103:1359-1365, 2007. First published Jul 5, 2007;  doi:10.1152/japplphysiol.00443.2007 J Appl PhysiolHunter and Arlin B. Blood Gordon G. Power, Shannon L. Bragg, Bryan T. Oshiro, Andre Dejam, Christian J.

You might find this additional information useful...

43 articles, 14 of which you can access free at: This article cites http://jap.physiology.org/cgi/content/full/103/4/1359#BIBL

2 other HighWire hosted articles: This article has been cited by

  [PDF]  [Full Text]  [Abstract]

, May 6, 2008; 105 (18): 6771-6776. PNASJ. Vogel, B. Pichler, R. S. Kasinathan, J. P. Nicolay, S. M. Huber, F. Lang and R. Feil M. Foller, S. Feil, K. Ghoreschi, S. Koka, A. Gerling, M. Thunemann, F. Hofmann, B. Schuler,

Anemia and splenomegaly in cGKI-deficient mice 

[PDF]  [Full Text]  [Abstract], February 1, 2009; 296 (2): H237-H246. Am J Physiol Heart Circ Physiol

D. B. Kim-Shapiro and G. G. Power A. B. Blood, M. Tiso, S. T. Verma, J. Lo, M. S. Joshi, I. Azarov, L. D. Longo, M. T. Gladwin,

Increased nitrite reductase activity of fetal versus adult ovine hemoglobin

including high-resolution figures, can be found at: Updated information and services http://jap.physiology.org/cgi/content/full/103/4/1359

can be found at: Journal of Applied Physiologyabout Additional material and information http://www.the-aps.org/publications/jappl

This information is current as of May 13, 2010 .  

http://www.the-aps.org/.ISSN: 8750-7587, ESSN: 1522-1601. Visit our website at Physiological Society, 9650 Rockville Pike, Bethesda MD 20814-3991. Copyright © 2005 by the American Physiological Society.those papers emphasizing adaptive and integrative mechanisms. It is published 12 times a year (monthly) by the American

publishes original papers that deal with diverse areas of research in applied physiology, especiallyJournal of Applied Physiology

on May 13, 2010

jap.physiology.orgD

ownloaded from

A novel method of measuring reduction of nitrite-induced methemoglobinapplied to fetal and adult blood of humans and sheep

Gordon G. Power,1 Shannon L. Bragg,3 Bryan T. Oshiro,2 Andre Dejam,4 Christian J. Hunter,4

and Arlin B. Blood1,3

1Center for Perinatal Biology and 2Department of Obstetrics and Gynecology, Division of Perinatal Medicine, and3Department of Pediatrics, Division of Neonatology, School of Medicine, Loma Linda University, Loma Linda, California;and 4Department of Medicine, Brigham and Women’s Hospital, Boston, Massachusetts

Submitted 24 April 2007; accepted in final form 3 July 2007

Power GG, Bragg SL, Oshiro BT, Dejam A, Hunter CJ, BloodAB. A novel method of measuring reduction of nitrite-induced met-hemoglobin applied to fetal and adult blood of humans and sheep.J Appl Physiol 103: 1359–1365, 2007. First published July 5, 2007;doi:10.1152/japplphysiol.00443.2007.—The reaction of nitrite withdeoxyhemoglobin results in the production of nitric oxide and met-hemoglobin, a reaction recently proposed as an important oxygen-sensitive source of vasoactive nitric oxide during hypoxic and anoxicstress, with several animal studies suggesting that nitrite may havetherapeutic potential. Accumulation of toxic levels of methemoglobinis suppressed by reductase enzymes present within the erythrocyte.Using a novel method of measuring methemoglobin reductase activityin intact erythrocytes, we compared fetal and adult sheep and humanblood. After nitrite-induced production of 20% methemoglobin, theblood was equilibrated with carbon monoxide, which effectivelystopped further production. Methemoglobin disappearance was firstorder in nature with specific rate constants (k � 1,000) of 12.9 � 1.3min�1 for fetal sheep, 5.88 � 0.26 min�1 for adult sheep, 4.27 � 0.34for adult humans, and 3.30 � 0.15 for newborn cord blood, allstatistically different from one another. The effects of oxygen ten-sions, pH, hemolysis, and methylene blue are reported. Studies oftemperature dependence indicated an activation energy of 8,620 �1,060 calories/mol (2.06 kJ/mol), appreciably higher than would becharacteristic of processes limited by passive membrane diffusion. Inconclusion, the novel methodology permits absolute quantification ofthe reduction of nitrite-induced methemoglobin in whole blood.

methemoglobin reductase; fetus; whole blood

NITRITE IS PRESENT IN THE plasma at concentrations ranging fromthe low- to mid-nanomolar range in most mammalian speciesstudied (9, 23). Circulating nitrite is derived primarily from theoxidation of endogenous nitric oxide (NO) produced by NOsynthase (23) and is cleared from the plasma by reaction withoxyhemoglobin to form nitrate or with deoxyhemoglobin toform NO (10, 25, 44). In both of these reactions, the iron inhemoglobin normally present in the ferrous form (Fe2�) isoxidized to the ferric form and methemoglobin is produced.The reaction of nitrite with deoxyhemoglobin has receivedmuch attention over the past 4–5 years because of its potentialas a physiologically important, hypoxia-sensitive, and NOsynthase-independent source of NO (13). Animal studies ofNO production from the reaction of nitrite with deoxyhemo-globin have demonstrated therapeutic potential for infusednitrite in models of myocardial infarction (11), intracerebralhemorrhage (33), and pulmonary hypertension (19).

The oxygen-carrying properties of hemoglobin depend onoxygen binding to ferrous iron at each of the four heme groups.Once iron is in the ferric (Fe3�) state, as in methemoglobin, itis unable to combine reversibly with oxygen and transport it inthe body. Clinically, methemoglobin concentrations �10–20%result in obvious cyanosis, with headaches, weakness, andbreathlessness becoming apparent at levels of 35% or greater(7). Approximately 2–3% of all hemoglobin in the body isconverted to methemoglobin each day as a result of endoge-nous oxidative stresses (15). Given the 120-day lifespan of theerythrocyte and endogenous rates of methemoglobin produc-tion, nearly all hemoglobin would be in an oxidized form wereit not for the activity of methemoglobin-reducing enzymeswithin the erythrocyte. By reducing iron back to its ferrousstate, these enzymes restore functional hemoglobin. Thus thecompeting processes of methemoglobin production and reduc-tion are in equilibrium such that methemoglobin levels aretypically �1% of total pigment.

Although an NADPH-dependent flavin reductase capable ofreducing methemoglobin is known to exist in erythrocytes, alarge majority of methemoglobin reduction is carried out by theenzyme cytochrome b5 reductase (cytb5r), a soluble 245-aminoacid protein found in the cytosol of the erythrocyte (4, 31).Although membrane-bound versions of the protein have beenfound, they normally confer only a minor portion of the overallmethemoglobin reductase activity. The cytb5r enzymes facili-tate the reduction of cytochrome b5 by transfer of hydrideanions from NADH (39). Cytochrome b5, in turn, rapidlyreduces methemoglobin (18). The predominant source ofNADH within the erythrocyte is via the metabolism of glucosein the Embden-Meyerhof pathway (15). Methemoglobin reduc-tase activities vary widely across species (24, 34, 35, 40) andwith age. Human newborns, for example, are known to be atgreater risk of methemoglobinemia than adults (15, 21). Thisincreased risk has been attributed to less methemoglobin re-ductase activity in neonatal erythrocytes (14, 21) and a greatertendency of fetal hemoglobin to assume forms that are stable inthe ferric state (15, 32).

As noted, the reaction of nitrite with either oxy- or deoxy-hemoglobin results in hemoglobin oxidation to methemoglobin(10, 25), but the chemistry in each instance is unusuallycomplex. The stoichiometry of these reactions is complicatedby formation of products, such as NO, that are themselvescapable of hemoglobin oxidation or regeneration of nitrite.

Address for reprint requests and other correspondence: G. G. Power, Centerfor Perinatal Biology, Loma Linda Univ. School of Medicine, Loma Linda, CA92350 (e-mail: gpower@llu.edu).

The costs of publication of this article were defrayed in part by the paymentof page charges. The article must therefore be hereby marked “advertisement”in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

J Appl Physiol 103: 1359–1365, 2007.First published July 5, 2007; doi:10.1152/japplphysiol.00443.2007.

8750-7587/07 $8.00 Copyright © 2007 the American Physiological Societyhttp://www. jap.org 1359

on May 13, 2010

jap.physiology.orgD

ownloaded from

Such reactions would result in measurements of a �1:1 ratio ofmethemoglobin production to nitrite metabolism. Conversely,recent evidence suggests that nitrite and NO may act togetherto reduce methemoglobin independent of methemoglobin re-ductase enzymes in a reaction known as reductive nitrosylation(12). Experiments that use whole blood with intact erythro-cytes are needed to determine the physiological relevance ofeither of these processes to rates of methemoglobin reduction.However, reported cytb5r activities are most commonly mea-sured indirectly by spectrophotometric measurement of theenzyme’s reduction of ferrocyanide-methemoglobin complexor NADH-ferricyanide in hemolysates as alternative substratesfor methemoglobin (3, 42). Alternatively, cytb5r activities havebeen studied in whole blood, but with the use of dilutesuspensions of red blood cells washed after pharmacologicalinduction of methemoglobinemia with hemoglobin-oxidizingagents (2, 40). Interestingly, however, the functional activity ofoverall methemoglobin reduction, i.e., the decline in methe-moglobin levels per unit time in whole blood at body temper-ature, does not appear to have been measured in any species,perhaps because of the complications introduced by the con-tinuous formation of methemoglobin that normally occurscoincident with its removal.

The present study was designed to test the hypothesis thatcarbon monoxide (CO), by converting hemoglobin to carboxy-hemoglobin, would block nitrite-induced methemoglobin pro-duction. We further hypothesized that this treatment wouldenable the measurement of methemoglobin-reducing processesin the absence of competing methemoglobin production. Usingthis simplified method, we compared rates of methemoglobinreduction in the presence of nitrite in the blood of fetal andadult sheep and humans.

MATERIALS AND METHODS

Collection of human blood samples was performed by followingprotocols that had been approved by the Loma Linda UniversityInstitutional Review Board. Collection of sheep blood samples wasperformed by following protocols that had been approved by theLoma Linda University Institutional Animal Care and Use Commit-tee. Venous blood samples were collected in heparinized syringesfrom in-dwelling catheters in fetal sheep at 130–140 days of gestation(term 145 days), from adult sheep and adult humans, and from humanumbilical cords immediately after separation from infants at parturi-tion of uncomplicated pregnancies.

In a typical study, 10.0 ml of blood were added to a sealed flaskmaintained in a water bath at body temperature (39 and 37°C forsheep and humans, respectively). Two baseline samples (0.3 ml) werecollected during a 10-min equilibration period for measurement ofmethemoglobin saturation (MetHb%), blood gases, oxyhemoglobinsaturation, total hemoglobin, and pH. One milligram of sodium nitritedissolved in 0.5 ml of saline was then added to the flasks, whichprovided an initial concentration of 1.4 mM nitrite. This level ofnitrite established an initial molar ratio of �1:2 nitrite to hemoglobin,equivalent to 1:8 nitrite to heme. After addition of nitrite, the increaseof MetHb% was followed with periodic sampling until MetHb%reached �20%, at which time the gas phase in the flask was replacedwith CO and the blood swirled for 2 min. The subsequent decline inMetHb% was followed with 0.1-ml sampling at 20-min intervalsduring the next 3–4 h. Blood gases and pH were measured at hourlyintervals. One to eight disappearance curves were recorded fromblood of each animal or subject.

Experiments were carried out to assess whether CO was effectivein completely preventing the formation of methemoglobin. In these

control experiments, blood was treated with CO before introduction ofnitrite and levels of MetHb% were measured with time. Results wereobtained after standard doses of nitrite and, in addition, after intro-ducing initial concentrations 5- and 10-fold higher than standard toamplify the effect of this potential error. To verify that methemoglo-bin was not being formed by xanthine oxidase during reduction, theexperiments were repeated after addition of 100 �M allopurinol.

To study the pH dependence of methemoglobin disappearance, pHwas varied from �7.2 to 7.6 by addition of 0.1 M HCl or NaOH to 10ml of swirled blood. To study the oxygen dependence of methemo-globin reduction, the oxygen tension of the blood was varied from 10to 100 Torr by equilibration of the blood with gas mixtures containing3% CO2 and varying oxygen concentrations in nitrogen. To study theimportance of the erythrocyte membranes and intact nature of the redblood cells, the experiments were repeated after the red blood cellswere lysed by one of three alternative methods. These were osmoticrupture with distilled water (2:1 water-blood), sonication, and re-peated cycles of freeze-thawing. The method of lysis did not appear tohave an effect on results, and the results of the freeze-thaw experi-ments are presented herein. The temperature dependence of methe-moglobin disappearance was studied by repeating experiments attemperatures ranging from 16 to 39°C.

To study the effects of methylene blue, a 1% solution was preparedand 0.025–0.1 mg of the dye was added to the equilibration flasks atthe time of equilibration with CO. This provided concentrationsranging from 7 to 30 �M.

Methemoglobin measurements. Methemoglobin, oxyhemoglobin,and carboxyhemoglobin saturations were measured to the nearest0.1% using a hemoximeter (model OSM3; Radiometer). The coeffi-cient of variation for methemoglobin saturation measurement was0.5% of measured values with MetHb% in the range studied and 13%of measured values with MetHb% at baseline levels. The accuracy ofhemoximetry measurements of methemoglobin in the presence ofsaturating amounts of CO was assessed by preparing and measuringmethemoglobin in mixtures of predetermined ratios of 100% carboxy-hemoglobin and 100% methemoglobin (created by exposing 100%oxygenated blood to 10 mM nitrite for �20 min followed by 3centrifuge-wash cycles with saline) using whole adult sheep blood sam-ples. Measured methemoglobin concentrations were plotted against ac-tual concentrations, and the resulting data points were fitted to a linearregression analysis. Blood gases and pH were measured with appropriatespecific electrodes (model ABL-5; Radiometer).

Calculation of methemoglobin reduction rates. Rates of methemo-globin reduction were determined from the linear regression analysesof semilog plots of methemoglobin concentration vs. time aftersaturation of the blood with CO. The relationship between the loga-rithm of methemoglobin concentration and time was highly linear, andthe rate constant (k) for methemoglobin reduction was taken as theslope of the linear relationships. For each set of experiments, the kvalue was calculated as a mean of k values determined for eachindividual experiment, rather than a linear regression analysis of allexperiments combined.

Statistical analysis. ANOVA was used to analyze differencesbetween experimental groups. When significant effects were found,the post hoc test used was Fisher’s protected least significant differ-ence. For all tests, was set at 0.05. Means are presented with theirSEs. The software DATAMSTR (courtesy of R. A. Brace), GraphPadPrism v4.0, and TableCurve were used for the analyses.

RESULTS

The accuracy of hemoximetry measurements of methemo-globin in saturating levels of carboxyhemoglobin in wholeadult sheep blood is shown in Fig. 1. Methemoglobin wasmeasured at levels ranging from �1% to 80%. The resultinglinear regression demonstrated an r2 value of 0.988, a slope of0.990, and a y-intercept of 3.455 � 1.329.

1360 FETAL AND ADULT METHEMOGLOBIN REDUCTION

J Appl Physiol • VOL 103 • OCTOBER 2007 • www.jap.org

on May 13, 2010

jap.physiology.orgD

ownloaded from

A representative curve showing changes in methemoglobin asa percentage of total pigment from cord blood of human newborn15 is shown in Fig. 2. After addition of nitrite at time 0, MetHb%increased from a baseline level of 0.7% to 19.6% during the next30 min. After equilibration with CO, MetHb% decreased, morerapidly at first and then more slowly, reaching 8.1% during thenext 230 min. When MetHb% was plotted semilogarithmicallyduring the decay phase, as shown in Fig. 2B, a linear relationshipwas observed, as is characteristic of first-order kinetics. A similarlinear relationship (average r2 of �0.98) was observed for each ofthe curves analyzed in this study. The specific k characterizingfirst-order kinetics and providing the fractional disappearance perunit time was calculated from the slope, and in this experiment itwas 0.00396 � 0.00007 min�1.

The summary results of fitting linear slopes to the decaycurves for the various experimental groups are shown in Fig. 3and summarized in Table 1. The results given are for 9 fetalsheep at 130–145 days gestation, 20 adult ewes, 8 adult humansubjects (4 men and 4 women with ages ranging from 16 to 70yr), and cord blood from 21 human term newborns. Thefractional loss per minute was highest in fetal sheep blood(averaging 0.0129 � 0.0012 min�1), which is more thantwofold higher than that for adult sheep (0.00552 � 0.00022min�1; P � 0.001). By way of contrast, k was least in humannewborns, averaging 0.00330 � 0.00015 min�1, and less thanthat in adult humans (0.00427 � 0.00034 min�1; P � 0.01).The apparent half-lives averaged 54 � 5 min for fetal sheepblood, 125 � 5 min for adult sheep blood, 162 � 13 min foradult human blood, and 210 � 10 min for cord blood fromterm newborns (Table 1).

The results of testing the extent to which CO preventsmethemoglobin formation are shown in Fig. 4. After hemoglo-bin had been converted to carboxyhemoglobin at the beginningof the experiment followed by nitrite addition, MetHb% in-creased 0.3% in 4 h. After nitrite was introduced at a dosefivefold higher than standard, MetHb% increased 1.2% in 4 hand after 10-fold excess it increased 2.2%. With the use ofnitrite concentrations typically employed, k values were recal-culated and found to alter 2.3% from values as reported here.These data indicate the near complete effectiveness of equili-bration with and continuous exposure of blood to 100% CO inpreventing methemoglobin formation in the presence of 1.4mM nitrite. After addition of allopurinol, a xanthine oxidaseinhibitor, the measured rate of methemoglobin reduction varied5% or less, indicating that reduction of nitrite to NO byxanthine oxidase does not contribute to continued methemo-globin production after CO saturation in these experiments(data not shown).

The influence of oxygen tension and pH in the blood duringthe reduction process was studied in adult human blood; k wasnot measurably affected by changes in oxygen tension from 7to 100 Torr (r 0.22, P 0.15, n 44). In addition, k wasnot affected by changes in pH from 7.2 to 7.6 (r 0.14, P0.38, n 44) (data not shown).

Fig. 1. Validation of hemoximetry measurements of methemoglobin (MetHb)in the presence of saturating concentrations of carbon monoxide (CO). Wholeadult sheep blood preparations of 100% MetHb and 100% carboxyhemoglobinwere admixed at varying ratios to achieve a range of MetHb concentrations.The resulting plot of MetHb dilutions vs. hemoximetry-reported MetHbconcentrations demonstrates a linear relationship with r2 0.989, slope of0.99 � .03, and y intercept of 3.46 � 1.33 (P � 0.0001).

Fig. 2. Time course of changes in cord blood of newborn 15. A: MetHb levelsin a 10.0-ml blood sample at 37°C after introduction of 1.0 mg of sodiumnitrite at time 0 and exposure to CO 30 min later. B: logarithm of MetHb%after conversion of the remaining non-MetHb to carboxyhemoglobin (COHb),stopping further oxidation by nitrite; 99% confidence limits are shown.C: changes in COHb% after exposure of the blood sample to CO.

Fig. 3. Box and whisker plots providing summary results for fetal and adultsheep and adult human and newborn cord blood. Mean, 25%, and 75%confidence intervals and largest and smallest measured values are shown.

1361FETAL AND ADULT METHEMOGLOBIN REDUCTION

J Appl Physiol • VOL 103 • OCTOBER 2007 • www.jap.org

on May 13, 2010

jap.physiology.orgD

ownloaded from

Figure 5 shows a summary of experimental results illustrat-ing the effect of erythrocyte lysis on rates of methemoglobinreduction. After lysis, k decreased to 9% of whole blood values(0.00054 � 0.00008 min�1; P � 0.001). A similar decrease ink was observed whether the lysis was carried out by distilledwater, sonication, or repeated freeze-thaw cycles. The pro-nounced decrease in rate was not restored by addition of themembrane-solubilizing agent Triton X-100 to the hemolysates,nor by replacement of the plasma by phosphate buffer beforelysis.

Addition of 0.025–0.10 mg of methylene blue to 10 ml ofblood in the equilibrium chamber increased k to 0.0487 �0.0065 min�1 (P � 0.001), some sixfold higher than in untreatedblood. Methylene blue had little effect when added to lysed blood.The combination of methylene blue and 1–5% Triton X-100 alsodid not alter the rate of reduction appreciably.

Figure 6 shows an Arrhenius plot in which the logarithm ofk is plotted against the reciprocal of the absolute temperature.The data shown are for the blood of adult sheep. The slope ofthe line is �1,883 � 231 using least squares regressionanalysis; after we multiplied the slope by the gas constant and2.303 to allow for base 10 logarithms, we were able tocalculate an activation energy of 8,620 � 1,060 cal/mol or2,060 J/mol. This value is two- to threefold higher than ischaracteristic of processes limited by passive diffusion throughmembranes (41), suggesting that these are not rate limiting, and

is more representative of chemical- or enzyme-mediated reac-tions with all necessary components contained within the redblood cell interior.

DISCUSSION

The present results provide data characterizing the rate thatmethemoglobin is reduced to functional hemoglobin in wholeblood at body temperature and are thus directly relevant to thesafety of administering nitrite to humans. Methemoglobin isreduced more slowly in humans than in sheep, and the processis measurably slower in cord blood of newborn humans than inadults. The effective half-life of methemoglobin in newbornhuman blood is 210 min, 30% longer than in adult humanblood. In both species and irrespective of age, the reduction ofmethemoglobin closely follows first-order kinetics in a 3-hperiod of observation. The reduction of methemoglobin istemperature dependent to an extent that is characteristic ofenzyme-catalyzed reactions rather than a process limited bypassive diffusion. Finally, methemoglobin is reduced manyfoldmore slowly after disruption of red cell membranes, and therate is not restored to normal by addition of methylene blue ordetergent to the hemolysates or by replacement of plasma withphosphate buffer before the lysis.

To our knowledge, the present study is the first to providethe specific rate constants that characterize the reduction ofelevated methemoglobin, without concomitant methemoglobinproduction, in undiluted whole blood at body temperature.Earlier work has established that this reductive process isbrought about largely by methemoglobin reductases, alterna-

Fig. 4. Time course of changes in MetHb% after pretreatment with CO. Smallincreases with time are seen at initial doses of nitrite 5- and 10-fold higher thanstandard (dashed lines) but not observed with standard doses (triangles),demonstrating near complete effectiveness of CO at inhibiting MetHb produc-tion. Also shown are the effects of CO treatment of whole blood (squares) andlysed blood (circles) approaching 20% MetHb, demonstrating the markedreduction of MetHb reduction by hemolysis.

Fig. 5. Box and whisker plots of adult sheep blood showing the first-order rateconstant k in whole blood, after lysis, after addition of methylene blue (0.025mg), and after Triton X-100 (final concentration of 5%) in combination withmethylene blue.

Table 1. Baseline methemoglobin levels, k (�1,000), and t1/2 of methemoglobin in sheep and human blood studied in vitroat body temperature

Blood SourceBasal Methemoglobin,

%saturation k � 1,000, min�1 t1/2, min

Fetal sheep (n 9/9) 1.9�0.2 12.9�1.2 54�5Adult sheep (n 24/41) 0.67�0.01 5.52�0.22 125�5Newborn humans (n 21/21) 0.54�0.02 3.30�0.15 210�10Adult humans (n 8/44) 0.49�0.05 4.27�0.34 162�13

Values are means � SE; n, no. of individuals studied/total number of curves analyzed. k, First-order rate constants; t1/2, effective half-life.

1362 FETAL AND ADULT METHEMOGLOBIN REDUCTION

J Appl Physiol • VOL 103 • OCTOBER 2007 • www.jap.org

on May 13, 2010

jap.physiology.orgD

ownloaded from

tively termed cytb5r, which are located in the interior of redblood cells. The rate of methemoglobin reduction in red bloodcells has also been found to be dependent on glycolytic rate andtherefore the rate at which NADH can be generated as a sourceof reducing equivalents for the methemoglobin reduction. It isworth noting that, although the processes of methemoglobinreduction are enzyme dependent, we have elected to describethe rates using first-order equations rather than Michaelis-Menten kinetics. This is because methemoglobin reduction is aresult of at least two distinct enzyme systems (cytb5r and flavinreductases) with two different cofactors (NADH and NADPH,respectively). In addition, the relative contribution of these twoenzyme systems to overall methemoglobin reduction is likelyto vary significantly with methemoglobin concentrations andbetween species (22). As a result, it was decided that the use offirst-order equations, which accurately described the experi-mental data, was more appropriate.

The NADH-methemoglobin reductase system comprises asoluble cytochrome b5 and cytb5r in the interior of erythrocytes(20). Six electrophoretically distinct variants of NADH-met-hemoglobin reductase have been found in studies of humanblood (16). The enzyme exists in two forms, one fraction beingtightly bound to the inner face of the red blood cell membraneand a second fraction dissolved in the cytosol (8). Earlier workhas examined the difference in methemoglobin reductase ac-tivity among species and at different ages (1, 26, 27, 36, 43).Typically, these earlier investigators have used alternativesubstrates such as NADH-ferricyanide and diluted hemolysatesin their analyses and have not sought to measure functionalactivity at body temperature. Nevertheless, where comparisonsare possible, results of the present study are in acceptableaccord with earlier work. For instance, methemoglobin reduc-tase activity in 1:20 dilutions of erythrocytes in a stabilizingsolution were comparable in adult sheep and humans (27), inagreement with measured first-order rate constants also beingof like magnitude in the present work (Fig. 3). Similarly,methemoglobin reductase activity has been reported to be 23%less in low-birth-weight human newborns than in adults (29) inaccordance with the 30% lower k values reported here. Thus, intwo species and at different ages, the rate of reduction reported

here appears to correlate reasonably well with indirectly mea-sured indicators of methemoglobin reductase activity.

In vivo rates of methemoglobin reduction have been previ-ously determined in humans after treatment with inhaled NOgas (45, 46). Comparison of rate constants calculated fromthese previous studies with the present study suggest thatmethemoglobin reduction may occur more rapidly in vivo (k �1,000 ranging from 11 to 25 s�1) than observed in the presentexperiments. The mechanism for this difference is not clear butmay be due to the increased presence of reducing equivalentsin vivo.

In the present study, the effective half-life of methemoglobinin human newborn blood was 30% longer than in adult humanblood (P � 0.01). This difference is in agreement with earlierwork that reported lower levels of methemoglobin reductaseactivity in the newborn human (26, 27). The slower rate innewborns clearly points out the greater care required in thetherapeutic use of NO, nitrite, and other methemoglobin form-ers in newborns than in adults. The difference is judged not tobe so large, however, that it would preclude the therapeutic useof nitrite. This conclusion is supported by clinical experiencein which methemoglobin levels seldom increased above 5% inhumans during experiments testing the safety of inhaled NO(45) or in newborns given NO.

Methylene blue is thought to be reduced both intra- andextracellularly by enzymes present in the erythrocyte mem-brane (28). Inside the erythrocyte, the reduced dye is capable ofreducing methemoglobin. Both NADPH (6, 38) and NADH(37) are required for reduction of methylene blue. After addi-tion of methylene blue in the present study, k increased aboutsixfold. This result is in agreement with the four- to sixfoldincreases in the rate of methemoglobin reduction observedwhen methylene blue is given to patients therapeutically (5)and the sevenfold increase in methemoglobin reduction that hasbeen reported in human erythrocytes studied in vitro (40). Meth-ylene blue was less effective when added to hemolysates (Fig. 5),and this is in agreement with the decline in methylene blueefficacy that has been noted in patients with hemolysis (5) andmay be due to disruption of the methylene blue-reducing enzymespresent in the red blood cell membrane (28) or cytosol (17).

The mechanism by which lysis causes a marked slowing inthe reduction of methemoglobin is not known, but somepossibilities can be largely discounted by the present work. Itseems unlikely to be inhibition induced by plasma chloride, aknown inhibitor of methemoglobin reductase (21) whose ex-tracellular concentration is �10-fold higher than intraerythro-cytic chloride concentration, because the rate was not restoredto normal by replacement of plasma with phosphate buffer. Itis also not likely to be via increased binding of the reductasesto red blood cell membranes after lysis because treatment with1–5% Triton X-100, an agent known to be effective in solubi-lization of membranes (30), did not restore k to normal values,although this treatment may not necessarily restore the protein-protein interactions of membrane-associated proteins. Rather,it seems more likely that lysis disrupted the assembly ofproteins that are closely associated with one another primarilyby electrostatic interaction in the cell interior and responsiblefor the reduction of methemoglobin. The disruption wouldbreak the close-linked chain that transfers electrons fromNADH to the flavin of cytb5r and then to cytochrome b5 andfinally on to methemoglobin. Other possibilities for loss of

Fig. 6. Graph of the logarithm of k vs. the reciprocal of the absolute temper-ature (1/T, in °K) for the reduction of MetHb in adult sheep blood studiedin vitro. The range of temperatures studied was from 16 to 39°C. The averageand range of values at 39°C are shown, and the 95% confidence limits of theregression are indicated. The calculated energy of activation is 8,620 � 1,060cal/mol (2,060 J/mol).

1363FETAL AND ADULT METHEMOGLOBIN REDUCTION

J Appl Physiol • VOL 103 • OCTOBER 2007 • www.jap.org

on May 13, 2010

jap.physiology.orgD

ownloaded from

function after lysis include dependency on the relatively highintraerythrocytic concentrations of components of the methe-moglobin reductase system that would be diluted after lysis ordecreased activity of glycolytic enzymes that supply NADH,which is known to occur with hemolysis (17).

In conclusion, the present study reports the rate constantsand effective half-times that characterize the disappearance ofexperimentally induced methemoglobinemia in blood samplesin vitro at body temperature. Rates of reduction are measurablyless in newborn human blood than in adult blood. The 30%longer half-life in newborns is not so extended, however, thatit should be taken to preclude the therapeutic use of nitrite inthe newborn period, provided caution is exercised with peri-odic monitoring and the ready availability of methylene blue.

ACKNOWLEDGMENTS

We express our appreciation to the obstetric staff of Loma Linda MedicalCenter for collection of the cord samples. We also thank Franklin H. Bunn,MD, for helpful criticisms and suggestions.

GRANTS

This study was supported in part by National Heart, Lung, and BloodInstitute Grant R01 HL-65494 to G. G. Power.

REFERENCES

1. Bartos HR, Desforges JF. Erythrocyte DPNH dependent diaphoraselevels in infants. Pediatrics 37: 991–993, 1966.

2. Beutler E, Baluda MC. Methemoglobin reduction. Studies of the inter-action between cell populations and of the role of methylene blue. Blood22: 323–333, 1963.

3. Board PG. NADH-ferricyanide reductase, a convenient approach to theevaluation of NADH-methaemoglobin reductase in human erythrocytes.Clin Chim Acta 109: 233–237, 1981.

4. Borgese N, Pietrini G, Gaetani S. Concentration of NADH-cytochromeb5 reductase in erythrocytes of normal and methemoglobinemic individ-uals measured with a quantitative radioimmunoblotting assay. J ClinInvest 80: 1296–1302, 1987.

5. Bradberry SM. Occupational methaemoglobinaemia. Mechanisms ofproduction, features, diagnosis and management including the use ofmethylene blue. Toxicol Rev 22: 13–27, 2003.

6. Brin M, Yonemoto RH. Stimulation of the glucose oxidative pathway inhuman erythrocytes by methylene blue. J Biol Chem 230: 307–317, 1958.

7. Bunn HF, Forget BG, Ranney HM. Human Hemoglobins. Philadelphia,PA: Saunders, 1977, p. 432.

8. Choury D, Leroux A, Kaplan JC. Membrane-bound cytochrome b5reductase (methemoglobin reductase) in human erythrocytes. Study innormal and methemoglobinemic subjects. J Clin Invest 67: 149–155,1981.

9. Dejam A, Hunter CJ, Pelletier MM, Hsu LL, Machado RF, Shiva S,Power GG, Kelm M, Gladwin MT, Schechter AN. Erythrocytes are themajor intravascular storage sites of nitrite in human blood. Blood 106:734–739, 2005.

10. Doyle MP, Pickering RA, DeWeert TM, Hoekstra JW, Pater D.Kinetics and mechanism of the oxidation of human deoxyhemoglobin bynitrites. J Biol Chem 256: 12393–12398, 1981.

11. Duranski MR, Greer JJ, Dejam A, Jaganmohan S, Hogg N, LangstonW, Patel RP, Yet SF, Wang X, Kevil CG, Gladwin MT, Lefer DJ.Cytoprotective effects of nitrite during in vivo ischemia-reperfusion of theheart and liver. J Clin Invest 115: 1232–1240, 2005.

12. Fernandez BO, Lorkovic IM, Ford PC. Mechanisms of ferrihemereduction by nitric oxide: nitrite and general base catalysis. Inorg Chem43: 5393–5402, 2004.

13. Gladwin MT, Raat NJ, Shiva S, Dezfulian C, Hogg N, Kim-ShapiroDB, Patel RP. Nitrite as a vascular endocrine nitric oxide reservoir thatcontributes to hypoxic signaling, cytoprotection, and vasodilation. Am JPhysiol Heart Circ Physiol 291: H2026–H2035, 2006.

14. Gruener N. Ontogenetic development of NADH-dependent methemoglo-bin reductase in erythrocytes of man and rat. J Toxicol Environ Health 1:787–791, 1976.

15. Herzog P, Feig SA. Methaemoglobinaemia in the newborn infant. ClinHaematol 7: 75–83, 1978.

16. Hsieh HS, Jaffe ER. Electrophoretic and functional variants of NADH-methemoglobin reductase in hereditary methemoglobinemia. J Clin Invest50: 196–202, 1971.

17. Huennekens FM, Liu L, Myers HA, Gabrio BW 3rd. Erythrocytemetabolism. Oxidation of glucose. J Biol Chem 227: 253–260, 1957.

18. Hultquist DE, Passon PG. Catalysis of methaemoglobin reduction byerythrocyte cytochrome b5 and cytochrome b5 reductase. Nature 229:252–254, 1971.

19. Hunter CJ, Dejam A, Blood AB, Shields H, Kim-Shapiro DB,Machado RF, Tarekegn S, Mulla N, Hopper AO, Schechter AN,Power GG, Gladwin MT. Inhaled nebulized nitrite is a hypoxia-sensitiveNO-dependent selective pulmonary vasodilator. Nat Med 10: 1122–1127,2004.

20. Jaffe ER. Methemoglobin pathophysiology. Prog Clin Biol Res 51:133–151, 1981.

21. Kiese M. Methemoglobinemia. A Comprehensive Treatise. Cleveland,OH: CRC, 1974.

22. Kinoshita A, Nakayama Y, Kitayama T, Tomita M. Simulation study ofmethemoglobin reduction in erythrocytes. Differential contributions oftwo pathways to tolerance to oxidative stress. FEBS J 274: 1449–1458,2007.

23. Kleinbongard P, Dejam A, Lauer T, Rassaf T, Schindler A, Picker O,Scheeren T, Godecke A, Schrader J, Schulz R, Heusch G, Schaub GA,Bryan NS, Feelisch M, Kelm M. Plasma nitrite reflects constitutive nitricoxide synthase activity in mammals. Free Radic Biol Med 35: 790–796,2003.

24. Klimmek R, Krettek C, Werner HW. Ferrihaemoglobin formation byamyl nitrite and sodium nitrite in different species in vivo and in vitro.Arch Toxicol 62: 152–160, 1988.

25. Kosaka H, Imaizumi K, Imai K, Tyuma I. Stoichiometry of the reactionof oxyhemoglobin with nitrite. Biochim Biophys Acta 581: 184–188,1979.

26. Lan F, Tang Y, Huang C, Zhu Z. Determination of concentration ofcytosolic NADH-cytochrome b5 reductase in erythrocytes from normalChinese adults, neonates and patients with hereditary methemoglobin-emia by double-antibody sandwich ELISA. Acta Haematol 100: 44 –48, 1998.

27. Lo SC, Agar NS. NADH-methemoglobin reductase activity in the erythro-cytes of newborn and adult mammals. Experientia 42: 1264–1265, 1986.

28. May JM, Qu ZC, Cobb CE. Reduction and uptake of methylene blueby human erythrocytes. Am J Physiol Cell Physiol 286: C1390 –C1398,2004.

29. Miyazono Y, Hirono A, Miyamoto Y, Miwa S. Erythrocyte enzymeactivities in cord blood of extremely low-birth-weight infants. Am JHematol 62: 88–92, 1999.

30. Neugebauer JM. Detergents: an overview. Methods Enzymol 182: 239–253, 1990.

31. Passon PG, Hultquist DE. Soluble cytochrome b5 reductase from humanerythrocytes. Biochim Biophys Acta 275: 62–73, 1972.

32. Percy MJ, McFerran NV, Lappin TR. Disorders of oxidised haemoglo-bin. Blood Rev 19: 61–68, 2005.

33. Pluta RM, Dejam A, Grimes G, Gladwin MT, Oldfield EH. Nitriteinfusions to prevent delayed cerebral vasospasm in a primate model ofsubarachnoid hemorrhage. JAMA 293: 1477–1484, 2005.

34. Reinke NB, O’Brien GM. High activity antioxidant enzymes protectflying-fox haemoglobin against damage: an evolutionary adaptation forflight? J Comp Physiol [B] 176: 729–737, 2006.

35. Rockwood GA, Armstrong KR, Baskin SI. Species comparison ofmethemoglobin reductase. Exp Biol Med (Maywood) 228: 79–83, 2003.

36. Ross JD. Deficient activity of DPNH-dependent methemoglobin diapho-rase in cord blood erythrocytes. Blood 21: 51–62, 1963.

37. Sass MD, Caruso CJ, Axelrod DR. Mechanism of the TPNH-linkedreduction of methemoglobin by methylene blue. Clin Chim Acta 24:77–85, 1969.

38. Sass MD, Caruso CJ, Farhangi M. TPNH-methemoglobin reductase defi-ciency: a new red-cell enzyme defect. J Lab Clin Med 70: 760–767, 1967.

39. Scott EM, Duncan IW, Ekstrand V. The reduced pyridine nucleotidedehydrogenases of human erythrocytes. J Biol Chem 240: 481– 485,1965.

40. Smith JE, Beutler E. Methemoglobin formation and reduction in man andvarious animal species. Am J Physiol 210: 347–350, 1966.

1364 FETAL AND ADULT METHEMOGLOBIN REDUCTION

J Appl Physiol • VOL 103 • OCTOBER 2007 • www.jap.org

on May 13, 2010

jap.physiology.orgD

ownloaded from

41. Stein WD. The Movement of Molecules Across Cell Membranes. NewYork: Academic, 1967.

42. Tanishima K, Fukuda N, Takeshita M, Takizawa Y, Kitamura T,Yoneyama Y. Automated determination of red cell methaemoglobinreductase activity by a continuous-flow system for screening hereditarymethaemoglobinaemia. J Clin Pathol 32: 584–589, 1979.

43. Vetrella M, Astedt B, Barthelmai W, Neuvians D. Activity of NADH- andNADPH-dependent methemoglobin reductases in erythrocytes from fetal toadult age. A parallel assessment. Klin Wochenschr 49: 972–977, 1971.

44. Wennmalm A, Benthin G, Petersson AS. Dependence of the metabolismof nitric oxide (NO) in healthy human whole blood on the oxygenation ofits red cell haemoglobin. Br J Pharmacol 106: 507–508, 1992.

45. Young JD, Dyar O, Xiong L, Howell S. Methaemoglobin production innormal adults inhaling low concentrations of nitric oxide. Intensive CareMed 20: 581–584, 1994.

46. Young JD, Sear JW, Valvini EM. Kinetics of methaemoglobin andserum nitrogen oxide production during inhalation of nitric oxide involunteers. Br J Anaesth 76: 652–656, 1996.

1365FETAL AND ADULT METHEMOGLOBIN REDUCTION

J Appl Physiol • VOL 103 • OCTOBER 2007 • www.jap.org

on May 13, 2010

jap.physiology.orgD

ownloaded from