JPM Vol 27. No. 2 April 1992:95-100

A Rapid Microtiter Plate Method for the Detection of Lysozyme Release From Human Neutrophils Rocio Moreira-Ludewig Research

Department,

Pharmaceuticals

and Cynthia T. Healy

Division, CUBA-GEIGY Corporation,

Summit, New Jersey,

U.S.A.

An improved method was devised to measure lysozyme secreted from human neutrophils [polymorphonuclear leukocyte (PMN)] using a microtiter plate reader capable of analyzing enzyme kinetics. The assay is an adaptation of the classical photometric method which detects changes in the turbidity of a bacterial suspension, Micrococcus lysodeikticus, caused by the enzymatic activity of lysozyme. A standard curve using chicken egg white lysozyme was generated, and activity was detectable between the range of I and 100 ng/ ml. Leukotriene B4 (LTB4)-induced lysozyme release from human PMN was comparable in both the standard assay and the microtiter plate adaptation with E&o values of 6.5 and 7.2 nM, respectively. Other select stimuli and their receptor antagonists were also used to evaluate the method. Dose-response curves for chemotactic hexapeptide (CHP), recombinant human CSa (rhCSa), and platelet-activating factor (PAF) resulted in EC50 values of 0.14, 0.80, and 542.00 nM, respectively. Inhibition of lysozyme release was studied using receptor antagonists N-r-Boc-L-methionyl-L-leucyl-L-phenylalanine (N-t-Boc), LY223982, and protamine, which are putative inhibitors of formyl peptides (i.e., CHP), LTB4, and C5a, respectively. N-t-Boc inhibited CHP-induced (0.2 nM) enzyme release with an I& of 2 PM; LY223982 blocked LTB4-induced (20 nM) release resulting in an I& of 52 nM; and protamine inhibited rhC5a-induced (1.5 nM) release with an I& of 2 PM. Further studies revealed that CHP, LTB4, and rhC5a were selectively inhibited by their respective antagonists, albeit LY223982 and protamine were also weak inhibitors of CHP and LTB4, respectively. The adaptation of this assay, as demonstrated here, to a microplate format provides a simple, reproducible, and high throughput method that is ideally suited for largescale drug screening. Keywords:

Muramidase;

Neutrophil;

LTB4; C5a; PAF; Formyl peptides

Introduction Fleming and Allison (1922) first described a crude method for detecting lysozyme activity in which the time required to dissolve part or all of a plate culture of Micrococcus lysodeikticus, using tissue rich in this enzyme, was measured. Further modification of this procedure was accomplished by many investigators, including Litwack (19.53, who used a more efficient and accurate optical method than his predecessors to measure the same phenomenon. Lysozyme activity, indicative of polymorphonuclear leukocyte (PMN) ly-

Address reprint requests to: Rocio Ludewig, CIBA-GEIGY Corporation, Building V-l 10, 556 Morris Avenue, Summit, NJ 07901, U.S.A. Received 1 August 1991; revised and accepted 24 October 1991. Journal of Pharmacological and Toxicological Methods 27, 95-100 (1992) 0 1992 Elsevier Science Publishing Co., Inc., 655 Avenue of the Americas, New York, NY 10010

sozyme release, represents one of several methods for the assessment of PMN function. However, extensive analyses performed with PMN have been complicated by their short lifespan in vitro. In addition, several readings must be taken per sample, thus limiting the total number of samples that can be run per day. Consequently, we now report a further adaptation of this lysozyme assay which overcomes these limitations. We developed a lysozyme microtiter plate assay predominantly for investigating PMN enzyme release. A comparison of the standard lysozyme release assay, using leukotriene B4 (LTB4) as the agonist (Lee et al., 1988; Shimazaki et al., 1990), and the microtiter plate assay was performed. The microtiter assay’s utility for other known agonists, and selective antagonists, of PMN activation was also demonstrated using chemo-

1056-8719/92/$5.00

%

JPM Vol 27. No. 2 April 1992:95-100

tactic hexapeptide (CHP) recombinant human C5a (rhCSa), and platelet-activating factor (PAF).

Materials and Methods Chemicals

and Reagents

LTB4 and PAF( I-0-Hexadecyl-2-O-acetyl-snglycero-3-phosphorylcholine) were purchased in ethanol from BIOMOL Research Laboratories (Plymouth Meeting, PA). Cytochalasin b (Cyto b), lysozyme from chicken egg white (20,000 U/mg), chemotactic hexapeptide (N-formyl-NLE-LEU-PHE-NLE-TYRLYS), lyophilized cells of Micrococcus lysodeikticus, protamine sulfate (Grade X), and N-t-Boc-L-methionyl-t_-leucyl-t_-phenylalanine were obtained from Sigma Chemical Co. (St. Louis, MO). LY223982 [(E)-5-(3carboxybenzoyl)-2-((6(4-methoxyphenyl)-5hexenyl)oxy)benzenepropanoic acid] and recombinant human C5a were prepared at CIBA-GEIGY (Summit, NJ). Apparatus All microtiter optical densities (O.D.) were measured using the Thermomax optical plate reader (Molecular Devices, Palo Alto, CA) with a 450 nm (20-nm band pass) filter. Results were compiled and analyzed using the SOFTmax program (Molecular Devices, Palo Alto, CA) from an IBM personal computer linked directly to the plate reader. Isolation of Human PMN Neutrophils were isolated based on an adaptation of the method by Boyum (1968). Approximately 175 ml of uncoagulated human venous blood was obtained for each experiment, dispersed into 50-ml polypropylene tubes containing 15 ml of HESPAN (DuPont, Wilmington, DE), and mixed. Tubes were allowed to stand at room temperature for 20 min or until most of the red blood cells had sedimented. The supernatants were removed and centrifuged for 5-10 min at 400 g. The remaining pellets were diluted in 70 ml of phosphatebuffered saline without calcium and magnesium (PBS w/o metals; GIBCO, Grand Island, NY), and 35 ml of this suspension were placed in each of two polypropylene tubes containing 15 ml of Ficoll-Paque (Sigma, St. Louis, MO). The tubes were then centrifuged for 15 min at 420 g. The mononuclear cell layer was discarded, and the remaining cell pellet was resuspended in 10 ml of PBS w/o metals. Filtered deionized water (20 ml) were added to the suspension for approximately 20 set followed by the same volume of PBS w/o metals at twice the normal concentration in order to lyse and

remove the red blood cells. The cell suspension was mixed and centrifuged for 5 min at approximately 200 g, followed by one wash with PBS w/o metals, and a final resuspension in PBS w/o metals was made at a concentration of 2 x IO’ cells/ml. The PMN remain viable after this isolation procedure and they are maintained at room temperature for immediate use. Caution must be taken to prevent cells from sitting for too long after isolation as this may cause the PMN to become preactivated, thus losing their sensitivity to various stimuli.

Microtiter Plate Lysozyme Release

Assay

A polypropylene microtiter plate (Costar, Cambridge, MA) was used for incubation of PMN (2 x 106/ well) in order to prevent their adherence to the plate and consequent preactivation. All stock reagents including ligands and antagonists were diluted in Hank’s Balanced Salt Solution (HBSS; GIBCO, Grand Island, NY). Additions to the plate, except ligands and cells, were made with a Perkin Elmer Cetus Pro/Pette (Norwalk, CT). Cyto b was incubated with the cells at a final concentration of 5 &ml at 37°C for 5 min prior to the addition of stimuli such as LTB4, rhCSa, CHP, or PAF, for an additional 15 min of incubation. LTB4 stock solutions were maintained in ethanol, rhC5a stocks were stored in 0.01% tween. PBS, CHP, and PAF were both maintained in a dimethyl sulfoxide (DMSO) stock solution. Antagonists, except protamine which is soluble in buffer, were first dissolved in DMSO and then diluted with Hank’s buffer. LY223982, N-t-Boc, and protamine were each coincubated with Cyto b for 5 min prior to cell activation by LTB4, CHP, or rhCSa, respectively. The final concentration of each of the various solvents in the wells was ~0.1%. After incubation, microtiter plates were centrifuged in a Sorvall Technospin R (DuPont, Chadds Ford, PA) tabletop centrifuge (equipped with microtiter plate rotor and carriers) at 4°C for 5 min at 200 g. The supernatants (50 ~1) were transferred to a second polystyrene microtiter plate containing 200 ~1 of a 0.36 mg/ml Micrococcus lysodeikticus suspension in potassium phosphate buffer (0.1 M; pH 6.24). For the egg white lysozyme standard curve, the dilutions of enzyme were made in HBSS, and 50 ~1 of each dilution were placed in the wells as previously described. The plate was immediately read at a wavelength of 450 nm, every minute, for a total of 10 min. Rates of lysis (V,,,; mOD/min) were calculated by SOFTmax from which %Inhibition, E&,, and I&, were determined. A paired t test was performed where statistical determinations were made.

MOREIRA-LUDEWIG AND HEALY MICROTITER PLATE LYSOZYME RELEASE ASSAY

Standard Lysozyme Assay Procedures for measuring lysozyme release from PMN have been described previously (i.e., Goldstein et al., 1975). Briefly, the conditions for the standard assay were as follows: 1) 3 x lo6 PMNs were used per sample; 2) the final volume per tube was 300 ~1; 3) 1.3 ml of a 0.3 mg/ml substrate concentration was placed in each cuvette with 0.2 ml of PMN supernatant; 4) the use of a Beckman DU6 Spectrophotometer for kinetic enzyme measurements; and 5) enzyme release was expressed as the percent of total activity released by PMN in the presence of 0.2% Triton X-100 (Sigma Chemical Co., St. Louis, MO).

Results Estimation of Total Releasable Human PMN

Lysozyme from

The rate of lysis induced by lysozyme was linear in the range of 3-70 &ml. Total release of the enzyme from 2 x lo6 PMN, induced by three freeze-thaw cycles and sonication, resulted in lo-20 ng/ml of lysozyme. Thus it would be expected that stimulus-induced lysozyme levels would not exceed the usable range of the assay, and that the degree of lysis would be proportional to the amount of enzyme released.

10.1 4.2 7.3 7.2 * 1.7”

Time Course of Lysozyme Release rhCSa, PAF, and CHP

Table 1. LTBCInduction of Human PMN Lysozyme Release: Standard Versus Microtiter Plate Assay ECso (nM)

a Paired t test showed methods.

effects of CHP, rhCSa, LTB4, and PAF on lysozyme release from human PMN. All points were measured in quadruplicate, and each curve is representative of three separate experiments performed on different donors. The mean 2 SEM EC50 of CHP, rhCSa, LTB4, and PAF was 0.14 2 0.04 nM, 0.8 5 0.1 nM, 7.2 * 1.4 nM, and 542 t 140 nM, respectively.

9.5 2.5 7.5

In addition to LTB4, CHP, rhCSa, and PAF are known promoters of neutrophil degranulation (Rossi et

Mean k SEM

Figure 1. Dose-dependent

Microtiter

Curves for CHP, rhCSa, LTB4,

1 2 3

(PM)

Standard

Standard

Both the standard and microtiter lysozyme assays were compared in LTBCinduced neutrophil degranulation. As shown in Table 1, there was no significant difference found between the ECso of LTB4 as determined using either method.

Trial Number

OF AGONIST

al., 1988; Goldstein et al., 1975; Smith et al., 1983). In the presence of Cyto b, all of these ligands stimulated PMN release of lysozyme in a dose-dependent manner (Figure 1). Lower concentrations of CHP and rhC5a induced degranulation (0.14 & 0.04 nM and 0.8 + 0.1 nM) more than did LTB4 and PAF (7.2 2 1.4 nM and 542 -+ 140 nM). Furthermore, it is evident that activation of PMN by maximal concentrations of CHP resulted in greater enzyme release than of that by any other chemoattractant. Interestingly enough, O’Flaherty and his coworkers (198 1) demonstrated a similar phenomenon when they compared the percent of total cellular lysozyme released due to fMet-Leu-Phe (FMLP), PAF, and LTB4. Although they did not note or discuss this in their report, it was suggested that one way in which FMLP differed from PAF and LTB4 was that it required extensive calcium uptake for its degranulating action on PMN. Because calcium is involved in the signal transduction mechanism of PMN, this may suggest that formyl peptides may produce a more effective signal. In addition, it has been reported by Jesaitis et al. (1982) that there are receptors for formy1 peptides on specific granules as well as the PMN plasma membrane. Presently, it is not clear if there are granular receptors for the other ligands. Using the standard curve illustrated on Figure 2, we estimated that these agonists induced the release of approximately 3- 10 ng/ml egg white lysozyme equivalents at maximum concentrations.

LTB4-Induced Lysozyme Release: versus Microtiter Assays

Dose-Response and PAF

CONCENTRATION

6.5 ” 2.1” no significant

difference

between

two

with LTB4,

Figure 3 illustrates that maximal lysozyme release occurred after 5-10 min of incubation with LTB4,

JPM Vol 27. No. 2 April 1992:95-100

--t U

1

IO

100

EGG WHITE LYSOZYME (ng/ml)

Figure 2. Chicken egg white lysozyme standard curve from the microtiter plate assay. Each measurement was performed in triplicate.

rhCSa, PAF, and CHP in the microtiter plate assay. Considerable activity was observed only 30 set postinduction. The expression of maximal release for each ligand as %total release was then determined and is as follows: 1) control: 9.7%; 2) 20 nM LTB4: 28.1%; 3) 1.5 nM rhC5a: 35.2%; 4) 1500 nM PAF: 30.2%; and 5) 0.2 nM CHP: 24.8%. The concentrations of ligands used are within the range of the ECTO obtained from each of their respective standard curves (data not shown).

Antagonist Protamine

Effects of N-t-Boc, LY223982, and

Receptor antagonists have been reported for formyl peptides, LTB4, and C5a. Korchak et al. (1984) used a synthetic Boc peptide to inhibit competitively the effects of FMLP. As shown on Table 2, N-t-Boc-methionyl-L-leucyl-L-phenylalanine inhibited 200 pM CHP with an ICsO of 2.1 + 0.6 FM. LY223982, a potent LTB4 antagonist (Jackson et al., 1988; Shappell et al., 1989), inhibited LTB4-induced lysozyme release with an IGo of 52.1 ? 16.5 nM. A basophil C5a receptor antagonist (Olsen et al., 1988), protamine, inhibited CSa-induced PMN lysozyme release with an I& of 2.1 ? 0.5 FM. The data in Table 3 suggest that these antagonists, within the range of concentrations tested,

Table 2. Inhibition Stimulus LTB4

CHP rhC5a

of Lysozyme

EC& Screening 7.20 -c 1.40 0.14 + 0.04 0.80 + 0.10

rhC% PAF

Figure 3. Time course of lysozyme release after stimulation of PMN by LTB4, PAF, rhCSa, and CHP. Enzyme release was measured after 0.5,3,5, 15,30, and 60 min of incubation at 37°C. Each measurement was done in triplicate.

were also more effective inhibitors of their respective agonist.

Discussion We have described a simple, rapid, and versatile adaptation of the standard PMN lysozyme release assay. The method is comparable with the standard assay (Table 1) based on equivalent Et& values for LTBCinduced enzyme release. Although we did not compare the methods using PAF, CHP, and rhCSa, as we will discuss, the results we found in our assay appear to agree with the findings of other investigators (who do not use microtiter assays) with regard to ligand-induced neutrophil degranulation. Omann et al. (1987), using elastase as an indicator, obtained an ECso of 0.1 nM for a formyl peptide and 10 nM for LTBC induced enzyme release. Bokoch and Reed (1981) reported an ECso of 6.5 nM for LTBCinduced lysozyme release in guinea pig PMN. No previous reports have been found in the literature on recombinant human CSa-induced lysozyme release. In our hands, PAF was the weakest stimulus tested. Sturm and Chang (1989) noted that to induce significant enzyme release, concentrations of 1 ~.LMor higher were required for PAF. Contrary to this, Smith et al. (1983) reported an E&I of 20 nM for the Cl6 analog of PAF, and 90 nM for the

Release by LY223982, AWBOC, and Protamine Concentration0 20.0 0.20 1.50

Note: Data show are Mean f SEM of six experiments. Q Concentrations in nanomolars.

Antagonist

Go”

LY223982 N-t-BOC Protamine

52.1 k

16.5

2091.6 2 578.0 2083.3 + 480.0

99

MOREIRA-LUDEWIG AND HEALY MICROTITER PLATE LYSOZYME RELEASE ASSAY

Table 3. Effect of Select Antagonists On Various Stimuli of PMN Lysozyme Release _ GO (nM) Stimulus LTB4 PAF rhC5a CHP

LY223982

Protamine

N-r-BOC

52.1 2 16.5 > 10000.0 > 10000.0 5066.7 ? 1838.7

9333.3 2 544.0 > 10000.0 2083.3 ? 480.0 > 10000.0

> 10000.0 > 10000.0 > 10000.0 2091.6 ? 578.0

Note: Data shown are Mean + SEM of at least three experiments.

Cl8 analog. In our studies, the ECso obtained for the Cl8 PAF analog was 542 nM. Shen et al. (1985) and Dewald and Baggiolini (1987) have observed that betaglucuronidase and elastase release by human PMN was augmented with increasing levels of PAF. Maximal release was not obtained with levels as high as 10 PM. It is clear that further studies would be needed to resolve this issue, however, variables to consider would include the specific PAF analog used and experimental differences in timing and/or methods. Lysozyme release appeared to occur very rapidly (as short as 30 set), as evidenced by the time course data on Figure 3. A 15min incubation with each stimulus for the agonist/antagonist experiments, however, ensured that the maximal concentration of enzyme was released and measured. There is currently much interest in finding inhibitors for various inflammatory mediators (i.e., LTB4 and C5a). In order to evaluate the lysozyme microtiter plate assay as a tool for screening potential inhibitors, we tested the effect of specific receptor antagonists. The LTB4 receptor inhibitor, LY223982, reported by Jackson et al. (1988) to be potent in nanomolar concentrations, produced similar results for LTB4-induced lysozyme release in our assay. A reported weak C5a receptor antagonist, protamine, inhibited PMN enzyme release produced by stimulating with 1.5 nM rhCSa, resulting in an I&,, of 2.1 FM (10.1 pg/ml). Olsen et al. (1988) observed that the ICsO for this compound in the presence of 3.0 nM C5a was 30 kg/ml in a human PMN beta-glucuronidase release assay. Because less C5a was used in our assay, both findings appear to be consistent. Finally, Korchak et al. (1984) used 100 FM of r-Boc-L-Phe-o-Leu-L-Phe-D-Leu-Leu-L-Phe to inhibit completely 50 nM of human PMN fMet-Leu-Phe-induced lysozyme release. In our studies, an ICsO of 2.1 PM for N-t-Boc-r_-Met-L-Leu-L-Phe was obtained using 0.2 nM of CHP to induce PMN activation. In order to assess the specificity of these compounds for their respective ligand receptor in our assay, we tested each antagonist against other stimuli (Table 3). LY223982 was about 100 times more potent in inhibiting LTB4-

induced lysozyme release than CHP-induced release. This is in agreement with the findings of Jackson et al. (1988). Although they were weak inhibitors, protamine and the Boc peptide were more effective in inhibiting the rhC5a and CHP PMN receptors, respectively. Olsen et al. (1988) demonstrated protamine’s selective inhibition of C5a compared with that of FMLP. Boc peptides were originally synthesized to bind to the FMLP receptor competitively, although they have not proven to be very potent inhibitors of formyl peptide stimulation. In conclusion, lysozyme is an effective marker of cellular degranulation of PMN. As demonstrated by the results discussed here, this method may be used in the evaluation of physiologic activators of PMN function, as well as the study of pharmacologic modulators of phagocytic cell activation. In addition, though not addressed in these experiments, the lysozyme microtiter plate assay may also provide a high-throughput screen for lysozyme levels in tissue or body fluids. The authors thank L. Wennogle, R. Armenti, W. Boyar. L. Conder, and N. Galakatos for providing us with recombinant human CSa; and to B. Seligmann, P. Simon, S. Psychoyos, and R. Patrick for their advice and critical review of this manuscript.

References Bokoch GM, Reed PW (1981) Effect of various lipoxygenase metabolites of arachidonic acid on degranulation of polymorphonuclear leukocytes. J Biol Chem 256( 1I):53 17-5320. Boyum A (1968) Isolation of mononuclear cells and granulocytes from human blood. Stand J C/in Lab Invest 97(Suppl):77-89. Dewald B, Baggiolini M (1987) Evaluation of PAF antagonists using human neutrophils in a microtiter plate assay. Biochem Pharmaco1 36(15):2505-2510. Fleming A, Allison VD (1922) Bacteriolytic substance (“lysozyme”) found in secretions and tissues. Br J Exper Pathol 3:252-260. Goldstein IM, Roos D, Kaplan HB, Weissmann G (1975) Complement and immunoglobulins stimulate superoxide production by human leukocytes independently of phagocytosis. J Clin Invest 56: 1155-I 163. Jackson WT, Boyd RJ, Froelich LL et al. (1988) Inhibition of LTB4 binding and aggregation of neutrophils by LY255283 and LY223982. Fed Proc 2(5):(Abstract)4730. Jesaitis AJ, Naemura JR, Painter RG, Sklar LA, Cochrane CG (1982) Intracellular localization of N-formyl chemotactic receptor and Mg2+-dependent ATPase in human granulocytes. Biochim Biophys Acra 719:556-568. Korchak HM, Wilkenfeld C, Rich AM, Radin AR, Vienne K, Rutherford LE (1984) Stimulus response coupling in the human neutrophil. Differential requirements for receptor occupancy in neutrophil responses to a chemoattractant. J Biol Chem 259(12):7439-45. Lee TH, Sethi T, Crea AEG et al. (1988) Characterization of Ieukotriene B3: Comparison of its biological activities with leukotriene B4 and leukotriene BS in complement receptor enhancement, lysozyme release and chemotaxis of human neutrophils. C/in Sci 741467-475.

100

Litwack G (1955) Photometric

JPM Vol 27. No. 2

April1992:95-100

determination

of lysozyme activity.

Proc Sot Exp Biol Med 89:401-403.

O’Flaherty JT, Wykle RL, Lees CJ, Shewmake T, McCall CE, Thomas MJ (1981) Neutrophil-degranulating action of 5,12-Dihydroxy-6,8,10,14-eicosatetraenoic acid and I-0-alkyl-2-O-acetylsn-glycero-3-phosphocholine. Am J Pathol 105:264-269. Olsen UB, Selmer J, Kahl JU (1988) Complement C5a receptor antagonism by Protamine and Poly-L-Arg on human leukocytes. Complement 5: 153-162. Omann GM, Traynor AE, Harris AL, Sklar LA (1987) LTB4-induced activation signals and responses in neutrophils are shortlived compared to formylpeptide. J Immunol 138(8):2626-2632. Rossi AG, McMillan RM, MacIntyre DE (1988) Agonist-induced calcium flux, phosphoinositide metabolism aggregation and enzyme secretion in human neutrophils. Agents Actions 24:272-282. Shappell SB, Smith CW, Gasic AC, Taylor AA, Mitchell JR (1989)

Effects of a specific LTB4 antagonist on human neutrophil function. Fed Proc 3(3):(Abstract)2145. Shen TY, Hwang SB, Chang MN et al. (1985) Characterization of a platelet-activating factor receptor antagonist isolated from haifenteng (Piper futokadsura): Specific inhibition of in vitro and in vivo platelet-activating factor-induced effects. Proc Nafl Acad Sci USA 82~672-676.

Shimazaki T, Kobayashi Y, Sato F, Iwama T, Shikada K (1990) Some newly synthesized leukotriene B4 analogs inhibit LTB4induced lysozyme release from rat polymorphonuclear leukocytes. Prostaglandins 39(4):459-467. Smith RJ, Bowman BJ, lden SS (1983) Characteristics of I-O-Hexadecyl- and I-Octadecyl-2-O Acetyl-Sn-Glyceryl-3-Phosphorylcholine-stimulated granule enzyme release from human neutrophils. C/in fmmuno/ lmmunoparhol28:13-28. Sturm RJ, Chang JY (1989) Neutrophil-derived lipid mediators of inflammation. In Immunopharmacology. Eds., SC Gilman and TJ Rogers. Caldwell, New Jersey: The Telford Press, pp. 89-103.