1-AMINO ACID OXIDASE IN LEUKOCYTES: A POSSIBLE - PNAS
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
1-AMINO ACID OXIDASE IN LEUKOCYTES: A POSSIBLE
D-AMINO-ACID-LINKED ANTIMICROBIAL SYSTEM*
BY MARTIN J. CLINE AND ROBERT I. LEHRER
CANCER RESEARCH INSTITUTE AND DEPARTMENT OF MEDICINE,
UNIVERSITY OF CALIFORNIA MEDICAL CENTER,
SAN FRANCISCO
Communicated by Julius H. Comroe, Jr., January 16, 1969
Abstract.-D-Amino acid oxidase has been identified within the granule frac-
tion of human neutrophilic leukocytes. Leukocyte homogenates and purified
kidney 1-amino acid oxidase can utilize either isolated D-amino acids or some
species of bacteria as substrates for the generation of hydrogen peroxide. When
linked to leukocyte myeloperoxidase in vitro, purified D-amino acid oxidase con-
stitutes a system lethal for certain bacteria. It is proposed that leukocyte
D-amino acid oxidase and myeloperoxidase constitute a biochemically specific
system for the recognition and killing of certain microorganisms.
Phagocytic leukocytes constitute a phylogenetically primitive system of re-
sistance to microbial infection. Reduced to its simplest elements, this system
must perform at least two integrated functions: the ingestion and the killing
of microorganisms. The requirements for particle uptake by leukocytes are
known in considerable detail." 2 The mechanisms of microbial killing are less
well defined, although some weapons in the arsenal of the phagocytic leukocyte
have been extensively studied.3' 4 A leukocyte microbicidal system consisting
of the leukocyte enzyme myeloperoxidase, a halide, and hydrogen peroxide or
a hydrogen peroxide-generating system has recently been described.5 The hy-
pothesis that this system plays a role in the normal defense function of the
human neutrophilic leukocyte is supported by the observation of impaired mi-
crobicidal activity in leukocytes genetically deficient in myeloperoxidase6 or in
postphagocytic hydrogen peroxide generation.7' 8
Several considerations led us to predict the existence in leukocytes of a hydro-
gen peroxide-generating system linked to D-amino acid oxidase. D-Amino acids
are an integral part of the cell walls of many bacterial species9 and have been
found in the products of certain fungi, but have not been conclusively identified
as a natural component of any mammalian tissues. Nevertheless, the activity
of 1-amino acid oxidase is known to occur in certain mammalian tissues and to
be abundant in the liver and kidney. This enzyme, in the presence of molecular
oxygen and the appropriate D-amino acid, catalyzes the generation of a corre-
sponding keto-acid and of hydrogen peroxide.
This communication reports the finding of D-amino acid oxidase in the granule
fraction of mammalian neutrophils. The enzyme utilized intact microorganisms
as a substrate for hydrogen peroxide generation and could be linked with human
myeloperoxidase in vitro to constitute a potent microbicidal system.
Materials and Methods.-Preparation of enzymes: (a) D-Amino acid oxidase: Popula-
tions of either mixed leukocytes or lymphocytes were prepared from the heparinized ve-
nous blood of normal subjects and patients with hematologic diseases.'0 Neutrophil-rich
756
Downloaded by guest on April 27, 2021VOL. 62, 1969 MICROBIOLOGY: CLINE AND LEHRER 757
exudates were induced in guinea pigs by intraperitoneal injection of 1% sodium caseinate
in saline and collected after 18-24 hr. The leukocytes from either the human or guinea
pig source were collected by centrifugation at 150 X g and washed in phosphate-buffered
saline; the contaminating red cells were removed completely by hypotonic lysis."1 The
packed white cells (at least 2 X 108) were disrupted by homogenization in 0.016 M pyro-
phosphate buffer, pH 8.3, for 60 sec in a Sorvall microhomogenizer. The homogenate was
centrifuged at 100,000 X g for 30 min. The protein content12 of the supernatant fraction
was adjusted to approximately 2 mg/ml with pyrophosphate buffer and assayed immedi-
ately for enzyme activity. Activity fell to less than 50% of the initial values after 6 hr at
40C. All the D-amino acid oxidase activity was found in the supernatant fraction.
Granules were isolated from humanl3 or guinea pig neutrophils'4 by homogenization in
0.30 or 0.34 M sucrose solution, respectively, and collected by centrifugation at 27,000 X g.
The isolated granules were disrupted and centrifuged, and the supernatant fraction was
tested for oxidase activity as described for whole leukocytes.
Liver, spleen, and kidneys from exsanguinated guinea pigs were homogenized, centri-
fuged, and tested for oxidase activity in a similar manner.
Purified D-amino acid oxidase (57 IU/mg) prepared from hog kidney was purchased
from Calbiochem.
(b) Myeloperoxidase: Purified human myeloperoxidase was the generous gift of Dr.
Julius Schultz. Its preparation and characteristics have been described previously.'5' 16
Enzyme assays: The basic D-amino acid oxidase assay system contained, in a final
volume of 1.63 ml, 15 ,umoles of pyrophosphate buffer, pH 8.3, 0.3 ,umole of flavin-adenine
dinucleotide (FAD) as the disodium salt (Calbiochem), 150 ,umoles of D-alanine, and
enzyme. Chromatographically homogeneous D-alanine (Calbiochem) and another prep-
aration containing no detectable i-alanine (the generous gift of Dr. D. M. Greenberg)
were used interchangeably. Various D- and -amino acids (Calbiochem) were used in
determining substrate specificity.
D-Amino acid oxidase activity was measured at 370C with a Clark oxygen electrode and
a Gilson model KM oxygraph by the method of Dixon and Kleppe.'7 Oxygen utilization
by enzyme in the absence of D-alanine, which occurred in some leukocyte enzyme prepara-
tions and which probably was attributable to peroxidation of lipids, was subtracted in
calculating enzyme activity. Activity was expressed as micromoles of oxygen consumed
per second per milligram of protein.
Myeloperoxidase activity was measured by recording the rate of oxidation of ortho-
anisidine to chromogenic compounds."8 A unit of enzyme activity was defined as the
amount that caused an increase in absorbency of 0.001/min.
Microbicidal assays: The killing of bacteria by systems containing myeloperoxidase,
halide, and hydrogen peroxide or a D-amino acid oxidase-linked hydrogen peroxide-generat-
ing system was assayed by a minor modification of the method of Klebanoff.19 Bacteria
(Escherichia coli ATCC 11775) were grown overnight in nutrient broth, quantitated with a
spectrophotometer, and resuspended in citrate-phosphate buffer, pH 7. The complete
system contained, in a final volume of 0.5 ml, 20 ,umoles of citrate-phosphate buffer, pH 5,
22 units of human myeloperoxidase, 0.033 jmole of potassium iodide, approximately 106
viable bacteria, and either 1 m/umole of hydrogen peroxide or a hydrogen peroxide-generat-
ing system consisting of 25 1Ag of hog kidney D-amino acid oxidase, 0.04 ,mole of FAD, and
2.5 ,umoles of D-alanine. The mixture was incubated with rotation at 370C for 30 min, at
which time three replicate samples were plated for colony counts.
Results.-Homogenates prepared from human or guinea pig leukocyte popu-
lations that were rich in mature neutrophils contained D-amino acid oxidase
activity as identified by the following criteria (Table 1): oxygen utilization
dependent on the presence of D-alanine or certain other D-amino acids, partial
requirement for FAD, reduced oxygen consumption in the presence of catalase,
inhibition by 2-hydroxybutyrate,20 and lack of activity with L-amino acids.
Downloaded by guest on April 27, 2021758 MICROBIOLOGY: CLINE AND LEHRER PROC. N. A. S.
TABLE 1. D-Amino acid oxidase activity of homogenates of human neutrophilic leukocytes.
Oxygen utilization
(ymoles/sec/mg
Conditions protein X 109)
Complete system* 6.30
" - D-alanine 1.22
" - D-alanine + L-alanine 1.42
" - FAD 4.50
c" + catalase 2.86
" + 2-hydroxybutyrate (0.01 M) 1.90
* The complete system consisted of pyrophosphate buffer, D-alanine, flavin-adenine dinucleotide,
and enzyme as described in Methods.
The substrate specificity of the D-amino acid oxidase activity of human leu-
kocytes and of guinea pig granulocytes was investigated with the methods de-
scribed by Dixon and Kleppe.20 The leukocyte enzymes were not highly puri-
fied; thus, any differences in activity in assays with different D-amino acid
substrates in the leukocyte system could reflect the presence of other enzymes
as impurities, as well as differences in pH optima for different amino acids.
With this limitation in mind, the activity of the human and guinea pig leukocyte
preparations with different D-amino acids was as shown in Table 2. D-Alanine,
D-phenylalanine, and D-threonine were active substrates for the leukocyte en-
zyme. D-Valine, D-serine, and D-asparagine were poor substrates. The most
notable difference between the leukocyte enzyme and the purified renal enzyme
studied by Dixon and Kleppe2e was seen with D-threonine. The white cell en-
zyme showed more activity with this amino acid than with D-alanine, whereas
D-threonine was a poor substrate for the purified enzyme from hog kidney.
DL-2-Hydroxybutyrate (0.01 M) was inhibitory both with the highly purified
renal enzyme (91% inhibition) and the leukocyte crude enzyme (70%). Potas-
sium cyanide (10-3 M) had no inhibitory effect.
Magnesium and manganese ions in concentrations of 10-4 to 10-2 M had no
effect on the D-amino acid oxidase activity of leukocyte homogenates. The
enzyme activity showed a broad optimum between pH 7.5 and 8.5 with D-alanine
as substrate at saturating concentrations. In this pH range, the homogenates
showed equal activity in Tris, pyrophosphate, and phosphate buffers.
The activity of homogenates of guinea pig granulocytes was compared with
that of homogenates of other tissues (Table 3). In other mammalian species,
TABLE 2. Substrates for leukocyte D-amino acid oxidase.*
Homogenate
Amino acid Human leukocyte Guinea pig leukocyte
D-Alanine 100 100
D-Phenylalanine 86 82
D-Threonine 168 135
D-Valine 12 20
D-Norvaline 59 66
D-Serine 10
D-Asparagine 0
* Results are expressed as the percentage of V (,umoles/sec/mg) with D-alanine as substrate.
Measurements were made at pH 8.3 with saturating amounts of substrate.
Downloaded by guest on April 27, 2021VOL. 62, 1969 MICROBIOLOGY: CLINE AND LEHRER 759
TABLE 3. D-Amino acid oxidase activity of homogenates of guinea pig tissues.
Oxygen utilization
Tissue (jAmoles/sec/mg protein X 109)
Kidney 8.20
Liver 2.35
Spleen 0
Leukocytes 1.88
Leukocyte granules 7.41
liver and kidney appear to be the richest sources of enzyme activity2l; the leuko-
cyte apparently has not been systematically examined.
Distribution of enzyme activity in leukocytes and subcellular fractions: Leuko-
cytes obtained from normal subjects and patients with acute and chronic lym-
phocytic leukemia and polycythemia vera were assayed to determine the activity
of 1-amino acid oxidase. As shown in Figure 1, significant activity was re-
stricted to mature granulocytes; little was found in lymphoblasts or in the
"mature" lymphocytes of a normal subject and of patients with chronic lympho-
cytic leukemia. Enzyme activity of leukocytes from patients with acute leu-
kemia in complete hematologic remission showed a return to normal levels.
Human and guinea pig neutrophils were disrupted so as to yield a granule-rich
fraction and a fraction of lower density. The efficiency of the separation was
determined by assaying each fraction for myeloperoxidase, a lysosomal enzyme.
Between 85 and 90 per cent of the activity of this enzyme was found in the
granule fraction. Similarly, 75 to 90 per cent of the activity of D-amino acid
oxidase was associated with the granules. Partial purification of the leukocyte
enzyme could be achieved simply by isolation of the granules, since the specific
activitv of this fraction was approximately four times that of homogenates of
whole cells (Table 3). Further purification could be achieved by precipitation
with 50 per cent acetone at -15'C (approximate fourfold purification with a
yield of 20-30%), followed by chromatography on Sephadex G-100 columns
, 8
x
FIG. 1.-D-Amino acid oxidase activ-
ity of homogenates of human leuko- ° 6 A
cytes. The mean 4 1 SD of the
activity of normal leukocyte popula- E
tions is shown. 0, Chronic lympho- I
cytic leukemia (CLL); *, normal 4 0
lymphocytes; A, acute lymphocytic A
leukemia (ALL) in relapse; A, acute As0
lymphocytic leukemia (ALL) in remis-
sion; 0, polycythemia vera. 2
0
L
Normal Lymph ALL P.Vora
(22 ) CLL
Downloaded by guest on April 27, 2021760 MICROBIOLOGY: CLINE AND LEHRER PROC. N. A. S.
(approximate 20-fold purification with a yield of 10%). Unfortunately, the
limited amount of starting material precluded further purification steps.
Activity with microorganisms: If D-amino acid oxidase activity is involved in
hydrogen peroxide generation after phagocytosis of microorganisms, then the
ability of the enzyme to interact with the D-amino acids of the ingested bacteria
is critical. A fraction of the D-amino acids in the cell wall of the bacteria
is not in polypeptide linkage and might be available to the enzyme. To
examine this question, several species of bacteria were grown overnight, washed
free of culture medium, and resuspended in pyrophosphate buffer at various
concentrations (usually with a protein content of 0.5-2 mg/ml). The organ-
isms were then either kept on ice until assayed as substrate for D-amino acid
oxidase or heat-killed at 60'C for 30 minutes. Neither the living nor the heat-
killed organisms in buffer containing FAD utilized oxygen during the period of
equilibration at 370C. The addition of leukocyte homogenates or of purified
kidney enzyme increased the consumption of oxygen (Table 4). The addition
TABLE 4. Oxygen utilization by D-amino acid oxidases and D-alanine or microorganisms.
Oxygen utilization
Enzyme source Substrate (umoles/sec)
Human leukocytes D-Alanine 3.50 X 10-i
Staphylococcus aureus 4.14 X 10-i
Hog kidney D-Alanine 2.85 X 10-'
Staphylococcus aureus 2.71 X 10-4
Escherichia coli 1.72 X 10-4
Proteus vulgaris 0.39 X 10-4
Serratia marcescens 1.77 X 10-4
of catalase resulted in changes identical to those reported for purified enzyme
and a single D-amino acid :17 an abrupt increase in oxygen content and a halving
of the rate of oxygen utilization.
Killing of microorganisms by a DAAO-MPO-halide system: If 1-amino acid
oxidase systems are involved in defenses against microorganisms, it should be
possible to demonstrate antimicrobial activity under appropriate conditions.
The test system consisted of DAAO, D-alanine, MPO, halide, and microorgan-
isms. D-Amino acid oxidase in the presence of microorganisms and of cell wall
constituents (i.e., single 1-amino acids) was used as a hydrogen peroxide-
generating system. The hydrogen peroxide thus produced served as the oxidant
substrate for purified leukocyte myeloperoxidase, which is assumed to oxidize a
halide (either chloride or iodide) to a biologically active form capable of killing
microorganisms. The results of tests of such systems with E. coli and purified
kidney D-amino acid oxidase are shown in Table 5. It is apparent that for maxi-
mal microbicidal activity the system required all three major constituents:
D-amino acid oxidase, myeloperoxidase, and the halide. Crude or partially
purified leukocyte homogenates could not be tested in such a system because
they contain myeloperoxidase and certain cationic proteins lethal for microorgan-
isms.4
Discussion.-Peroxidase enzymes are found widely distributed in nature, from
protozoa to higher mammals.22' 23 Three mammalian peroxidases-lactoper-
Downloaded by guest on April 27, 2021VOL. 62, 1969 MICROBIOLOGY: CLINE AND LEHRER 761
TABLE 5. D-Amino acid oxidase (DAAO)-myeloperoxidase (MPO) antibacterial system.
Components Added Escherichia colt
Flavin-adenine (colony count
MPO Iodide H202 DAAO dinucleotide D-Alanine X 106)
0 0 0 0 0 0 2.69
o + + 0 0 0 2.33
+ 0 0 0 0 0 2.14
+ + + 0 0 0 0.04
+ + 0 0 0 0 1.36
+ + 0 + 0 0 0.61
+ + 0 + + 0 0.03
+ + 0 + + + 0
0 + 0 + + + 0.73
+ + 0 0 + + 2.29
oxidase, myeloperoxidase, and salivary peroxidase-have been shown to possess
bactericidal activity under appropriate conditions.5' 19 The three phagocytic
leukocytes found in the peripheral blood of man-the neutrophil, monocyte,
and eosinophil-all contain peroxidases. The neutrophil is a particularly rich
source. Its peroxidase, myeloperoxidase, constitutes between 1 and 5 per cent
of the dry weight of the cell'5 and accounts for the green color of purulent exu-
dates and chloromas. The importance of this enzyme in neutrophil defense
function was established by the identification of a patient with an inherited lack
of this enzyme.6
Each of the three mammalian peroxidases requires a halide and a hydrogen
peroxide-generating system for maximal antimicrobial activity in vitro. Hydro-
gen peroxide is known to be produced by neutrophils after particle ingestion.24
The mechanism of generation, however, is not well defined. We suggest that
D-amino acid oxidase may serve as one of the sources of production. A signifi-
cant fraction of the D-amino acids in the teichoic acid portion of the bacterial
cell wall is not in peptide linkage. Amine groups of these D-amino acids are
titratable with 1-fluoro-2,4-dinitrobenzene9 and presumably would be vulnerable
to attack by the oxidase. There are undoubtedly additional sources of hydrogen
peroxide within the phagocytic leukocyte since this compound is also generated
after ingestion of inert polystyrene particles.
Our data indicate that the granule fraction of the normal mature human
granulocyte is the principal source of leukocyte D-amino acid oxidase activity.
Our preliminary results suggest that the enzyme is similar in its major character-
istics to the D-amino acid oxidase that has been crystallized from mammalian
kidney.25
The D-amino acid oxidase of phagocytic leukocytes could provide a biochemi-
cally specific system for the recognition of the "foreignness" of phagocytized
microorganisms. Such a system for the generation of hydrogen peroxide, keyed
to the recognition of the D-amino acids of microorganisms and linked to the
abundant myeloperoxidase of the neutrophilic granulocyte, might constitute a
potent antimicrobial system. Our data indicate that in fact D-amino acid oxi-
dase appropriately linked to myeloperoxidase does form a system lethal for certain
bacteria. The elucidation of this system may explain, in part, why microbial
killing is impaired under anaerobic conditions.26
Downloaded by guest on April 27, 2021762 MICROBIOLOGY: CLINE AND LEHRER PROC. N. A. S.
Both myeloperoxidase and leukocyte D-amino acid oxidase are granule enzymes
and have pH optima near neutrality. Certain evidence indicates that peroxi-
datic enzymes and D-amino acid oxidase may be segregated into a specific class of
granules distinct from those that contain the acid hydrolases.'7 Baudhuin and
collaborators,28 employing differential centrifugation techniques, have established
the existence in rat liver of a distinct group of cytoplasmic particles containing
the enzymes D-amino acid oxidase, urate oxidase, and catalase. These particles,
termed microbodies or peroxisomes, have not yet been identified in granulocytes.
We propose the following sequence of events in the interaction of mature
neutrophils and certain microorganisms: (1) particle is ingested and vacuole
forms; (2) lysosomal enzymes enter vacuole; (3) D-amino acid oxidase and D-
amino acids -- H202; (4) myeloperoxidase + halide ion + H202 O, killing; (5)
pH falls and acid hydrolases are activated; (6) particle is digested. In the pres-
ence of serum factors, including specific opsonins and complement and an intact
leukocyte glycolytic pathway,'-3 microorganisms are ingested and isolated within
a phagocytic vacuole. Within seconds or minutes after its formation, leukocyte
granules discharge their contents into the vacuole. Whether peroxisomes and
granules containing acid hydrolases rupture simultaneously is not certain, but
presumably the enzymes effective at neutral pH act first. Thus, D-amino acid
oxidase may produce hydrogen peroxide directly after rupture of the granules and
exposure of the leukocyte enzyme to the bacterial D-amino acids. The hydrogen
peroxide so produced constitutes the oxidant substrate for myeloperoxidase (also
active at neutrality) to activate chloride ions present within the vacuoles. The
system thus generated is lethal for at least those strains of bacteria studied here.
Subsequently, intracellular pH falls, possibly as a result of an increased rate of
leukocyte glycolysis.' The stage is then set for the activity of lysosomal acid
hydrolases, whose primary function may be degradation of the ingested and
killed microorganisms.
Abbreviations used: MPO, myeloperoxidase. DAAO, D-amino acid oxidase; FAD,
flavin-adenine dinucleotide.
*
This work was supported by U.S. Public Health Service grant CA-07723 and by cancer
research funds of the University of California.
' Karnovsky, M. L., Physiol. Rev., 42, 143 (1962).
2Cline, M. J., Physiol. Rev., 45, 674 (1965).
3 Hirsch, J. G., in The Inflammatory Process, ed. B. W. Zweifach, L. Grant, and R. T.
McCluskey (New York: Academic Press, 1965), p. 245.
4 Zeya, H. I., and J. K. Spitznagel, J. Bacteriol., 91, 755 (1966).
5 Klebanoff, S. J., J. Exptl. Med., 126, 1063 (1967).
6 Lehrer, R. I., and M. J. Cline, Clin. Res., 16, 331 (1968) (abstract).
7 Holmes, B., A. R. Page, and R. A. Good, J. Clin. Invest., 46, 1422 (1967).
8 Quie, P. G., J. B. White, B. Holmes, and R. A. Good, J. Clin. Invest., 46, 668 (1967).
9 Salton, M. R. J., The Bacterial Cell Wall (Amsterdam: Elsevier Publishing Co., 1964).
10 Cline, M. J., and V. C. Swett, J. Exptl. Med., 128, 1309 (1968).
11 Fallon, H. J., E. Frei, III, J. D. Davidson, J. S. Trier, and D. Burk, J. Lab. Clin. Med.,
59, 779 (1962).
12 Lowry, 0. H., N. J. Rosebrough, A. L. Farr, and R. J. Randall, J. Biol. Chem., 193, 265
(1951).
'3 Williams, H. E., Biochim. Biophys. Acta, 124, 34 (1966).
14 Cohn, Z. A., and J. G. Hirsch, J. Exptl. Med., 112. 983 (1960)
Downloaded by guest on April 27, 2021VOL. 62, 1969 MICROBIOLOGY: CLINE AND LEHRER 763
15 Schultz, J., and K. Kaminker, Arch. Biochem. Biophys., 96, 465 (1962).
16Schultz, J., and H. W. Shmukler, Biochemistry, 3, 1234 (1964).
17 Dixon, M., and K. Kleppe, Biochim. Biophys. Acta, 96, 357 (1965).
18 KIebanoff, S. J., Endocrinology, 76, 301 (1965).
19 Klebanoff, S. J., J. Bacteriol., 95, 2131 (1968).
20 Dixon, M., and K. Kleppe, Biochim. Biophys. Acta, 96, 368 (1965).
21 Dixon, M., and E. C. Webb, Enzymes (New York: Academic Press, 1964), 2nd ed., p. 638.
22 Saunders, B. C., A. G. Holmes-Siedle, and B. P. Stark, in Peroxidase (London: Butter-
worths & Co., Ltd., 1964), p. 40.
23Seaman, G. R., in Biochemistry and Physiology of Protozoa, ed. S. H. Hutner and A. Lwoff
(New York: Academic Press, 1955), vol. 2, p. 91.
'4 Paul, B., and A. J. Sbarra, Biochim. Biophys. Acta, 156, 168 (1968).
25 Massey, V., G. Palmer, and R. Bennett, Biochim. Biophys. Acta, 48, 1 (1961).
'6 McRipley, R. J., and A. J. Sbarra, J. Bacteriol., 94, 1425 (1967).
27Enomoto, T., and T. Kitani, Acta Haematol., Japon., 29, 554 (1966).
28 Baudhuin, P., H. Beaufay, and C. De Duve, J. Cell Biol., 26, 219 (1965).
Downloaded by guest on April 27, 2021You can also read
