1442

Effect of Catalase on Regional Cerebral Blood Flow and Brain Edema during the Early Phase of Experimental Pneumococcal Meningitis H. Walter Pfister, Uwe Ktidel, Ulrich Dirnagl, Roman L. Haberl, Gotthard Ruckdeschel, and Karl M. Einhaupl

Neurologische Klinik, Klinikum Grobhadern. Max von Pettenkofer Institute for Hygiene and Microbiology. University of Munich. Germany

Despite marked progress in antimicrobial chemotherapy and improved medical intensive care medicine, the outcome of bacterial meningitis has not significantly improved [I]. The mortality rate of meningitis associated with Streptococcus pneumoniae, the most common causative agent of'bacterial meningitis in adults, is 20%- 30%. In addition, neurologic sequelae frequently occur in survivors of bacterial meningitis [I]. Animal models of bacterial meningitis have improved our understanding ofthe pathophysiology ofthe disease, offering the hope of developing new therapeutic approaches. Recent studies in rat and rabbit models of bacterial meningitis have indicated that the pathophysiologic mechanisms resulting in cerebral complications and brain injury are complex. In animal models, several pathophysiologic alterations during bacterial meningitis have been demonstrated, including brain edema formation and elevation ofintracranial pressure (ICP) [2], changes in cerebrospinal fluid (CSF) outflow resistance [3], morphologic alterations of the blood-brain barrier [4], and changes in cerebral blood flow (CBF) [5-7]. In addition, a series of mediators ofmajor pathophysiologic changes have been identified, including cytokines [8], cyclooxygenase metabolites [9], and platelet-activating factor [10]. We have recently shown in a rat model of pneumococcal

Received 12 March 1992; revised 26 June 1992. Grant support: Deutsche Forschungsgemeinschaft (project Pf 246/3-1) and Friedrich Baur Stiftung (project 62/91) to H.W.P. Reprints or correspondence: Dr. Hans-Walter Pfister. Neurologische Klinik, Klinikum Groflhadern, Marchioninistr. 15. W-8000 Munich 70, Germany. The Journal of Infectious Diseases 1992;166:1442-5 © 1992 by The Universityof Chicago. All rights reserved. 0022-1899/92/6606-0039$01.00

meningitis that continuous intravenous (iv) administration of the free radical scavenger superoxide dismutase (SOD) completely inhibited the increase in regional CBF (rCBF), ICP, and brain water content as observed in infected untreated rats during the early phase of disease [7]. These findings suggested that reactive oxygen species playa major role in the pathogenesis of experimental pneumococcal meningitis. In the present study, we tested the effect of catalase, which eliminates hydrogen peroxide, to obtain further information on the reactive oxygen species involved.

Materials and Methods The experimental procedure used has been described in detail [7]. Briefly, adult male Wistar 'rats weighing 250-350 g were anesthetized with an intraperitoneal injection ofthiobutabarbiturate (Inactin: Byk Gulden, Konstanz, Germany), 100 rug/kg, tracheotomized, and kept under anesthesia throughout the experiment. They were artificially ventilated using a small animal ventilator (model 683; Harvard, South Natick, MA). The endexpiratory CO z was continuously monitored with an infrared CO z analyzer (model 2200; Heyer, Bad Ems, Germany) and was maintained at a constant level of - 30 mm Hg. Arterial blood pressure was measured with a Statham pressure transducer connected to a catheter introduced into the right femoral artery. The right femoral vein was cannulated for drug administration. A rat was placed in a stereotaxic frame, and a craniotomy was done in the right parietal bone for placement of the laserDoppler probe. The dura was left intact. A burr hole was drilled in the occipital bone to place the cisterna magna catheter. All rats were maintained at a constant rectal temperature of -38°C. Arterial blood was analyzed for Pco-, Paz, pH, and hematocrit on samples taken before and every 2 h after intracisternal (ic) injection. rCBF, as measured by laser-Doppler flowmetry (model BPM 403a; TSI, St. Paul, MN), ICP (via the cisterna magna catheter), arterial blood pressure, end-expiratory

Downloaded from http://jid.oxfordjournals.org/ at East Carolina University on September 12, 2015

Previous studies have demonstrated that the radical scavenger superoxide dismutase completely blocked the increase of regional cerebral blood flow (rCBF), intracranial pressure (ICP), and brain water content during the early phase of experimental pneumococcal meningitis in the rat. To obtain information on the nature of the reactive oxygen species involved, the effect of catalase, a hydrogen peroxide scavenger, was tested. Rats injected intracisternally with livepneumococci were either untreated or received intravenous catalase. The increase of rCBF and brain water content in infected untreated rats was significantlyattenuated by catalase 6 h after intracisternal challenge. ICP increased in both infected groups, with a trend toward lower ICP with catalase treatment. Cerebrospinal fluid white blood cell counts were not significantly different between infected groups. These results and previous experiments using superoxide dismutase suggest that the increase ofrCBF, ICP, and brain water content is mainly caused bysuperoxide or superoxide reaction products.

JID 1992; 166 (December)

Concise Communications

Results Physiologic variables. Arterial blood pressure, Po2 , Pco., pH, hematocrit, and body temperature were within normal ranges throughout the experiments in ali groups (data not shown). rCBF. The increase of rCBF observed in infected untreated animals was attenuated by catalase, with a significant difference at 6 h after infection (figure 1, table 1). There was no significant change in rCBF in rats treated with free catalase and injected ic with PBS (table I). [CPo lCP was markedly elevated in both infected groups, with a trend toward lower ICP in catalase-treated infected rats (figure 1, table I). ICP data of infected untreated rats differed significantly 3.5 h after infection from that of catalase-treated infected rats (figure 1). Catalase alone did not alter ICP (table 1). Brain edema. The increase in brain water content as observed in untreated infected rats was significantly attenuated by treatment with catalase (table 1). CSF WBC and bacterial titers. CSF WBC counts were slightly lower in the catalase-treated infected rats (2360 ± 628 cells/ul., n = 5) than in untreated infected rats (3419 ± 797 cells/ul., n = 8); however, this difference was not significant. CSF bacterial titers did not differ significantly between catalase-treated infected rats (5.7 ± 0.6 log cfu/rnl., n = 5)

240 o Infected untreated • Infected calala_treated o unlnfected controls

.... 220

'#.

; ' 200

0

oJ IL

II:

W

180 160

oJ IL IL 140

0

C I

II:

W

~

120 100 80 14

m

J:

E E

W

o Infected untreated • Infected catala_treated unlnfecled controls

12

o

10

,

II:

::;)

en en

Ie

8

w

II: IL oJ

6

Z

4

Cl:

Cl: II:

~

2

I-

!:

0 -1

o

2

3

4

5

6

TIME [h]

Figure 1. Time course of regional cerebral blood flow (rCBF) and intracranial pressure in untreated infected, catalase-treatedinfected, and control rats within 6 h after intracisternal (ic) injection of pneumococci and PBS. rCBF was expressed as % of baseline value (100%). Data were compared every half hour from point ofic injection until 6 h later. Data are,mean ± SE. P < .05 compared with uninfected controls; t, P < .05 compared with infected untreated rats.

*,

and untreated infected rats (6.1 ± 0.6 log cfu/rnl., n = 5). These data excluded the possibility that catalase may have an effect on bacterial growth in CSF.

Discussion Reactive oxygen species are capable ofinducing cell injury and playa major role in the pathogenesis of an array of conditions, including experimental fluid percussion brain injury, cold-induced brain edema, cerebral ischemia-reperfusion injury, cerebral arteriolar abnormalities after acute hypertension in rats, and inflammation [11, 12]. It has been shown in animal models that reactive oxygen species may cause vasodilation of cerebral arterioles, an abnormal CO 2 reactivity of these vessels, and an increase in blood-brain barrier permeability [11, 13]. The results of our previous studies using free SOD have shown that reactive oxygen species also are in-

Downloaded from http://jid.oxfordjournals.org/ at East Carolina University on September 12, 2015

CO2 , and rectal body temperature were continuously monitored for 6 h after infection by using a personal computer system (ASYST program; MacMillan Software, New York). At the end of the experiment, CSF samples were collected for determination of white blood cell (WBC) counts and bacterial titers. After rats were killed by exsanguination, brain water content of both hemispheres was calculated by the formula [(wet weight - dry weightj/wet weight] X 100 [7]. After 75 ~L of CSF was removed through the cisterna magna catheter, meningitis was induced by ic injection of 75 ~L (_10 5_106 cfu) of pneumococci (strain type 6b) [7J. Rats were divided into three treatment groups: 6 rats untreated and injected ic with 75 ~L of PBS (controls); 10 rats untreated and injected ic with pneumococci; and 8 rats treated continuously iv with 25,000 units/kg/h free catalase (from bovine liver; Sigma, Deisenhofen, Germany) and injected ic with pneumococci (catalase administration was started just before ic challenge). To exclude an effect of catalase alone on rCBF, ICP, and brain water content, a separate group of3 rats treated continuously with iv catalase and injected ic with PBS was tested. Statistical methods. To detect significant changes within each group, repeated-measures analysis of variance was done. The three groups were compared for rCBF, ICP, and brain water content by one-way analysis of variance and Student-NewmanKeuls multiple comparisons. Data of rCBF and ICP were compared every half hour from the point ofic injection for 6 h. CSF WBC counts in infected untreated rats and catalase-treated infected rats were compared by using two-tailed Student's t test. Data are expressed as mean ± SE. Differences were considered significant when P < .05.

1443

Concise Communications

1444

Regional cerebral blood flow (rCBF), intracranial pressure(ICP), and brain watercontent in rats with Streptococcus pneumoniae meningitis.

Table 1.

rCBF. Group (n)

99.7±5.1 (6) 213.9 ± 15.3 (10)* 164.6± 11.6 (8)*t 85.7 ± 9.4 (3)

3.5 ± 1.0 (6) 9.9± 1.8 ( 10)* 6.8 ± 2.6 (6) 2.4 ± 0.2 (3)

Brain water content, % (n)

78.94 ± 0.07 (5) 79.64 ± 0.08 (7)* 79.32 ± 0.10 (7)*t 78.91 ±O.IO (3)

NOTE. First three groups were compared using one-way analysis of variance andStudent-Newman-Keuls multiple comparisons. Data aremean ± SE. * P < .05 compared with uninfected controls. t p < .05 compared with infected untreated rats.

volved in the pathophysiologic mechanisms during the early phase of experimental pneumococcal meningitis [10]. SOD dismutates superoxide to hydrogen peroxide and thus prevents the generation of hydroxyl radical by removing one reactant of the iron-catalyzed Haber-Weiss reaction. SOD may also act by preventing the formation of peroxynitrite, which is toxic because it decomposes to form a strong oxidant with reactivity similar to hydroxyl radical [14]. Catalase, which was used in this study, scavenges hydrogen peroxide, thereby removing one reagent of the HaberWeiss reaction and preventing the formation ofhydroxyl radical via this pathway. The results of the current study show that catalase attenuated the increase ofrCBF, ICP, and brain water content in the untreated infected rats. Catalase did not prevent the influx of granulocytes into the subarachnoid space. An effect of catalase per se on rCBF, ICP, or brain water content was excluded by our experiments investigating catalase in uninfected rats. The failure of a complete blocking effect of catalase compared with that of SOD raises the question of different pharmacology of these enzymes. Both free SOD and free catalase injected iv have very short circulating half-lives of 8 and 20 min, respectively [15]. Therefore, both enzymes were given in our experiments by continuous iv infusion. The molecular weight of SOD is 31,000 compared with -210,000 for catalase, thus explaining the slightly shorter half-life of SOD due to its more rapid renal clearance. Because oftheir high molecular weights, both enzymes nearly cannot permeate cell membranes or the blood-brain barrier [16]. The catalase dosage of 150,000 units/kg/6 h in our study exceeds that commonly used in animal models [17]. In addition, it was shown in a model of postischemic myocardial dysfunction in dogs that a high-dose regimen of catalase (240,000 units/kg) did not provide an additional attenuating

effect compared with a low-dose regimen (18,000 units/kg) [18]. Thus, we believe that the 150,000 units/kg dose may be sufficient to show an effect, provided that H 2 0 2 or H 20 2-related reactions are involved. In previous experiments, we have shown that iv administration of free SOD, which is unable to sufficiently cross the blood-brain barrier, completely prevented an increase in rCBF, ICP, and brain water content, suggesting that reactive oxygen species may be generated within the cerebral vessels during the early phase of pneumococcal meningitis. From these findings, one would also expect an effect of free catalase, provided that hydroxyl radical generated via the HaberWeiss reaction participates in the increase in rCBF during the early phase of the disease. The marginal effects of catalase may be explained by the fact that superoxide radical mainly produces the microvascular changes and hydroxyl radical generated via the Haber-Weiss reaction does not play a major role. Alternatively, it may be possible that another hydroxyl radical-producing pathway that is inhibited by SOD, for example, the formation of peroxynitrite by superoxide and nitric oxide, contributes to the changes in rCBF, ICP, and brain water content. Further studies are warranted to investigate whether inhibitors of nitric oxide synthase prevent the alterations ofrCBF, ICP, and brain water content. Acknowledgment

We thank Mary J. Finke for manuscript preparation. References I. Swartz MN. Bacterial meningitis: more involved than justthe meninges. N Engl J Med 1984;311 :912-4. 2. Tauber MG. Khayam-Bashi H, Sande MA. Effects of ampicillin and corticosteroids on brain water c9ntent. cerebrospinal fluid pressure, andcerebrospinal fluid lactate levels in experimental pneumococcal meningitis. J Infect Dis 1985;151:528-~4. 3. Scheid WM. Dacey RG Jr, Winn HR. Welsh JE,Jane JA. Sande MA. Cerebrospinal fluid outflow resistance in rabbits with experimental meningitis. Alterations with penicillin and methylprednisolone. J Clin Invest 1980;66:243-53. 4. Quagliarello VJ. Long WJ. Scheid WM. Morphological alterations of theblood brain barrier with experimental meningitis inthe rat. J Clin Invest 1986;77: 1084-95. 5. Smith AL. Pathogenesis of Haemophilus influenzae type b meningitis. In: Keusch G,Wadstriim T,eds. Experimental bacterial andparasitic infections. New York: Elsevier. 1983:295-301. 6. Tureen JH.Dworkin RJ, Kennedy SL. Sachdeva M. Sande MA. Loss of cerebrovascular autoregulation in experimental meningitis in rabbits. J Clin Invest 1990;85:577-81. 7. Pfister HW, Koedel U.Haberl RL. etal. Microvascular changes during theearly phase ofpneumococcal meningitis inthe rat. J Cereb Blood Flow Metab 1990;10:914-22. 8. Mustafa M. Ramilo O. Olsen KD, et al. Tumor necrosis factor in mediating experimental Haemophilus influenzae type b meningitis. J Clin Invest 1989;84: 1253-9. 9. Tuomanen E. Hengstler B, Rich R. Bray MA, Zak O.Tomasz A. Nonsteroidal antiinflammatory agents in the therapy for experimental pneumococcal meningitis. J Infect Dis 1987; 155:985-90.

Downloaded from http://jid.oxfordjournals.org/ at East Carolina University on September 12, 2015

Uninfected control (6) Infected untreated (10) Infected catalase-treated (8) Uninfected catalase-treated (3)

% (n)

ICP. mm Hg (n)

JID 1992; 166 (December)

JID 1992;166 (December)

Concise Communications

10. Arditi M, Manogue KR, Caplan M, Yogev R. Cerebrospinal fluid cachectinftumor necrosis factor-a and platelet-activating factor concentrations and severity of bacterial meningitis in children. J Infect Dis 1990;162:139-47, II. Kontos HA. Oxygen radicals in experimental brain injury. In: Hoff JT, Betz AL, eds. Intracranial pressure. 7th ed. Berlin: Springer Verlag, 1989:787-98. 12. McCord JM, Wong K, Stokes SH. Petrone WF. English D. Superoxide and inflammation: a mechanism for the anti-inflammatory activity of superoxide dismutase. Acta Physiol Scand Suppl 1980;492:25-30. 13. Chan PH, Longar S, Fishman RA. Protective effects of liposome-entrapped superoxide dismutase on posttraumatic brain edema, Ann NeuroI1987;21:540-7. 14. Beckman JS. Beckman TW, Chen J. Marshall MA, Freeman BA, Apparent hydroxyl radical production by peroxynitrite: implications for

1445

endothelial injury from nitric oxide and superoxide. Proc Natl Acad Sci USA 1990;87:1620-4. 15. Turrens JF, Crapo JD, Freeman BA. Protection against oxygen toxicity by intravenous injection of liposome-entrapped catalase and superoxide dismutase. J Clin Invest 1984;73:87-95. 16. Yusa T, Crapo JD, Freeman BA. Liposome-mediated augmentation of brain SOD and catalase inhibits CNS O 2 toxicity. J Appl Physiol 1984;57:1674-81. 17. Myers ML, Bolli R, Lekich RF, Hartley CJ, Roberts R. Enhancement of recovery of myocardial function by oxygen free-radical scavengers after reversible regional ischemia, Circulation 1985;72:915-21. 18, Jeroudi MO, Triana FJ, Patel BS, Bolli R. Effect ofsuperoxide dismutase and catalase given separately, on myocardial "stunning." Am J Physiol I 990;259:H889-90 I.

Kimberly A. Workowski, Robert J. Suchland, Mary B. Pettinger, and Walter E. Stamm

Division of Infectious Diseases. Department of Medicine. Harborview Medical Center. University of Washington. Seattle

Black race is an important risk marker for Chlamydia trachomatis genital infection. To define whether C. trachomatis serovars differ by ethnic distribution, a panel of monoclonal antibodies was used to serotype 934 urethral and 581 cervical isolates from patients attending a sexually transmitted diseases clinic over 2 years. The overall serovar distribution in cervical and urethral infections was comparable, with B class serovars predominating. Significantly higher inclusion counts were observed both in younger women and in nonblacks regardless ofserovar. Serovar D was less frequent among blacks at the urethral site (P = .001), while serovar Ia was more frequent in blacks at both sites (urethral, P < .001; cervical, P = .02). These associations remained significant after adjusting for age and number of inclusion-forming units by multivariate analysis. Thus, specific serovars may be associated with particular racial groups; either behavioral or biologic factors could explain these findings.

Chlamydia trachomatis is an important sexually transmitted pathogen with worldwide distribution. Using type-specific polyclonal antisera in a microimmunofluorescence assay, Wang et al. [1] differentiated C. trachomatis into 15 serovars and showed that serovars A, B, Ba, and C were associated with endemic trachoma, serovars D to K with genital tract disease, and Ll-L3 with lymphogranuloma venereum. However, few studies have examined more detailed epidemi-

Received 24 February 1992; revised 6 July 1992. Presented in part: ninth meeting, International Society for Sexually Transmitted Diseases Research. Banff. Canada, October 1991 [P-06-144]. Grant support: National Institutes of Health (AI-27456), Reprints: Dr. Walter E, Stamm, 325 9th Ave, Mailstop ZA-89, Department of Infectious Diseases, Seattle, WA 98104. Correspondence (present address): Dr. Kimberly A. Workowski, Division of Infectious Diseases. Crawford Long Hospital, Emory University, 550 Peachtree St. NE, Atlanta. GA 30365. The Journal of Infectious Diseases 1992;166:1445-9 © 1992 by The University of Chicago. All rightsreserved. 0022-1899/92/6606-0040$01.00

ologic and clinical correlates ofgenital infection with specific chlamydial serovars due to the technical difficulty in serotyping large numbers of individual isolates. Young age and black race [2, 3], for example, have been repeatedly identified as risk markers for chlamydial infection, but the association of these variables with specific infecting serovars has not been evaluated. Recently, the availability of monoclonal antibody-based typing assays has made serotyping oflarge numbers ofstrains more practical. Previous data obtained using such methods have suggested that specific serovars (D, G, Ll , and L2) are associated with rectal infection in homosexual men [4], that the infecting serovar may influence the clinical manifestations of cervicitis [5], and that recurrences of chlamydial infection with the same serovar are more often associated with concurrent gonorrhea [6]. We undertook the present study to ascertain whether specific C. trachomatis serovars were associated with patient age, race, gender, or inclusion count in primary culture.

Downloaded from http://jid.oxfordjournals.org/ at East Carolina University on September 12, 2015

Association of Genital Infection with Specific Chlamydia trachomatis Serovars and Race