Role of oxygen in phagocyte microbicidal action.

Immune information in the form of inflammatory mediators directs phagocyte locomotion and increases expression of opsonin receptors such that contact with an opsonized microbe results in receptor ligation and activation of microbicidal metabolism. Carbohydrate dehydrogenation and O2 consumption feed reactions that effectively lower the spin quantum number (S) of O2 from 1 to 1/2 and finally to 0. Oxidase-catalyzed univalent reduction of O2 (S = 1; triplet multiplicity) yields hydrodioxylic acid (HO2) and its conjugate base superoxide, O2- (S = 1/2; doublet multiplicity). Acid or enzymatic disproportionation of superoxide yields H2O2 (S = 0; singlet multiplicity). Haloperoxidase catalyzes H2O2-dependent oxidation of Cl- yielding HOCl (S = 0), and reaction of HOCl with H2O2 yields singlet molecular oxygen, 1O2 (S = 0; singlet multiplicity). The Wigner spin conservation rule restricts direct reaction of S = 1 O2 with S = 0 organic molecules. Lowering the S of O2 overcomes this spin restriction and allows microbicidal combustion. High exergonicity dioxygenation reactions yield electronically excited carbonyl products that relax by photon emission, i.e., phagocyte luminescence. Addition of high quantum yield substrates susceptible to spin allowed dioxygenation, i.e., chemiluminigenic substrates, greatly increases detection sensitivity and defines the nature of the oxygenating agent. Measurement of luminescence allows high sensitivity, real-time, and substrate-specific differential analysis of phagocyte dioxygenating activities. Under assay conditions where immune mediator and opsonin exposure are controlled, luminescence analysis of the initial phase of opsonin-stimulated oxygenation activity allows functional assessment of the opsonin receptor expression per circulating phagocyte and can be used to gauge the in vivo state of immune activation.


Phagocyte Reduction and Oxygenation Activities
Phagocyte microbicidal action is a dynamic, metabolically linked process whereby reduction potential provides the driving force to convert molecular oxygen (02) into effective oxygenating agents. Increased phagocyte metabolism of glucose through the dehydrogenases of the hexose monophosphate shunt in combination with increased nonmitochrondrial 02 consumption is collectively referred to as phagocyte respiratory burst metabolism (1)(2)(3). Glucose dehydrogenation generates the equivalents for univalent reduction of 02' that in turn initiate the transducing reactions required for phagocyte generation of oxygenating agents. Microbe killing results from the reactions of these metabolically generated oxygenating agents with molecular components of the target microbe.
On initial consideration, the mobilization of reduction potential might seem an unlikely initial step in the direction of phagocyte generation of oxygenating agents. This apparent paradox is resolved by taking a quantum mechanical perspec-tive. As predicted by the Pauli exclusion principle, the 02 we breathe is a diradical molecule with one electron occupying each of its two pi antibonding (7*) orbitals, and according to Hund's maximum multiplicity rule, both electrons have the same spin quantum number (s), i.e., both electrons will have an s value equal to 1/2 or -1/2 (4). As such, the total or net spin quantum number (S) of 02 is either 1 or -1. Multiplicity is a spectroscopy expression related to S by the equation 12S1+1. Thus, in its lowest energy or ground state, 02 is a triplet multiplicity diradical molecule, i.e., 12(1)1+1=3, and is highly paramagnetic (5). Spin Conservation Symmetry must be conserved in all chemical reactions. The Wigner spin conservation rules define the tendency of a reacting system to conserve spin angular momentum; i.e., the total orbital angular momentum of all the electrons of a reacting system is conserved (5)(6)(7). This conservation principle is illustrated by the Wigner-Witmer correlation rules presented in Table 1 (8). The more familiar Woodward-Hoffman rules define the conservation of orbital angular momentum of each electron separately. Reaction allowedness and feasibility require conservation of both total and individual symmetries (9).

Chemical Combustion
We all have a first-hand experience with the phenomenon of burning. Hold a lit match to a piece of paper and the paper will burn. In this case, the match provides the activation energy required to spread the fire to the paper, i.e., to initiate combustion. Although ethylene does not spontaneously react with 02, combustion is readily initiated with a spark. Even highly exergonic chemical combustion can require relatively large activation energies. Once initiated, the energy liberated by reaction maintains propagation.
Chemical combustion can be best understood as a radical (S > 0) propagation reaction. Initiation of burning requires the application of energy sufficient to produce homolytic cleavage of the organic molecular bonds, i.e., to convert the S= 0 substrate molecule ('substrate) into S=1/2 reactants 2reactants): substrate + heat -> 2 2reactants. [5] The superscript preceding the symbol indicates the multiplicity of the symbolized reactant. The doublet multiplicity (S= 1/2) radicals produced by homolytic cleavage can react with triplet multiplicity (S= 1) 302 by radical-radical orbital overlap to yield covalent bonding, 2reactant + 302 o 2product + AG [6] as described in Table 1. The S= 1/2 radical product of this oxygenation plus the free energy liberated ensure additional reaction with S= 1 triplet multiplicity 302, 2product +302 2product + AG [7] to produce S=112 radical products as required for radical propagation, i.e., continued burning, until the substrate or 302 is depleted.
Reaction with another S= 1/2 radical, 2product + 2reactant --'product + AG [8] i.e., radical-radical annihilation (S = 1/2 -1/2 = 0), terminates the reaction and yields a singlet multiplicity product as shown in Table 1 Phagocytes exert a broad and lethal microbicidal action using the simple ingredients of metabolically generated reduction potential and 02' The character of respiratory burst metabolism suggests the possibility that phagocyte microbicidal action might involve combustionlike oxygenation reactions, but are phagocytes energetically and mechanistically capable of changing the spin quantum number of 02 from S= 1 to S= 1/2, and finally to S= O? Doing so would overcome the symmetry restrictions to 02 reactivity and, as such, the electrophilic reactive potential for direct 02 reactivity would be realized. Chemical combustion (burning) requires radicalization of substrate, i.e., increasing the value of S, to guarantee reaction with triplet multiplicity 02. Phagocyte microbicidal combustion requires deradicalization of 02, i.e., decreasing the value of S, to guarantee its reaction with singlet multiplicity organic molecules (13). Note that reaction by either pathway satisfies the spin conservation requirements.
Activation of phagocyte respiratory burst metabolism is linked to lowering the Michaelis constant of membrane-associated NADPH:02 oxidoreductase for its reduced substrate NADPH (2). The univalent reduction of 02 catalyzed by this flavoprotein oxidase (14-17): 2 2H (NADPH + H+) + 2 302 2 2 2HO2 <-+2 0 + 2 H+ [9] is the first step in lowering the S value or multiplicity of 02. Hydrodioxylic acid (2HO2; perhydroxylic acid) and its conjugate base superoxide (2O), the products of univalent reduction, are doublet multiplicity S = 1/2 molecules. Hydrodioxylic acid has a plA of 4.8, and its generation by the activated oxidase may be a major factor in the dynamic acidification of the relatively small phagolysosomal space (18). When the pH of the space is 4.8, the ratio of HO2 to°2 is unity. As the pH approaches the 2~~~~~~~~~p K), the anionic repulsion that prevents direct disproportionation of 20is no 2 longer a barrier to reaction, and as such, the doublet-doublet annihilation reaction, 2HO +20 + H+ H2O2 + 2 [10] approaches maximum rate, i.e., 8 x I07 M-1sec-1 (19). Note that Equation 10 is the same type of radical termination reaction described by Equation 8. If the reaction is a direct annihilation through an S= 0 surface, both H202 and 02 will be in the S= 0, singlet multiplicity state as described in Table 1 (20). On the other hand, an indirect stepwise disproportionation involving high multiplicity catalysts, e.g., superoxide dismutase (21), can yield the mixed spin products IH 0 and ground state 30 2 2 2 (22). Superoxide (2O-) can also participate in doublet-doublet annihilation reaction with nitric oxide (2NO), another S= 1/2 reactant, 2NO 20-OON0- [11] to yield the peroxynitrite anion. As predicated by the spin symmetry conservation rules, neither 20 nor 2NO reacts readily with S= 0 organic or biologic molecules, but since both have a common radical S= 1/2 symmetric character, they readily react with each other. The reported rate constant for Reaction 1 1 is 4 x 107 M-'sec-' (23,24). Radical annihilation via a S= 0 surface generates a product, i.e., peroxynitrite, with S= 0 symmetry in common with most organic and biologic molecules, and as such, symmetry is not a barrier to direct reaction of peroxynitrite with these S= 0 molecules (25).

Haloperoxidase Activity
Although metabolically costly, the reduction of 02 to H202 is necessary to eliminate the S= 1 character that limits its direct reactivity. H202, a S= 0 molecule, can participate in spin allowed reactions with S= 0 organic and biologic molecules. However, H202 is a weak acid with a pKa of 11.65, and in its protonated form, H202 is relatively unreactive. The direct reactive capacity of H202 increases with alkalinity, e.g., the Dakin reaction. At a physiologic pH of Environmental Health Perspectives PHAGOCYTE OXYGENATING ACTIV TY 7, the HOT/H202 ratio is approximately 1/100,000 and, as such, the actual concentration of HO-available as a microbicidal oxygenating agent is modest.
Granulocytic leukocytes and blood monocytes contain haloperoxidases, i.e., myeloperoxidase and eosinophil peroxidase, that exert a lethal microbicidal action (26). These H202:halide oxidoreductases catalyze the oxidation of halide (X7) to hypohalous acid (HOX) and the reduction of H202 to H20 (27,28). This oxidation-reduction reactive coupling can be described in the Nernst equation: [12] AE is the change in potential (volts), R is the gas constant, T is the absolute temperature, n is the number of electrons per gram-equivalent transferred, F is a faraday, and In is the natural log of the reactants and products as shown (29). The influence of pH and type of halide on the AE and AG of the reaction is shown in Table 2.
The exergonicity of either chloride (CF) or bromide (Br-) oxidation, the primary reaction catalyzed by haloperoxidase, increases directly with proton availability. In the secondary reaction of Table 2, HOX, the product of the primary reaction, reacts with a second molecule of H202 to yield singlet molecular oxygen (30). Note that for this secondary reaction, exergonicity is directly related to halogen electronegativity and inversely related to proton availability. When [H202] is relatively low, some HOX can accumulate and react with biologic molecules, e.g., chloramine formation. Since both HOX production and consumption are dependent on [H202], and since the reaction of HOX with H202 is essentially diffusion controlled, it is likely that most HOX will be consumed in the secondary reaction. It is also likely that chloramines can react with H202 to yield 102.
The first-order relaxation of 'Ag 02 to ground state 302 yields a 1268-nm photon (30), and measurement of this infrared emission has been taken as direct spectroscopic proof of 102 generation by haloperoxidase (31). A 1268-nm emission has also been reported from phorbol myristate acetate (PMA) stimulated eosinophil leukocytes but not from neutrophil leukocytes (32 (34). Using Equations 9 and 10, there is the possibility that one 102 and Luminescence is an energy product of phagocyte microbicidal activity (13). In fact, phagocyte luminescence was first observed during experiments to test the hypothesis that changing the S of 02 from 1 to 1/2 to 0 was required for effective 02dependent microbicidal action. Lowering the S of 02 from 1 to 0 would involve transductions driven by respiratory burst metabolism. If the resulting low multiplicity reactants participate in microbicidal oxygenation reactions, then it is likely that a portion of the reactions proceed via endoperoxide or dioxetane intermediates to Volume  Although of theoretical importance, there are several limitations to using native luminescence to quantify phagocyte oxygenation activity. Typically, native substrates are of relatively low luminescence quantum yield, and the nature of the substrate varies with the microbe and the test condition. A chemiluminigenic substrate (C) is an organic molecule susceptible to dioxygenation with high quantum (luminescence) yield. Introducing a C to the phagocyte test systems increases the luminescence yield magnitudinally (35,36). Using a C with relatively well-established reactive characteristics also defines the nature of the oxygenating agent (X) measured (37). As such, the luminescence intensity (dLldt) is functionally and quantitatively linked to both the concentration of oxygenating agent ([X]) generated by the phagocyte and the concentration of C ([C ]) available for reaction: [13] where k is the proportionality constant.
When the [C] is much greater than the [X], the reaction order approximates zero with respect to C, i.e., C is not rate limiting, and the relationship simplifies to: dLIdt = k[X] [14] Phagocyte oxidase function can be measured as the luminescence product of dimethylbiacridinium (DBA++; lucigenin) dioxygenation (38). Under alkaline conditions, i.e., as the pH approaches the pKa of H202, lucigenin reacts directly with peroxide, IDBA2+ + 1HO2 (102 ) 2 IN-methylacridone + photon [15] Under neutral to mildly acid conditions, DBA++ can be univalently reduced resulting in change from the S= 0 to the S= 1/2 state, IDBA2+ + e-(2Q-) _> 2DBA + (02) In its radicalized form, 2DBA' can directly react with 20in a doublet-dou-2 blet annihilation via a S = 0 surface, 2DBA+ +202_< 2 lN-methylacridone + photon [17] One of the two N-methylacridone produced will have an excited ketone carbonyl function that can relax to ground state by photon emission. Note that both the onestep peroxidation reaction and the two-step DBA++ does not react with HOC1 or 102 to yield luminescence and, as such, DBA++ is not a substrate for the measurement of haloperoxidase activity. In fact, these oxidizing and dioxygenating agents will tend to competitively inhibit oxidasedependent DBA++ luminescence. The net reaction by any of the above considered pathways is: DBA2+ +02 +2e -2 lN-methylacridone + photon [18] and, as such, can be considered as a reductive dioxygenation (RDOX) reaction. The net dioxygenation of a cyclic hydrazide, e.g., luminol, to yield luminescence does not require electrons or reducing equivalents. Haloperoxidase-containing phagocytes can catalyze dioxygenation of S= 0 luminol via the sequential two equivalent oxidation: 'luminol + 'HOClI diazaquinone + 'Cl + 1H20 [19] followed by a two equivalent reductive dioxygenation: 'diazaquinone + 'H202 < aminophthalate + N2 + photon [20] The possibility also exists that 102 or a related singlet multiplicity dioxygenating agent, e.g., 0-0-Cl-, can directly dioxygenate luminol in a nonradical, spinallowed manner to yield electronically excited aminophthalate. This pathway is suggested by the electrophilic reactivity of '02 with it-systems and its tendency to form dioxetanes and endoperoxides (36). Although presently unproved, the haloperoxidase catalyzed dioxygenation of luminol might also proceed by the reaction: 'luminol + I02 aminophthalate + 'N2 + photon [21] The essential relatedness of the luminol luminescence reactions catalyzed by either mechanistic pathway can be appreciated by recalling that the reaction of HOC1 with H202, the sequential reactants of Equations 19 and 20, yields 02, the reactant of Equation 21. Note the secondary reaction of Table 2.

Estimating Pagocyte Opsonin
Receptor Reserve: Theory The humoral-phagocyte axis of the inflammation response is essentially an information-effector system directed to the task of killing pathogenic microbes. Tissue injury or infection elicits the synthesis and release of immunologic information in the form of cytokines and other inflammatory mediators. These mediators affect the expression of membrane receptors as required for effective phagocyte-endothelial contact, diapedesis and locomotion through the interstitial space to the site of injury or infection. Expression of phagocyte opsonin, i.e., complement and immunoglobulin, receptors increases with the level of exposure to a diverse range of inflammatory mediators, including complement-derived C5a (complement anaphylatoxin), cell-derived platelet activator factors (PAF), leukotrienes, tumor necrosis factors (TNF), colony-stimulating factors (CSF), and certain interleukins such as IL-8, as well as microbe-derived factors such as N-formylmethionyl peptides. It should be emphasized that opsonin receptors are rapidly expressed on exposure to mediators and that this increased expression does not require neosynthesis or prolonged incubation. Opsonin receptor expression can be increased with little or no activation of respiratory burst metabolism. However, opsonin-opsonin receptor ligation triggers respiratory burst metabolism. Under properly controlled conditions of testing, phagocyte opsonin receptor expression can be functionally linked to the acceleration phase of the respiratory burst and measured as luminescence (42)(43)(44).
Phagocyte opsonin receptor expression is dependent on exposure to inflammatory mediators, R/P = k[P] [I] [26] where R/P is the opsonin receptor (R) expressed per phagocyte (P), k is the proportionality constant, [P] is the concentration of phagocytes and [I] is the concentration of inflammatory mediators capable of increasing opsonin receptor expression.
[Imax] is a concentration of inflammatory mediator sufficient to produce maximum opsonin receptor expression Rmax/P.
Phagocyte receptor ligation of opsonins, such as complement and/or immunoglobulin-coated microbes, activates respiratory burst metabolism yielding oxygenating agents (X), where dXldt is the rate of oxygenating agent generation, [0] is the concentration of opsonin, and the other components are as previously described. If the conditions of testing are set such that [0] is much greater than [R], the equation simplifies to, dXldt = k[R]. [28] Since dLIdt is a function of [X], it follows that dLIdt = k[R]. [29] A commercially available luminescence system is now available for measurement of basal and stimulated phagocyte oxidase and oxidase-driven haloperoxidase activities. The MORE value is obtained by simultaneously testing an equivalent quantity of the same blood specimen with the same nonlimiting concentrations of 0 and C but with sufficient inflammatory mediator, i.e., Imax, to ensure maximum opsonin receptor expression but minimum azurophilic degranulation and minimum direct activation of respiratory burst metabolism. The resulting luminescence response to 0 represents maximum receptor activity, dLm xldt = k[Rmax] = k[P][Imax] [31] The luminescence responses L and Lma reflect the CORE and MORE, respectively, and can be presented as a ratio, the CORE/MORE ratio or inflammatory index (42)(43)(44). Low ratio values reflect minimum in vivo exposure to inflamma-tory mediators. The ratio increases to reflect the level of in vivo exposure to inflammatory mediators. The ratio can also be presented in reciprocal form as the percentage opsonin receptor reserve (%ORR) per phagocyte, 1-(L/L) x 100 = %ORR [32] The %ORR decreases in proportion to the degree of in vivo stimulation.
Estimating Phagocyte Opsonin Receptor Reserve: The Application Analysis of Phagocyte Opso

Reeptor Expression
Materials and Methods. Basal and stimulated phagocyte oxidase and oxidase-driven haloperoxidase activities were measured with the AXIS luminescence system and reagents (ExOxEmis Inc., San Antonio, TX). Both the CORE and the MORE per phagocyte were also measured using this system. The K EDTA-anticoagulated whole blood coleected for complete blood count with differential was kept at ambient temperature (20-230C) and tested within 2 hr of venipuncture. Just prior to measurement, 100 pl of whole blood was added to 9.9-ml blood-diluting medium (BDM, an AXIS reagent) for a 1:100 dilution. The diluted blood was loaded into the luminometer (Berthold LB953 AXIS modified, Wildbad, Germany) for automatic injection into prefabricated test tubes containing the indicated coating of inflammatory mediator or PMA. Tubes coated with PMA were used for opsonin receptorindependent measurement of oxidase and oxidase-driven haloperoxidase activities.
On initiation of measurement, the luminometer injected 600 pl of either luminol balanced salt solution (LBSS, an AXIS reagent) or dimethylbiacridinium (DBA++, i.e., lucigenin) balanced salt solution (DBSS, an AXIS reagent), and 100 pl of diluted blood, i.e., 1 pl equivalent of whole blood, into each tube. LBSS and DBSS contain non-rate-limiting quantities of luminol and DBA++ as the respective chemiluminigenic substrate (C) to ensure that the luminescence response reflects the quantity of oxygenating agent generated by the activated leukocyte and not the availability of C. All of the luminescence measurements were in triplicate and were taken over a 20-min interval (approximately 1 measurement/1.5 min). The absolute polymorphonuclear neutrophil leukocyte (PMN) count was used to calculate the Volume 102, Supplement 10, December 1994 specific luminescence activity expressed as total counts/PMN/20 min. Alternatively, specific activity can be expressed as counts/phagocyte/20 min. The total phagocyte count is the total leukocyte count minus the lymphocyte count.
Opsonin receptor-dependent oxidasedriven MPO activity was measured with LBSS as the C medium. Human complement opsonized zymosan (hC-OpZ, an AXIS reagent) was used as the opsonin.
Response to this opsonin was tested in the absence and presence of sufficient inflammatory agent to produce maximum neutrophil opsonin receptor expression. The data presented in Table 3 were obtained using uncoated tubes, recombinant human C5a-coated tubes (C5a, 20 pmole tubes, an AXIS reagent) and PAFcoated tubes (PAF, 10 pmole tubes, an AXIS reagent) with 100 pl complementopsonized zymosan (108 zymosan particles per test). Figure 1 presents the plot of luminescence velocity, dLldt, expressed as megacounts per minute (CPMx 1 06) (the ordinate) versus time in minutes (the abscissa) for both CORE (lower curve) and C5a-MORE (upper curve) responses to 0, hC-OpZ. The leukocyte count was 5100/p1 with an absolute differential count of 3200 segmented PMN, 205 monocytes, 51 eosinophils and 1630 lymphocytes/pl. Each microliter of blood tested contained 3200 PMN and 3456 phagocytes. The %ORR is calculated from the initial 10-min integral response. During this initial period, activation is causally linked to opsonin receptor expression. After this initial period, phagocyte metabolic capacity begins to exert a limiting effect on the luminescence response. The initial integral CORE (i.e., L) response is 2.6 x 106 counts/10 min, and the C5a-MORE (L max) response is 13.0 x 106 counts/10 min. Therefore, the inflammatory index is 0.20 and the %ORR is 80. This %ORR value is within the normal range of 77±10 (SD) for healthy subjects (n = 20). Table 3 presents oxidase-dependent and oxidase-haloperoxidase-dependent (both opsonin receptor dependent and independent activities) integral 20-min specific luminescence responses for the same blood specimen. Imax primed, hC-OpZ-stimulated phagocytes typically show peak dLldt by 15 min. Once peak dLldt is reached, phagocyte metabolic capacity exerts a limiting effect. As such, the 20-min MORE response reflects metabolic capacity. The MORE responses are essentially the same with either C5a or PAF as immunologic modulator.   Table 3. The effects of azide on the CORE and C5a-MORE raw luminescence responses are presented in Figure 2. Although a final concentration of 6 ,iM of azide inhibits luminescence by approximately 80%, the %ORR is increased slightly from 80 to 86%. This small increase may reflect a greater susceptibility of CORE activity to azide inhibition. As shown in Table 3, azide also inhibits PMA-stimulated LBSS luminescence by approximately 80%. Azide does not inhibit DBSS luminescence in response to PMA (10 pmole/test). As shown in Table 3, azide produces a mild augmentation that probably reflects inhibition of MPO-mediated oxidation activity which can interfere with DBA++ Environmental Health Perspectives reductive dioxygenation. The effect of azide on the raw DBSS luminescence responses from PMA-stimulated blood is illustrated in Figure 3. The lower and upper curves depict the responses to PMA (10 pmole/test) in the absence and presence of azide, respectively. All of the luminescence data presented were obtained from less than 0.5 ml blood and required less than 1 hr for setup, measurement, and analysis. Luminescence testing allows rapid, high sensitivity differential analysis of phagocyte oxygenation activities. If the conditions of analysis are properly controlled, opsonin receptor expression can also be measured, and since opsonin receptor expression is in proportion to the degree of inflammatory mediator exposure, analysis of the circulating phagocyte provides a gauge of in vivo inflammation. The AXIS system is ultimately designed for use in patient diagnosis and management (46,47). However, the system also has broad potential for assessing inflammation and antiinflammatory activity using in vivo and in vitro models (48).