When blood is brought in contact with foreign surfaces, i.e., during hemodialysis or cardiopulmonary bypass (CPB), a variety of complex and interrelated events leading to an inflammatory response occur. In particular, activation of mononuclear cells and concomitant complement activation induce the release of an array of inflammatory mediators into the extracellular environment, including cytokines.1 Heparin coating of foreign materials in several instances has been shown to improve clinical outcomes involving procedures requiring contact of human tissue with foreign materials. Thus, CPB with heparin-coated oxygenators in several studies2 has reduced the systemic inflammatory reaction. Similarly, a significant reduction of postoperative inflammatory complications has been reported with the use of heparin-coated intraocular eye lenses.3 The known affinity of tumor necrosis factor (TNF) to heparin4–6 has also been used in several therapeutic approaches. For instance, Pevni et al. 7 used heparin in a cardioplegic solution to inhibit TNF production and attenuate myocardial ischemic-reperfusion injury.
Although heparin is a glycosaminoglycan exclusively found in the granules of mast cells, the structurally related heparin sulfate-containing proteoglycans are present on the surface of nearly all adherent mammalian cells and are ubiquitous components of cell surfaces and extracellular matrix in eukaryotic organisms.4 The carbohydrate portions of heparin and heparan sulfate are complex and composed of alternating d-glucosamine and hexuronic acid (d-glucuronic or l-iduronic) residues, thus allowing for very variable interactions with proteins, which likely are the key to their many and incompletely understood biologic activities.4 On cellular membranes, heparin sulfates react with chemokines, cytokines, and growth factors, likely facilitating specific receptor binding.4
Recent studies have indicated a potential therapeutic role for heparin as an inhibitor of cytokine signaling,8,9 and we hypothesized that some of these putative antiinflammatory properties could be mediated as a result of the molecule acting as a receptor for selected cytokines, analogous with heparin sulfate glycosaminoglycans. Because immobilized heparin should thus have a higher affinity for these proteins, we tested the hypothesis that passage through a heparin-coated column could selectively remove cytokines from human blood while not activating the immune response.
Materials and Methods
Preparation of Heparin Column
Polyethylene (PE) beads, with an average diameter of 0.3 mm (lot no. 180153), were supplied by the Polymer Technology Group (Berkeley, CA) and columns (Mobicol, 1 ml) were obtained from MoBiTec (Göttingen, Germany). Heparin and polyethyleneimine (PEI) were purchased from Scientific Protein Laboratories (Waunakee, WI) and BASF (Ludwigshafen, Germany), respectively. All chemicals used were of analytical grade or better.
The PE-beads were etched with an oxidizing agent (potassium permanganate in sulfuric acid). These hydrophilized beads, inter alia containing OH-groups and double bonds, were later used as controls. Immobilization of heparin onto the beads was performed as described by Larm et al.10 The resulting PE-beads, with covalently end-point attached heparin, were sterilized with ethylenoxide and rinsed with 0.9% sodium chloride and ultra pure water. The amount of heparin was determined to be 2.0 mg heparin/g bead with the 3-methyl-2-benzothiazolinone hydrochloride (MBTH) method.10,11
The study protocol was approved by the local ethics committee at the Karolinska University Hospital, and signed informed consent was obtained from each patient. Arterial blood was drawn from the hemodialyzers of three septic (fever >38°C, chills, leukocytes >12 × 109 cells/L) patients (2 men and 1 woman, aged 43, 56, and 68 years, respectively; Table 1). The patients were previously administered with broad-spectrum antibiotics (ceftazidime or cefuroxime together with an aminoglycoside; one dose of each) and heparin (200 IU/kg body weight at the start of the dialysis). The blood was collected in ethylenediaminetetraacetic acid (EDTA) vacuum tubes and immediately transferred to an adjacent room, where 1 ml was applied to the previously prepared columns and passed through using a roller pump at intervals of 1, 5, and 10 ml/min. Blood that had passed through the columns was immediately collected at the other end and cold centrifuged (4500g). The supernatants were subsequently collected and frozen at −80°C for later analysis.
The amounts of cytokines were determined using photoluminescence with a plate reader (Multiskan Ascent, Thermo Scientific). Each sample was measured in three wells, and the geometric mean was used for analysis. The intraassay coefficient of variation was below 8% for all kits. We used Coamatic antithrombin kit (Hemochrom, Cat 8211991), Quantikine Human interleukin (IL)-6 (R&D Systems, Cat D6050, Minneapolis, MN), Quantikine Human IL-10 (R&D Systems, Cat D1000B, Minneapolis, MN), Protein C Antigen Test 96 (HL Scandinavia AB, product H5285), Human CCL5/RANTES Quantikine (R&D Systems, Cat DRN00B, Minneapolis, MN), Quantikine Human soluble vascular cell adhesion molecule (sVCAM)-1 (CD106; R&D Systems, Cat DVC00, Minneapolis, MN), Quantikine Human interferon (IFN)-γ (R&D Systems, Cat DIF50, Minneapolis, MN), Quantikine HS TNF-α/TNFSF1A (R&D Systems, Cat HSTA00D, Minneapolis, MN), and BD OptEIA Human C5a enzyme-linked immunosorbent assay (ELISA) Kit II (BD Biosciences, Cat 557965).
We used paired Kruskal-Wallis tests to compare the before and after column blood concentrations of each cytokine, with a two-tailed p-value below 0.05 indicating significance. The results are summarized in Table 2.
Precolumn and postcolumn concentrations of analyzed cytokines are given in Table 2. Briefly, passage through the heparinized beads resulted in a significantly bigger decrease in blood of VCAM and TNF when compared with nonheparinized beads.
Impact of Bead Volume
Data obtained with a 1:1 and 1:10 blood-to-bead volume did not vary significantly (Table 2).
Impact of Flow Rate
Varying the speed of the blood flow between 1 and 10 ml/min did not significantly influence the postcolumn concentrations (data not shown).
In this study, we demonstrate the binding capability of an end-point heparinized surface to multiple cytokines in the complex protein mixture, i.e., whole blood. The PE beads that were used had a mean diameter of 0.3 mm and were heparinized with a technology that guaranteed that the heparin molecules were covalently end-point attached to the surface, thereby making the carbohydrate chains more accessible for proteins with affinity for heparin/heparan sulfate. The mean molecular weight of the immobilized heparin was about 8 kDa, while 2 mg (equal to approximately 360 IU) was coupled to each gram of beads. The integrity of this surface was verified by the expected removal of 75% of antithrombin concentrations from the blood passed over heparinized, but not nonheparinized, beads. These data corresponds well with the previous observations from extracorporeal lung assistance (ECLA) on septic patients using surface heparinized oxygenators, published by Bindslev et al.20 Furthermore, varying the blood flow rate from 1 to 10 ml/min did not significantly affect the amount removed of the respective cytokines, indicating that the observed binding to the immobilized heparin molecules is a very rapid event.
Our study used immobilized heparin to semiselectively bind cytokines, and these data should be compared with other approaches to extracorporeal cytokine removal. Weber et al. 5 reported significant in vitro binding of TNF using cellulose microparticles coated with a monoclonal anti-TNF antibody, whereas Haase et al. 12 reported a significant reduction in IL-1ra, but not in IL-6, using a similar ex vivo methodology as ours but with a porous adsorption device. In vivo, Mariano et al. 13 were able to significantly reduce several circulating cytokines with hemoperfusion and a high cutoff polysulphone membrane, but also reported a loss of serum albumin. The putative clinical relevance of these findings were demonstrated by Schefold et al.,14 who in a randomized study of 33 patients with septic shock were able to simultaneously reduce circulating endotoxin, IL-6, and C5a levels by selective immunoadsorption, resulting in improved organ function.
The present study shows a strong affinity for TNF for heparin, which is not surprising given recent studies showing that the N-terminus of the trimeric TNF molecule comprises the short amino-acid sequence VRSSSR, which is known to be essential for TNF binding to heparin.15 There is also evidence that the TNF-binding protein (TNF-BP)-I is responsible for the dissociation of TNF from heparin.16 Furthermore, the presence of heparin in plasma may significantly decreases the level of TNF, measured by ELISA and bioassays.17
However, it is probable that many, if not most, pro-inflammatory cytokines show affinity for heparan sulfate proteoglycans and thus heparin.4,18 In all heparan sulfates and in heparin, glycosaminoglycans are covalently linked to a protein backbone via their reducing, terminal monosaccharide units (4). Proteins thus bind to the negatively charged heparan sulfates and heparin by electrostatic forces.4 However, to induce biological activity changes in these proteins, binding also has to be conducted through semi-specific monosaccharide sequence binding in the glycosaminoglycans inducing conformational changes of target proteins.4,18 This fact was utilized by Kenig et al. in a recent study detailing a method for the purification of heparin-binding TNF.15 A similar, if non-specific, approach is used in the current study, where the large number of heparin molecules guarantees a large variety of semi-specific binding sites, thus accommodating multiple cytokines.
Indeed, heparan sulfates and heparin glycosaminoglycans have a remarkable structural diversity owing to the multiple possibilities for positioning sulfate groups, as well as their variable proportions of L-iduronic and D-glucuronic acids.4,18 This structural diversity, which is concentrated in areas of high sulfatation along the glycosaminoglycan chain, makes it possible to bind with high affinity to a large number of functionally diverse proteins.4,18 When immobilizing heparan sulfates or heparin onto a solid surface, it is thus important to utilize a technology where the coupling procedure does not induce any structural alterations in the glycosaminoglycan sequence, sulfatations or their positioning. In this study, polyethylene beads were surface heparinised with the heparin molecules covalently end-point attached to the surface, thereby making the carbohydrate chains accessible for proteins with affinity for heparin/heparan sulfate.19
Also of interest, blood concentrations of RANTES rose significantly in blood exposed to control beads, but stayed constant in blood exposed to heparinized beads. This finding suggests an improved biocompatibility of the heparinized surface in comparison with the non-heparinized surface and is in consistence with previous reports.3,20 It is also consistent with a recent report21 revealing a significant conformational change in RANTES upon heparan sulfate binding, suggesting that blocking of oligomerisation and activation of RANTES may contribute to the observed effects. This should be taken into consideration when septic patients are subjected to extracorporeal treatment, especially long-term and continuous therapies such as CPB and hemodialysis.
In summary, we demonstrate that by passing blood from septic patients through a column packed with surface heparinized beads, several elevated proinflammatory cytokines may be semiselectively removed. This innovative technology may thus offer a hope for reducing the very high mortality rate seen in hyper-inflammatory conditions such as sepsis and also improve the health of dialysis patients.
The authors thank George Pitarra at Emergence, Inc., for financial support and valuable discussions.
1. Tang L, Eaton JW: Inflammatory responses to biomaterials. Am J Clin Pathol
103: 466–471, 1995.
2. Jessen ME: Pro: Heparin-coated circuits should be used for cardiopulmonary bypass. Anesth Analg
103: 1365–1369, 2006.
3. Trocme SD, Li H: Effect of heparin-surface-modified intraocular lenses on postoperative inflammation after phacoemulsification: A randomized trial in a United States patient population. Heparin-Surface-Modified Lens Study Group. Ophthalmology
107: 1031–1037, 2000.
4. Bishop JR, Schuksz M, Esko JD: Heparan sulphate proteoglycans fine-tune mammalian physiology. Nature
446: 1030–1037, 2007.
5. Weber V, Linsberger I, Ettenauer M, et al
: Development of specific adsorbents for human tumor necrosis factor-α: Influence of antibody immobilization on performance and biocompatibility. Biomacromolecules
6: 1864–1870, 2005.
6. Baram D, Rashkovsky M, Hershkoviz R, et al
: Inhibitory effects of low molecular weight heparin on mediator release by mast cells: Preferential inhibition of cytokine production and mast cell-dependent cutaneous inflammation. Clin Exp Immunol
110: 485–491, 1997.
7. Pevni D, Frolkis I, Shapira I, et al
: Heparin added to cardioplegic solution inhibits tumor necrosis factor-α production and attenuates myocardial ischemic-reperfusion injury. Chest
128: 1805–1811, 2005.
8. Ellyard JI, Simson L, Bezos A, et al
: Eotaxin selectively binds heparin. An interaction that protects eotaxin from proteolysis and potentiates chemotactic activity in vivo. J Biol Chem
282: 15238–15247, 2007.
9. Culley FJ, Fadlon EJ, Kirchem A, et al
: Proteoglycans are potent modulators of the biological responses of eosinophils to chemokines. Eur J Immunol
33: 1302–1310, 2003.
10. Larm O, Larsson R, Olsson: A new non-thrombogenic surface prepared by selective covalent binding of heparin via a modified reducing terminal residue. Biomater Med Devices Artif Organs
11: 161–173, 1983.
11. Riesenfeld J, Roden L: Quantitative analysis of N
-acetylated, and unsubstituted glucosamine amino groups in heparin and related polysaccharides. Anal Biochem
188: 383–389, 1990.
12. Haase M, Bellomo R, Baldwin I, et al
: The effect of three different miniaturized blood purification devices on plasma cytokine concentration in an ex vivo model of endotoxinemia. Int J Artif Organs
31: 722–729, 2008.
13. Mariano F, Fonsato V, Lanfranco G, et al
: Tailoring high-cut-off membranes and feasible application in sepsis-associated acute renal failure: In vitro studies. Nephrol Dial Transplant
20: 1116–1126, 2005.
14. Schefold JC, von Haehling S, Corsepius M, et al
: A novel selective extracorporeal intervention in sepsis: Immunoadsorption of endotoxin, interleukin 6, and complement-activating product 5a. Shock
28: 418–425, 2007.
15. Kenig M, Gaberc-Porekar V, Fonda I, Menart V: Identification of the heparin-binding domain of TNF-alpha and its use for efficient TNF-alpha purification by heparin-sepharose affinity chromatography. J Chromatogr B Analyt Technol Biomed Life Sci
867: 119–125, 2008.
16. Lantz M, Thysell H, Nilsson E, Olsson I: On the binding of tumor necrosis factor (TNF) to heparin and the release in vivo of the TNF-binding protein I by heparin. J Clin Invest
88: 2026–2031, 1991.
17. Hoffmann JN, Hartl WH, Faist E, et al
: Tumor necrosis factor measurement and use of different anticoagulants: Possible interference in plasma samples and supernatants from endotoxin-stimulated monocytes. Inflamm Res
46: 342–347, 1997.
18. Hamel DJ, Sielaff I, Proudfoot AE, Handel TM: Chapter 4. Interactions of chemokines with glycosaminoglycans. Methods Enzymol
461: 71–102, 2009.
19. Larm O LR, Olsson P: Surface-immobilized heparin, in Lane A and Lindau U (eds). Heparin: Chemical and Biological Properties, Clinical Applications.
London, UK, Edward Arnold Press, 1989, pp. 597–608.
20. Bindslev L, Eklund J, Norlander O, et al
: Treatment of acute respiratory failure by extracorporeal carbon dioxide elimination performed with a surface heparinized artificial lung. Anesthesiology
67: 117–120, 1987.
21. Rek A, Brandner B, Geretti E, Kungl AJ: A biophysical insight into the RANTES-glycosaminoglycan interaction. Biochim Biophys Acta
1794: 577–582, 2009.