- Email: [email protected]
In vivo and in vitro effects of sodium azide on mouse complement
TOXICOLOGY
AND APPLIED
PHARMACOLOGY
73, 559-563 (1984)
SHORT COMMUNICATIONS In Vivo and in Vitro Effects of Sodium Azide on Mouse Complement’ Effects of Sodium Azide on Mouse Complement. JOHNSON, K. W., A. E., AND WHITE, K. L., JR. (1984). Toxicol. Appl Pharmacol. 73,559-563. A microtiter hemolytic assay was utilized to determine sodium azide (NaN,) modulation of B6C3Fl and C3H mouse serum complement levels in vivo and in vitro. Functional complement was expressed in CH50 units per milliliter. Experiments were performed to determine the in vitro effect of NaN, on complement mediated lysis of IgM sensitized rabbit erythrocytes. Concentrations of 5, 10, 20, 30, 40, 60, and 80 mM NaN, were added to microtiter wells containing Tris buffer, IgM sensitized rabbit erythrocytes, and serum complement from naive female C3H mice. Although NaCl and KCl controls had an inhibitory effect, NaNs demonstrated a significant dose-dependent inhibition of complement-mediated lysis. In the three in vivo experiments, female B6C3Fl mice were exposed to NaNr and physiological saline (vehicle control). Complement hemolytic ability was evaluated after a l-day, single iv injection of 0.2, 2.0, and 20.0 mg/kg NaN,; at Days 1, 2, 3, 4, and 6 of a 6day time course study after ip administration of 20 mg/kg NaNS; and at the end of an 1l-day study involving daily injections of 10, 15, and 20 mg/kg NaNr given ip. No significant changes in complement-mediated hemolysis were observed in the in vivo experiments. These studies indicate that NaN, does not affect mouse complement levels in viva. However, NaNr suppresses in vitro complement hemolytic ability. In
Vivo
ANDERSON,
and in Vitro
A. C., MUNSON,
Sodium azide has found prominent use as a Although the metabolic and cardiovascular bacteriostatic agent to preserve biological effects of sodium azide are well documented, specimens and commercial immunoglobulin little work has been done concerning its acpreparations. Sodium azide is also used in tions on the immune system. Shaw et al. rocket fuels and in agriculture as a biode- (1980) reported that NaN3 at concentrations gradable herbicide and insecticide (Khayamfound in immunoglobulin preparations ( 15 to Bashi and Sims, 1978). As a result of the di- 30 IIIM) was capable of suppressing in vitro verse uses of sodium azide, many workers are human and guinea pig complement activity. potentially exposed to the toxic effects of the By mediating immune and allergic reactions, chemical. Sodium azide is best known as a complement helps defend against invading metabolic inhibitor via inhibition of cyto- microorganisms. Consequently, agents that chrome oxidase (Stannard, 1939). It also in- decrease complement levels may compromise hibits nuclear phosphorylation (Mettler, 1972) host defense mechanisms. and intra-granulocyte cytotoxicity (Koch, The objectives of these studies were twofold 1974). Sodium azide in an acidic environment to determine whether or not sodium azide is a potent mutagen (Sideris and Argyrakis, suppressed mouse complement in vitro and 1974). It has also been described as an activator to determine if the in vitro effects were bioof adenylate cyclase in rat adipocyte plasma logically relevant by assessing complement membrane (Rahmanian and Jarrett, 1974). In activity of mice treated with NaN3. rabbits and cats, sodium azide causes tachycardia, hyperventilation, and hypotension METHODS (Graham, 1949). The chemical induces basal ganglia necrosis in rats and monkeys (Mettler, Animals. (C57BL/6 X C3H)Fl (B6C3Fl) and C3H fe1972). male mice were purchased from Litton (Frederick, Md.). ’ Supported by NIEHS Contract ES-l-500 1 and Training Grant lT32ESO7087.
The mice arrived at 5 to 6 weeks of age (17 to 20 g) and were quarantined 1 week prior to use. Mice were randomized to cage and treatment group, identified by earpunch, and housed and fed ad lib&m. 559
0041-008X/84
$3.08
Copyright Q 1984 by Academic PRS, Inc. All rights of reproduction in any form reserved.
560
SHORT COMMUNICATIONS
Exposure. Sodium azide (NaN3, Sigma Chemical Co., St. Louis, MO.) was prepared weekly in physiological saline (0.15 M NaCl). Caution was observed in working with NaNs since it is a mutagen and may also explode when heated or when contacting lead. NaNs, diluted in saline, was prepared at concentrations of 0.02, 0.2, 1.O, 1.5, and 2.0 mg/ml. B6C3Fl mice were used for the in vivo experiments. Each mouse received 0.01 ml of NaNr or NaCl (control) per g of body weight. Animals in the 24-h, single exposure study (six animals per condition) received 0.2, 2.0, and 20.0 me/kg NaNr iv. The 1 Iday study mice (eight animals per group) were injected daily with doses of 10,15, and 20 mg/kg NaNr ip. The 6day coutse animals (six mice per group) received 20 mg/kg NaN3 iv. Complement determinations. Based on Van Dijk’s work (Van Dijk, 1980) the following modifications were made for a microtiter assaywith a homologous system of mouse IgM and mouse complement. Following in vivo NaNJ treatment, blood was collected from chloroform-anesthetized mice by cardiac puncture. After allowing the blood to clot at room temperature for 2 h, serum was harvested and diluted 15 in ice cold Tris buffer (0.01 M Tris; 0.14 M NaCI; 0.15 mM CaC&; 0.5 mM MgCl,; 0.1% gelatin; pH 7.3). Diluted serum was added in volumes of 20, 30, 50, 100, and 200 *I to 96 well V-bottom microtiter plates (Dynatech) and placed on ice. Each well contained 1.5 X 10’ rabbit erythrocytes sensitized with a proper dilution of heat inactivated mouse anti-rabbit sera. Tris buffer was added to each well such that a total volume of 250 pi/ well was obtained. After incubating at 30°C for 30 min, the plates were centrifuged at 650g at 4°C for 10 min to pellet the erythrocytes and halt complement-mediated lysis. From each well, 200 ~1 of supematant fraction was placed in ABA-100 (bichromatic analyzer; Abbott Laboratories, North Chicago, Ill.) multicuvettes. The bichromatic analyzer measured amounts of hemoglobin present in two wavelengths (550 and 650 nm). The percentage of lysis was determined for each sample via comparison to a 1m water-lysed control. A Tris buffer blank containing opsonized erythrocytes and no serum was automatically subtracted from each reading to correct for spontaneous lysis. The ABA-100 data were analyzed by a program based on the Von Krogh equation (Mayer, 1961) which derives the CH50 value from the percentage lysis data for each sample. The functional complement activity was expressed in CH50 units per milliliter. The CH50 referred to the amount of complement lysing 50% of sensitized erythrocytes under standard conditions. Expressing functional complement activity in CH50 units was first proposed by Mayer ( 196 1) and is the most acceptable method in both clinical and research settings. Modifications for in vitro experiments: The functional hemolytic assay for in vitro NaN, studies was performed as described above with the following modifications: I .5 X 10’ IgM-sensitized rabbit erythrocytes were added to each well, however, in a reduced voiume of 25 pl of buffer, NaNJ, NaCl, and KC1 final concentrations of 5, 10, 20,
30,40,60, and 80 mM were obtained by adding 25 ~1 of IO-fold higher concentrations of the salt diluted in Tris to each well; serum from 13- to 15-week-old female C3H mice was used as the source of complement; and quadruphcates of each condition were performed. Since NaNx dissociates in solution to Na+ and NT (Graham, 1949), there is no osmotic difference between KCI, NaCI, and NaN, at equal concentrations. Statistical analysis. CH50 values were first tested for homogeneity by Bartlett’s test (Bartlett, 1937). Homogeneous data were evaluated with a parametric analysis of variance When significant differences occurred, treatment groups were compared to the control by Dunnett’s T test. Nonhomogeneous data were analyzed with a nonparametric analysis of variance. When significant differences occurred, treatment groups were compared to the control by Wilcoxan Rank Test. Jonckheer’s test for dose dependency among treatments was also performed. For all analyses, treatment values differing from controls at the level of p < 0.05 are considered significant.
RESULTS In vitro experiments. Sodium azide caused a marked, dose-related suppression of complement-mediated lysis in the in vitro assay. The results, as shown in Fig. 1, demonstrate
0
5
IO
20
30
40
ADDED SALT (mM)
FIG. 1. In vitro effects of NaN, on C3H complement. Effect of NaN3, NaCl, and KC1 on the ability of pooled C3H complement to lyse IgM opsonized rabbit erythrocytes in vitro. CH50 value for Tris control was 36.7 + 1.6. All values significantly different (p < 0.01) from control except 5 mM NaCl, 5 mM NaN,, and 10 mM NaCI. In addition, 20,30, and 40 mM concentrations of NaNs were significantly diGrent from equal concentrations of NaCI and KCI. NaCl and KC1 values were not significantly different at 20, 30, and 40 mM concentrations.
561
SHORT COMMUNICATIONS
that increasing concentrations of NaN3 cause an increasing suppression of functional complement. NaCl and KCl also had a significant inhibitory effect compared to controls. However, the suppression by KC1 and NaCl was significantly less than NaNS at 20, 30, and 40 mM concentrations. Although NaCl values were slightly dose dependent, KCl treatments were not. Previous experiments with NaNj, NaCl, and KC1 and pooled C3H mouse complement established that the inhibitory effect of the salts on functional complement reached a maximum at 40 mM. CH50 values at 60 and 80 mM concentrations of the three salts were not significantly different from their respective 40 mrvr value. In addition, heat inactivated (56°C 30 min) mouse serum controls yielded a mean CH50 value of 0.04, which is much less than the lowest CHSO value for the treatment groups (9.97 for the 40 mM NaN3 group, Fig. 1). Naive control values for TABLE
the same strain of mice and same reagents ranged from 65 to 112, with a mean of 84 CH50 units/ml. In vivo studies. As seen in Table 1, the effect of NaN,, on B6C3Fl complement hemolytic activity was limited. The CH50 values for NaN3-treated animals in the 1day acute study were neither dose related nor significantly different from saline controls. Animals exposed to 20 mg/kg NaNS in the 6-&y time course experiment displayed no significant changes in CH50 values. Of 36 treated animals, 3 died shortly after dosing during the 6&y study. In the 1 l-day study, animals exposed to 10 mg/kg NaNJ exhibited significantly higher CH50 values. However, this effect was not considered biologically significant due to low control values and the lack of a dose-dependent relationship. Lethality also occurred among animals in the 1 l-day study. Two mice in the 15-mg/kg group and all animals in the I
THE EFFECTOF NaNr ON R6C3Fl MOUSE COMPLEMENT LEVELS FOLLOWING In Vivo EXFQSURE Experiment l-day, single injection
Route iv
N 6 6
6. 6
Exposure group
CHSO value”
Vehicle (NaCl) mg/kg mg/kg mg/kg
107 + 11.5 144 + 20.6
0.2 2.0 20.0
98 k
2.5
114 f 11.2
Dose response , . . nonsignificant 6day time course after single dose (20 wVW
iv
6 6 6 6 6 6
Vehicle Day 1 Day 2 Day 3 Day 4 Day 6
106 136 112 102 124 117
1 l-day, daily injections
ip
8 8
Vehicle 10 mg/kg
101 + 7.5d
8 8
15 w/kg
19 k 89 k
20 w/k Dose response . . . nonsignificant
’ The CH50 value reflects functional complement levels. * One animal died (within 24 hr of dosing). ’ Two animals died (within 24 hr of dosing). d Value differed significantly from control (p < 0.05). ‘Two animals died by Day 11. ‘All animals died by Day 4.
+ + f * ? 2
-f
6.6 9.5 9.66 7.0 15.9’ 4.5 2.0 9.9’
562
SHORT
20-mg/kg group died of NaN3-induced ratory paralysis.
COMMUNICATIONS
respi-
DISCUSSION The objective of this study was to evaluate changes in in vivo and in vitro mouse complement hemolytic activity after exposure to NaN3. B6C3Fl and C3H mice were chosen as subjects because they are well characterized immunologically. The in vitro effects of sodium azide on guinea pig and human complement were reported by Shaw et al. (1980). In their studies sodium azide caused a marked dose-dependent suppression of complement-mediated hemolysis. Since equal concentrations (10,20, 30, 40 mM) of NaCl had no effect, they implicated azide ion (N;) as the inhibitory agent. The results of our studies confirmed a strong dose-related suppression of functional complement by NaN3. The inhibition was dose dependent to 40 I’nM and then plateaued. At concentrations of NaN3 from 40 to 80 mM, CH50 values did not change significantly and were still well above values for heat-inactivated serum controls. NaCl and KC1 also had an inhibitory effect on mouse complement hemolysis. Although the suppression was less drastic and effects of KC1 were not dose dependent, the inhibition was still significant when compared to Tris buffer controls. The suppression of complement hemolysis by sodium azide was primarily, but not exclusively, due to the azide ion. Our interpretation of the aforementioned experiments is that NT together with Na+ ions suppresses complement hemolytic ability. The increased ionic strength of the reaction environment resulting from the addition of salts (NaN3, NaCl, or KCl) contributes to a decrease in complement-mediated hemolysis. The latter hypothesis is supported by the inhibitory effect of NaCl and KC1 controls. Since effects of KC1 were qualitatively similar to NaCl, the sodium ion was not acting specifically as an inhibitor, but was contributing to an increased ionic
strength which appears to suppress erythrocyte lysis by complement. Therefore, suppression of complement hemolysis in vitro by sodium azide may be due to an ionic effect of Na+ and N; ions and specific inhibition of complement function by N;. The specific effects of azide are evidenced by the data showing that NaN3 treatments were significantly more inhibitory and dose dependent than KC1 or NaCl groups. The phase of the investigation dealing with the in vivo effects of NaN3 was based on three studies: a 1day acute experiment, a 6-day time course study following a single exposure, and an 1 l-day daily exposure experiment. In the acute study, animals were dosed with NaN3 at three log intervals up to the reported LD50 of 20 mg/kg iv (Graham, 1949). No significant changes in serum complement levels were observed. The 6day time course experiment was performed to determine whether or not the effect of a single high dose injection of NaN3 on complement was delayed. Since sodium azide is hepatotoxic (Mettler, 1972) it may suppress synthesis of complement proteins in the liver. Decreased serum complement levels would result after a delay period. The results of the study indicated no significant difference in the CH50 values at any day. The range of doses for the 11day daily injection experiment was 10, 15, and 20 mg/kg. These concentrations were selected based on the premise that at high doses, bordering lethality, immunological effects would be observed. Lethal endpoints were observed. All the mice at the high dose and two of the animals at the intermediate dose died by Day 11. Three of the dead mice did not exhibit any gross pathological changes upon necropsy. Evaluation of complement levels indicated that the low-dosed ( 10 mg/kg) NaN3 animals displayed significantly higher CH50 values than vehicle controls. However, this effect is not considered biologically relevant because the effect was not dose related and the lo-mg/kg NaN3 group CH50 value of 10 1 is still lower than most of the CHSO’s in the acute and 6-day experiments.
SHORT COMMUNICATIONS
Certain compounds such as Cobra Venon Factor (CVF) and sodium cyanate (NaNCO) are capable of depleting mouse serum complement in both in viva and in vitro situations (Cochrane et al., 1970; Schultz and Arnold, 1975; Megel et al., 1978). Our data indicate that sodium azide is unlike CVF and NaNCO in that NaN3 suppressed complement activity only in the in vitro assay. Sodium azide had no apparent effect on mouse complement hemolysis in vivo. To our knowledge, there are no other in vivo studies to confirm our findings. As a result of these investigations, it does not appear that NaN3 represents a hazard to the complement system of an intact animal. However, NaN3 markedly inhibits complement activity in vitro. Consequently, immunoglobulin preparations containing sodium azide as a preservative should be avoided when performing in vitro assays involving complement-mediated hemolysis. REFERENCES
MAYER, M. M. (1961). Complement and complement fixation. In Experimental Immunochemistry (Ed. by E. Kabut and M. Mayer, eds.), 2nd ed. pp. 135-145. Thomas, Springfield, 111. MEGEL, H., RAYCHAUDHVRI, BEAVER, T. H. (1978). The
409. KOCH,
C. (1974). Effect of sodium azide upon normal and pathological granulocyte function. Acta Pathol. Microbial. Stand. Sect, B 82, 136-142.
A.,
BAYER,
M.,
AND
immunopharmacologic and anti-inflammatory properties of RM 9563 with special reference.to its effect on the complement system. Agents Actions 8, 3, 2 18-228. METTLER, F. A. (1972). Neuropathological effects of sodium azide administration in primates. Fed. Proc., Fed. Amer. Sot. Exp. Biol. 31, 5, 1504-1507. RAHMANIAN, M., AND JARRETT, L. (1974). Activation of rat adipocyte plasma membrane adenylate cyclase by sodium azide. B&hem. Biophys. Res. Commun. 61, 1051-1056. SCHULTZ, D. R., AND ARNOLD, P. I. (1975). Cyanate as an inactivator of complement proteins. J. Immunol. 115, 1558-1565. SHAW, D. R., SHAW, M. W., HICKMAN, S. E., LAMON, E. W., AND GRIFFIN, F. M., JR. (1980). Sodium azide
inhibition
of complement-mediated
functions. Im-
munology39,
53-56. E. G., AND ARGYRAKIS,
M. (1974). Chemical alterations induced in DNA and DNA components by the mutagenic agent azide. Biochem. Biophys. Acta.
SIDERIS,
366, 367-373. STANNARD, J.
N. (1939). Separation of the resting and activity oxygen consumptions of frog muscle by means of sodium azide. Amer. J. Physiol. 126, 196-213.
VAN
BARTLETT, M. S. (1937). Sub-sampling for attributes. J. R. Stat. Sot., Suppl. 4, 131. COCHRANE,C. G., MULLER-EBERHARD,H. J., AND AKIN, B. S. (1970). Depletion of plasma complement in vivo by a protein of cobra venom: Its effect on various immunologic reactions. J. Immunol. 105, 1, 55-69. GRAHAM, J. D. P. (1949). Actions of sodium azide. Brit. J. Pharmacol. 4, 1-6. KHAYAM-BASHI, H., AND SIMS, C. (1978). Effects of sodium azide on the quantitation of the chemical constituents of serum. Amer. J. Clin. Pathol. 69, 4, 405-
563
DIJK,
H.,
RADEMAKER,
P. M.,
AND
WILLERS,
J. M. N. (1980). Estimation of classical pathway of mouse complement activation by use of sensitized rabbit erythrocytes. J. Immunol. Methods 39, 257-268. KIRK
W. JOHNSON
ALICE C. ANDERSON ALBERT KIMBER
E. MUNSON L. WHITE, JR.
Department of Pharmacology and Toxicology Department of Microbiology and Immunology Medical College of Virginia Virginia Commonwealth University Richmond, Virginia 23298 Received July 7, 1983; accepted November 8, I983
AND APPLIED
PHARMACOLOGY
73, 559-563 (1984)
SHORT COMMUNICATIONS In Vivo and in Vitro Effects of Sodium Azide on Mouse Complement’ Effects of Sodium Azide on Mouse Complement. JOHNSON, K. W., A. E., AND WHITE, K. L., JR. (1984). Toxicol. Appl Pharmacol. 73,559-563. A microtiter hemolytic assay was utilized to determine sodium azide (NaN,) modulation of B6C3Fl and C3H mouse serum complement levels in vivo and in vitro. Functional complement was expressed in CH50 units per milliliter. Experiments were performed to determine the in vitro effect of NaN, on complement mediated lysis of IgM sensitized rabbit erythrocytes. Concentrations of 5, 10, 20, 30, 40, 60, and 80 mM NaN, were added to microtiter wells containing Tris buffer, IgM sensitized rabbit erythrocytes, and serum complement from naive female C3H mice. Although NaCl and KCl controls had an inhibitory effect, NaNs demonstrated a significant dose-dependent inhibition of complement-mediated lysis. In the three in vivo experiments, female B6C3Fl mice were exposed to NaNr and physiological saline (vehicle control). Complement hemolytic ability was evaluated after a l-day, single iv injection of 0.2, 2.0, and 20.0 mg/kg NaN,; at Days 1, 2, 3, 4, and 6 of a 6day time course study after ip administration of 20 mg/kg NaNS; and at the end of an 1l-day study involving daily injections of 10, 15, and 20 mg/kg NaNr given ip. No significant changes in complement-mediated hemolysis were observed in the in vivo experiments. These studies indicate that NaN, does not affect mouse complement levels in viva. However, NaNr suppresses in vitro complement hemolytic ability. In
Vivo
ANDERSON,
and in Vitro
A. C., MUNSON,
Sodium azide has found prominent use as a Although the metabolic and cardiovascular bacteriostatic agent to preserve biological effects of sodium azide are well documented, specimens and commercial immunoglobulin little work has been done concerning its acpreparations. Sodium azide is also used in tions on the immune system. Shaw et al. rocket fuels and in agriculture as a biode- (1980) reported that NaN3 at concentrations gradable herbicide and insecticide (Khayamfound in immunoglobulin preparations ( 15 to Bashi and Sims, 1978). As a result of the di- 30 IIIM) was capable of suppressing in vitro verse uses of sodium azide, many workers are human and guinea pig complement activity. potentially exposed to the toxic effects of the By mediating immune and allergic reactions, chemical. Sodium azide is best known as a complement helps defend against invading metabolic inhibitor via inhibition of cyto- microorganisms. Consequently, agents that chrome oxidase (Stannard, 1939). It also in- decrease complement levels may compromise hibits nuclear phosphorylation (Mettler, 1972) host defense mechanisms. and intra-granulocyte cytotoxicity (Koch, The objectives of these studies were twofold 1974). Sodium azide in an acidic environment to determine whether or not sodium azide is a potent mutagen (Sideris and Argyrakis, suppressed mouse complement in vitro and 1974). It has also been described as an activator to determine if the in vitro effects were bioof adenylate cyclase in rat adipocyte plasma logically relevant by assessing complement membrane (Rahmanian and Jarrett, 1974). In activity of mice treated with NaN3. rabbits and cats, sodium azide causes tachycardia, hyperventilation, and hypotension METHODS (Graham, 1949). The chemical induces basal ganglia necrosis in rats and monkeys (Mettler, Animals. (C57BL/6 X C3H)Fl (B6C3Fl) and C3H fe1972). male mice were purchased from Litton (Frederick, Md.). ’ Supported by NIEHS Contract ES-l-500 1 and Training Grant lT32ESO7087.
The mice arrived at 5 to 6 weeks of age (17 to 20 g) and were quarantined 1 week prior to use. Mice were randomized to cage and treatment group, identified by earpunch, and housed and fed ad lib&m. 559
0041-008X/84
$3.08
Copyright Q 1984 by Academic PRS, Inc. All rights of reproduction in any form reserved.
560
SHORT COMMUNICATIONS
Exposure. Sodium azide (NaN3, Sigma Chemical Co., St. Louis, MO.) was prepared weekly in physiological saline (0.15 M NaCl). Caution was observed in working with NaNs since it is a mutagen and may also explode when heated or when contacting lead. NaNs, diluted in saline, was prepared at concentrations of 0.02, 0.2, 1.O, 1.5, and 2.0 mg/ml. B6C3Fl mice were used for the in vivo experiments. Each mouse received 0.01 ml of NaNr or NaCl (control) per g of body weight. Animals in the 24-h, single exposure study (six animals per condition) received 0.2, 2.0, and 20.0 me/kg NaNr iv. The 1 Iday study mice (eight animals per group) were injected daily with doses of 10,15, and 20 mg/kg NaNr ip. The 6day coutse animals (six mice per group) received 20 mg/kg NaN3 iv. Complement determinations. Based on Van Dijk’s work (Van Dijk, 1980) the following modifications were made for a microtiter assaywith a homologous system of mouse IgM and mouse complement. Following in vivo NaNJ treatment, blood was collected from chloroform-anesthetized mice by cardiac puncture. After allowing the blood to clot at room temperature for 2 h, serum was harvested and diluted 15 in ice cold Tris buffer (0.01 M Tris; 0.14 M NaCI; 0.15 mM CaC&; 0.5 mM MgCl,; 0.1% gelatin; pH 7.3). Diluted serum was added in volumes of 20, 30, 50, 100, and 200 *I to 96 well V-bottom microtiter plates (Dynatech) and placed on ice. Each well contained 1.5 X 10’ rabbit erythrocytes sensitized with a proper dilution of heat inactivated mouse anti-rabbit sera. Tris buffer was added to each well such that a total volume of 250 pi/ well was obtained. After incubating at 30°C for 30 min, the plates were centrifuged at 650g at 4°C for 10 min to pellet the erythrocytes and halt complement-mediated lysis. From each well, 200 ~1 of supematant fraction was placed in ABA-100 (bichromatic analyzer; Abbott Laboratories, North Chicago, Ill.) multicuvettes. The bichromatic analyzer measured amounts of hemoglobin present in two wavelengths (550 and 650 nm). The percentage of lysis was determined for each sample via comparison to a 1m water-lysed control. A Tris buffer blank containing opsonized erythrocytes and no serum was automatically subtracted from each reading to correct for spontaneous lysis. The ABA-100 data were analyzed by a program based on the Von Krogh equation (Mayer, 1961) which derives the CH50 value from the percentage lysis data for each sample. The functional complement activity was expressed in CH50 units per milliliter. The CH50 referred to the amount of complement lysing 50% of sensitized erythrocytes under standard conditions. Expressing functional complement activity in CH50 units was first proposed by Mayer ( 196 1) and is the most acceptable method in both clinical and research settings. Modifications for in vitro experiments: The functional hemolytic assay for in vitro NaN, studies was performed as described above with the following modifications: I .5 X 10’ IgM-sensitized rabbit erythrocytes were added to each well, however, in a reduced voiume of 25 pl of buffer, NaNJ, NaCl, and KC1 final concentrations of 5, 10, 20,
30,40,60, and 80 mM were obtained by adding 25 ~1 of IO-fold higher concentrations of the salt diluted in Tris to each well; serum from 13- to 15-week-old female C3H mice was used as the source of complement; and quadruphcates of each condition were performed. Since NaNx dissociates in solution to Na+ and NT (Graham, 1949), there is no osmotic difference between KCI, NaCI, and NaN, at equal concentrations. Statistical analysis. CH50 values were first tested for homogeneity by Bartlett’s test (Bartlett, 1937). Homogeneous data were evaluated with a parametric analysis of variance When significant differences occurred, treatment groups were compared to the control by Dunnett’s T test. Nonhomogeneous data were analyzed with a nonparametric analysis of variance. When significant differences occurred, treatment groups were compared to the control by Wilcoxan Rank Test. Jonckheer’s test for dose dependency among treatments was also performed. For all analyses, treatment values differing from controls at the level of p < 0.05 are considered significant.
RESULTS In vitro experiments. Sodium azide caused a marked, dose-related suppression of complement-mediated lysis in the in vitro assay. The results, as shown in Fig. 1, demonstrate
0
5
IO
20
30
40
ADDED SALT (mM)
FIG. 1. In vitro effects of NaN, on C3H complement. Effect of NaN3, NaCl, and KC1 on the ability of pooled C3H complement to lyse IgM opsonized rabbit erythrocytes in vitro. CH50 value for Tris control was 36.7 + 1.6. All values significantly different (p < 0.01) from control except 5 mM NaCl, 5 mM NaN,, and 10 mM NaCI. In addition, 20,30, and 40 mM concentrations of NaNs were significantly diGrent from equal concentrations of NaCI and KCI. NaCl and KC1 values were not significantly different at 20, 30, and 40 mM concentrations.
561
SHORT COMMUNICATIONS
that increasing concentrations of NaN3 cause an increasing suppression of functional complement. NaCl and KCl also had a significant inhibitory effect compared to controls. However, the suppression by KC1 and NaCl was significantly less than NaNS at 20, 30, and 40 mM concentrations. Although NaCl values were slightly dose dependent, KCl treatments were not. Previous experiments with NaNj, NaCl, and KC1 and pooled C3H mouse complement established that the inhibitory effect of the salts on functional complement reached a maximum at 40 mM. CH50 values at 60 and 80 mM concentrations of the three salts were not significantly different from their respective 40 mrvr value. In addition, heat inactivated (56°C 30 min) mouse serum controls yielded a mean CH50 value of 0.04, which is much less than the lowest CHSO value for the treatment groups (9.97 for the 40 mM NaN3 group, Fig. 1). Naive control values for TABLE
the same strain of mice and same reagents ranged from 65 to 112, with a mean of 84 CH50 units/ml. In vivo studies. As seen in Table 1, the effect of NaN,, on B6C3Fl complement hemolytic activity was limited. The CH50 values for NaN3-treated animals in the 1day acute study were neither dose related nor significantly different from saline controls. Animals exposed to 20 mg/kg NaNS in the 6-&y time course experiment displayed no significant changes in CH50 values. Of 36 treated animals, 3 died shortly after dosing during the 6&y study. In the 1 l-day study, animals exposed to 10 mg/kg NaNJ exhibited significantly higher CH50 values. However, this effect was not considered biologically significant due to low control values and the lack of a dose-dependent relationship. Lethality also occurred among animals in the 1 l-day study. Two mice in the 15-mg/kg group and all animals in the I
THE EFFECTOF NaNr ON R6C3Fl MOUSE COMPLEMENT LEVELS FOLLOWING In Vivo EXFQSURE Experiment l-day, single injection
Route iv
N 6 6
6. 6
Exposure group
CHSO value”
Vehicle (NaCl) mg/kg mg/kg mg/kg
107 + 11.5 144 + 20.6
0.2 2.0 20.0
98 k
2.5
114 f 11.2
Dose response , . . nonsignificant 6day time course after single dose (20 wVW
iv
6 6 6 6 6 6
Vehicle Day 1 Day 2 Day 3 Day 4 Day 6
106 136 112 102 124 117
1 l-day, daily injections
ip
8 8
Vehicle 10 mg/kg
101 + 7.5d
8 8
15 w/kg
19 k 89 k
20 w/k Dose response . . . nonsignificant
’ The CH50 value reflects functional complement levels. * One animal died (within 24 hr of dosing). ’ Two animals died (within 24 hr of dosing). d Value differed significantly from control (p < 0.05). ‘Two animals died by Day 11. ‘All animals died by Day 4.
+ + f * ? 2
-f
6.6 9.5 9.66 7.0 15.9’ 4.5 2.0 9.9’
562
SHORT
20-mg/kg group died of NaN3-induced ratory paralysis.
COMMUNICATIONS
respi-
DISCUSSION The objective of this study was to evaluate changes in in vivo and in vitro mouse complement hemolytic activity after exposure to NaN3. B6C3Fl and C3H mice were chosen as subjects because they are well characterized immunologically. The in vitro effects of sodium azide on guinea pig and human complement were reported by Shaw et al. (1980). In their studies sodium azide caused a marked dose-dependent suppression of complement-mediated hemolysis. Since equal concentrations (10,20, 30, 40 mM) of NaCl had no effect, they implicated azide ion (N;) as the inhibitory agent. The results of our studies confirmed a strong dose-related suppression of functional complement by NaN3. The inhibition was dose dependent to 40 I’nM and then plateaued. At concentrations of NaN3 from 40 to 80 mM, CH50 values did not change significantly and were still well above values for heat-inactivated serum controls. NaCl and KC1 also had an inhibitory effect on mouse complement hemolysis. Although the suppression was less drastic and effects of KC1 were not dose dependent, the inhibition was still significant when compared to Tris buffer controls. The suppression of complement hemolysis by sodium azide was primarily, but not exclusively, due to the azide ion. Our interpretation of the aforementioned experiments is that NT together with Na+ ions suppresses complement hemolytic ability. The increased ionic strength of the reaction environment resulting from the addition of salts (NaN3, NaCl, or KCl) contributes to a decrease in complement-mediated hemolysis. The latter hypothesis is supported by the inhibitory effect of NaCl and KC1 controls. Since effects of KC1 were qualitatively similar to NaCl, the sodium ion was not acting specifically as an inhibitor, but was contributing to an increased ionic
strength which appears to suppress erythrocyte lysis by complement. Therefore, suppression of complement hemolysis in vitro by sodium azide may be due to an ionic effect of Na+ and N; ions and specific inhibition of complement function by N;. The specific effects of azide are evidenced by the data showing that NaN3 treatments were significantly more inhibitory and dose dependent than KC1 or NaCl groups. The phase of the investigation dealing with the in vivo effects of NaN3 was based on three studies: a 1day acute experiment, a 6-day time course study following a single exposure, and an 1 l-day daily exposure experiment. In the acute study, animals were dosed with NaN3 at three log intervals up to the reported LD50 of 20 mg/kg iv (Graham, 1949). No significant changes in serum complement levels were observed. The 6day time course experiment was performed to determine whether or not the effect of a single high dose injection of NaN3 on complement was delayed. Since sodium azide is hepatotoxic (Mettler, 1972) it may suppress synthesis of complement proteins in the liver. Decreased serum complement levels would result after a delay period. The results of the study indicated no significant difference in the CH50 values at any day. The range of doses for the 11day daily injection experiment was 10, 15, and 20 mg/kg. These concentrations were selected based on the premise that at high doses, bordering lethality, immunological effects would be observed. Lethal endpoints were observed. All the mice at the high dose and two of the animals at the intermediate dose died by Day 11. Three of the dead mice did not exhibit any gross pathological changes upon necropsy. Evaluation of complement levels indicated that the low-dosed ( 10 mg/kg) NaN3 animals displayed significantly higher CH50 values than vehicle controls. However, this effect is not considered biologically relevant because the effect was not dose related and the lo-mg/kg NaN3 group CH50 value of 10 1 is still lower than most of the CHSO’s in the acute and 6-day experiments.
SHORT COMMUNICATIONS
Certain compounds such as Cobra Venon Factor (CVF) and sodium cyanate (NaNCO) are capable of depleting mouse serum complement in both in viva and in vitro situations (Cochrane et al., 1970; Schultz and Arnold, 1975; Megel et al., 1978). Our data indicate that sodium azide is unlike CVF and NaNCO in that NaN3 suppressed complement activity only in the in vitro assay. Sodium azide had no apparent effect on mouse complement hemolysis in vivo. To our knowledge, there are no other in vivo studies to confirm our findings. As a result of these investigations, it does not appear that NaN3 represents a hazard to the complement system of an intact animal. However, NaN3 markedly inhibits complement activity in vitro. Consequently, immunoglobulin preparations containing sodium azide as a preservative should be avoided when performing in vitro assays involving complement-mediated hemolysis. REFERENCES
MAYER, M. M. (1961). Complement and complement fixation. In Experimental Immunochemistry (Ed. by E. Kabut and M. Mayer, eds.), 2nd ed. pp. 135-145. Thomas, Springfield, 111. MEGEL, H., RAYCHAUDHVRI, BEAVER, T. H. (1978). The
409. KOCH,
C. (1974). Effect of sodium azide upon normal and pathological granulocyte function. Acta Pathol. Microbial. Stand. Sect, B 82, 136-142.
A.,
BAYER,
M.,
AND
immunopharmacologic and anti-inflammatory properties of RM 9563 with special reference.to its effect on the complement system. Agents Actions 8, 3, 2 18-228. METTLER, F. A. (1972). Neuropathological effects of sodium azide administration in primates. Fed. Proc., Fed. Amer. Sot. Exp. Biol. 31, 5, 1504-1507. RAHMANIAN, M., AND JARRETT, L. (1974). Activation of rat adipocyte plasma membrane adenylate cyclase by sodium azide. B&hem. Biophys. Res. Commun. 61, 1051-1056. SCHULTZ, D. R., AND ARNOLD, P. I. (1975). Cyanate as an inactivator of complement proteins. J. Immunol. 115, 1558-1565. SHAW, D. R., SHAW, M. W., HICKMAN, S. E., LAMON, E. W., AND GRIFFIN, F. M., JR. (1980). Sodium azide
inhibition
of complement-mediated
functions. Im-
munology39,
53-56. E. G., AND ARGYRAKIS,
M. (1974). Chemical alterations induced in DNA and DNA components by the mutagenic agent azide. Biochem. Biophys. Acta.
SIDERIS,
366, 367-373. STANNARD, J.
N. (1939). Separation of the resting and activity oxygen consumptions of frog muscle by means of sodium azide. Amer. J. Physiol. 126, 196-213.
VAN
BARTLETT, M. S. (1937). Sub-sampling for attributes. J. R. Stat. Sot., Suppl. 4, 131. COCHRANE,C. G., MULLER-EBERHARD,H. J., AND AKIN, B. S. (1970). Depletion of plasma complement in vivo by a protein of cobra venom: Its effect on various immunologic reactions. J. Immunol. 105, 1, 55-69. GRAHAM, J. D. P. (1949). Actions of sodium azide. Brit. J. Pharmacol. 4, 1-6. KHAYAM-BASHI, H., AND SIMS, C. (1978). Effects of sodium azide on the quantitation of the chemical constituents of serum. Amer. J. Clin. Pathol. 69, 4, 405-
563
DIJK,
H.,
RADEMAKER,
P. M.,
AND
WILLERS,
J. M. N. (1980). Estimation of classical pathway of mouse complement activation by use of sensitized rabbit erythrocytes. J. Immunol. Methods 39, 257-268. KIRK
W. JOHNSON
ALICE C. ANDERSON ALBERT KIMBER
E. MUNSON L. WHITE, JR.
Department of Pharmacology and Toxicology Department of Microbiology and Immunology Medical College of Virginia Virginia Commonwealth University Richmond, Virginia 23298 Received July 7, 1983; accepted November 8, I983