Clinical Hemorheology and Microcirculation 61 (2015) 657–665 DOI 10.3233/CH-152026 IOS Press
657
Adsorption capacity of poly(ether imide) microparticles to uremic toxins Sarada D. Tetalia,∗ , Vera Jankowskib , Karola Luetzowc,e , Karl Kratzc,e , Andreas Lendleinc,d,e and Joachim Jankowskib,f,∗ a
Department of Plant Sciences, School of Life Sciences, University of Hyderabad, Hyderabad, India RWTH Aachen University, University Hospital, Aachen, Germany c Institute of Biomaterial Science and Berlin-Brandenburg Center for Regenerative Therapies, Helmholtz-Zentrum Geesthacht, Teltow, Germany d Institute of Chemistry, University Potsdam, Potsdam, Germany e Helmholtz Virtual Institute – Multifunctional Biomaterials for Medicine, Teltow and Berlin, Germany f Department of Pathology, Cardiovascular Research Institute Maastricht (CARIM), University of Maastricht, Maastricht, The Netherlands b
Abstract. Uremia is a phenomenon caused by retention of uremic toxins in the plasma due to functional impairment of kidneys in the elimination of urinary waste products. Uremia is presently treated by dialysis techniques like hemofiltration, dialysis or hemodiafiltration. However, these techniques in use are more favorable towards removing hydrophilic than hydrophobic uremic toxins. Hydrophobic uremic toxins, such as hydroxy hipuric acid (OH-HPA), phenylacetic acid (PAA), indoxyl sulfate (IDS) and p-cresylsulfate (pCRS), contribute substantially to the progression of chronic kidney disease (CKD) and cardiovascular disease. Therefore, objective of the present study is to test adsorption capacity of highly porous microparticles prepared from poly(ether imide) (PEI) as an alternative technique for the removal of uremic toxins. Two types of nanoporous, spherically shaped microparticles were prepared from PEI by a spraying/coagulation process. PEI particles were packed into a preparative HPLC column to which a mixture of the four types of uremic toxins was injected and eluted with ethanol. Eluted toxins were quantified by analytical HPLC. PEI particles were able to adsorb all four toxins, with the highest affinity for PAA and pCR. IDS and OH-HPA showed a partially non-reversible binding. In summary, PEI particles are interesting candidates to be explored for future application in CKD. Keywords: Adsorption of uremic toxins, chronic kidney disease (CKD), hydrophobic uremic toxins, poly(ether imide), microparticles, uremia
1. Introduction The alchemists already knew: “Similia similibus solvuntur” (“like dissolves like”). However, we still try to remove hydrophobic uremic toxins by hydrophilic solutions – like dialysate. Therefore, chronic kidney disease (CKD) is characterized by the progressive retention of a multitude of uremic toxins, since recent concepts of conventional dialysis techniques like dialysis, hemofiltration and hemodiafiltration ∗
Corresponding author: Dr. Sarada D. Tetali, Associate Professor, Department of Plant Sciences, School of Life Sciences, University of Hyderabad, Hyderabad 500 046, India. Tel.: +1 91 40 23134512; Fax: +1 91 40 23010120; E-mails: [email protected], [email protected] and Univ.-Prof. Dr. Joachim Jankowski, Director, Institute for Cardiovascular Research, RWTH Aachen University, University Hospital, Pauwelsstraße 30, D-52074 Aachen, Germany. Tel.: +49 241 80 80580; Fax: +49 241 80 82716; E-mail: [email protected]. 1386-0291/15/$35.00 © 2015 – IOS Press and the authors. All rights reserved
658
S.D. Tetali et al. / Adsorption of uremic toxins by polyetherimide
focus mainly on the removal of small water-soluble uremic toxins by techniques based on diffusion- and convection processes in aqueous solvents [14]. Based on their hydrophobicity, these uremic toxins bind to plasma proteins with a molecular weight higher than the cut-off of conventional dialysis membranes. For example, 30% of phenylacetic acid (PAA) [10], >90% of indoxyl sulfate (IDS) [18] and about 90% of p-cresyl sulfate (pCRS) [12] are protein-bound. Therefore, these conventional dialysis therapies do not sufficiently remove hydrophobic uremic toxins from the plasma in patients suffering from chronic kidney disease (CKD). In consequence, conventional dialysis therapies are time-consuming and still associated with side effects and complications, despite the high quality of treatment. Cardiovascular diseases are the most common complications in CKD patients treated by dialysis, resulting in an increased mortality of the patients. These uremic toxins are characterized by their strong impact on genesis and progression of cardiovascular diseases, a major cause of mortality and morbidity in CKD patients [7]. For example, PAA [10], IDS [15, 19] and pCRS contribute significantly to the development of atherosclerotic vascular lesions, affecting erythrocyte, endothelial cell, leukocyte, platelet and/or vascular smooth muscle cell function in CKD patients [6, 16]. Since only the solved, non-protein bound part of these uremic toxins is readily removable by conventional hemodialysis techniques, treatment of uremia remains strongly limited by this approach. Hence, there is a strong need for new therapeutic strategies to remove hydrophobic these uremic toxins. A reasonable approach might be therapeutic strategies based on adsorption of hydrophobic uremic toxins on hydrophobic adsorbers. The objective of the present study was the quantification of the adsorption capacity of poly(ether imide) (PEI) particles for hydrophobic uremic toxins. The primary task of the kidneys is the excretion of urinary substances. The kidneys of patients suffering from chronic renal failure cannot fulfil this function. The disease can be alleviated by means of a dialysis, and dialysis is able to bridge the time to find a suitable donor organ. Dialysis is based on the principle of diffusion and filtration. The dialysis membranes currently act as pure filter membranes. The reason for the high mortality is mainly caused by low molecular weight, hydrophobic, aromatic uremic toxins. A sufficient removal of these uremic toxins is not possible by the available filtration membranes since diffusion and convection are the driving forces of this process. This limitation is caused by the molecular size of these substances. Since these toxins are present in the plasma protein-bound, they have a molecular size which prevents passage through the pores of the membrane of the dialyser. With increasing molecular size substances are removed significantly worse. Conventional dialysis membranes have a cut-off of approximately 14–17 kDa. It could be demonstrated that fractionated plasma separation, adsorption, and dialysis utilizing hydrophobic and cationic adsorbers was more effective for removal of protein-bound, hydrophobic uremic toxins than conventional high-flux hemodialysis [5]. Recently, electrospun nanofibrous polymer composite membranes prepared from on poly(ethylene-co-vinyl alcohol) and zeolites have been reported to adsorb uremic solutes like creatinine in a blood purification system [13]. A good in vitro absorption capacity for protein bound middle and high molecular weight uremic toxins could be achieved with a nanoporous activated carbon monolith [17]. Another strategy for the removal of low molecular weight uremic toxins from the blood of patients might be the utilization of porous polymeric microparticles, prepared from hydrophobic polymers such as poly(ether imide) (PEI) in a spraying/coagulation process [1]. The hydrophobic characteristics of PEI microparticles comprising aromatic moieties combined with a high surface area and a good permeability should ensure a good adsorption efficiency for uremic toxins. The pore sizes of such PEI microparticles can be tailored by adaptation of the processing conditions of e.g. by variation of the solvent [1]. These particles are mechanical robust, steam-sterilisable and can be
S.D. Tetali et al. / Adsorption of uremic toxins by polyetherimide
659
easily surface modified by reaction of nucleophiles with the imide bond of PEI either in an integrated or two-step procedure [3]. In this study we investigated the in vitro adsorption capacity of highly porous poly(ether imide) microparticles for hydroxy hipuric acid, phenylacetic acid, indoxyl sulfate and p-cresylsulfate. Two types of PEI microparticles were prepared via a spraying/precipitation process from either a 1-methyl2-pyrrolidone/water or a dimethyl sulfoxide/1-methyl-2-pyrrolidone solvent mixture, which should lead to microparticles with different pore sizes. 2. Materials and methods 2.1. Particle preparation Particles were prepared from PEI (Ultem® 1000, General Electric, USA; Mw = 30000 ± 10000 g·mol–1 , Mn = 12000 ± 4000 g·mol–1 ) by a spraying/coagulation process described in [1, 17]. Briefly, PEI was dissolved in a solvent mixture of either 1-methyl-2-pyrrolidone (NMP; Merck, Darmstadt, Germany; purity >99%) and water mixture (96:4 wt.%) leading to particle PEI-NW or in a mixture of dimethyl sulfoxide (DMSO; Merck, Darmstadt, Germany; purity = 99.9%) and NMP (65:35 wt.%) leading to particle PEI-DN. Each polymer solution was sprayed into a coagulation bath of water containing 0.1 wt.% sodium dodecyl sulfate (Merck, Darmstadt, Germany; purity >99%) at room temperature. Afterwards the particles were washed thoroughly with water and then heated in water for 1 h at 100◦ C. After sieving of the particles (Analysette 3 Pro, Fritsch, Idar-Oberstein, Germany, with sieves having a mesh sizes of 50 m and 180 m) the unmodified PEI microparticles were obtained. In a second step both types of particle were treated with an aqueous solution of branched polyethylene imine (Pei, purity >99%, SigmaAldrich, Taufkirchen, Germany; Mw = 800 g·mol–1 ; Mn ∼600 g·mol–1 according to provider) with a polymer concentration of 4 wt% for 5 min at 90◦ C and finally washed thoroughly with water according to the method described in [3]. The particles were stored in aqueous sodium azide (Sigma-Aldrich, Taufkirchen, Germany; purity >99%) solution of 0.02 wt% in the fridge. 2.2. Particle characterization The average particle diameter was estimated from manual measurement of n = 200 wet particles with a stereomicroscope (Stemi 2000, Zeiss, Oberkochem, Germany). The cross sections of the particles were investigated by scanning electron microscopy (SEM; Zeiss Supra 40 VP, Zeiss, Oberkochem, Germany). Particle cross-sections were prepared by cryoultramicrotomy SteREO Discovery.V12 (Leica Mikrosysteme Vertrieb GmbH, Wetzlar, Germany). The particles were suspended in 2-propanol and frozen at –1 160◦ C. The frozen droplet was cut with a speed of 80 mm·s–1 into slices of 500 nm. Sectioned surfaces were sputtered (EAP 101, Gatan Inc., Pleasanton, CA, USA) with a thin gold/palladium (80:20 wt.%) layer of 5 nm and investigated by SEM with an acceleration voltage of 3 kV. The particle porosity (n = 2) was determined by gravimetric analysis from the difference in weight between the moist particles and the dried particles divided by the weight of the moist particles (method described in [1]). The influence of the treatment with Pei was checked by Fourier transform attenuated total internal reflection infrared spectroscopy (FT-ATR-IR; MagnaIR 550, Nicolet, Thermo Fischer Scientific, Waltham, MA, USA).
660
S.D. Tetali et al. / Adsorption of uremic toxins by polyetherimide
2.3. Assay of particle binding capacity to uremic toxin by preparative HPLC About 1.5 g of non-sterile absorber either PEI-NW or PEI-DN were packed into a preparative HPLC (Merck) column and equilibrated with water. Mixture of uremic toxins, namely hydroxy hipuric acid (OH-HPA), phenylacetic acid (PAA), indoxyl sulfate (IDS) and p-cresylsulfate (pCR), each of 0.1 mg per 1 g of PEI microparticles were injected into the column, which is calibrated with water and subsequently eluted with 100% ethanol. Gradient time period was 20 min with a flow rate of 1 ml·min–1 in a linear gradient 0 to 100% B with eluent ‘A’ as water/0.01% TFA and ‘B’ as 100% ethanol by monitoring UV absorbance of uremic toxins in the UV region, at λ = 280 nm. Fractions of 1 ml were collected, i.e. for every one min and subjected to analytical HPLC for quantification of uremic toxins. Dionex - Chromeleon chromatography data system (v. 6.6) was used for data acquisition. 2.4. Quantitation of uremic toxins by analytical HPLC Unbound as well as bound fractions collected from the above mentioned preparative HPLC were subjected to analytical reverse-phase HPLC (Merck), Chromolith column, to quantify the amount of the uremic toxins using the extinction coefficient derived from the standard plots generated using synthetic OH-HPA, PAA, IDS, pCR by subjecting the samples to analytical HPLC. Solvent ‘A’ was tetra-n-butyl ammonium hydrogen sulphate (TBA) in buffer K2 HPO4 (pH 6.5); solvent ‘B’ was 100 % ethanol (pH in the range of 7.5 and 2.0). Sample volume of 20 l mixed with 130 l, mixed with solvent A and 100 l of it was injected on to the column for each assay. Gradient time period was 26 min in a linear gradient of 0 to 60% B with a flow rate of 1 ml·min–1 by recording absorbance of uremic toxins in the UV region, at λ = 220 nm. Unbound and bound amounts of each uremic toxin to the absorber were estimated from the O.D. values of fractions obtained from the above mentioned preparative HPLC. Dionex - Chromeleon chromatography data system (v. 6.6) was used for data acquisition. The amount of uremic toxin that irreversibly bound to the PEI absorber was estimated by calculating the difference between the total injected amount of uremic toxin, i.e. 0.1 mg of each uremic toxin on to the preparative column and sum of bound and unbound amounts. 2.5. Statistics All data were expressed as mean value of minimum of three independent experiments ± standard deviation. 3. Results 3.1. Processing and Characterization of PEI particles PEI particles have been prepared by a spraying/coagulation process [1]. The polymer PEI was dissolved either in solvent mixture of NMP and water (96:4 wt%), finally leading to particles PEINW, or in a mixture of DMSO and NMP (65:35 wt.%), leading to particle PEI-DN. Two different solvent mixtures have been used in order to achieve differences in pore size as discussed in an earlier publication [1]. Both PEI-NW and PEI-DN particles show a similar finger-pore like morphology (Fig. 1a and b) and are open-porous as depicted in Fig. 1c, while different mean pore diameters of 136 nm (PEI-NW) and
S.D. Tetali et al. / Adsorption of uremic toxins by polyetherimide
661
Fig. 1. Scanning electron microscopy images visualizing the microparticles’ cross section of PEI-NW (a), PEI-DN (b) and the outer part of the PEI-DN cross-section at higher magnification (c).
41 nm (PEI-DN) have been reported earlier [1]. Mean particle diameters of 78 ± 51 for PEI-NW and 132 ± 39 m for PEI-DN were determined by light microscopy (see Experimental section). In order to enhance the wettability of the particles, the particles were reacted with low-molecular weight polyethylene imine (Pei) for a short time of 5 min at 90◦ C according to the method described in [3]. A low-molecular weight Pei was chosen to ensure that also the inner pores are contacted and modified. The functionalization on the surface of the particles was verified by ATR-FT-IR spectroscopy. An increase of the carbonyl stretching vibrations at 1660 cm–1 (amide I band) and 1550 cm–1 (amide II) due to the formation of an amide bond, which is the result of the imide moiety reaction of the PEI with an amine group of Pei, was found. This is in agreement with the results reported earlier [2, 3, 11]; however the short-time Pei functionalization was only carried out to enhance the wettability of the particles and is not considered to be a complete surface coverage. The total porosity of the particles was calculated from the difference in weight of the moist particles and dried particles and was determined to be 78 ± 1% for PEI-NW and 79 ± 1 % for PEI-DN, and thus is similar to the ones reported earlier [1]. 3.2. Binding capacity of absorber PEI particles to uremic toxins Preparative HPLC experiment using columns filled with PEI-NW or PEI-DN absorber with sample loaded uremic toxins separated unbound uremic toxins (pool of all four toxins used in the sample) from bound fraction. Based on the absorbance at λ = 280 nm, unbound toxins were eluted within first 3 min of the gradient and bound fractions were eluted at higher concentration of ethanol i.e. after 10 and within 16 min of the linear gradient (Figs. 2a, 2b) as indicated by absorbance of the elution profile. Therefore, fractions of 1 to 3 min were pooled (unbound fraction) and fractions collected between 10 and 16 min (bound fraction) were pooled. 3.3. Affinity of PEI particles to uremic toxins Elution profile of four toxins from RP-HPLC is shown in Figs. 3a and 3b. Using standard compounds, it is noted that OH-HPA elutes within 10 min, followed by PAA, IDS and pCR (Figs. 3a, 3b). Unbound and bound fractions collected from the above mentioned preparative HPLC were subsequently subjected to analytical RP-HPLC to quantify affinity of each of the four uremic toxins to PEI absorber particles. Results revealed that both PEI-NW and PEI-DN particles had high affinity to all four toxins (Figs. 4a, 4b).
662
S.D. Tetali et al. / Adsorption of uremic toxins by polyetherimide
Fig. 2. Preparative HPLC profile of uremic toxin mixtures passed through PEI-NW (a) and PEI-DN (b). First peak within 5 minutes indicates unbound fraction of uremic toxins to the PEI column material where as a second peak in between 10 to 15 min indicates bound fraction of uremic toxin eluted. Absorption of the toxins was recorded in UV region, at λ = 280 nm.
Fig. 3. Separation and detection of four uremic toxins, OH-HPA, PAA, IDS and pCR by analytical HPLC with PEI-NW (a) and PEI-DN (b). Absorption of the toxins was recorded in UV region at λ = 220 nm.
Among the absorbers tested, PEI-NW showed high affinity to pCR, IDS followed by PAA and OH-HPA (Figs. 4a). 85 wt.% of the pCR loaded to the PEI particles was reversibly bound, 7.5 wt.% irreversibly bound and remaining 7.5 wt.% eluted as unbound fraction. Similarly, 60 wt.% of PAA loaded was reversibly bound, 25 wt.% as irreversibly bound and 15 wt.% of it was unbound indicating 85 wt.% as total bound fraction. In case of IDS, both unbound (5 wt.%) and reversibly bound (2 wt.%) amounting to about only 7 wt.%, suggesting that about 93 wt.% was irreversibly bound. For OH-HPA 30% was unbound, 15 wt.% reversibly bound giving rise to 70 wt.% of bound and 30 wt.% on unbound. PEI-DN particles (Fig. 4b) showed higher affinity to PAA and pCR followed by IDS and OH-HPA. 15 wt.% of the loaded PAA came into unbound fraction and 40 wt.% of it was eluted as bound fraction implicating rest of the 45 wt.% as irreversibly bound fraction. This data suggested that 85 wt.% of loaded PAA was bound to the absorber particles. Similarly, only 5 wt.% of the pCR loaded came into the unbound fraction, out 95 wt.% of bound pCR, 60 wt.% was eluted into bound fraction and rest of 35 wt.% showed irreversible binding to the column. Whereas both IDS and OH-HPA showed 70 wt.% binding. 30 wt.% of loaded IDS came into unbound fraction, 10 wt.% of it was eluted into bound fraction and rest
S.D. Tetali et al. / Adsorption of uremic toxins by polyetherimide
663
Fig. 4. Analysis of binding capacity of the four uremic toxins to PEI-NW (a) and PEI-DN (b) was determined independently by HPLC by using absorption coefficients generated using pure synthetic toxins. The adsorbed uremic toxin quantity was differentiated by unbound (UB) and bound (Elu): PAA UB (1), PAA Elu (2), IDS UB (3), IDS Elu (4), pCR UB (5), pCR Elu (6), OH-HPA UB (7), OH HPA Elu (8).
60 wt.% was irreversibly bound to the column. In case of OH-HPA, 30 wt.% of the loaded toxin came into unbound fraction and out of bound fraction, the distribution between reversibly and irreversibly bound were 20 wt.% and 50 wt.% respectively. 4. Discussion Presently, the assessment of dialysis quality is exclusively based on the elimination of small, watersoluble molecules. However, increasing dialysis adequacy in terms of Kt/Vurea index has not improved mortality or morbidity in dialysis patients [9]. Hemodialysis helps patients in compensating their impaired kidney function, however, limitation of existing strategies is removal hydrophobic uremic toxins, PAA, IDS, pCR and etc. These uremic toxins accumulate in the blood plasma over the period of time despite of regularly hemodialysis treatments three times per week at least. Accumulated hydrophobic uremic toxins show adverse effects on the cardiovascular system, thus CVD becomes one of the major causes of premature death of CKD patients. Most of these uremic toxins are characterized by aromatic structures, which are less water-soluble and therefore associated to plasma proteins. Due to their binding to proteins, these substances should be considered as high Mw substances [8], which are thus not eliminated by diffusion but by convection. However, since the filtration rate in dialysis is small, the removal of protein-bound, hydrophobic uremic toxins from plasma of CKD patients is limited. The current dialysis techniques still are far away from replacing the natural elimination of uremic toxins by the kidneys. Dialysis is based on the principles of diffusion and/or filtration; the dialysis membranes currently used solely act as filtering membranes. The major drawbacks of current dialysis are related to these principles of action: By filtration and diffusion with a cut-off in the range of 18 kDa preferably low molecular weight hydrophilic substances are eliminated, whereas protein-bound hydrophobic small molecules are largely retained in the blood plasma. In the present study the adsorption capacity of highly porous microparticles prepared from PEI was investigated. Both particle types PEI-NW and PEI-DN showed high affinity of adsorption to all four types of toxins tested. The results are promising in terms of uremic toxin adsorption capacity, and also indicate that the majority of the toxins are reversibly bound except for IDS by PEI-NW, which offers the hope of recycling of the particles, by repeated washings with alcohol and subsequent steam sterilization. As
664
S.D. Tetali et al. / Adsorption of uremic toxins by polyetherimide
identical surface chemistries can be anticipated for both PEI-NW and PEI-DN microparticles and a larger mean pore diameter around 136 nm was reported for PEI-NW compared to PEI-DN exhibited (41 nm) before [1], we have currently no reasonable explanation for the non-reversible interaction of IDS and PEI-NW. Therefore, new treatment options have to be developed for improving the outcome of hemodialysis patients. Recently, hydrophobic and cationic adsorbers were used in a fractionated plasma separation, adsorption, and dialysis (FPAD) system for removal of protein-bound, hydrophobic uremic toxins [5]. Removal rates of FPAD treatment in comparison to conventional high-flux hemodialysis were increased by 130% for phenylacetic acid, 187% for indoxyl sulfate, and 127% for p-cresol compared to conventional high-flux hemodialyzer. The strongest limitation of the FPAD system for clinical application was the necessary isolation of plasma before the uremic toxins were able to interact with the adsorber(s). This requirement necessitates new dialysis machines in the clinics, causing high cost for the investment as well as additional disposables for their operation. This requirement hampers the introduction of the technique in the clinical routine presently and most likely in the future. The separation of plasma from the blood cells is necessary to prevent any interaction of corpuscular blood compounds with the hydrophobic or anionic surfaces of the adsorber. These side effects cause the need for new adsorber materials, which have a high capacity of the adsorption of hydrophobic uremic toxins on the one hand and which can be modified to stop the interaction of corpuscular blood components. In the current study we were able to demonstrate the characteristics of PEI particles for adsorption of uremic toxins. These particles are characterized by their strong affinity to hydrophobic uremic toxins and the feasibility of hydrophilic surface modification to prevent interaction of corpuscular blood components with the adsorber surface [3]. The latter described for membranes [2, 4, 11]. The data of the current study demonstrate the high capacity of the PEI particles for the adsorption of hydrophob uremic toxins like PAA, IDS and pCRS.
5. Conclusion PEI particles showed high capacity of adsorption to hydrophobic uremic toxins and most of the investigated toxins were reversibly bound, which offers the possibility of recycling these particles provided the particles are biocompatible. After demonstrating the high adsorption capacity of the PEI microparticles, future investigations will focus on the hemocompatibility of the adsorbers.
Acknowledgments The authors acknowledge the experimental assistance by Mario Rettschlag and characterization support by Yvonne Pieper and Regine Apostel. DBT-CREST overseas fellowship (2010-11) awarded to Dr. Sarada D. Tetali, Indo-German Science and Technology Centre (Grant No. IGSTC/NPORE/SDT/2012) and German Federal Ministry for Education and Research (BMBF), (Grant No.s 01DQ13006A, 01DQ13006B and 01DQ13006C) are acknowledged for financial support.
S.D. Tetali et al. / Adsorption of uremic toxins by polyetherimide
665
References [1] W. Albrecht, K. Lutzow, T. Weigel, T. Groth, M. Schossig, and A. Lendlein, Development of highly porous microparticles from poly(ether imide) prepared by a spraying/coagulation process, Journal of Membrane Science 273(1-2) (2006), 106–115. [2] W. Albrecht, B. Seifert, T. Weigel, M. Schossig, A. Hollander, T. Groth and R. Hilke, Amination of poly(ether imide) membranes using di- and multivalent amines, Macromolecular Chemistry and Physics 204(3) (2003), 510–521. [3] S. Basu, M. Heuchel, T. Weigel, K. Kratz and A. Lendlein, Integrated process for preparing porous, surface functionalized polyetherimide microparticles, Polymers for Advanced Technologies accepted on 24.08.2015 (2015). [4] S. Braune, M. von Ruesten-Lange, C. Mrowietz, K. Lutzow, T. Roch, A.T. Neffe, A. Lendlein and F. Jung, Dynamic in vitro hemocompatibility testing of poly(ether imide) membranes functionalized with linear, methylated oligoglycerol and oligo(ethylene glycol), Clinical Hemorheology and Microcirculation 54(3) (2013), 235–248. [5] F. Brettschneider, M. Tolle, M. von der Giet, J. Passlick-Deetjen, S. Steppan, M. Peter, V. Jankowski, A. Krause, S. Kuhne, W. Zidek and J. Jankowski, Removal of Protein-Bound, Hydrophobic Uremic Toxins by a Combined Fractionated Plasma Separation and Adsorption Technique, Artificial Organs 37(4) (2013), 409–416. [6] P. Brunet, L. Dou, C. Cerini and Y. Berland, Protein-bound uremic retention solutes, Advances in Renal Replacement Therapy 10(4) (2003), 310–320. [7] L. Dou, M. Sallee, C. Cerini, S. Poitevin, B. Gondouin, N. Jourde-Chiche, K. Fallague, P. Brunet, R. Calaf, B. Dussol, B. Mallet, F. Dignat-George and S. Burtey, The Cardiovascular Effect of the Uremic Solute Indole-3 Acetic Acid, Journal of the American Society of Nephrology 26(4) (2015), 876–887. [8] F. Duranton, G. Cohen, R. De Smet, M. Rodriguez, J. Jankowski, R. Vanholder, A. Argiles and G. European Uremic Toxin Work, Normal and pathologic concentrations of uremic toxins, J Am Soc Nephrol 23(7) (2012), 1258–1270. [9] G. Eknoyan, G.J. Beck, A.K. Cheung, J.T. Daugirdas, T. Greene, J.W. Kusek, M. Allon, J. Bailey, J.A. Delmez, T.A. Depner, J.T. Dwyer, A.S. Levey, N.W. Levin, E. Milford, D.B. Ornt, M.V. Rocco, G. Schulman, S.J. Schwab, B.P. Teehan, R. Toto and G. Hemodialysis Study, Effect of dialysis dose and membrane flux in maintenance hemodialysis, N Engl J Med 347(25) (2002), 2010–2019. [10] J. Jankowski, M. van der Giet, V. Jankowski, S. Schmidt, M. Hemeier, B. Mahn, G. Giebing, M. Tolle, H. Luftmann, H. Schluter, W. Zidek and M. Tepel, Increased plasma phenylacetic acid in patients with end-stage renal failure inhibits iNOS expression, J Clin Invest 112(2) (2003), 256–264. [11] M. Lange, S. Braune, K. Luetzow, K. Richau, N. Scharnagl, M. Weinhart, A.T. Neffe, F. Jung, R. Haag and A. Lendlein, Surface functionalization of poly(ether imide) membranes with linear, methylated oligoglycerols for reducing thrombogenicity, Macromol Rapid Commun 33(17) (2012), 1487–1492. [12] A.W. Martinez, N.S. Recht, T.H. Hostetter and T.W. Meyer, Removal of P-cresol sulfate by hemodialysis, J Am Soc Nephrol 16(11) (2005), 3430–3436. [13] K. Namekawa, M.T. Schreiber, T. Aoyagi and M. Ebara, Fabrication of zeolite-polymer composite nanofibers for removal of uremic toxins from kidney failure patients, Biomaterials Science 2(5) (2014), 674–679. [14] N. Neirynck, R. Vanholder, E. Schepers, S. Eloot, A. Pletinck and G. Glorieux, An update on uremic toxins, Int Urol Nephrol 45(1) (2013), 139–150. [15] T. Niwa and M. Ise,Indoxyl sulfate, A circulating uremic toxin, stimulates the progression of glomerular sclerosis, J Lab Clin Med 124(1) (1994), 96–104. [16] D. Polak-Jonkisz, D. Zwolinska and L. Purzyc, The PMA action via metabolism of uremic erythrocytes, Clin Hemorheol Microcirc 59(3) (2015), 283–285. [17] S.R. Sandeman, C.A. Howell, G.J. Phillips, Y.S. Zheng, G. Standen, R. Pletzenauer, A. Davenport, K. Basnayake, O. Boyd, S. Holt and S.V. Mikhalovsky, An adsorbent monolith device to augment the removal of uraemic toxins during haemodialysis, Journal of Materials Science-Materials in Medicine 25(6) (2014), 1589–1597. [18] H. Watanabe, Y. Miyamoto, M. Otagiri and T. Maruyama, Update on the pharmacokinetics and redox properties of protein-bound uremic toxins, J Pharm Sci 100(9) (2011), 3682–3695. [19] H. Yamamoto, S. Tsuruoka, T. Ioka, H. Ando, C. Ito, T. Akimoto, A. Fujimura, Y. Asano and E. Kusano, Indoxyl sulfate stimulates proliferation of rat vascular smooth muscle cells, Kidney Int 69(10) (2006), 1780–1785.
657
Adsorption capacity of poly(ether imide) microparticles to uremic toxins Sarada D. Tetalia,∗ , Vera Jankowskib , Karola Luetzowc,e , Karl Kratzc,e , Andreas Lendleinc,d,e and Joachim Jankowskib,f,∗ a
Department of Plant Sciences, School of Life Sciences, University of Hyderabad, Hyderabad, India RWTH Aachen University, University Hospital, Aachen, Germany c Institute of Biomaterial Science and Berlin-Brandenburg Center for Regenerative Therapies, Helmholtz-Zentrum Geesthacht, Teltow, Germany d Institute of Chemistry, University Potsdam, Potsdam, Germany e Helmholtz Virtual Institute – Multifunctional Biomaterials for Medicine, Teltow and Berlin, Germany f Department of Pathology, Cardiovascular Research Institute Maastricht (CARIM), University of Maastricht, Maastricht, The Netherlands b
Abstract. Uremia is a phenomenon caused by retention of uremic toxins in the plasma due to functional impairment of kidneys in the elimination of urinary waste products. Uremia is presently treated by dialysis techniques like hemofiltration, dialysis or hemodiafiltration. However, these techniques in use are more favorable towards removing hydrophilic than hydrophobic uremic toxins. Hydrophobic uremic toxins, such as hydroxy hipuric acid (OH-HPA), phenylacetic acid (PAA), indoxyl sulfate (IDS) and p-cresylsulfate (pCRS), contribute substantially to the progression of chronic kidney disease (CKD) and cardiovascular disease. Therefore, objective of the present study is to test adsorption capacity of highly porous microparticles prepared from poly(ether imide) (PEI) as an alternative technique for the removal of uremic toxins. Two types of nanoporous, spherically shaped microparticles were prepared from PEI by a spraying/coagulation process. PEI particles were packed into a preparative HPLC column to which a mixture of the four types of uremic toxins was injected and eluted with ethanol. Eluted toxins were quantified by analytical HPLC. PEI particles were able to adsorb all four toxins, with the highest affinity for PAA and pCR. IDS and OH-HPA showed a partially non-reversible binding. In summary, PEI particles are interesting candidates to be explored for future application in CKD. Keywords: Adsorption of uremic toxins, chronic kidney disease (CKD), hydrophobic uremic toxins, poly(ether imide), microparticles, uremia
1. Introduction The alchemists already knew: “Similia similibus solvuntur” (“like dissolves like”). However, we still try to remove hydrophobic uremic toxins by hydrophilic solutions – like dialysate. Therefore, chronic kidney disease (CKD) is characterized by the progressive retention of a multitude of uremic toxins, since recent concepts of conventional dialysis techniques like dialysis, hemofiltration and hemodiafiltration ∗
Corresponding author: Dr. Sarada D. Tetali, Associate Professor, Department of Plant Sciences, School of Life Sciences, University of Hyderabad, Hyderabad 500 046, India. Tel.: +1 91 40 23134512; Fax: +1 91 40 23010120; E-mails: [email protected], [email protected] and Univ.-Prof. Dr. Joachim Jankowski, Director, Institute for Cardiovascular Research, RWTH Aachen University, University Hospital, Pauwelsstraße 30, D-52074 Aachen, Germany. Tel.: +49 241 80 80580; Fax: +49 241 80 82716; E-mail: [email protected]. 1386-0291/15/$35.00 © 2015 – IOS Press and the authors. All rights reserved
658
S.D. Tetali et al. / Adsorption of uremic toxins by polyetherimide
focus mainly on the removal of small water-soluble uremic toxins by techniques based on diffusion- and convection processes in aqueous solvents [14]. Based on their hydrophobicity, these uremic toxins bind to plasma proteins with a molecular weight higher than the cut-off of conventional dialysis membranes. For example, 30% of phenylacetic acid (PAA) [10], >90% of indoxyl sulfate (IDS) [18] and about 90% of p-cresyl sulfate (pCRS) [12] are protein-bound. Therefore, these conventional dialysis therapies do not sufficiently remove hydrophobic uremic toxins from the plasma in patients suffering from chronic kidney disease (CKD). In consequence, conventional dialysis therapies are time-consuming and still associated with side effects and complications, despite the high quality of treatment. Cardiovascular diseases are the most common complications in CKD patients treated by dialysis, resulting in an increased mortality of the patients. These uremic toxins are characterized by their strong impact on genesis and progression of cardiovascular diseases, a major cause of mortality and morbidity in CKD patients [7]. For example, PAA [10], IDS [15, 19] and pCRS contribute significantly to the development of atherosclerotic vascular lesions, affecting erythrocyte, endothelial cell, leukocyte, platelet and/or vascular smooth muscle cell function in CKD patients [6, 16]. Since only the solved, non-protein bound part of these uremic toxins is readily removable by conventional hemodialysis techniques, treatment of uremia remains strongly limited by this approach. Hence, there is a strong need for new therapeutic strategies to remove hydrophobic these uremic toxins. A reasonable approach might be therapeutic strategies based on adsorption of hydrophobic uremic toxins on hydrophobic adsorbers. The objective of the present study was the quantification of the adsorption capacity of poly(ether imide) (PEI) particles for hydrophobic uremic toxins. The primary task of the kidneys is the excretion of urinary substances. The kidneys of patients suffering from chronic renal failure cannot fulfil this function. The disease can be alleviated by means of a dialysis, and dialysis is able to bridge the time to find a suitable donor organ. Dialysis is based on the principle of diffusion and filtration. The dialysis membranes currently act as pure filter membranes. The reason for the high mortality is mainly caused by low molecular weight, hydrophobic, aromatic uremic toxins. A sufficient removal of these uremic toxins is not possible by the available filtration membranes since diffusion and convection are the driving forces of this process. This limitation is caused by the molecular size of these substances. Since these toxins are present in the plasma protein-bound, they have a molecular size which prevents passage through the pores of the membrane of the dialyser. With increasing molecular size substances are removed significantly worse. Conventional dialysis membranes have a cut-off of approximately 14–17 kDa. It could be demonstrated that fractionated plasma separation, adsorption, and dialysis utilizing hydrophobic and cationic adsorbers was more effective for removal of protein-bound, hydrophobic uremic toxins than conventional high-flux hemodialysis [5]. Recently, electrospun nanofibrous polymer composite membranes prepared from on poly(ethylene-co-vinyl alcohol) and zeolites have been reported to adsorb uremic solutes like creatinine in a blood purification system [13]. A good in vitro absorption capacity for protein bound middle and high molecular weight uremic toxins could be achieved with a nanoporous activated carbon monolith [17]. Another strategy for the removal of low molecular weight uremic toxins from the blood of patients might be the utilization of porous polymeric microparticles, prepared from hydrophobic polymers such as poly(ether imide) (PEI) in a spraying/coagulation process [1]. The hydrophobic characteristics of PEI microparticles comprising aromatic moieties combined with a high surface area and a good permeability should ensure a good adsorption efficiency for uremic toxins. The pore sizes of such PEI microparticles can be tailored by adaptation of the processing conditions of e.g. by variation of the solvent [1]. These particles are mechanical robust, steam-sterilisable and can be
S.D. Tetali et al. / Adsorption of uremic toxins by polyetherimide
659
easily surface modified by reaction of nucleophiles with the imide bond of PEI either in an integrated or two-step procedure [3]. In this study we investigated the in vitro adsorption capacity of highly porous poly(ether imide) microparticles for hydroxy hipuric acid, phenylacetic acid, indoxyl sulfate and p-cresylsulfate. Two types of PEI microparticles were prepared via a spraying/precipitation process from either a 1-methyl2-pyrrolidone/water or a dimethyl sulfoxide/1-methyl-2-pyrrolidone solvent mixture, which should lead to microparticles with different pore sizes. 2. Materials and methods 2.1. Particle preparation Particles were prepared from PEI (Ultem® 1000, General Electric, USA; Mw = 30000 ± 10000 g·mol–1 , Mn = 12000 ± 4000 g·mol–1 ) by a spraying/coagulation process described in [1, 17]. Briefly, PEI was dissolved in a solvent mixture of either 1-methyl-2-pyrrolidone (NMP; Merck, Darmstadt, Germany; purity >99%) and water mixture (96:4 wt.%) leading to particle PEI-NW or in a mixture of dimethyl sulfoxide (DMSO; Merck, Darmstadt, Germany; purity = 99.9%) and NMP (65:35 wt.%) leading to particle PEI-DN. Each polymer solution was sprayed into a coagulation bath of water containing 0.1 wt.% sodium dodecyl sulfate (Merck, Darmstadt, Germany; purity >99%) at room temperature. Afterwards the particles were washed thoroughly with water and then heated in water for 1 h at 100◦ C. After sieving of the particles (Analysette 3 Pro, Fritsch, Idar-Oberstein, Germany, with sieves having a mesh sizes of 50 m and 180 m) the unmodified PEI microparticles were obtained. In a second step both types of particle were treated with an aqueous solution of branched polyethylene imine (Pei, purity >99%, SigmaAldrich, Taufkirchen, Germany; Mw = 800 g·mol–1 ; Mn ∼600 g·mol–1 according to provider) with a polymer concentration of 4 wt% for 5 min at 90◦ C and finally washed thoroughly with water according to the method described in [3]. The particles were stored in aqueous sodium azide (Sigma-Aldrich, Taufkirchen, Germany; purity >99%) solution of 0.02 wt% in the fridge. 2.2. Particle characterization The average particle diameter was estimated from manual measurement of n = 200 wet particles with a stereomicroscope (Stemi 2000, Zeiss, Oberkochem, Germany). The cross sections of the particles were investigated by scanning electron microscopy (SEM; Zeiss Supra 40 VP, Zeiss, Oberkochem, Germany). Particle cross-sections were prepared by cryoultramicrotomy SteREO Discovery.V12 (Leica Mikrosysteme Vertrieb GmbH, Wetzlar, Germany). The particles were suspended in 2-propanol and frozen at –1 160◦ C. The frozen droplet was cut with a speed of 80 mm·s–1 into slices of 500 nm. Sectioned surfaces were sputtered (EAP 101, Gatan Inc., Pleasanton, CA, USA) with a thin gold/palladium (80:20 wt.%) layer of 5 nm and investigated by SEM with an acceleration voltage of 3 kV. The particle porosity (n = 2) was determined by gravimetric analysis from the difference in weight between the moist particles and the dried particles divided by the weight of the moist particles (method described in [1]). The influence of the treatment with Pei was checked by Fourier transform attenuated total internal reflection infrared spectroscopy (FT-ATR-IR; MagnaIR 550, Nicolet, Thermo Fischer Scientific, Waltham, MA, USA).
660
S.D. Tetali et al. / Adsorption of uremic toxins by polyetherimide
2.3. Assay of particle binding capacity to uremic toxin by preparative HPLC About 1.5 g of non-sterile absorber either PEI-NW or PEI-DN were packed into a preparative HPLC (Merck) column and equilibrated with water. Mixture of uremic toxins, namely hydroxy hipuric acid (OH-HPA), phenylacetic acid (PAA), indoxyl sulfate (IDS) and p-cresylsulfate (pCR), each of 0.1 mg per 1 g of PEI microparticles were injected into the column, which is calibrated with water and subsequently eluted with 100% ethanol. Gradient time period was 20 min with a flow rate of 1 ml·min–1 in a linear gradient 0 to 100% B with eluent ‘A’ as water/0.01% TFA and ‘B’ as 100% ethanol by monitoring UV absorbance of uremic toxins in the UV region, at λ = 280 nm. Fractions of 1 ml were collected, i.e. for every one min and subjected to analytical HPLC for quantification of uremic toxins. Dionex - Chromeleon chromatography data system (v. 6.6) was used for data acquisition. 2.4. Quantitation of uremic toxins by analytical HPLC Unbound as well as bound fractions collected from the above mentioned preparative HPLC were subjected to analytical reverse-phase HPLC (Merck), Chromolith column, to quantify the amount of the uremic toxins using the extinction coefficient derived from the standard plots generated using synthetic OH-HPA, PAA, IDS, pCR by subjecting the samples to analytical HPLC. Solvent ‘A’ was tetra-n-butyl ammonium hydrogen sulphate (TBA) in buffer K2 HPO4 (pH 6.5); solvent ‘B’ was 100 % ethanol (pH in the range of 7.5 and 2.0). Sample volume of 20 l mixed with 130 l, mixed with solvent A and 100 l of it was injected on to the column for each assay. Gradient time period was 26 min in a linear gradient of 0 to 60% B with a flow rate of 1 ml·min–1 by recording absorbance of uremic toxins in the UV region, at λ = 220 nm. Unbound and bound amounts of each uremic toxin to the absorber were estimated from the O.D. values of fractions obtained from the above mentioned preparative HPLC. Dionex - Chromeleon chromatography data system (v. 6.6) was used for data acquisition. The amount of uremic toxin that irreversibly bound to the PEI absorber was estimated by calculating the difference between the total injected amount of uremic toxin, i.e. 0.1 mg of each uremic toxin on to the preparative column and sum of bound and unbound amounts. 2.5. Statistics All data were expressed as mean value of minimum of three independent experiments ± standard deviation. 3. Results 3.1. Processing and Characterization of PEI particles PEI particles have been prepared by a spraying/coagulation process [1]. The polymer PEI was dissolved either in solvent mixture of NMP and water (96:4 wt%), finally leading to particles PEINW, or in a mixture of DMSO and NMP (65:35 wt.%), leading to particle PEI-DN. Two different solvent mixtures have been used in order to achieve differences in pore size as discussed in an earlier publication [1]. Both PEI-NW and PEI-DN particles show a similar finger-pore like morphology (Fig. 1a and b) and are open-porous as depicted in Fig. 1c, while different mean pore diameters of 136 nm (PEI-NW) and
S.D. Tetali et al. / Adsorption of uremic toxins by polyetherimide
661
Fig. 1. Scanning electron microscopy images visualizing the microparticles’ cross section of PEI-NW (a), PEI-DN (b) and the outer part of the PEI-DN cross-section at higher magnification (c).
41 nm (PEI-DN) have been reported earlier [1]. Mean particle diameters of 78 ± 51 for PEI-NW and 132 ± 39 m for PEI-DN were determined by light microscopy (see Experimental section). In order to enhance the wettability of the particles, the particles were reacted with low-molecular weight polyethylene imine (Pei) for a short time of 5 min at 90◦ C according to the method described in [3]. A low-molecular weight Pei was chosen to ensure that also the inner pores are contacted and modified. The functionalization on the surface of the particles was verified by ATR-FT-IR spectroscopy. An increase of the carbonyl stretching vibrations at 1660 cm–1 (amide I band) and 1550 cm–1 (amide II) due to the formation of an amide bond, which is the result of the imide moiety reaction of the PEI with an amine group of Pei, was found. This is in agreement with the results reported earlier [2, 3, 11]; however the short-time Pei functionalization was only carried out to enhance the wettability of the particles and is not considered to be a complete surface coverage. The total porosity of the particles was calculated from the difference in weight of the moist particles and dried particles and was determined to be 78 ± 1% for PEI-NW and 79 ± 1 % for PEI-DN, and thus is similar to the ones reported earlier [1]. 3.2. Binding capacity of absorber PEI particles to uremic toxins Preparative HPLC experiment using columns filled with PEI-NW or PEI-DN absorber with sample loaded uremic toxins separated unbound uremic toxins (pool of all four toxins used in the sample) from bound fraction. Based on the absorbance at λ = 280 nm, unbound toxins were eluted within first 3 min of the gradient and bound fractions were eluted at higher concentration of ethanol i.e. after 10 and within 16 min of the linear gradient (Figs. 2a, 2b) as indicated by absorbance of the elution profile. Therefore, fractions of 1 to 3 min were pooled (unbound fraction) and fractions collected between 10 and 16 min (bound fraction) were pooled. 3.3. Affinity of PEI particles to uremic toxins Elution profile of four toxins from RP-HPLC is shown in Figs. 3a and 3b. Using standard compounds, it is noted that OH-HPA elutes within 10 min, followed by PAA, IDS and pCR (Figs. 3a, 3b). Unbound and bound fractions collected from the above mentioned preparative HPLC were subsequently subjected to analytical RP-HPLC to quantify affinity of each of the four uremic toxins to PEI absorber particles. Results revealed that both PEI-NW and PEI-DN particles had high affinity to all four toxins (Figs. 4a, 4b).
662
S.D. Tetali et al. / Adsorption of uremic toxins by polyetherimide
Fig. 2. Preparative HPLC profile of uremic toxin mixtures passed through PEI-NW (a) and PEI-DN (b). First peak within 5 minutes indicates unbound fraction of uremic toxins to the PEI column material where as a second peak in between 10 to 15 min indicates bound fraction of uremic toxin eluted. Absorption of the toxins was recorded in UV region, at λ = 280 nm.
Fig. 3. Separation and detection of four uremic toxins, OH-HPA, PAA, IDS and pCR by analytical HPLC with PEI-NW (a) and PEI-DN (b). Absorption of the toxins was recorded in UV region at λ = 220 nm.
Among the absorbers tested, PEI-NW showed high affinity to pCR, IDS followed by PAA and OH-HPA (Figs. 4a). 85 wt.% of the pCR loaded to the PEI particles was reversibly bound, 7.5 wt.% irreversibly bound and remaining 7.5 wt.% eluted as unbound fraction. Similarly, 60 wt.% of PAA loaded was reversibly bound, 25 wt.% as irreversibly bound and 15 wt.% of it was unbound indicating 85 wt.% as total bound fraction. In case of IDS, both unbound (5 wt.%) and reversibly bound (2 wt.%) amounting to about only 7 wt.%, suggesting that about 93 wt.% was irreversibly bound. For OH-HPA 30% was unbound, 15 wt.% reversibly bound giving rise to 70 wt.% of bound and 30 wt.% on unbound. PEI-DN particles (Fig. 4b) showed higher affinity to PAA and pCR followed by IDS and OH-HPA. 15 wt.% of the loaded PAA came into unbound fraction and 40 wt.% of it was eluted as bound fraction implicating rest of the 45 wt.% as irreversibly bound fraction. This data suggested that 85 wt.% of loaded PAA was bound to the absorber particles. Similarly, only 5 wt.% of the pCR loaded came into the unbound fraction, out 95 wt.% of bound pCR, 60 wt.% was eluted into bound fraction and rest of 35 wt.% showed irreversible binding to the column. Whereas both IDS and OH-HPA showed 70 wt.% binding. 30 wt.% of loaded IDS came into unbound fraction, 10 wt.% of it was eluted into bound fraction and rest
S.D. Tetali et al. / Adsorption of uremic toxins by polyetherimide
663
Fig. 4. Analysis of binding capacity of the four uremic toxins to PEI-NW (a) and PEI-DN (b) was determined independently by HPLC by using absorption coefficients generated using pure synthetic toxins. The adsorbed uremic toxin quantity was differentiated by unbound (UB) and bound (Elu): PAA UB (1), PAA Elu (2), IDS UB (3), IDS Elu (4), pCR UB (5), pCR Elu (6), OH-HPA UB (7), OH HPA Elu (8).
60 wt.% was irreversibly bound to the column. In case of OH-HPA, 30 wt.% of the loaded toxin came into unbound fraction and out of bound fraction, the distribution between reversibly and irreversibly bound were 20 wt.% and 50 wt.% respectively. 4. Discussion Presently, the assessment of dialysis quality is exclusively based on the elimination of small, watersoluble molecules. However, increasing dialysis adequacy in terms of Kt/Vurea index has not improved mortality or morbidity in dialysis patients [9]. Hemodialysis helps patients in compensating their impaired kidney function, however, limitation of existing strategies is removal hydrophobic uremic toxins, PAA, IDS, pCR and etc. These uremic toxins accumulate in the blood plasma over the period of time despite of regularly hemodialysis treatments three times per week at least. Accumulated hydrophobic uremic toxins show adverse effects on the cardiovascular system, thus CVD becomes one of the major causes of premature death of CKD patients. Most of these uremic toxins are characterized by aromatic structures, which are less water-soluble and therefore associated to plasma proteins. Due to their binding to proteins, these substances should be considered as high Mw substances [8], which are thus not eliminated by diffusion but by convection. However, since the filtration rate in dialysis is small, the removal of protein-bound, hydrophobic uremic toxins from plasma of CKD patients is limited. The current dialysis techniques still are far away from replacing the natural elimination of uremic toxins by the kidneys. Dialysis is based on the principles of diffusion and/or filtration; the dialysis membranes currently used solely act as filtering membranes. The major drawbacks of current dialysis are related to these principles of action: By filtration and diffusion with a cut-off in the range of 18 kDa preferably low molecular weight hydrophilic substances are eliminated, whereas protein-bound hydrophobic small molecules are largely retained in the blood plasma. In the present study the adsorption capacity of highly porous microparticles prepared from PEI was investigated. Both particle types PEI-NW and PEI-DN showed high affinity of adsorption to all four types of toxins tested. The results are promising in terms of uremic toxin adsorption capacity, and also indicate that the majority of the toxins are reversibly bound except for IDS by PEI-NW, which offers the hope of recycling of the particles, by repeated washings with alcohol and subsequent steam sterilization. As
664
S.D. Tetali et al. / Adsorption of uremic toxins by polyetherimide
identical surface chemistries can be anticipated for both PEI-NW and PEI-DN microparticles and a larger mean pore diameter around 136 nm was reported for PEI-NW compared to PEI-DN exhibited (41 nm) before [1], we have currently no reasonable explanation for the non-reversible interaction of IDS and PEI-NW. Therefore, new treatment options have to be developed for improving the outcome of hemodialysis patients. Recently, hydrophobic and cationic adsorbers were used in a fractionated plasma separation, adsorption, and dialysis (FPAD) system for removal of protein-bound, hydrophobic uremic toxins [5]. Removal rates of FPAD treatment in comparison to conventional high-flux hemodialysis were increased by 130% for phenylacetic acid, 187% for indoxyl sulfate, and 127% for p-cresol compared to conventional high-flux hemodialyzer. The strongest limitation of the FPAD system for clinical application was the necessary isolation of plasma before the uremic toxins were able to interact with the adsorber(s). This requirement necessitates new dialysis machines in the clinics, causing high cost for the investment as well as additional disposables for their operation. This requirement hampers the introduction of the technique in the clinical routine presently and most likely in the future. The separation of plasma from the blood cells is necessary to prevent any interaction of corpuscular blood compounds with the hydrophobic or anionic surfaces of the adsorber. These side effects cause the need for new adsorber materials, which have a high capacity of the adsorption of hydrophobic uremic toxins on the one hand and which can be modified to stop the interaction of corpuscular blood components. In the current study we were able to demonstrate the characteristics of PEI particles for adsorption of uremic toxins. These particles are characterized by their strong affinity to hydrophobic uremic toxins and the feasibility of hydrophilic surface modification to prevent interaction of corpuscular blood components with the adsorber surface [3]. The latter described for membranes [2, 4, 11]. The data of the current study demonstrate the high capacity of the PEI particles for the adsorption of hydrophob uremic toxins like PAA, IDS and pCRS.
5. Conclusion PEI particles showed high capacity of adsorption to hydrophobic uremic toxins and most of the investigated toxins were reversibly bound, which offers the possibility of recycling these particles provided the particles are biocompatible. After demonstrating the high adsorption capacity of the PEI microparticles, future investigations will focus on the hemocompatibility of the adsorbers.
Acknowledgments The authors acknowledge the experimental assistance by Mario Rettschlag and characterization support by Yvonne Pieper and Regine Apostel. DBT-CREST overseas fellowship (2010-11) awarded to Dr. Sarada D. Tetali, Indo-German Science and Technology Centre (Grant No. IGSTC/NPORE/SDT/2012) and German Federal Ministry for Education and Research (BMBF), (Grant No.s 01DQ13006A, 01DQ13006B and 01DQ13006C) are acknowledged for financial support.
S.D. Tetali et al. / Adsorption of uremic toxins by polyetherimide
665
References [1] W. Albrecht, K. Lutzow, T. Weigel, T. Groth, M. Schossig, and A. Lendlein, Development of highly porous microparticles from poly(ether imide) prepared by a spraying/coagulation process, Journal of Membrane Science 273(1-2) (2006), 106–115. [2] W. Albrecht, B. Seifert, T. Weigel, M. Schossig, A. Hollander, T. Groth and R. Hilke, Amination of poly(ether imide) membranes using di- and multivalent amines, Macromolecular Chemistry and Physics 204(3) (2003), 510–521. [3] S. Basu, M. Heuchel, T. Weigel, K. Kratz and A. Lendlein, Integrated process for preparing porous, surface functionalized polyetherimide microparticles, Polymers for Advanced Technologies accepted on 24.08.2015 (2015). [4] S. Braune, M. von Ruesten-Lange, C. Mrowietz, K. Lutzow, T. Roch, A.T. Neffe, A. Lendlein and F. Jung, Dynamic in vitro hemocompatibility testing of poly(ether imide) membranes functionalized with linear, methylated oligoglycerol and oligo(ethylene glycol), Clinical Hemorheology and Microcirculation 54(3) (2013), 235–248. [5] F. Brettschneider, M. Tolle, M. von der Giet, J. Passlick-Deetjen, S. Steppan, M. Peter, V. Jankowski, A. Krause, S. Kuhne, W. Zidek and J. Jankowski, Removal of Protein-Bound, Hydrophobic Uremic Toxins by a Combined Fractionated Plasma Separation and Adsorption Technique, Artificial Organs 37(4) (2013), 409–416. [6] P. Brunet, L. Dou, C. Cerini and Y. Berland, Protein-bound uremic retention solutes, Advances in Renal Replacement Therapy 10(4) (2003), 310–320. [7] L. Dou, M. Sallee, C. Cerini, S. Poitevin, B. Gondouin, N. Jourde-Chiche, K. Fallague, P. Brunet, R. Calaf, B. Dussol, B. Mallet, F. Dignat-George and S. Burtey, The Cardiovascular Effect of the Uremic Solute Indole-3 Acetic Acid, Journal of the American Society of Nephrology 26(4) (2015), 876–887. [8] F. Duranton, G. Cohen, R. De Smet, M. Rodriguez, J. Jankowski, R. Vanholder, A. Argiles and G. European Uremic Toxin Work, Normal and pathologic concentrations of uremic toxins, J Am Soc Nephrol 23(7) (2012), 1258–1270. [9] G. Eknoyan, G.J. Beck, A.K. Cheung, J.T. Daugirdas, T. Greene, J.W. Kusek, M. Allon, J. Bailey, J.A. Delmez, T.A. Depner, J.T. Dwyer, A.S. Levey, N.W. Levin, E. Milford, D.B. Ornt, M.V. Rocco, G. Schulman, S.J. Schwab, B.P. Teehan, R. Toto and G. Hemodialysis Study, Effect of dialysis dose and membrane flux in maintenance hemodialysis, N Engl J Med 347(25) (2002), 2010–2019. [10] J. Jankowski, M. van der Giet, V. Jankowski, S. Schmidt, M. Hemeier, B. Mahn, G. Giebing, M. Tolle, H. Luftmann, H. Schluter, W. Zidek and M. Tepel, Increased plasma phenylacetic acid in patients with end-stage renal failure inhibits iNOS expression, J Clin Invest 112(2) (2003), 256–264. [11] M. Lange, S. Braune, K. Luetzow, K. Richau, N. Scharnagl, M. Weinhart, A.T. Neffe, F. Jung, R. Haag and A. Lendlein, Surface functionalization of poly(ether imide) membranes with linear, methylated oligoglycerols for reducing thrombogenicity, Macromol Rapid Commun 33(17) (2012), 1487–1492. [12] A.W. Martinez, N.S. Recht, T.H. Hostetter and T.W. Meyer, Removal of P-cresol sulfate by hemodialysis, J Am Soc Nephrol 16(11) (2005), 3430–3436. [13] K. Namekawa, M.T. Schreiber, T. Aoyagi and M. Ebara, Fabrication of zeolite-polymer composite nanofibers for removal of uremic toxins from kidney failure patients, Biomaterials Science 2(5) (2014), 674–679. [14] N. Neirynck, R. Vanholder, E. Schepers, S. Eloot, A. Pletinck and G. Glorieux, An update on uremic toxins, Int Urol Nephrol 45(1) (2013), 139–150. [15] T. Niwa and M. Ise,Indoxyl sulfate, A circulating uremic toxin, stimulates the progression of glomerular sclerosis, J Lab Clin Med 124(1) (1994), 96–104. [16] D. Polak-Jonkisz, D. Zwolinska and L. Purzyc, The PMA action via metabolism of uremic erythrocytes, Clin Hemorheol Microcirc 59(3) (2015), 283–285. [17] S.R. Sandeman, C.A. Howell, G.J. Phillips, Y.S. Zheng, G. Standen, R. Pletzenauer, A. Davenport, K. Basnayake, O. Boyd, S. Holt and S.V. Mikhalovsky, An adsorbent monolith device to augment the removal of uraemic toxins during haemodialysis, Journal of Materials Science-Materials in Medicine 25(6) (2014), 1589–1597. [18] H. Watanabe, Y. Miyamoto, M. Otagiri and T. Maruyama, Update on the pharmacokinetics and redox properties of protein-bound uremic toxins, J Pharm Sci 100(9) (2011), 3682–3695. [19] H. Yamamoto, S. Tsuruoka, T. Ioka, H. Ando, C. Ito, T. Akimoto, A. Fujimura, Y. Asano and E. Kusano, Indoxyl sulfate stimulates proliferation of rat vascular smooth muscle cells, Kidney Int 69(10) (2006), 1780–1785.
