_4). Whether CO binding to myoglobin can be a mechanism of CO toxicity depends on the function of myoglobin and how critical myoglobin is to cell oxygenation. C1NDU0 • S.NTORIUS- 2.0 CO-s ~ COnn LS - 1.0 100 200 300 400 I _ _ ARTER'.LL Ol TENSION (NY NQ) - _ - - I FIG. 7. Effect of altering arterial pO_ on the ratio of CO binding to myoglobin to CO binding lo hemoglobin. Data obtained in an aneslhetized dog wfiere arterial pO. chances were caused by allering pO: of inspired gas. Carboxymyoglobin uas determined in muscle biopsy specimens- Data previuu.ly published (10. 11). r

WORKSHOP; CARBON MONOXIDE AND CVD hypoxem.ia. {. 1" mnr meiabolism, distribution and action of hexobarhiial.J. Phurrrrnr.,l. E.p. Tfrer. 199, 53-!A ( 1976). - 27. Tamura, N1., Oshino. N.,.and Chance, B. The myoglobin probed optical studies of myocardial energy mctabolism. AJ.an. Erp. Bi"l. 94, 85-91 ( 1978). . 28. Warburg, 0., and Negelein. E. fJber das Absorptionssp<klrum des Almungsfermcnts. Bio<henr. Z. 214. 64-106 (1929). _ 29. Wiuenbcrg, J. B. Myoglobin-facilitatcd diffusion of oxygcn. J. Gen. Ph?siol. 49, 57-74 (1966). 30_ Witlcnberg, B. A.. and Wittenbcrg, J. B. Role of myoglobin in the oxygen supply to red skeletal muscle. J. 8iol. Chem. 250, 9038-9043 (1975). 31. Yoshikawa. S.. Choc, M. G., OToole, M. C.. and Caughey, W. S. An infrared study of CO binding to hean cyiochrome r oxidase and hcmoglobin A. J. 8inf. Chem. 252. 5498-5508 (1977). 32. Zom, H. The panial oxygen pressure in the brain and liver at subroxic ronccniraiions of carbon monoxide. SrmrM1 Reinhnfi. Lng 6r,c1. Fd. 3d.24-29 (1972). I rop)' 1tnt

I Mechanisms of Carbon Monoxide Toxicityt= - RONAt-o F. COBURN Oepnrrnrenrr nfPhcsidfqer'. Unirersirv ufPenrrsthnrriu Sthonf rrf llerlirinr. PdilrrJtlphra. Pennsdrania 19104 This reviesr rcexamines Ihe possibility that some of the toxic effects of CO exposure in man could be due to CO binding ssithin cells, as aell as decreases in tissue pO; resulting from the presence of carbozyhemoglobin /HbCOI. Data are reviewed from esprriments designed to measure adverse effects of CO on isolated smooth muscle. Thne data show that CO tensions greater than 1.000 times those seen in intact tissues when-HbCO-is 5 to I0-c saturation have only a small effect on oxygen uptake. Thus it is unlikely thm in tiru CO toxicity occurs in this tissue as a result of binding to cytochrome oxidase. 1 reviewed existing data which indicate that CO binding to mTogJobin, in heart and skeletal muscle. is signifi- cant, even at low CO exposure giving HbCO less than S:r saturation. vfhether myo- globin-CO binding is a significant ouse of CO toxicity during exerci_se, or of the sensitivny of the myocardium to CO, is unknown. In this review the topics chosen relate to possible mechanisms of CO toxicity to mammalian cells, with special reference to the heart and great vessels. In particu- lar, I wanted to reexamine the possibility of an "intracellular" toxic effect of CO in peripheral tissues. Although there is still no direct evidence of CO toxicity mediated by a mechanism other than CO binding to hemoglobin and resultant effects on oxygen availability in peripheral tissues, recent data strongly suggest that CO toxicity also may be due to CO binding to other compounds. This evi- dence includes the following: (a) There is evidence that intracellular pO.. in prox- imity to mitochondria and cytochrome oxidase may be lower than previously suspected (6, 25, 27). As will be discussed later in this review. CO and O, compete for CO binding to cytochrome oxidase and a low pO, promotes CO binding at a given pCO. (b) There is evidence that cytochrome oxidase may be partially re- duced in intact tissues (21). Since CO binds only to the reduced state, this finding makes it more likely that enough CO could bind to cytochrome oxidase to inhibit its function. Previous studies with isolated mitochondria had indicated that cyto- chrome oxidase is almost entirely in the oxidized form in state 111 mitochondria. (c) Several different approaches to the study of CO toxicity have demonstrated ad- verse biological effects with small increases in the body CO stores, given HbCO as low as 4 to 5 c saturation (2, 2D). Calculations of the effect of this HbCO level on capillary pO, give such a small decrease that the question naturally arises whether the only CO toxic effect is due to binding to hemoglobin. In addition, the impor- tance of shifts of the oxyhemoglobin dissociation curve as a determinant of tissue ' Presented at a w-orkshop on Carbon Monoxide and Cardiovascular Disease, sponsored by the Ameriaan Health Foundation and the Federal Health ORce, Federal Republic of Germany, Berlin, October 10-12. 1978. r Supported by Grant HL 19737 from the National Institutes of Health. Bethesda. \Id. l 0091-7435,79 008 3-170005 0_.lxFO eoqnehr ~. w-a b Aroeemc Aeu. t~c. AG net•r. d rtp~odurtion in amr lorm rrvnrJ ~(O~ 1

WORKSfION: CARBON MONOXIDE AND CVD t-I `~ ! .. What fa P02 in 'Cori ? P02 Outside Geometry Do2 V02 Boundary canditians I 02mm _ - Uniformify Fte. 7. Factors influencing tissuepOt gradients. This is drawn for isolated strips of tissue oxygen- ated by ddTusion from a bathing solution. O; is diffusing into a metabolizing tissue and the decrease in pO, in Ihe lissue as a funetion of distance from the surface is a resull of diffusion and metabolic consumption of Or. - I tion is that mechanical tension is oxygen tension dependent, so-that energy re- quirement varies as a function of oxygen tension. Our previous study was an attempt to gain insight about the mechanisms of oxygen-tension-dependent mechanical tension ( 12). Figure 4 shows the typical effects of increasing the organ bath pCO from 0 to 540 mm Hg on isometric tension of rabbit aorta at constant orgarn bath pOe. Figure 5 shows a similar experiment except that pCO was increased after VO, was corn- pletely inhibited by NaCN. Figure 6 shows the effects a CO pressure of 570 mm - Hg on VOi. These data suggest several conclusions. First, since CO at a tension >1,000 times that ever seen in tissue irr riro had only a small effect on oxygen uptake, significant CO binding to cytochrome oxidase is unlikely to be an in tiro CO 530 mm Hg 1 Lsc 5,MIN FiG.4. EfkctofincreasingpCOonisometnctensionofrabbitaonaslrips.OrganhaihpOiscaskept constant at 160 before and after pCO was increased by adding CO to the gas bubbling the solution. Arrows indicate changes in pOr, keepingpCO at 530 mm Hg. Prior to these measurements. Ihe strip was contracted by adding I µtP norepinephrine to the baihing solution. tnterrupred lines indicate resting tension prior to norepinephrine. I-

-i1 o t-+i WORI:SHOP: CARBON AtONOXIUE AND CVD An approach to the question of whether there is significant CO binding to various hemoproteins to inhibit their functions is to compute binding, using values forpCO, pOi, and binding constants. In these calculations there is some certainty about values used for tissue pCO since tissue pCO must be in equilibrium with mean capillary pCO which_can be computed using the Haldane Equation (9, 18) and has been more directly measured using gas-bubbling techniques ( 17). Data are available regarding CO binding constants, but the great uncertainty in this type of calculation is the p0: in proximity to the hemoprotein. This review mainly considers CO binding to cytochrome oxidase and to myo- globin. In the case of ytochrome oxidase the pCO giving 50% binding in the absence of O: is approximately 10-`.S1 (=0.1 mm Hg) (31); the K, for O, in actively respiring mitochondria is usually given as 3 x 10-'ti/ (5). Studies ef CO binding to isolated cytochrome oxidase suggest the CO/0z for 50% CO binding is slightly above unity (3, 28). In isolated mitochondria the COIOi for 50r7e inhibition of respiratory chain activity was found to be lower during a transient from state 4 to state 3(4). These data suggest that pO, in the mitochondria would have to be less than =0.1 mm Hg for much CO binding to occur under conditions where HbCO is in the range 10 to 15% saturation and tissue pCO =-0.01 mm Hg. IntrecellularpO_ data in the literature are dominated by polarographic electrode studies where, in most tissues, most penetrations give values of 10 to 40 mm Hg and very few penetrations give values less than 3 mm Hg. It is, however, possible that compounds other than oxygen can react with these electrodes and give falselyy high values. A recent approach (6, 25. 27) is to monitor the spectrum of myoglobin 0 8 M.un Ccp;llcry Co OZ \\ e Mitochondrion 61ood .__. - ___ _. Pa.v. -~ Fw. I. Competition of O: and CO for binding to myoglobin and cytochrome oxidase. Heavy arro%.s indicate diffusion of oxygan frum capillary to mirochondrion and emphasize ihe pOr gradient that occurs because oxygen is consumed by the mitochondria. "CO" arrows indicate the pCO in lissue is equilibrated u{Ih a mean capillary pCO. Since myoglobin is presumably locaied in the qloplacm, the pO, in prosimity to myoglobin molecules must be considerably higher than p0, in proximiiy to cytochrome oxidase. 02

11o ___ - WORKSHOP: CARBON MONOXIDE AND CVD I Clearly, myoglobin is involved in some way in helping supply the mitochondria of skeletal and heart muscle cells with oxygen. ]vlyoglobin may facilitate oxygen diffusion in the cytoplasm (29) or it may provide stores of accessible oxygen very close and available to mitochondria. In a recent experiment on isolated strips of skeletal muscle, chemical oxidation of myoglobin to metmyoglobin, which cannot bind 0, inhibited oxygen uptake of the strip (30). This is the first direct evidence of the role of myoglobin in a cell preparation. It should be pointed out that we think it is most likely that myoglobin is dissolved in cytosol but this has not been proven. If myoglobin is bound to intracellular membranes it is not likely that this compound facilitates oxygen transport within the cell. It has been shown that increases in HbCO to levels as low as 59o saturation can cause a decrease in maximal oxygen consumption during exercise. The sensitivity of the heart to elevated HbCO is also documented. But whether binding to myo- globin is a cause of toxicity is unknown.. The evidence. however, is not much weaker than that for the hypothesis that CO binding to hemoglobin is the cause of clinical CO toxicity. The evidence is that CO is bound to myoglobin and that this has to have an effect on the myoglobin function with regard to respiration at the cell level. OTHER CELL HEMOPROTfINS . It is possible that CO may bind to hemoproteins other than cytochrome oxidase, hemoglobin, or myoglobin in sufficient amounts to inhibit their function. Cyto- ,chrome P-450, a mixed-function oxidase, probably does not bind sufficient CO to ,cause inhibition of drug hydroxylation, even at HbCO 15-2017c saturation (26). 'Tryptophan dioxygenase and catalase have high affinities for CO (14) and a possi- ,ble role of these enzymes in CO toxicity should be studied. Tryptophan dioxygenase catalyzes conversion of tryptophane to 1-formyl kynurenine. Inacti- vation of this enzyme in the liver would result in increased scratonin levels in 'other tissues including the brain. Catalase and peroxidase catalyze degradation of _HiOz which is a toxic oxidant. CO cottcentrnrron in ltrpo.ric cells. Since CO and O_ compete for binding to hemoproteins one might expect that CO binding would increase during tissue hypoxia. This, in fact, has been demonstrated to occur for the case of myo- globin_-CO binding in skeletal muscle and myocardium (10, 11, 24). Figure 8 illus- trates effects of decreasing arterial pO: on [HbCO] under conditions where CO could not be lost via the lung. The decrease in blood HbCO is explained by increases in bibCO as determined by biopsies in skeletal muscle and myocardium. Whether this phenomenon plays an aggravating role in CO toxicity depends on whether increased intracellular CO binding is of importance. Another argument is that influx of CO into muscle lowers blood HbCO and protects brain and other tissues from adverse effects of increased HbCO. CO SHIFTS DURING EXERCISE AT MAXIMAL OXYGEN UPTAKE During exercise at maximal oxygen uptake, it is known that oxygen uptake is decreased by increasing blood [HbCO] to levels as low as 4 to 5% saturation (20). Since it seemed likely that oxygen uptake during strenuous exercise may be lim- ited by availability of oxygen (and that this results in a fall in intmcellularpO_), we t. ~ ------------- L.,.__--------------------~ i o~ri3ht I I E P ~c~C^~

RONALD F. COBURN wondered whether CO shifts into exercising skeletaF muscle could be playing a role in the toxic effects of CO. Figure 9 shows data obtained in an experiment designed to look for CO shifts in an exercising young human subject (7). In this subject, large shifts of CO out of blood were seen during "maximal" exercise with the subject breathing 13% O„ but at maximal exercise breathing 2i5''r O.. only relatively small shifts of CO were observed. The lack of a shift out of blood suggests that myoglobin pO_ did not fall during maximal exercise and that there was another limiting factor for aerobic metabolism, rather than oxygen supply to tissues. This type of experiment does not exclude the possibility that CO shifis occur during maximal exercise with elevated HbCO where there may be an inter- play of factors: a fall in intracellularpO_ resulting from binding of CO to hemoglo- bin which, in turn, caused a shift of CO into the cell and an increase in M1IbCO, limiting oxygen supply to exercising skeletal muscle cell mitochondria. Disnessinrr of rlree esperimenu of L. R. Grldbamn et nl. Dr. Goldbaum and colleagues have published data which have challenged some of the basic concepts in our understanding of CO toxicity (15, 16). They have compared CO toxicity during CO inhalation with CO toxicity during infusion of blood containing HbCO, giving equivalent values of HbCO. When anesthetized dogs breathed 13% CO, giving 54 to 90% HbCO saturation, all animals died within 15-60 min. Acute removal of hemoglobin by bleeding and reinfusion of plasma, giving hemoglobin concentrations of 26-369bQ of control did not result in death of the animals. When blood was removed and replaced with transfused HbCO giving HbCO saturation of 57-64% the animals survived indefinitely. When CO was injected in- traperitoneally giving HbCO 40-80% saturation, the animals also survived indefi- nitely. These data were interpreted as indicating that CO toxicity is independent of [HbCOj but is a result of dissolved CO. It was postulated that HbCO and pCO are not in chemical equilibrium in blood in the various segments of the circulation. Thus when CO is given as transfused HbCO or absorbed from the peritoneum. these investigators suggest that dissolved CO is cleared in the lung and is absent or ~ I ARTERIeE iD} P0l \ ~ r- ~ MM/:G KK COHB 96 SAT Ar i 0 0 TIME IN HOURS I FrG. g. Shifts of CO oul of blood during hypoxic hypoxcmia. Inspired p0, uas progressively decreased in this anesthetized dog and effects on mixed venous carbos1hemog_lobin were deiermined. Taken from previously published data (24).