International Journal of Laboratory Hematology The Official journal of the International Society for Laboratory Hematology
ORIGINAL ARTICLE
INTERNAT IONAL JOURNAL OF LABORATO RY HEMATO LOGY
Is there a role of C-reactive protein in red blood cell aggregation? D. FLORMANN*, E. KUDER † , P. LIPP † , C. WAGNER*, L. KAESTNER †
*Experimental Physics, Saarland University, Saarbr€ ucken, Germany † Institute for Molecular Cell Biology and Research Centre for Molecular Imaging and Screening, School of Medicine, Saarland University, Homburg/ Saar, Germany Correspondence: Dr Lars Kaestner, Institute for Molecular Cell Biology, Research Cente for Molecular Imaging and Screening, Saarland University, Building 61, 66421 Homburg/Saar, Germany. Tel.: +49 6841 1626149; Fax: +49 6841 1626104; E-mail: [email protected]
doi:10.1111/ijlh.12313
S U M M A RY Introduction: Numerous clinical studies related the plasma level of C-reactive protein (CRP) to the erythrocyte sedimentation rate (ESR) independent of the kind of disease. The molecular regulation of the process is unknown. Methods: We performed a meta-analysis of 10 previous studies and experimentally probed for a direct action of CRP on red blood cells (RBCs) by different methods including determination of a microscopic aggregation index, Ca2+ imaging and analysis of sedimentation experiments. Results: The meta-analysis revealed a statistically significant correlation (Pearson coefficient of 0.37; P < 0.0001), but we could not find any experimental evidence for a direct CRP–RBC interaction. Instead, we could confirm a correlation between fibrinogen level and ESR. Conclusion: Therefore, we concluded that CRP and ESR cannot account for nor replace each other as a diagnostic measure. The correlation between CRP level and ESR is most probably caused by fibrinogen, because its increase coincides with elevated CRP levels.
Received 10 September 2014; accepted for publication 13 October 2014 Keywords C-reactive protein, erythrocyte sedimentation rate, aggregation index, thrombotic risk
INTRODUCTION C-reactive protein (CRP) was originally found by Tillett and Francis [1] and the ‘C’ originates from the fact that it was the third fraction derived from pneumococci – at that time believed to be a pathogenic secretion [2]. It was identified as an acute-phase protein [3]. Later, it was shown to be an endogenous 474
protein [4], which is produced predominantly in the liver [5]. Nowadays CRP is the classical acute-phase protein, whose plasma level is very frequently determined for patients in clinical practice. Its entire regulation and effectors are still far from being fully understood (PubMed lists 740 papers with ‘C-reactive protein’ in the title just for papers published in 2013). CRP has © 2014 John Wiley & Sons Ltd, Int. Jnl. Lab. Hem. 2015, 37, 474–482
D. FLORMANN ET AL. | PROBING C-REACTIVE PROTEIN ON RED CELLS
been connected to thrombotic events, for example myocardial infarction [6], stroke [7], thrombosis associated with injury [8] or thrombotic effects in general [9]. Aside all correlations between CRP levels and pathophysiological incidents, there is more direct evidence based on the availability of a specific small-molecule inhibitor of CRP [6]. Administration of this inhibitor to rats undergoing acute myocardial infarction abrogated the increase in infarct size and cardiac dysfunction produced by injection of human CRP [6]. A recent investigation associated CRP related thrombosis with prostaglandins [8]. Interestingly, prostaglandin E2 (PGE2) was proposed to induce red blood cell (RBC) aggregation [10]. The concept of an active participation of RBC in clot formation and thrombosis was emerging over almost 100 years and was reviewed by Andrews and Low in 1999 ([11] and references therein). Ever since this concept was further developed [12] and additional evidence came from experimental [13–15] and clinical observations, including sickle cell disease [16], malaria [17], dialysis patients under erythropoietin treatment [18], from very low birthweight infants during the first week of life [19] or as side effects in photodynamic therapy [20], just to name a few. Concomitantly, CRP and erythrocyte sedimentation rate (ESR) were related to each other in a number of studies of different clinical background ranging from orthopaedic studies (rheumatoid arthritis) [21–23] and postsurgery monitoring [24, 25] to autoimmune disease [26] and several forms of inflammation (osteomyelitis [27], endocarditis [28], pancreatitis [29] and ulcerative colitis [30]). Despite all of this evidence, we are not aware of a report probing the direct effect of CRP on RBCs.
M AT E R I A L S A N D M E T H O D S RBC preparation Venous blood of healthy donors was drawn into standard EDTA tubes (S-Monovette; Sarstedt, N€ umbrecht, Germany). After that it was washed three times by centrifugation (70 g; 3 min) with phosphate-buffered solution (PBS, Life Technologies, Waltham, MA, USA) supplemented with bovine serum albumine (BSA, Sigma-Aldrich, St Louis, IL, USA) at a concentration of 1 mg/mL. The RBCs were stored at 4 °C and used © 2014 John Wiley & Sons Ltd, Int. Jnl. Lab. Hem. 2015, 37, 474–482
475
for experiments within 4 h. For fluorescence measurements, 0.1 lL CellMask (Life Technologies) was added to 200 lL packed RBCs in 1 mL PBS. After a waiting time of 10 min, the RBCs were centrifuged again (704 g; 3 min), and the supernatant was discarded. Sedimentation measurements All sedimentation measurements were performed at a temperature of 23 °C with an haematocrit of 10%. CRP from human plasma was purchased from Hoelzel Diagnostika (Cologne, Germany) and fibrinogen from Sigma-Aldrich. Both substances were diluted with PBS to different concentrations. Sedimentation experiments were performed in 170-QS cuvettes (Hellma Analytics, M€ ullheim, Germany) with a filling level of 120 lL. Visualization of the sedimentation front was performed with a consumer camera (EOS 550D with a EF-S 18–200 mm lens; Canon, Tokyo, Japan). The spectroscopic measurements were performed with the Spectronic Genesys 6 (Thermo Scientific, Waltham, MA, USA) at a wavelength of 940 nm. The inflection point was determined as the maximum of the third derivative of the absorption curves over time. Microscopic measurements All microscopic measurements were performed at a temperature of 23 °C with an haematrocrit of 10%, unless otherwise stated. The RBCs were labelled with CellMask (see above) and l-Slides 18 well (ibidi, Munich, Germany) were used as measuring chambers. Imaging was performed using a confocal microscope (TSC SP5 II; Leica Microsystems, Mannheim, Germany) as recently described [31] using a 633 nm laser as the excitation source. RBCs were classified as part of an aggregate, if the membrane of 2 cells was optically not distinguishable over a distance of at least 3 lm. The number of RBCs in a field of view refers to the above-mentioned condition in an area of 144 9 144 lm. Calcium imaging Calcium imaging was performed using the small-molecule indicator Fluo-4 (Life Technologies) at loading conditions as previously described [32]. Measurements were performed in Tyrode solution containing (in
476
D. FLORMANN ET AL. | PROBING C-REACTIVE PROTEIN ON RED CELLS
Table 1. Overview of clinical studies relating C-reactive protein (CRP) to erythrocyte sedimentation rate (ESR) and their meta-analysis
Gaussian distributed
Pearson coefficient (Spearman coefficient)
Significance of correlation
56 (5)
Yes
0.76
n.s.
40 (14) 150 (26)
Yes Yes
0.35 0.94
n.s.
ORIGINAL ARTICLE
INTERNAT IONAL JOURNAL OF LABORATO RY HEMATO LOGY
Is there a role of C-reactive protein in red blood cell aggregation? D. FLORMANN*, E. KUDER † , P. LIPP † , C. WAGNER*, L. KAESTNER †
*Experimental Physics, Saarland University, Saarbr€ ucken, Germany † Institute for Molecular Cell Biology and Research Centre for Molecular Imaging and Screening, School of Medicine, Saarland University, Homburg/ Saar, Germany Correspondence: Dr Lars Kaestner, Institute for Molecular Cell Biology, Research Cente for Molecular Imaging and Screening, Saarland University, Building 61, 66421 Homburg/Saar, Germany. Tel.: +49 6841 1626149; Fax: +49 6841 1626104; E-mail: [email protected]
doi:10.1111/ijlh.12313
S U M M A RY Introduction: Numerous clinical studies related the plasma level of C-reactive protein (CRP) to the erythrocyte sedimentation rate (ESR) independent of the kind of disease. The molecular regulation of the process is unknown. Methods: We performed a meta-analysis of 10 previous studies and experimentally probed for a direct action of CRP on red blood cells (RBCs) by different methods including determination of a microscopic aggregation index, Ca2+ imaging and analysis of sedimentation experiments. Results: The meta-analysis revealed a statistically significant correlation (Pearson coefficient of 0.37; P < 0.0001), but we could not find any experimental evidence for a direct CRP–RBC interaction. Instead, we could confirm a correlation between fibrinogen level and ESR. Conclusion: Therefore, we concluded that CRP and ESR cannot account for nor replace each other as a diagnostic measure. The correlation between CRP level and ESR is most probably caused by fibrinogen, because its increase coincides with elevated CRP levels.
Received 10 September 2014; accepted for publication 13 October 2014 Keywords C-reactive protein, erythrocyte sedimentation rate, aggregation index, thrombotic risk
INTRODUCTION C-reactive protein (CRP) was originally found by Tillett and Francis [1] and the ‘C’ originates from the fact that it was the third fraction derived from pneumococci – at that time believed to be a pathogenic secretion [2]. It was identified as an acute-phase protein [3]. Later, it was shown to be an endogenous 474
protein [4], which is produced predominantly in the liver [5]. Nowadays CRP is the classical acute-phase protein, whose plasma level is very frequently determined for patients in clinical practice. Its entire regulation and effectors are still far from being fully understood (PubMed lists 740 papers with ‘C-reactive protein’ in the title just for papers published in 2013). CRP has © 2014 John Wiley & Sons Ltd, Int. Jnl. Lab. Hem. 2015, 37, 474–482
D. FLORMANN ET AL. | PROBING C-REACTIVE PROTEIN ON RED CELLS
been connected to thrombotic events, for example myocardial infarction [6], stroke [7], thrombosis associated with injury [8] or thrombotic effects in general [9]. Aside all correlations between CRP levels and pathophysiological incidents, there is more direct evidence based on the availability of a specific small-molecule inhibitor of CRP [6]. Administration of this inhibitor to rats undergoing acute myocardial infarction abrogated the increase in infarct size and cardiac dysfunction produced by injection of human CRP [6]. A recent investigation associated CRP related thrombosis with prostaglandins [8]. Interestingly, prostaglandin E2 (PGE2) was proposed to induce red blood cell (RBC) aggregation [10]. The concept of an active participation of RBC in clot formation and thrombosis was emerging over almost 100 years and was reviewed by Andrews and Low in 1999 ([11] and references therein). Ever since this concept was further developed [12] and additional evidence came from experimental [13–15] and clinical observations, including sickle cell disease [16], malaria [17], dialysis patients under erythropoietin treatment [18], from very low birthweight infants during the first week of life [19] or as side effects in photodynamic therapy [20], just to name a few. Concomitantly, CRP and erythrocyte sedimentation rate (ESR) were related to each other in a number of studies of different clinical background ranging from orthopaedic studies (rheumatoid arthritis) [21–23] and postsurgery monitoring [24, 25] to autoimmune disease [26] and several forms of inflammation (osteomyelitis [27], endocarditis [28], pancreatitis [29] and ulcerative colitis [30]). Despite all of this evidence, we are not aware of a report probing the direct effect of CRP on RBCs.
M AT E R I A L S A N D M E T H O D S RBC preparation Venous blood of healthy donors was drawn into standard EDTA tubes (S-Monovette; Sarstedt, N€ umbrecht, Germany). After that it was washed three times by centrifugation (70 g; 3 min) with phosphate-buffered solution (PBS, Life Technologies, Waltham, MA, USA) supplemented with bovine serum albumine (BSA, Sigma-Aldrich, St Louis, IL, USA) at a concentration of 1 mg/mL. The RBCs were stored at 4 °C and used © 2014 John Wiley & Sons Ltd, Int. Jnl. Lab. Hem. 2015, 37, 474–482
475
for experiments within 4 h. For fluorescence measurements, 0.1 lL CellMask (Life Technologies) was added to 200 lL packed RBCs in 1 mL PBS. After a waiting time of 10 min, the RBCs were centrifuged again (704 g; 3 min), and the supernatant was discarded. Sedimentation measurements All sedimentation measurements were performed at a temperature of 23 °C with an haematocrit of 10%. CRP from human plasma was purchased from Hoelzel Diagnostika (Cologne, Germany) and fibrinogen from Sigma-Aldrich. Both substances were diluted with PBS to different concentrations. Sedimentation experiments were performed in 170-QS cuvettes (Hellma Analytics, M€ ullheim, Germany) with a filling level of 120 lL. Visualization of the sedimentation front was performed with a consumer camera (EOS 550D with a EF-S 18–200 mm lens; Canon, Tokyo, Japan). The spectroscopic measurements were performed with the Spectronic Genesys 6 (Thermo Scientific, Waltham, MA, USA) at a wavelength of 940 nm. The inflection point was determined as the maximum of the third derivative of the absorption curves over time. Microscopic measurements All microscopic measurements were performed at a temperature of 23 °C with an haematrocrit of 10%, unless otherwise stated. The RBCs were labelled with CellMask (see above) and l-Slides 18 well (ibidi, Munich, Germany) were used as measuring chambers. Imaging was performed using a confocal microscope (TSC SP5 II; Leica Microsystems, Mannheim, Germany) as recently described [31] using a 633 nm laser as the excitation source. RBCs were classified as part of an aggregate, if the membrane of 2 cells was optically not distinguishable over a distance of at least 3 lm. The number of RBCs in a field of view refers to the above-mentioned condition in an area of 144 9 144 lm. Calcium imaging Calcium imaging was performed using the small-molecule indicator Fluo-4 (Life Technologies) at loading conditions as previously described [32]. Measurements were performed in Tyrode solution containing (in
476
D. FLORMANN ET AL. | PROBING C-REACTIVE PROTEIN ON RED CELLS
Table 1. Overview of clinical studies relating C-reactive protein (CRP) to erythrocyte sedimentation rate (ESR) and their meta-analysis
Gaussian distributed
Pearson coefficient (Spearman coefficient)
Significance of correlation
56 (5)
Yes
0.76
n.s.
40 (14) 150 (26)
Yes Yes
0.35 0.94
n.s.
