PHAGOCYTE REDOX METABOLISM. Robert C. Allen. U.S. Army Institute of Surgical Research and Clinical Investigation Service. Brooke Army Medical Center.
CHEMILUMINESCENCE AND THE STUDY OF PHAGOCYTE REDOX METABOLISM Robert C. Allen U.S. Army Institute of Surgical Research and Clinical Investigation Service Brooke Army Medical Center Fort Sam Houston, Texas 78234 U. S. A. INTRODUCTION The phenomenon of polymorphonuclear leukocyte (PMNL) chemiluminescence (CL) is a natural consequence of the redox metabolism that follows phagocytosis or chemical stimulation. These metabolic alterations are collectively referred to as the respiratory burst. As the term implies, there is a rapid increase in non-mitochondrial 02 consumption, and a corresponding increase in glucose metabolism via the hexose monophosphate (HMP) shunt-(Karnovsky, 1968; Rossi et al., 1971). Of the mechanisms advanced to explain these metabolic changes, the proposal of Rossi et al., has best stood the test of time (Rossi and Zatti, 1964, 1966). Simply stated, phagocytic or chemical stimulation of the PMNL effects activation of NADPH:02 oxidoreductase (NADPH oxidase) as measured by decrease in Km. Activity of the HMP shunt dehydrogenases is controlled by the availability of oxidized cofactor, NADP+. Therefore, increase in the NADP+/NADPH ratio following oxidase activation results in increase in dehydrogenase activity (Patriarca et al., 1971). The consumption of 02 results from its reduction by the oxidase. In this regard, Iyer et al., (1961) proposed that H202 is a product of the respiratory burst. The generation of H202 was later confirmed by the work of Paul and Sbarra (1968) and Zatti et al., (1968). *The oplnlons or assertions contained herein are the private views of the author and are not to be construed as reflecting the views the Department of the Army or the Department of Defense. 411
F. Rossi et al. (eds.), Biochemistry and Function of Phagocytes © Plenum Press, New York 1982
R.C.ALLEN
412
The H202 could serve as substrate for myeloperoxidase (MPO), a PMNL lysosomal enzyme. Cell-free preparations of MPO are highly microbicidal when presented with H202 and an oxidizable halide cofactor in a mild acid environment. (Klebanoff 1968, 1971; McRipley and Sbarra, 1967).
PMN
+
Bacteria - (opsonized)
I
PMN HMP - Shunt - Activity
NADPH - Oxidase • . 02
NADPH
11
t
0i
i
NADPH
8/or HOOH
Myeloperoxidase. ' 02 CI-
-H
in VivOll!
02
boctericidol octivity fungicidol octivity lipid peroxidotions oryl- hydroxylotions nucleic ocid oxidations chemiluminescence ~ { posslbly thermal relaxation
via diox.ela"n~
Intermedlotes -+ ,...c= 0
• "
in vitro
*
* ~Dye 302 1 + 102
1 hl-' 1 3 Dye ---+Dye - D y e
Fig. 1.
Initially proposed mechanism for PMNL-CL (Allen et al., 1972).
These antecedent observations provided the basis for the following inferential construction. Phagocyte microbicidal activity is an energy requlrlng process. The primary event in this process is utilization of the potential of NADPH for reduction and electronic alteration of 02. This enzymatic action could best be explained by a flavoprotein oxidase, and the evidence available at the time suggested that this was the case. Firstly, dehydrogenation of NAD(P)H is commonly effected by flavoproteins. The coupling of NAD(P)H oxidation to flavoprotein reduction is consistent with the re duction potentials of these redox reactants. Secondly, unlike most metalloprotein oxidases, flavoproteins are not s~nsitive to inhibition by cyanide. Lastly, flavoproteins are capable of directly catalyzing the univalent reduction of 02 to yeild the superoxide anion, ·02 (Knowles et al., 1969). The schema of figure 1 was proposed as a possible mechanistic explanation of the oxidative microbicidal action of the PMNL. Note that superoxide, ·02, was proposed as a product of NADPH oxidase activity (Allen et al., 1972, 1973), but for reasons that will be
CHEMILUMINESCENCE AND PHAGOCYTE REDOX METABOLlSM
413
discussed, it was not considered as a direct microbicidal reactant. Iristead it was assigned an intermediate function in microbicidal action. Acid disproportionation of ·Oi had been reported to yield HZOZ and singlet molecular oxygen, 10Z (Khan, 1970; Sander and Stauff, 1971). The electrophilic reactivity of 10Z might be directed to microbicidal oxygenations. Additionally, HZO Z ' the other product of disproportionation, could serve as substrate for the MPO-catalyzed oxidation of Cl- as originally suggested by Agner (1958). The products of halide oxidation, such as OCl- or ClZ, might further react with HZOZ to yield 10Z (Kasha and Khan, 1970). Thus, 10Z might be generated by two different PMNL mechanisms. Photodynamic oxidations were known to effect potent microbicidal action since the turn of the century (Raab, 1900; Jodlbauer and von Tappeiner, 1905; Blum, 1941). In 1939, Kautsky proposed that photodynamic oxygenation was effected by the action of 10Z. Chemical support for this 10Z mechanism was presented by Foote and Wexler (1964) and Corey and Taylor (1964). Thus if 10Z were generated by the PMNL, it could serve as a microbicidal agent. 10Z is an electrophilic reactant capable of reacting with the substituted double bonds and electron rich regions of a substrate. In many cases, dioxygenation of substrate results in the generation of a dioxetane or dioxetanone product. These dioxygen-ring compounds break down to yield electronically excited carbonyl containing products in relatively high yield. The excited products can relax to ground state by photon emission or chemiluminescence (McCapra, 1968). Therefore, if such electrophilic dioxygenation reactions are involved in phagocyte microbicidal action, there should be an associated light emission or chemiluminescence (CL). A CL response is observed from phagocytosing PMNL. This CL can be quantified by single photon counting techniques, and the integral CL response correlates with the extent of metabolic activation as measured by radiolabeled glucose studies (Allen et al., 197Z). The then proposed mechanism for PMNL-CL was summarized in the following statemerit taken from the original report. "The 10Z may react as an electrophile at certain sites of high electron density, TI systems, within the cell and/or bacteria to form labile dioxetane structures. These structures may then cleave with the formation of electronically excited carbonyl groups which may then relax with light emission," (Allen et al., 1972). Although additional data support the suggested role of 10Z as a microbicidal reactant (Krinsky, 1974; Rosen and Klebanoff, 1976), evidence is now emerging which suggests that the mechanism of dioxygenation may be more complex, and that the typical, gas-phase '6g0 Z may not be the reactant. For example, the preliminary results of Foote (1980) and the unpublished research of N. Krinsky suggest
R.C.ALLEN
414
that one of the major products of PMNL oxygenation of cho1estero1 cannot be adequate1y exp1ained by either a radicalor a 102 me chanism. Alternative mechanisms, both radica1 and non-radica1, for effecting dioxygenations yie1ding light emission will be considered. NADPH OXIDASE:
SUPEROXIDE SYNTHETASE AND PROTON PUMP
Babior et a1. (1973) demonstrated that activation of PMNL does resu1t in the generation of '02 as measured by superoxide dismutaseinhibitab1e cytochrome c reduction. Nitrob1ue tetrazo1ium reduction by '02 has also been reported in the PMNL (Allen et a1., 1974). Superoxide is generated as a product of NAD(P)H oxidase (Allen et a1., 1974; Johnston et a1., 1975; Babior et a1., 1975, 1976). Evidence also suggests that the oxidase is membrane associated, and that the generation of '02 is extracytop1asmic (Goldstein et a1., 1975).
Radioal Redox Reac-tione
Dj! - REOOX REACTlIIIS
Fig. 2. Redox chain of the PMNL with the oxidase as a proton pump. The schema presented in figure 2 depicts the centra1 ro1e of the oxidase in phagocyte metabolism. Note that 02 reduction proceeds by one e1ectron plus one proton; that is, the oxidase cata1yzes the one equiva1ent reduction of 02. As such, the product of reduction is hydrodioxy1ic acid, the conjugate acid of '02' Therefore, in generating '02, the oxidase can function as a proton pump (Allen, 1979) • This oxidase mechanism predicts that phago1ysosoma1 acidification wii1 be rapid and metabo1ica11y driven. In accordance with the pKa of ·02H, and with the pH dependence of '02 disproportionation, the pH of the phago1ysosoma1 space shou1d approach a steady-state va1ue of 4.8 (Allen, 1979). These theoretica1 predictions are in agreement with the experimental resu1ts reported by Ohkuma and Poo1e (1978).
CHEMILUMINESCENCE AND PHAGOCYTE REDOX METABOLlSM
415
RADICAL AND NON-RADICAL MECHANISMS FOR DIOXYGENATION With respect to the chemistry of '02, it is important to appreciate that '02 is not a potent oxidant. In fact, it is a relatively good reductant, and is commonly quantified by its reducing activity (McCord and Fridovich, 1969). It is likewise important to appreciate that radicals are not necessarily reactive. For example, atmospheric 02 is a triplet multiplicity, diradical molecule. It is the relative electronegativity of 02' not its radical character, that allows it to serve as the terminal electron acceptor of redox metabolism. Radicals tend to react with radicals. The radical character of atmospheric 02 imposes a mechanistic barrier to thermodynamically allowed reactions of 02 with non-radical biological molecules. Rowever, organic and inorganic radicals readily react with atmospheric 02 (Allen, 1979). The hydroxyl radical, ·OR has been proposed as a PMNL-generated microbicidal agent by Johnston et al., 1975. These authors suggested that 'OR could be generated by the reaction of '02 and R202 as originally proposed by Raber and Weiss (1934). Although this reaction is improbable in "clean" chemical systems, it can proceed by metal catalysis (Ralliwell, 1976; Koppenol and Butler, 1977). The hydroxyl radical is a very potent oxidizing agent. Its reactivity is a consequence of its great affinity for electrons. As such ·OR can effect dehydrogenation of many biological substrates, and assuming that it is generated in close proximity to the target microbe, 'OR could serve weIl as a microbicidal agent. It seems probable that ·OR does participate in PMNL microbicidal action. Rowever, certain chemical realities must be kept in mind when designing experiments to "prove" the involvement of ·OR. One cannot realistica1ly employ ·OR-scavengers in studies of cellu1ar systems. These scavengers work weIl on1y in weIl defined systems, where the availabi1ity of susceptib1e substrates is 1imited. Perusua1 of the data collected by Dorfman and Adams (1973) will demonstrate that the common substrates of cellu1ar systems are as susceptible or more susceptible to reaction with ·OR than commonly employed scavengers. For example, the rate constant for the reaction of ·OR with '02 is 1010M-1s -1 (Sehested et al., 1968). Furthermore, it is difficult to determinewhether inhibition of microbicidal action is due to scavenger action or derangement of physiology when scavengers are used at 10- 1 to 10- 2M concentrations. The reaction of ·OR with '02 has been proposed to yield ORplus the relatively low energy exicted state, 102, but the reaction of ·OR with organic substrate is not sufficiently exergonic to yield a high energy, e1ectronica1ly excited carbonyl product direct1y. Rowever, ·OR, as a radical dehydrogenation agent, could participate with '02 in substrate dioxygenation reactions yielding excited carbonyl functions by the mechanism proposed in figure 3.
416
R. C. ALLEN
RADICAL DIOXYGENATION: Rodicol Dehydrogehotioh: H
e
I
\
/
101 e
C=C
/
---...,~,. IO-H
+
\ /
€El l C-e-C
\
\
Rodicol (Ahioh-Cotioh) Ahhihilotioh: 101 I 10Ie \
/
-----;.~
• $1
10-01
I $ -~ e -C-CI
C-e-C
----)l~
10-01 I I
-C-CI
I
I
\
NON-RADICAL DIOXYGENATION: _ _ _.~
I?-Qle
-c~r!I
Fig. 3.
I
---~~
10-01 I
I
I
I
-C-C-
Radica1 and non-radica1 mechanism of dioxygenation.
There are also non-radica1, non- 1 02 mechanisms for effecting substrate dioxygenation. It is 1ike1y that such a mechanism is operative in MPO-cata1yzed dioxygenations. When presented with H2ü2 and C1-, iso1ated MPO can effect a potent microbicidal action with an associated CL (Klebanoff, 1968; Allen, 1975a,b). MPO is ahalide peroxidase, and the reaction of oxidized halogens with H202 is known to yield 102 probab1y through the generation of OOC1- as an intermediate (Kasha and Khan, 1970). A1though the liberation of 00C1- is not favored in acid conditions, e1ectrophi1ic dioxygenation might proceed by a ch1oroperoxy mechanism as depicted in figure 3. CHEMILUMIGENIC PROBING The native CL of phagocytes provides a window for obse~ving the energentics of microbicida1 oxygenations. The pheonomenon of CI. also demonstrates that human cells are capable of liberating energies greater than 50 kca1/mo1e in a single step (Allen, 1979). Chemi1umigenic probing provides an expanded approach to the study of ce11u1ar 02-redox reactions. In comparison to native CL studies, the use of chemi1umigenic probes prov~des a greater than three order of magnitude increase in sensitivity of measurement, and in the same instance, a1lows a more specific measurement of 02-redox activity (Allen and Loose, 1976; Allen, 1980a, 1980b; Wi1son et a1., 1978).
CHEMILUMINESCENCE AND PHAGOCYTE REDOX METABOLlSM
417
A ohemilumigenic probe (CLP) is defined as an exogenous organic moleeule with a high CL quantum yield that can serve as a substrate for CL measurement of certain types of 02-redox activities. This bystander substrate must be non-toxic at the concentrations used for testing. Specific probes can be employed for assay of relatively specific oxygenation activities. For example, when luminol, 5-amino2,3-dihydro-l,4-phthalazinedione, is employed as a CLP, CL is a result of oxidative dioxygenation; that is, luminol undergoes two equivalent dehydrogenation plus 02 incorporation to yield electronically excited aminophthalate as product. However, when lucigenin, lO,lO'-dimethyl-9,9' -biacridinium dinitrate (DBA++), is employed as a CLP, CL is a result of reductive dioxygenation; that is, DBA++ undergoes a two equivalent reduction plus 02 incorporation to yield electronically excited N-methyl acridone. Reaction can involve the radicalor univalent reducted ·DBA+ as intermediate. The
,....
8~
y
r;-... I
~ ...... x ......,
7~
0')
6~
6 -+l
=
7.57 * x - 1.3
r2 =
o
~.g2 o
0
..c a..
.. ....... 0 > 'cu -+l s: ......
':t:
5~ 4~
o
3~
o
C\J
'-.J
2~
CJ
....... 0
'tn
cu -+l s: ......
10 ~ !SI
.....
(Tl
Ln
co
Granulooy~ee (xl~t-3)
Fig. 4.
co
/
~.5
ul
Integral CL as a function of the number of granulocytes in O.5pl of whole blood. Conditions: luminol 5.0pM; serumopsonified zymosan, 50pg; albumin-veronal buffer complete with divalent cations and glucose to a final volume of 2.Oml. CL signal was detected by a single photon counting using an EMI 9829A bialkali photomultiplier tube.
R.C.ALLEN
418
radical-radical, anion-cation annihilation reaction of ·DBA+ with '02 can directly yield the dioxetane (Allen, 1980b). Unlike the luminol-CL response DBA++-CL can be inhibited by unit quantities of superoxide dismutase (manuscript in preparation). At this meeting, a great interest has been expressed regarding methods for quantifying phagocytic function. The data presented in figure 4 is offered as a preview of a CLP technique designed for routine quantification of phagocyte 02-redox response to phagocytosis. The blood donors for this study were healthy laboratory personnel and healthy pregnant females in early labor. The latter group allowed investigation of a physiologic leukocytosis in the absence of infection. Microliter aliquots of EDTA-anticoagulated whole blood were removed from blood specimens collected from routine blood counts. The whole blood aliquots were then diluted one to one-hundred with normal saline, and the equivalent of one-half microliter of whole blood was added to counting vial containing the CLP luminol in albumin-veronal buffer complete with glucose and divalent cations. Stimulation was initiated by addition of serum-opsonified zymosan, and the CL response was automatically monitored every thirteen minutes over a two hour period. The data shown was obtained in three days of routine testing. The post-venipuncture age of the blood specimens at the time of testing varied from two to eight hours, and the contribution of monocytes to CL was not taken into account. No CL response was obtained from any preparation following heat treatment at 56 0 C for 10 minutes. This and other microtechniques are presently being developed and tested as objective clinical laboratory procedures for quantifying the specific dioxygenation activity of phagocytes. ACKNOWLEDGEMENTS I wish to thank the U. S. Army Medical Research and Development Command for allowing my participation in this conference. The research data presented is a product of R&D project 1f3A16ll01A91C and D.C.I. project IfC-5-79. REFERENCES Agner, Allen, Allen, Allen,
K., 1958, Proc. Int. Congr. Biochem. 4th Vienna 15:64. R. C., 1975a, Biochem. Biophys. Res. Commun. 63:675. R. C., 1975b, Biochem. Biophys. Res. Commun. 63:684. R. C., 1979, in: "Frontiers in Biology, Vol. 48", J.T. Dingle, P. J. Jacques, and I. H. Shaw, eds., p. 197, North Holland, Amsterdam.
CHEMILUMINESCENCE AND PHAGOCYTE REDOX METABOLlSM
419
Allen, R. C., 1980a, in: "Liquid Scintillation Counting: Recent Applications and Developments, Vol.· 2," C. T. Peng, D. 1. Horrocks, and E. L. Alpen, eds., p. 37, Academic Press, NY. Allen, R. C., 1980b, "Bioluminescence and Chemiluminescence," M. DeLuca and W. D. McElroy, eds., (in press) Academic Press, New York. Allen, R. C., and Loose, L. D., 1976, Biochem. Biophys. Res. Commun. 69:245 • .Allen, R. C., Stjernholm, R. L., and Steele, R. H., 1972, Biochem. Biophys. Res. Commun. 47:679. Allen, R. C., Stjernholm, R. L., Benerito, R. R., and Steele, R. H., 1973, in: "Chenliluminescence and Bioluminescence," M. J. CormieI: D. M. Hercules, and J. Lee, eds., p. 498, Plenum Press, New York. Arneson, R. M., 1970, Arch. Biochem. Biophys. 136:352. Babior, B. M., Kipnes, R. S., and Curnutte, J. T., 1973, J. Clin. lnvest. 52:74l. Babior, B. M., Curnutte, J. T., and Kipnes, R. S., 1975, J. Clin. lnvest. 56:1035. Babior, B. M., Curnutte, J. T., and McMurrich, B. J., 1976, J. C1in. lnvest. 58:989. B1um, H. F., 1941, "Photodynamic Action and Diseases Caused by Light," Reinhold, New York. Corey, E. J., and Tay1or, W. C., 1964, J. Amer. Chem. Soc. 86:3881. Dorfman, 1. M., and Adams, G. L., 1973, "Reactivity of the Hydroxyl Radica1 in Aqueous Solution," Natl. Bureau Stds., NSRDS-46. Foote, C. S., and Wex1er, S., 1964, J. Amer. Chern. Soc. 86:3879. Foote, C. S., Abaker1i, R. B., Clough, R. L., and Shook, F. C., 1980, in: "Bio1ogica1 and C1inica1 Aspects of Superoxide and Superoxide dismutase," W. H. Bannister and J. V. Bannister, eds., p. 222, E1sevier/North Holland, New York. Goldstein, I. M., Roos, D., Kaplan, H. B., and Weissmann, G., 1975, J. C1in. lnvest. 56:1155. Haber, F., and Weiss, J., 1934, Proc. Roy. Soc., Ser A 147:332. Ha11iwe11, B~, 1976, FEBS Lett. 72:8. lyer, G. Y. N., Islam, M. F., and Quastei, J. H., 1961, Nature 192:535. Jod1bauer, A., and von Tappeiner, H., 1905, Deut. Arch. K1in. Med. 82:520. Johnston, R. B., Kee1e, B. B., Misra, H. P., Lehmeyer, J. E., Webb, L. S., Bachner, R. L., and Rajagopa1an, K. V., 1975, J. C1in. lnvest. 55:1357. Karnovsky, M. L., 1968, Semin. Hemato1. 5:156. Kasha, M., and Khan, A. U., 1970, Ann. N.Y. Acad. Sei. 171:5. Kautsky, H., 1939, Trans. FaradaySoc. 35:216. Khan, A. U., 1970, Science 168:476. K1ebanoff, S. J., 1968, J. Bacteriol. 95:2131. K1ebanoff, S. J., 1971, Annu. Rev. Med. 22:39. Know1es, P. J., Gibson, J. F., P1ick, F. M., and Bray, R. C., 1969, Biochem. J. 111:53. Koppeno1, W. H., and Butler, J., 1977, FEBS Lett. 83:1.
420
R.C.ALLEN
Krinsky, N. I., 1974, Science 186:363. McCapra, F., 1968, J. C. S. Chem. Commun. 155. McCord, J. M., and Fridovich, I., 1969, J. Bio1. Chem. 244:6049. McRip1ey, R. J., and Sbarra, A. J., 1967 J. Bacterio1. 94:1425. Ohkuma, S., and Poo1e, B., 1978, Proc. Nat1. Acad. Sei. USA. 75:3327. Patriarca, P., Cramer, R., Monca1vo, S., Rossi, F., and Romeo, D., 1971, Arch. Biochem. Biophys. 145:255. Pau1, B., and Sbarra, A. J., 1968, Biochem. Biophys. Acta. 156:168. Raab, 0., 1900, Z. Bio1. 39:524. Rosen, H., and K1ebanoff, S. J., 1977, J. Bio1. Chem. 252:4803. Rossi, F., and Zatti, M., 1964, Brit. J. Exp. Patho1. 45:548. Rossi, F., and Zatti, M., 1966, Biochem. Biophys. Acta. 121:110. Rossi, F., Patriarca, P., and Romeo, D., 1971, in: "The Reticu1oendothe1ia1 System and Immune Phenomena," N: R. DiLuzio, ed., 191, Plenum Press, New York. Sander, U., and Stauff, J., 1971, Anales Asoc. Quim. Argent. 59:149. Sehested, J., Rasmussen, O. L., and Fricke, H., 1968, J. Phys. Chem. 72:626. Wi1son, M. E., Trush, M. A., Van Dyke, K., Ky1e, J. M., Mu11ett, ~ Immuno1. Meth. 23:315.
CHEMILUMINESCENCE AND PHAGOCYTE REDOX METABOUSM
421
DISCUSSION Edited by: R.C. Alle~ ROMEO: I have a very short technical question. How long does it take to run one of the CL assays you have just described? ALLEN: Using previously prepared vials, the assay requires approximately thirty minutes to set up fifty samples. The reaction is allowed to continue automatically for two hours; however, good correlations between CL and the number of functional granulocytes are obtained in one hour. Using one technician with two counters, we perform more than two hundred assays per day. SEGAL: You showed that azide and cyanide inhibited luminol luminescence and enhanced that of lucigenin. You also showed that SOD inhibited lucigenin luminescence. This suggests that myeloperoxidase might be the natural substance that promotes the conversion of 02 to H2 0. ALLEN: Ohkuma and Poole (see references) and others have estimated the intralysosomal pH as approximately 4.8. At this pH, the rate of proton-disproportionation of O~'yielding H2 0 2 plus 02 is greater than 7 -1 -1 -. 10 M s . Therefore, 02 can be non-enzymatically converted to H202, and then MPO can effect reduction of H202 to H20. MPO is in many ways analogous to cytochrome oxidase of the mitochondria. Both enzymes contain two a-type cytochromes, each with a different reactivity, and both enzymes yield H20 as a product. WEVER: In one of your slides you showed that NADPH oxidase contains two flavins. Do you have any empirical evidence for that proposal? ALLEN: There is no direct empirical evidence for this proposal. it is very difficult to obtain the enzyme in quantities that would allow analysis. However, a two flavin system is consistent with a semiquinone mechanism of catalysis. Similar flavoprotein systems have this type of two flavin structure (Massey and Gibson, 1964, Fed. Proc. 23: 18; Masters et al., 1965, J. Biol. Chem. 240: 291). A two flavin structure might also obscure the detection of the semiquinone signal by electron spin resonance.
