Short Communication

Mediators of Inflammation, 7, 417420 (1998)

WE in ve s tigated th e in vitro e ffe ct of dif fe re nt for m s of acido s is (p H 7.0) on th e for m ation of anap h ylato x in s C3a an d C5a. Me tabo lic acido s is due to addition of h ydr och loric acid (10 m m o l/m l bloo d) or lactic acid (5.5 m m o l/m l) to h e p arin bloo d (N=12) caus ed s ign ifican t activation of C3a an d C5a com p ar ed to con tr ol (bo th p =0.002). Re s p irator y acidos is activated C3a (p =0.007) an d C5a (p =0.003) com p ar ed to n or m o cap n ic con tro ls. Ma kin g blo od s am p les w ith lactic acido s is h yp ocap n ic re s ulted in a m edia n p H o f 7.37. In th is re s p iratory com pe ns ated m etabo lic acidos is , C3a and C5a w er e n ot in cr eas ed. Th es e ex p e rim en ts s h ow th at acidos is its elf and n ot lactate tr igge r fo r activation o f com p le m en t co m p on en ts C3 an d C5. Key w or ds : Ac idosis, Alcalosis, Complement syste m, Comple me nt activation, Anaphylatox ins

Acidosis activates complement system in vitro

Michael Emeis,1 Josef Sonntag,1.CA Carsten Willam,2 Evelyn Strauss,1 Matthias M. Walka and Michael Obladen1 Departments of 1Neonatology and 2Nephrology, Charit Virchow-Hospital, Humboldt-University e, Berlin, 13353 Berlin, Germany

CA

Corresponding Author Te l: (+49) 30 4506646 3 Fax : (+49) 30 4506692 2

Introduction
Hypox ia and re perfusion c ause compleme nt activation in animal ex perime nts and in clinical studie s.1 4 It c ould be show n, that after c ytological damage, contac t w ith cellular compone nts such as mitochondria, or ex c ess of hydrox yl radicals, w as re spons ible for the activation of the compleme nt syste m.5 ,6 On the othe r hand, in vitro studies show ed a compleme nt ac tivation by ac idosis or hypox ia only.7 9 In a pre vious study, w e found complement ac tivatio n induc ed by lactic acid.10 The aim of the pre sent in v itro study w as to inve stigate: (1) w he the r lac tate itself or metabolic acidosis is re sponsible for activation, and (2) w hethe r re spiratory acidosis is also able to ac tivate the complement syste m.

Material and Methods

Blood sample s (10 ml) from 12 healthy volunte ers (6 male and 6 female, age : 28 40 years) w e re c ollec te d and 50 IE Heparin w as adde d. Each sample w as divide d into e ight portions and plac ed in polypropylen tube s. To inve stigate the influe nce of re sp iratory as w ell as me tabolic changes of pH on anaphylatox in formation, the p ortions w e re e quilibrated w ith differe nt gas mix tures and supplemented w ith acids re sulting in marked changes in pH, pCO2 and base de ficit (Table 1). Probe s used as c ontrols are No. 13. Four portions w ere e quilibrate d w ith differe nt gas mix tures at 37C
0962-9351/98/060417-04 $9.00 1998 Carfax Publishing Ltd

to achie ve normocapnia and normox e mia in c ontrol portions (No. 3; Table 1), re spiratory acidosis (No 6; Table 1) or alcalosis (No. 7,8; Table 1). For this purpose w e modified the me thod of Siriw ardhana e t a l. 11 A 6 ml polyprop ylen tube (6 ml) w as filled 1.5 to 2.0 ml blood and the end of a polyprop yle n tube w ith an inne r diame te r of 3 mm w as place d into the blood a few millimetre s above the bottom of the tube . Gas mix tures (O2 , N2 , CO2 ) c omposed from me dic al gas es by f low me te rs w e re put through the tube into the blood sample. Gas flow w as adjuste d to one bubble pe r second for 20 min. Foam caused by bubbling w as c ontinuou sly re move d by a suction cathete r place d above the blood level. After e quilibration, the blood samples w e re transferred into different tubes for further proce ssing. For metabolic ac idosis 10 m mol hydrochloric ac id (No. 4; Sigma-Aldrich, De isenhofe n, Ge rmany) or 5.5 m mol lactate (No: 5; Sigma-Aldrich, Deise nhofen, Ge rmany) w ere adde d per ml of blood. To achie ve compe nsated me tabolic ac idosis , the sample w as equilibrate d w ith a hypoc apnic gas mix ture for 20 min and the n 5.5 m mol lac tate per ml of blood w as added (No. 8). Blood gas analysis, potassium and lactate c onc entrations w ere me as ure d by a Radiome te r Copenhagen ABL 505 (Willich, Ge rmany) using he parinise d syringes after all samples w ere incubated at 37C for 1 h. To stop c omplement activation after inc ubation, w e added 1 mg EDTA dissolve d in purified w ater to each sample. Plasma w as se parated by ce ntrifugation at
417

M. Em e is e t al. Table 1. Sample characteristics with different forms of acidosis after gas equilibration, addition of acids and 1h incubation time at 37C (N=12). Values are given as medians with quartiles No Name 1 2 3 Treatment Sample procedure No No No No 78% N2 , 5% CO2 , 17% O2 , 20min No No 10 m mol HCl per 1ml sample No 5.5 m mol lactate per 1ml sample 81% N2 , 19% CO2 , 20min No 83% N2 , 17% O2 , 20min No 83% N2 , 17% O2 , 20min 5.5 m mol lactate per 1ml sample pH value 7.36 7.35/7.36 7.30 7.27/7.33 7.30 pCO2 (mmHg) 50 46/51 50 44/53 51 Base excess Potassium (mEq/L) (mmol/L) 0.0 0.0/0.0 3.0 2.4/4.4 2.2 4.1 4.0/4.3 3.5 3.4/3.7 4.2 C3a (m g/L) 128 89/144 1595 972/2070 1025 C5a (m g/L) 0.13 0.10/0.42 1.4 0.9/2.6 1.4

baseline value bubbling: (without incubation) additive: incubation control bubble control bubbling: additive: bubbling:

additive: 4 HCI acidosis bubbling: additive:

7.25/7.35 7.01 7.00/7.05

45/57 78 71/85

0.8/3.0 13.8 12.4/14.9

3.9/4.4 5.2 4.5/5.7

802/1257

0.8/2.8

3360 21.8 2860/3850 17.8/35.2

lactic acidosis

bubbling: additive:

6.98 6.90/7.00

57 50/76

17.1 16.2/19.6

6.4 6.4/6.8

2390 1975/2670

4.4 3.2/6.6

respiratory acidosis

bubbling: additive:

7.01 6.97/7.04 7.54 7.49/7.57 7.37 7.32/7.38

137 123/158 20 17/23 28 24/34

5.2 3.9/7.3 3.5 2.4/4.0 8.3 7.3/10.5

4.3 3.9/4.7 4.4 4.1/4.7 4.2 4.1/4.5

1625 1080/1868 917 825/1425 935 760/1150

3.3 1.8/4.2 1.4 0.8/2.8 1.8 1.3/5.4

respiratory alcalosis bubbling: additive:

respiratory alcalosis bubbling: + lactate additive:

4C at 3000 g for 10 min and froze n immediate ly at 80C. Anaphylatox ins C3a and C5a w ere dete rmined as pre viously desc ribed.1 2 C3a enzyme immunoassay (EIA, Fa. Progen Biote chnik GmbH, Heide lberg, Ge rmany) selec tively de te c ts C3a-desArg using monoclonal antib odie s.13 C5a w as dete rmine d w ith a spe cific sandw ich EIA (Fa. Behring, Marburg, Ge rmany). 14

Results
Influence of incubation and equilibration procedures
The inc ubation (No. 2) and equilibration (No. 3) proce dure led to a measurable comple ment activation compared to the base line value (No. 1; Table 1). There fore the re sults w e re compared to controls of the same handling method but w ithout pH change .

Statistical analysis
As most of the data w e re not normally distribute d, re sults w ere show n as me dians w ith quartiles. Diffe renc es be tw een c ontrols and study samples w ere asse sse d using the Wilcox on te st. Statistical significance w as assume d at p< 0.05. All c alc ulations and te sts w e re performed w ith the SPSS-PC softw are (Chicago, Illinois, USA).
418 Mediators of Inflammation Vol 7 1998

Influence of metabolic and respiratory acidosis

Addition of hydrochloric as w ell as lactic acid c ause d a marked metabolic ac idosis. Conc entrations of C3a, C5a, and potassium w ere higher after inc ubation in acidic blood c omp ared to inc ubation of c ontrol samples (No. 4 and 5 versus 2; Table 1). The potassium conc entration w as low e r in hydrochloric acidosis and the anaphylatox in c once ntration s w e re highe r than

Acido s is a ctiva te s co m ple m e n t s y s te m

those from lactic acidosis (No. 4 ve rsus 5). The anaphylatox ins w ere also higher in the hype rc ap nic samples c ompared to the normoc apnic c ontrols (No. 6 ve rsus 3; Table 1).

Influence of respiratory alkalosis

Respiratory alkalosis had no influenc e on anaphylatox in c onc entration (No. 7 versus 3; Table 1).

Influence of respiratory compensated metabolic acidosis

Although measured median lac tate (13.1 ve rsus 11.6 mmol/l) w as not diffe re nt in the lac tic samples (No. 5 ve rsus 8; p = 0.75), anap hylatox ins w ere higher in ac idic sample s compared to re spiratory c ompensate d normac idic samples (No. 5 ve rsus 8; Table 1).

me mbrane fragments of damage d erythroc ytes 6 is improbable, as w e could show that activation also occ urs in plasma w ithout ce ll compone nts.1 0 There are other arguments against membrane fragme ntinduc ed complement ac tivation in ac idosis, incre ase d potassium value s and a slight visible hemolysis occ ure d only for metabolic but not for re spiratory acidosis , but both forms of ac idosis activate the comple ment syste m. A probable me chanism is that acidification of plasma inactivate s compleme nt protease inhibitors, this inhibition then re sults in compleme nt ac tivation in the absenc e or dec re as ed pre senc e of functional inhibitors. The activation of the compleme nt syste m by acidosis may ex p lain the ele vation of anap hylatox ins in diffe re nt dise as es, such as pe rinatal asphyx ia, myoc ardial infarction and shock.3,4 ,1 5 1 7 The se proce sse s probably partic ipate in the pathoge ne sis of re pe rfusion injury.4 ,17 1 9

Discussion
The p re se nt in vitro study c onfirms the re sults of our pre vious study that ac idosis ac tivates compleme nt syste ms in heparin blood. The study show s that the acidosis itself is the trigger for ac tivation, be c ause all thre e forms of ac idosis (hydrochloric or lac tic acid or carbon diox ide) le ad to a signific ant incre ase of anaphylatox in conc entrations. Lac tate did not lead to an ac tivation, w he n ac idosis w as pre ve nted by pre vious re spiratory alcalosis (No. 8; Table 1). Heparin w as chose n as antic oagulant, be c ause the comple ment may be markedly spontaneously activated in se rum and both EDTA and sodium c itrate possibly influe nce pH value s and impair compleme nt activation by c alc ium binding. Heparin how ever did not complete ly pre vent spontaneous compleme nt activation even after incubation for 1 h at 37C and using a bubble ox yge nator for gas e quilibration. There fore w e could only c ompare the ex te nt of comple ment activation by ac idosis to the spontane ous activation that oc curred during the sample pre paration proce dure. As pH le ve ls w e re similar in lac tic and hydrochloric acidosis, the ac tivation of the comple ment syste m w as stronger for hydrochloric acid. The e ffec ts of the tw o acids on the blood gas analysis w ere differe nt. Whe re as lac tate led to a large r inc re as e in base de fic it, hydrochloric ac id c aused a stronge r re lease of carbon diox ide . This differe nc e may ex plain the stronger compleme nt activation by hydrochloric acid. Respiratory acidosis follow ing equilibration of blood samples w ith e le vated c arbon diox ide conc entrations also le d to an ac tivation of the comple ment syste m compared to e quilibration control. In contrast to ac idosis, re spiratory alcalosis did not have an influenc e on the compleme nt syste m. With the use d in v itro setting, w e could not clarify the mechanism of anaphylatox in formation by ac idosis. Complement activation in acidosis mediate d by

Conclusion
We have show n evidenc e that acidosis ac tivates comple ment factors C3 and C5. This is indepe nde nt from the type of ac idosis w hich occurs.

References
1. Hill J, Lindsay TF Ortiz F Ye h CG, He c htman HB, Moore FD. Soluble , , c omp le me nt rec e p tor type 1 ame liorate s th e loc al and remote organ injury afte r inte stinal isc he mia re p e rfus ion in the rat. J Im m u n o l 1992; 149 : 1723 1728. 2. Brus F, van Oeve re n W, Okke n A, Oe tomo SB. Activation of c irc ulating p olymorp honuc le ar le ukoc yte s in p re te rm infants w ith se ve re idiop athic re spiratory distre s s syndrome . Pe dia tr Re s 1996; 39: 456 463. 3. Sonnta g J, Wagne r MH, Straus s E, Oblade n M. Comp le me nt and c onta c t activation in te rm ne w b or ns afte r fe tal acidos is. Ar ch Dis Child Fe ta l Ne o n a ta l Ed 1998; 78 : F125 F128. 4. Hors tic k G, He imann A, G tz e O, Hofne r G, Be rg O, Boe hme r P, e t a l. o Intracoronary app lication of C 1 e ste ras e inhib itor imp rove s c ardiac func tion and reduc e s myoc ardial ne c rosis in an ex p e rime ntal mode l of isc he mia and rep e rfusion. Cir cu la tio n 1997; 95: 701708. 5. Turnage RH, Mage e JC, Guic e KS, Myers SI, Oldham KT. Comp le me nt activation b y hydrox yl radicals during inte stinal rep e rfusion. Sh o c k 1994; 2: 445 450. 6. Ros se n RD, Michael LH, Ka giyima A, Savage HE, Hanson G, Re isbe rg MA, e t a l. Me c hanis m of c omp le me nt ac tivation afte r c oronary arte ry oc c lusion: Evide nc e that myoc ardial isc he mia in dogs c ause s rele ase of c ons titue nts of myoc ardial subc e llular origin th at complex w ith human C1q in v iv o . Cir cu la tio n Re s e a rc h 1988; 62: 72 584. 7. Fishe ls on Z, Horstman RD, M lle r-Ebe rhard HJ. Re gulation of the u alte mative pathw ay of c omp le me nt by p H. J Im m u n o l 1987; 138: 3392 3395. 8. Hamme r CH, Hans ch G, Gre s ham HD, Shin ML. Activation of th e fifth and six th c omp one nts of the human comp le me nt s yste m: C6-de p e nde nt c le avage of C5 in acid and the formation of a bimole cular lytic c omp lex . J Im m u n o l 1983; 131: 892 898. 9. Collard CD, Vake va A, Bukusoglu C, Zund G, Sp e rati CJ, Colgan SP, e t a l. Re ox yge nation of hyp ox ic human umbilical ve in e ndothe lial c e lls activate s th e class ic c omp le me nt p athw ay. Circ u la tio n 1997; 96: 226. 10. Sonnta g J, Eme is M, Strauss E, Oblade n M. In v itro activation of c omp le me nt and c onta c t syste m by lac tic acidosis. Me dia t In fla m m 1998; 7: 49 51. 11. Siriw ardhana SA, Kaw as A, Yate s S, Gulde n H, Lip ton JM, Gie s e c ke AH. Ox yge n me diate d comp le me nt activation: an in-vitro study. Ca n J An a e s th 1990; 37 : 116. 12. Sonnta g J, Stille r B, Walka MM, Maie r R. Anaphylatox ins in fre s h froze n p lasma. Tra n s fu s io n 1997; 37: 798 803. 13. Burge r R, Zilow G, Bade r A, Frie dle in A, Nas e r W. The C te rminus of th e anap hylatox in C3a ge ne rate d upon c omp le me nt activation rep rese nts a ne oantige nic de te rminant w ith diagnos tic pote ntial. J Im m u n o l 1988; 14: 553 558.

Mediators of Inflammation Vol 7 1998

419

M. Em e is e t al.
14. Klos A, Ihrig V Me s sme r M, Grabbe J, Bitte r-Sue rmann D. De te c tion of , native human c omp le me nt comp one nts C3 and C5 and the ir p rimary activation pe ptide s C3a and C5a (anap hylatox ic p e p tide s) by ELISAs w ith mono clonal antibodie s . J Im m u n o l Me tho ds 1988; 111: 241252. 15. He ide man M, Hugli TE. Anap hylatox in gene ration in multisyste m organ failure. J Tra u m a 1984; 24: 1038 1043. 16. We ise r MR, Williams JP, Moor e FD, Kob zik L, Ma M, He c htman HB, e t a l. Re p e rfus ion injury of ischemic s kele tal mus cle is me diate d b y natural antibody and comp le me nt. J Ex p Me d 1996; 183: 2343 2348. 17. Bue rke M, Murohora T, Lefer AM. Cardiop rote ctive e ffe c ts of a C1 e s te rase inhibitor in myocardial isc hemia and rep e rfusion. Circ u la tio n 1995; 91: 393 402. 18. Ikai M, Itoh M, Joh T, Yokoyama Y, Okada N, Okada H. Comple me nt p lays an e sse ntial role in shoc k follow ing inte stinal ischemia in rats. Clin Ex p Im m u n o l 1996; 106: 156 159. 19. Craw ford MH, Grove r ML, Kolb WP, Mc Mahan A, ORourke RA, Mc Manus LM, e t a l. Comp le me nt and ne utrop hil activation of the p athoge ne sis of isc he mic myocardial injury. Cir cu la tio n 1988; 78: 144 148. ACKNOWLEDGEMENTS. This w ork w as s upporte d b y the Ge rman Re s e arch Soc iety (DFG-0b 43/6 2).

Received 10 September 1998; accepted 5 October 1998

420

Mediators of Inflammation Vol 7 1998