doi: 10.1161/01.RES.0000196746.75724.8b
© 2005 American Heart Association, Inc.
Letter to the Editor |
Department of Vascular Medicine, Academic Medical Center, Amsterdam, The Netherlands
To the Editor:
We have diligently read the research commentary by Pepys et al,1 which among others addresses the in vivo effects of C-reactive protein (CRP)- infusion into humans, as recently reported by us in a previous issue of this journal.2 We would like to clarify some misconceptions surrounding our study, as generated in their report. Pepys et al observed, using recombinant human (rh)CRP that still contained large quantities of endotoxin, that this rhCRP solution induced an inflammatory reaction both in vitro as well as in mice, whereas CRP from human resources had no such effect. Pepys et al then extrapolate these findings to our study in humans and conclude that the in vivo effects of rhCRP we observed must have been caused by contaminants rather than rhCRP itself. In fact, these authors arrive at the wrong conclusion, and they fail to acknowledge several shortcomings in their own experiments.
First, their dialysed commercial rhCRP displayed residual endotoxin activity of 46·6 endotoxin units (EU) per mg of rhCRP and would for that sake never have been allowed for human use. In comparison, endotoxin activity of our rhCRP was 30x lower (<1·5 EU/mL), which resulted in less than 1·6 EU per mg of rhCRP (at a CRP concentration of 0·91 mg/mL). Interestingly, the trace amounts in our rhCRP are very similar to those reported by Pepys et al in their natural human (nh)CRP solution, ie, 0·9 EU/mg of nhCRP. Needless to state that extrapolation of their findings, using highly contaminated rhCRP, to our study results does not make any sense.
Second, Pepys et al argued that contaminants, rather than CRP, are responsible for the inflammatory effects in humans and tested this hypothesis in a mouse model. Using rhCRP with very high endotoxin activity, they observe a substantial acute phase response in accord with previous reports on the clinical sequelae of endotoxin.3 Nevertheless, Pepys et al fail to acknowledge that the residual endotoxin levels in our rhCRP have been proven insufficient to cause any bioactivity in vivo4 and in vitro. Notably, data on nhCRP now provided by Pepys et al, containing comparable amounts of endotoxin as our rhCRP, further corroborate this observation. To even further substantiate this, we infused lipopolysaccharide (LPS; E coli lipopolysaccharide, lot G2B274, United States Pharmacopeial Convention Inc, Rockville, Md) into two healthy volunteers at a dose (1·5 EU/kg) that equaled the mean coinfused dose during the CRP infusion experiments. Compared with the reference values of higher dose LPS infusion (10 EU/kg; n=4) and rhCRP infusion (n=4) groups, none of the subjects receiving the low dose LPS infusion showed any change in TNF-
, as assayed by cytometric bead array analysis (BD Biosciences; Figure). Despite the heterogeneity in potency among endotoxins of different sources, these data demonstrate that the trace amounts of endotoxin present in our rhCRP-solution cannot have contributed to the inflammatory reactions we observed in our human study subjects.
|
Third, it should be noted that assessing CRP biology in a mouse model has been criticized. As emphasized by Reifenberg et al, the mouse is not a suitable model for assessment of the consequences of human CRP on atherogenesis.5 The latter is most markedly illustrated by the complete absence in this rodent model of one of the main biological functions of CRP, ie, lipoprotein-dependent complement activation. In line, recent observations in genetically engineered mice that regard the effect of human CRP on atherosclerosis are rather contradictory.6,7
Finally, the statement by Pepys et al that the massive SAA response to rhCRP infusion may even increase risk of AA amyloidosis is completely misplaced. In reality, the SAA levels observed in our study are comparable to SAA levels seen in adults infected with uncomplicated influenza.8
In conclusion, we concur with Pepys et al that data on the downstream proinflammatory consequences of CRP must be interpreted with caution. However, the data currently generated by Pepys et al bear no relevance with respect to our findings in humans. This is attributable, among others, to the 30-fold higher endotoxin activity in their solution as well as to the use of a rodent model in evaluating the effects of human CRP. Hence, we still whole-heartedly support Pepys in his conclusion that CRP represents an attractive target for cardiovascular prevention.
References
1. Pepys MB, Hawkins PN, Kahan MC, Tennant GA, Gallimore JR, Graham D, Sabin CA, Zychlinsky A, Diego J de. Pro-inflammatory effects of bacterial recombinant human C-reactive protein are caused by contamination with bacterial products not by C-reactive protein itself. Circ Res. 2005; 97: e97–e103.
2. Bisoendial RJ, Kastelein JJ, Levels JH, Zwaginga JJ, van den BB, Reitsma PH, Meijers JC, Hartman D, Levi M, Stroes ES. Activation of inflammation and coagulation after infusion of C-reactive protein in humans. Circ Res. 2005; 96: 714–716.
3. Haudek SB, Natmessnig BE, Redl H, Schlag G, Hatlen LE, Tobias PS. Isolation, partial characterization, and concentration in experimental sepsis of baboon lipopolysaccharide-binding protein. J Lab Clin Med. 2000; 136: 363–370.[Medline] [Order article via Infotrieve]
4. Levi M, Ten Cate H, Bauer KA, van der PT, Edgington TS, Buller HR, Van Deventer SJ, Hack CE, ten Cate JW, Rosenberg RD. Inhibition of endotoxin-induced activation of coagulation and fibrinolysis by pentoxifylline or by a monoclonal anti-tissue factor antibody in chimpanzees. J Clin Invest. 1994; 93: 114–120.[Medline] [Order article via Infotrieve]
5. Reifenberg K, Lehr HA, Baskal D, Wiese E, Schaefer SC, Black S, Samols D, Torzewski M, Lackner KJ, Husmann M, Blettner M, Bhakdi S. Role of C-reactive protein in atherogenesis:can the apolipoprotein E knockout mouse provide the answer? Arterioscler Thromb Vasc Biol. 2005; 25: 1641–1646.
6. Paul A, Ko KW, Li L, Yechoor V, McCrory MA, Szalai AJ, Chan L. C-reactive protein accelerates the progression of atherosclerosis in apolipoprotein E-deficient mice. Circulation. 2004; 109: 647–655.
7. Trion A, de Maat MP, Jukema JW, van der LA, Maas MC, Offerman EH, Havekes LM, Szalai AJ, Princen HM, Emeis JJ. No effect of C-reactive protein on early atherosclerosis development in apolipoprotein E*3-leiden/human C-reactive protein transgenic mice. Arterioscler Thromb Vasc Biol. 2005; 25: 1635–1640.
8. Falsey AR, Walsh EE, Francis CW, Looney RJ, Kolassa JE, Hall WJ, Abraham GN. Response of C-reactive protein and serum amyloid A to influenza A infection in older adults. J Infect Dis. 2001; 183: 995–999.[CrossRef][Medline] [Order article via Infotrieve]
This article has been cited by other articles:
 |
|
 |
|
  G. Huang, C. Luo, X. Gu, Z. Wu, Z. Wang, Z. Du, C. Hu, and L. Tang Mechanical Strain Induces Expression of C-Reactive Protein in Human Blood Vessels J. Pharmacol. Exp. Ther., July 1, 2009; 330(1): 206 - 211. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W.-D. Docke, K. Asadullah, G. Belbe, M. Ebeling, C. Hoflich, M. Friedrich, W. Sterry, and H.-D. Volk Comprehensive biomarker monitoring in cytokine therapy: heterogeneous, time-dependent, and persisting immune effects of interleukin-10 application in psoriasis J. Leukoc. Biol., March 1, 2009; 85(3): 582 - 593. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. J. Bisoendial, J. J. P. Kastelein, S. L. M. Peters, J. H. M. Levels, R. Birjmohun, J. I. Rotmans, D. Hartman, J. C. M. Meijers, M. Levi, and E. S. G. Stroes Effects of CRP infusion on endothelial function and coagulation in normocholesterolemic and hypercholesterolemic subjects J. Lipid Res., April 1, 2007; 48(4): 952 - 960. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. M. Scirica, D. A. Morrow, S. Verma, S. Devaraj, I. Jialal, B. M. Scirica, D. A. Morrow, S. Verma, S. Devaraj, and I. Jialal The Verdict Is Still Out Circulation, May 2, 2006; 113(17): 2128 - 2151. [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||