WO2014058359A1

WO2014058359A1 - Method of coating and encapsulation of cells and cell aggregates with thick and stable polymer membrane

Info

Publication number: WO2014058359A1
Application number: PCT/SE2012/051079
Authority: WO
Inventors: Yuji Teramura; Bo Nilsson
Original assignee: Yuji Teramura; Bo Nilsson
Priority date: 2012-10-09
Filing date: 2012-10-09
Publication date: 2014-04-17

Abstract

The present invention relates to a cross-linked polymer for surface modification of a cell or cell aggregate, said cross-linked polymer comprising micelles of a maleimide-conjugated amphiphilic first polymer, and a second polymer having multiple thiol groups. It further relates to cell and cell aggregates encapsulated by such polymers, and theire use in therapy, as well as kits and methods for preparing such polymers.

Description

Title

Method of coating and encapsulation of cells and cell aggregates with thick and stable polymer membrane Field of the invention

The present invention relates to the field of immune-isolation of cells by encapsulation, particularly for transplantation.

Background of the invention

Globally, it is estimated that more than 366 million people suffer from diabetes (type 1 and 2) in 2010 and the patients will increase continuously. In Sweden, there are more than 700,000 people (out of the whole population of 9,000,000) with diabetes. Type 1 is one of the chronic diseases of children. It is caused by autoimmune destruction of insulin- producing β cells in islet of Langerhans (islet) of pancreas, which leads to insulin deficiency. For the therapy of Type 1 diabetes, islet transplantation is promising since this surgery is less invasive and dangerous than pancreas transplantation. However, the use of immunosuppressive drugs is a major problem to be solved.

There are three therapies for type 1 diabetes: insulin therapy, pancreas transplantation, and islet transplantation. Insulin therapy is subcutaneous injection, which is the most common. But, the control of glucose level is still difficult especially during sleeping and gradually patients have various complications. Pancreas transplantation can restore proper glucose regulation, however, the surgery and immunosuppressive drugs are considered to be more dangerous than insulin therapy. Islet transplantation can also regulate glucose level, which is infusion of islet cells suspension into liver of patients. Since this surgery is less invasive and dangerous than pancreas transplantation, it is becoming more common. However, the use of immunosuppressive drugs is still a problem to be solved in islet transplantation. Their side effects are not fully understood during the medication for the long term.

Therefore, researchers have worked with immuno-isolation device/membrane of islets, which can reduce the use immunosuppressive drugs. The surface modification and microencapsulation of cells or cell aggregates e.g. pancreatic islets (islets) with polymer membrane has been an important subject in cell transplantation because transplanted cells can be isolated from the immune system of the host, reducing the need for immunosuppressive drugs. This membrane coating results in the improvement of graft survival by regulating immune response reactions. Patients might be able to stop taking immuno suppressive drugs or reduce these drugs when the coating with membrane would inhibit immune response reactions.

Microencapsulation of cells or islets with hydrogel such as alginate and agarose has been successfully reported so far. In clinical trials, microcapsules of islets with alginate have been transplanted into human body without immunosuppressive drugs, and some improvement could be achieved in their glucose control. First cells/islets suspension is mixed with alginate solution. Then it is dropped into calcium (or barium) chloride solution to make a microcapsule. For agarose hydrogel, cells/islets suspension is mixed with agarose solution which is warmed at 40°C. Then it is shaken on ice to make a

microcapsule. These methods have been often used since it is easy to prepare a

microcapsule of cells or islets. Also the microcapsule is quite stable since it is not degraded. However, it is difficult to control the membrane thickness. The diameter of the capsule is quite large. Oxygen supply is not enough if the membrane becomes thicker around islets. And the total volume of microencapsulated cells or islets would dramatically increase after encapsulation with thicker membrane because it will increase by a third power of the radius. For these reasons, it is difficult to use them in clinical setting.

Much effort has been made to reduce the size of the capsules. Now the cell coating of thinner membrane is performed by cell surface modification with polymers membrane. In most of cases, assembling of cationic polymers and anionic polymers (layer-by-layer membrane) has been used to make membrane on cell surface. Since cationic polymers can interact with negatively charged cell surface, it is easy to deposit cationic polymers on cell surface and add further anionic polymer on that surface. But cationic polymers are known to be cytotoxic, and it is not straightforward to safely coat cells with cationic polymers. For the membrane thickness, it is at most tens of nanometer level because layer-by-layer membrane is made of several polymers. Therefore, there is no substantial increase of the membrane thickness. Other reports have used amphiphilic polymers such as poly(ethylene glycol)-conjugated phospholipid (PEG-lipid) and covalently binding polymers or materials. Cytotoxicity has been improved after cell coating with these polymers, however, the membrane thickness is at nanometer level. Polymer membrane should be much thicker, up to micrometer level to completely isolate the cell surface from the attack of the host immune system. Also the stability of those polymer membranes is quite low.

Most of polymer membranes on cell surface will be detached or disappear from the cell surface or taken up into the cell within one or two days.

Summary of the invention

The present invention aims to address problems of the prior art by providing the compositions of matter, methods and kits according to the appended claims.

In one aspect, the invention relates to a cross-linked polymer for surface modification of a cell or cell aggregate, said cross-linked polymer comprising micelles of a maleimide- conjugated amphiphilic first polymer, and a second polymer having multiple thiol groups. In one embodiment of this aspect, the cross-linked polymer further comprises a third polymer having multiple maleimide groups.

Preferred embodiments are set forth in the dependent claims. The invention provides a unique surface coating of living cells in order to improve the engraftment in cell-based therapy by controlling immune reactions. When cells such as islets of Langerhans (islets), embryonic stem (ES) cells, and induced pluripotent stem (iPS) cells are transplanted into human body, there are two major problems to be solved: (1) One is that most of the cells are immediately destroyed by activation of innate immunity just after transplantation (called instant blood mediated inflammatory reaction, IBMIR), which results in lower engraftment levels. To inhibit IBMTR is an important factor for improving the engraftment. (2) Another is immune-rejection reaction. Transplanted cells are constantly attacked by the host immune system after transplantation. Therefore, recipients need to take

immunosuppressive drugs to suppress the immune-rejection reaction for the whole life. Since the side effects are not fully understood, it is preferred to avoid taking medication for the long term. An approach without the use of immunosuppressive drugs or less is desired.

The invention addresses these issues by providing

1. Cell coating with bioactive molecules to inhibit IB MIR involving coagulation and complement activation

2. Cell coating with stable ultra-thin polymeric membrane to avoid immune-rejection reaction

In presently preferred embodiments, the invention relates to methods, composition and kits for preparation of thick and stable polymer membrane, which is made of maleimide- conjugated PEG-phospholipid (Mal-PEG-lipid) micelle, polymers having multiple maleimide groups (e.g. 4-arm PEG-Mal) and polymers having multiple SH groups (e.g. 8- arm PEG-SH), on the surface of cells and cell aggregates, e.g. pancreatic islets. The polymer membranes on cell surface can be used for microencapsulation of cells or islets.

When Mal-PEG-lipid is mixed with cells or islets, Mal-PEG-lipid is spontaneously inserted into lipid bilayer membrane of cells by hydrophobic interaction. Since Mal-PEG- lipid forms a micelle in an aqueous solution above critical micelle concentration (cmc), the interaction between Mal-PEG-lipid on cell surface and micelle in an aqueous solution is at equilibrium state. When a polymer having multiple SH groups (e.g. 8-arm PEG-SH) is added to cells or islets in the presence of Mal-PEG-lipid micelle, these polymers react with both Mal-PEG-lipids on the cell surface and the micelles through thiol-maleimide reaction. These polymers cross-link between cells and micelles, between micelles and micelles, which results in the formation of polymer membrane on cell surface. The thiol- maleimide reaction forms covalent bonding, so the membrane becomes stable. A polymer having multiple maleimide groups (e.g. 4-arm PEG-Mal) is added to cross-link between polymers having multiple SH groups in the membrane. And then, the membrane becomes more stable and the outer surface of the membrane is modified with maleimide groups.

Cells and islets are not damaged by this method because there is no direct interaction between polymers and cell surface.

While lower-molecular weight molecules such as glucose and insulin can rapidly permeate through the membrane on cell surface, higher-molecular weight molecules such as C3 and IgG are not easy to permeate through the membrane. Therefore, this polymer membrane can protect cells or islets from the attack of immune response reaction.

In order to increase the membrane thickness, Mal-PEG-lipid is added to the cells or islets coated with the polymers membrane again. Then polymers having multiple SH groups is added to cells or islets in the presence of Mal-PEG-lipid micelle. Finally, polymer having multiple maleimide groups is added. The membrane thickness can be increased by repeat of this procedure. Furthermore, unreacted maleimide groups exist on the outer surface membrane which are fabricated on cell surface. Therefore various bioactive substances having SH or maleimide groups can be immobilized onto the membrane through thiol- maleimide reaction. The immobilization of various bioactive substances (e.g. apyrase, heparin, factor H, etc.) is useful to regulate immune response reactions.

One aspect of the invention also relates to the methods, composition and kits for preparation of thick and stable polymer membrane, which is made of Mal-PEG-lipid micelle, polymers having multiple maleimide groups (e.g. 4-arm PEG-Mal) and polymers having multiple SH groups (e.g. 8-arm PEG-SH), on substrate surface (e.g. glass, plastic). The surface of substrate is coated with maleimide groups. After Mal-PEG-lipid is added to the surface, polymers having multiple SH groups is mixed in the presence of Mal-PEG- lipid micelle. Then, polymers having multiple maleimide groups are added. The membrane thickness can be increased by repeat of this procedure. Various bioactive substances having SH or maleimide groups can be also immobilized onto the membrane through thiol-maleimide reaction. Short description of the appended drawings

Figure 1 shows a schematic illustration of polymer membrane formation on

cells/pancreatic islets surface. The membrane is formed by polymers having SH groups (e.g. 8-arm PEG-SH) and polymers having maleimide groups (e.g. 4-arm PEG-maleimide) in the presence of Mal-PEG-lipid micelle.

Figure 2 A-C show a chemical structure of typical materials which can be used for the membrane as shown in Fig. 1. A: Maleimide-conjugated PEG-lipid, B: polymers having maleimide groups, 4-arm PEG-Mal, and C: polymers having SH groups, 8-arm PEG-SH.

Figure 3 A-D show confocal images of rabbit erythrocytes which are coated with

Alexa488-apyrase and polymer membranes: A: Erythrocyte was mixed with Mal-PEG- lipid, followed by washing with buffer. And then they were reacted with Alexa488- apyrase-SH. B: Erythrocyte was mixed with Mal-PEG-lipid, followed by washing with buffer. After that, 8-arm PEG-SH was added and then 4-arm PEG-Mal was added after washing with buffer. Finally Alexa488-apyrase-SH was added. C: Erythrocyte was mixed with Mal-PEG-lipid and then the mixture was added into 8-arm PEG-SH solution with agitation. After washing with buffer, 4-arm PEG-Mal was added. After washing with buffer, Alexa488-apyrase-SH was added. D: Erythrocyte was mixed with Mal-PEG-lipid and then the mixture was added into 8-arm PEG-SH solution with agitation. After washing with buffer, 4-arm PEG-Mal was added. After washing with buffer, these procedures were repeated again. After that, Alexa488-apyrase-SH was added. Figure 4 A and B show confocal images of rabbit erythrocytes which were coated with Alexa488-apyrase and polymer membranes as shown in Figure 3 D, A: after lysis with pure water and B: after 30 days incubation in buffer.

Figure 5 A and B show confocal images of porcine aortic endothelial (PAE) cells which were coated with Alexa488-apyrase and polymer membranes. A: PAE cells was mixed with Mal-PEG-lipid, followed by washing with buffer. And then they were reacted with Alexa488-apyrase-SH. B: PAE cells were mixed with Mal-PEG-lipid and then the mixture was added into 8-arm PEG-SH solution with agitation. After washing with buffer, 4-arm PEG-Mal was added. After washing with buffer, these procedures were repeated again. After that, Alexa488-apyrase-SH was added. Figure 6 A-C show confocal images of pancreatic mouse islets which were coated with Alexa488-apyrase and polymer membranes. Islets were mixed with Mal-PEG-lipid and then the mixture was added into 8-arm PEG-SH solution (A: 5mg/mL, B: 7.5 mg/mL, C: 10 mg/mL) with agitation. After washing with buffer, 4-arm PEG-Mal was added. After washing with buffer, these procedures were repeated again. After that, Alexa488-apyrase- SH was added.

Figure 7 A-D show confocal images of mouse islets which were coated with Alexa488- apyrase and polymer membranes at 1, 5, 8, and 15 days culture after preparation. Islets were coated with Alexa488-apyrase and polymer membranes as shown in Fig. 6 A.

Figure 8 shows confocal images of A: rabbit erythrocytes and B: mouse islets which were coated with Alexa488-apyrase and polymer membranes using Mal-PEG-lipid, 4-arm PEG- Mal and alginate having SH instead of 8-arm PEG-SH. Figure 9 A and B show insulin release in response to a glucose challenge: insulin release from control islets and islets coated with polymer membrane as shown in Fig. 6 A, which were cultured for 1 day and 7 days (n=3). The islets were sequentially stimulated with 1.67 (0 to 24 min), 16.7 (24 to 60 min) and 1.67 mM (60 to 96 min) glucose. Data shown are means ± SEM.

Figure 10 A and B show membrane permeability test of A: glucose, insulin, human serum albumin (HSA), and IgG, and B: serum proteins (insulin, C3, IgG, and IgM). The membrane was prepared by mixing 4-arm PEG-Mal, 8-arm PEG-SH, and Mal-PEG-lipid. A: When 4-arm PEG-Mal, 8-arm PEG-SH, and Mal-PEG-lipid were mixed, each of proteins (glucose, insulin, HSA, and IgG) was added. Then absorbance of the supernatant (at 280 nm) was measured with time. B: When 4-arm PEG-Mal, 8-arm PEG-SH, and Mal- PEG-lipid were mixed, human serum which was supplemented with insulin was added. Then the supernatant was analyzed to determine the concentration of insulin, insulin, C3, IgG, and IgM with time. IgM was not detected in the supernatant during the time because the concentration was below the limit of detection (0.0417 g/L). Figure 11 shows hemolytic assay of rabbit erythrocytes in human serum. Rabbit erythrocyte was coated with PEG-lipid (single PEG layer) or polymer membranes made of 4-arm PEG-Mal, 8-arm PEG-SH in the presence of Mal-PEG-lipid as shown in Fig. 3 D.

Figure 12 A-D show blood compatibility test for polymer membranes in human whole blood. Whole blood without heparin was incubated on single PEG-coated (single PEG chain) and polymer membrane-coated (made of 4-arm PEG-Mal, 8-arm PEG-SH, and Mal-PEG-lipid) substrate for 60 min at 37°C, (n=6). The figures show A: platelet consumption and B: generation of TAT, C: C3a, and D: sC5b-9. Data shown are means ± SEM.

Detailed description of preferred embodiments of the invention

The present invention is not limited to the above-described preferred embodiments.

Various alternatives, modifications and equivalents may be used. Therefore, the above embodiments should not be taken as limiting the scope of the invention, which is defined by the appending claims.

The invention relates to methods, composition and kits for preparation of thick and stable polymer membrane, which is made of maleimide-conjugated PEG-phospholipid (Mal- PEG-lipid) micelle, polymers having multiple maleimide groups (e.g. 4-arm PEG-Mal) and polymers having multiple SH groups (e.g. 8-arm PEG-SH), on the surface of cells and cell aggregates, e.g. pancreatic islets. The polymer membranes on cell surface can be used for encapsulation of cells or islets.

As used herein, "encapsulation" is used to refer to a coating which completely enclose and physically isolate the encapsulated cells or islets from the surrounding environment. An encapsulation with polymer membrane can have sufficient permeability of nutrients, waste products, and oxygen for maintenance of an encapsulated cell or islet while preventing from the attack of immune response reactions of the host in which the encapsulated cells or islets is transplanted.

As used herein, "cells and cell aggregates" are used to refer to functional cells or cell aggregates such as erythrocyte, culture cell line, primary cell, pancreatic islets, stem cells (e.g. mesenchymal stem cells (MSC), embryoid stem (ES) cells, induced pluripotent stem (iPS) cells) or differentiated cells from stem cells.

The term "multiple", as used in connection with polymers having multiple thiol or maleimide groups, shall be construed as more than one. This includes at least 3, and at least 4, as well as at most 4, 8, 12, 18, or 30.

Mal-PEG-lipid is a conjugate of hydrophilic maleimide-PEG and hydrophobic phospholipid as shown in Fig. 2A, which is an amphiphilic polymer. An amphiphilic polymer forms a micelle in an aqueous solution above critical micelle concentration (cmc). The molecular weight of PEG can be 600 to 40,000 Daltons. Exemplary hydrophilic polymers include PEG but are not limited to PEG and copolymers thereof. The acyl chain length (n) of phospholipid can be 12 to 16.

As used herein, "multiple" is used to refer to a polymer having more than two groups of maleimide or SH in a polymer chain. The molecular weight of the polymer can be 5,000 to 50,000 Daltons. Exemplary hydrophilic polymers include, but are not limited to PEG, alginate and copolymers thereof.

When Mal-PEG-lipid is mixed with cells or islets, Mal-PEG-lipid is spontaneously inserted into lipid bilayer membrane of cells by hydrophobic interaction. Since Mal-PEG- lipid forms a micelle in an aqueous solution above critical micelle concentration (cmc), the interaction between Mal-PEG-lipid on cell surface and micelle in an aqueous solution is at equilibrium state. When polymers having multiple SH groups (e.g. 8-arm PEG-SH) is added to cells or islets in the presence of Mal-PEG-lipid micelle, these polymers react with both Mal-PEG-lipid on the cell surface and micelle through thiol-maleimide reaction. These polymers cross-link between cells and micelles, and between micelles and micelles, which results in the formation of polymer membrane on cell surface. The thiol-maleimide reaction forms covalent bonding, so the membrane becomes stable. After that, polymer having multiple maleimide groups (e.g. 4-arm PEG-Mal) is added to cross-link between polymers having multiple SH groups in the polymer membrane. And then, the membrane becomes more stable and the outer surface of the membrane is modified with maleimide groups which are available for further modification.

In order to increase the membrane thickness, Mal-PEG-lipid is added to the cells or islets coated with polymers membrane again. Then polymers having multiple SH groups is added to cells or islets in the presence of Mal-PEG-lipid micelle. Finally, polymer having multiple maleimide groups is added. The membrane thickness can be increased by repeat of this procedure. Furthermore, unreacted maleimide groups exist on the outer surface membrane which are fabricated on cell surface. Therefore various bioactive substances having SH or maleimide groups can be immobilized onto the membrane through thiol- maleimide reaction. The immobilization of various bioactive substances (e.g. apyrase, heparin, factor H, etc.) is useful to regulate immuno response reactions.

Thus the polymer membrane formed on cell surface by the invention has the substantial increase of thickness at micrometer level (Fig. 3C, D, 5, 6). The membrane thickness is actually increased by repeat of the procedure. The membrane thickness after the second procedure is larger than that after the first procedure.

Figure 3A shows a picture of cells which were mixed with Mal-PEG-lipid and washed with buffer, followed by labeling with Alexa488-apyrase-SH, indicating of the single PEG layer. Obviously the membrane thickness of the invention is larger than the single PEG chain.

The substantial increase of membrane thickness could not be observed in the absence of Mal-PEG-lipid micelle during addition of 8-arm PEG-SH (Fig. 3B). After cells were mixed with Mal-PEG-lipid and washed with buffer to remove the micelle, 8-arm PEG-SH was added to the cells. After washing with buffer, 4-arm-PEG-Mal was added to the cells. The procedure was repeated twice. Fluorescence from the polymer membrane could be observed at the periphery of cells, however, there was no substantial increase of membrane thickness. The polymer membrane is almost the same as single PEG chain as seen in Fig. 3A. Deposition of polymer layer with 8-arm PEG-SH and 4-arm-PEG-Mal onto the cell surface can reach nanometer level, but cannot reach micrometer level. Therefore polymer membrane can be substantially increased in the presence of Mal-PEG- lipid micelle.

Polymer membrane of the invention is stable because thiol-maleimide reaction forms stable covalent bonding. The polymer membrane structure surrounding cells was still maintained even after encapsulated cells were destroyed by lysis with pure water (Fig. 4A). The polymer membrane could be still remained on the cell surface after 30 days (Fig. 4B). These results show the polymer membrane of the invention is thick and stable.

The invention can be used for various sizes of cells (rabbit erythrocyte: 5μιη, porcine aortic endothelial cell: 12μιη and islets: 100-300 μιη) as shown in Fig. 3, 5, and 6. Since the invention is based on the use of Mal-PEG-lipid micelle, the size of cells or cell aggregates is not important in comparison with microencapsulation with hydrogel using agarose and alginate. For porcine aortic endothelial cells, the polymer membrane at micrometer level could be formed on the cell surface (Fig. 5). In addition, polymer membrane at micrometer level could be formed on the islet surface (Fig. 6). The polymer membrane could exist on islet surface for more than 15 days (Fig. 7), indicating that the polymer membrane was more stable than single PEG layer (completely disappear within a few days).

Alginate having SH groups (1%) can be also available to make polymer membranes on cell surface instead of 8-arm PEG-SH (Fig. 8A, B). It is possible to make polymer membrane at micrometer level on the surface of erythrocyte and islets when Mal-PEG- lipid micelle, alginate having SH groups, and 4-arm PEG-Mal were used. The polymer membrane is stable for 30 days.

Generally cells or islets are destroyed when positively charged polymers interact with cell surface. Deposition of polymers onto cell surface such as layer-by-layer membrane results in the destruction of cell structure. In the invention, encapsulated cells or islets are not damaged by the polymer membrane because there is no direct interaction between polymers and cell surface. Erythrocytes and porcine aortic endothelial cells are still intact after encapsulation with the polymer membrane (Fig. 3, 5, 6). Figure 9A, B show insulin release from encapsulated islets in response to a glucose challenge after 1 day and 7 days. The insulin release of encapsulated islets was almost the same as control islets, indicating that the function of encapsulated islet was well maintained.

The polymer membrane can be used for encapsulation of cells or islets which can protect from the immune response reactions. Figure 10 A, B show the permeability test of glucose, insulin albumin, IgG, C3 and IgM through the polymer membrane. While lower- molecular weight molecules such as glucose and insulin can rapidly permeate through the polymer membrane, higher-molecular weight molecules such as C3 and IgG are not easy to permeate through the membrane. IgM cannot permeate through the polymer membrane because of the larger molecular size (900 kDa). It was considered that the membrane has molecular cut-off between 190 kDa and 900 kDa approximately. Since IgM is mainly involved in the complement activation, the suppression of the permeation of IgM is promising for attenuation of IBMIR in xenotrasnplantation as well as allotransplantation. When encapsulated rabbit erythrocyte within the polymer membrane was incubated in human serum, the lysis of erythrocyte can be protected while control erythrocyte is completely destroyed by the attack from the immune response reaction as shown in Fig. 11. Single PEG chain can suppress the lysis, however, the effect is limited. Therefore, this polymer membrane of the invention can protect cells from the attack of immune response reaction.

The polymer membrane of the invention is more biocompatible than single PEG chain as shown in Fig. 12. When human whole blood is incubated on the polymer membrane, platelet aggregation and coagulation activation are suppressed (Fig. 12 A, B). However, platelet aggregation and coagulation activation were induced on single PEG chain surface. For complement system, the activation markers, C3a, sC5b-9 were the same level between single PEG chain surface and the polymer membrane (Fig. 12 C, D). These results show the polymer membrane of the invention is higher biocompatibility. One aspect of the invention also relates to the methods, composition and kits for preparation of thick and stable polymer membrane, which is made of Mal-PEG-lipid micelle, polymers having multiple maleimide groups (e.g. 4-arm PEG-Mal) and polymers having multiple SH groups (e.g. 8-arm PEG-SH), on substrate surface (e.g. glass, plastic, etc). The surface of substrate is coated with maleimide groups. After Mal-PEG-lipid is added to the surface, polymers having multiple SH groups is mixed in the presence of Mal-PEG-lipid micelle. Then, polymers having multiple maleimide groups are added. The membrane thickness can be increased by repeat of this procedure. Various bioactive substances having SH or maleimide groups can be also immobilized onto the membrane through thiol-maleimide reaction.

EXAMPLES

The below examples are provided to illustrate and further explain the general concept of the immune-isolation technology according to the invention by describing specific embodiments in detail. The individual details set out below are generally applicable to the general concept of the invention, if not otherwise indicated.

The examples show the use of the invention in relation to erythrocytes and islet cells. The examples shall however not be construed as limiting the scope of the invention.

Method of encapsulation of cells (erythrocyte, porcine aortic endothelial cell islet) Encapsulation of rabbit erythrocyte with polymer membrane

Mal-PEG-conjugated phospholipid (Mal-PEG-lipid) was synthesized as described in the following. Briefly, a-N-hydroxysuccinimidyl-w-maleimidyl poly(ethylene glycol)

(200mg, HS-PEG-Mal, Mw: 5000, NOF Corporation, Tokyo, Japan), triethylamine (50 μί, Sigma-Aldrich Chemical Co, St. Louis, Missouri), and 1, 2-dipalmitoyl-s«-glycerol- 3-phosphatidylethanolamine (20mg, DPPE, NOF Corporation) were dissolved in dichloromethane (Sigma-Aldrich Chemical Co) and stirring for 48 h at room temperature (RT). After precipitation with diethyl ether, Mal-PEG-lipid was obtained as a white powder (190 mg, yield 80%). Rabbit erythrocyte (5.0xl0⁵ cells) was mixed with Mal-PEG-lipid (50μΕ, 50 mg/mL in PBS) and incubated for 30min at room temperature with gentle mixing. The erythrocyte suspension was put on ice for lOmin. The erythrocyte suspension was added into 8-arm PEG-SH solution (50μΙ., 5mg/mL, in PBS, pH7.4, hexaglycerol octa(mercaptoethyl) polyoxy ethylene, Mw: 20,000) with vigorous mixing. The mixture was left for 30min at room temperature. After washed with PBS containing bovine serum albumin (lOmg/mL, BSA, Sigma- Aldrich Chemical Co) (BS A/PBS) by centrifugation twice, 4-arm PEG-Mal solution (50μΙ., 50mg/mL, in PBS, pH7.4, pentaerythritol tetra{[3-(3-maleimido-l- oxopropyl)amino]propyl} -polyoxy ethylene) was added to the erythrocyte and incubated for lOmin on ice with gentle mixing. After washed with BS A/PBS twice, the polymer membrane-encapsulated erythrocyte was prepared. This process is the first procedure. In order to increase the thickness of the polymer membrane, the above-mentioned procedure was repeated. Again, the erythrocyte was mixed with Mal-PEG-lipid (50μΙ., 50 mg/mL in PBS) and incubated for 30min at room temperature with gentle mixing. The erythrocyte suspension was put on ice for lOmin. The erythrocyte suspension was added into 8-arm PEG-SH solution (50μΕ, 5mg/mL, in PBS) with vigorous mixing. The mixture was left for 30min at room temperature. After washed with BS A/PBS by centrifugation twice, 4-arm PEG-Mal solution (50μΙ., 50mg/mL, in PBS) was added to the erythrocyte and incubated for lOmin on ice with gentle mixing. After washed with BS A/PBS twice, the polymer membrane-encapsulated erythrocyte was prepared.

For further increase of the membrane thickness, the above-mentioned procedure was repeated for the erythrocyte.

Immobilization of apyrase on polymer membrane

For the immobilization of bioactive substances (e.g. apyrase, heparin, factor H, etc), they may need to be modified with SH or maleimide group previously. For example, apyrase was modified with thiol groups by thiolation using Traut's reagent (Thermo Fisher Scientific, Waltham, MA, USA): Apyrase solution (10 mg/mL, 400 μΐ., from potato, Sigma- Aldrich) was mixed with Traut's reagent (10 mg/mL, 66 μΕ). The solution was incubated with gentle mixing at RT for 1 h. Thiolated apyrase (apyrase-SH) was purified using a spin column (Thermo Fisher Scientific). In order to visually examine the immobilization of apyrase, apyrase-SH was labeled with Alexa Fluor® 488 by using a labeling kit (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's protocol. Alexa488-apyrase-SH (20μΙ., 3mg/mL in PBS) was added to the erythrocyte and incubated for 5min on ice. After washed with BSA/PBS twice, Alexa488-apyrase immobilized polymer membrane-encapsulated erythrocyte was prepared.

For the lysis, Alexa488-apyrase immobilized polymer membrane-encapsulated erythrocyte (5.0xl0⁵ cells in ΙΟΟμΙ. PBS) was mixed with pure water (500 μΕ) for lOmin at room temperature.

All samples were analyzed by confocal laser scanning microscopy (LSM510 META, Carl Zeiss, Jena, Germany).

Control experiments

As a control experiment, rabbit erythrocyte (5.0xl0⁵ cells) was mixed with Mal-PEG-lipid (50μΙ., 50 mg/mL in PBS) and incubated for 30min at room temperature with gentle mixing. The erythrocyte suspension was put on ice for lOmin. After washed with

BSA/PBS twice, Alexa488-apyrase-SH (20μΙ., 3mg/mL in PBS) was added to the erythrocyte and incubated for 5min on ice. After washed with BSA/PBS twice, Alexa488- apyrase immobilized-PEG-erythrocyte (single PEG layer) was prepared. As a control experiment, rabbit erythrocyte (5.0xl0⁵ cells) was mixed with Mal-PEG-lipid (50μΙ., 50 mg/mL in PBS) and incubated for 30min at room temperature with gentle mixing. The erythrocyte suspension was put on ice for lOmin. After washed with

BSA/PBS twice, the erythrocyte suspension was added into 8-arm PEG-SH solution (50μΕ, 5mg/mL, in PBS) with vigorous mixing. The mixture was left for 30min at room temperature. After washed with BSA/PBS by centrifugation twice, 4-arm PEG-Mal solution (50μΙ., 50mg/mL, in PBS) was added to the erythrocyte and incubated for lOmin on ice with gentle mixing. After washed with BSA/PBS twice, the erythrocyte was mixed with Mal-PEG-lipid (50μΙ., 50 mg/mL in PBS) again, and incubated for 30min at room temperature with gentle mixing. The erythrocyte suspension was put on ice for lOmin. After washed with BSA/PBS twice, the erythrocyte suspension was added into 8-arm PEG-SH solution (50μΙ., 5mg/mL, in PBS) with vigorous mixing. The mixture was left for 30min at room temperature. After washed with BSA/PBS by centrifugation twice, 4- arm PEG-Mal solution (50μΙ., 50mg/mL, in PBS) was added to the erythrocyte and incubated for lOmin on ice with gentle mixing. After washed with BS A/PBS twice, Alexa488-apyrase-SH (20μΙ., 3mg/mL in PBS) was added to the erythrocyte and incubated for 5min on ice. After washed with BSA/PBS twice, Alexa488-apyrase immobilized-PEG membrane-encapsulated erythrocyte (in the absence of Mal-PEG-lipid micelle) was prepared.

All samples were analyzed by confocal laser scanning microscopy.

Encapsulation of porcine aortic endothelial cells with polymer membrane

Porcine aortic endothelial cells (a kind gift from Prof Lena Claesson- Welsh, Uppsala University) were cultured in F-12+GlutaMAX™ (Invitrogen) supplemented with 10% fetal bovine serum, 100 U/mL penicillin, and 100 μg/mL streptomycin at 37°C in 95% air/5% C0₂. Those cells (3 10⁶ cells) were collected by centrifugation (100 *g, 5 min) after incubation with trypsin/EDTA (Invitrogen) for 3 min at 37°C. After washed with PBS twice, cells were mixed with Mal-PEG-lipid (50μΙ., 50 mg/mL in PBS) and incubated for 20min at room temperature with gentle mixing. The cell suspension was put on ice for 5min. The cell suspension was added into 8-arm PEG-SH solution (50μΙ., 5mg/mL, in PBS, pH7.4) with vigorous mixing. The mixture was left for 20min at room temperature. After washed with BSA/PBS by centrifugation twice, 4-arm PEG-Mal solution (50μΙ., 50mg/mL, in PBS, pH7.4) was added to the cell and incubated for lOmin on ice with gentle mixing. After washed with BSA/PBS twice, the polymer membrane- encapsulated cell was prepared. And then, again, the cells were mixed with Mal-PEG-lipid (50μΙ., 50 mg/mL in PBS) and incubated for 20min at room temperature with gentle mixing. The cell suspension was put on ice for 5min. The cell suspension was added into 8-arm PEG-SH solution (50μΙ., 5mg/mL, in PBS) with vigorous mixing. The mixture was left for 20min at room temperature. After washed with BSA/PBS by centrifugation twice, 4-arm PEG-Mal solution (50μΙ., 50mg/mL, in PBS) was added to the cell and incubated for lOmin on ice with gentle mixing. After washed with BSA/PBS twice, the polymer membrane-encapsulated cell was prepared. Alexa488-apyrase-SH (20μΕ, 3mg/mL in PBS) was added to the cells and incubated for 5min on ice. After washed with BS A/PBS twice, Alexa488-apyrase immobilized polymer membrane-encapsulated cell was prepared.

As a control experiment, porcine aortic endothelial cells (3 ^χ 10⁶ cells) was mixed with Mal-PEG-lipid (50μΕ, 50 mg/mL in PBS) and incubated for 20min at room temperature with gentle mixing. The cell suspension was put on ice for 5min. After washed with BSA/PBS twice, Alexa488-apyrase-SH (20μί, 3mg/mL in PBS) was added to the cells and incubated for 5min on ice. After washed with BSA/PBS twice, Alexa488-apyrase immobilized-PEG-cells (single PEG layer) were prepared.

All samples were analyzed by confocal laser scanning microscopy.

Encapsulation of pancreatic islets with polymer membrane

Islets were isolated from the pancreases of C57BL/6 mice (males, M&B Research and Breeding Center, Ry, Denmark) by the collagenase digestion method. Islets (150) were cultured free floating in 5 mL RPMI-1640 medium (Invitrogen) supplemented with 10% fetal bovine serum, 100 U/mL penicillin, and 100 μg/mL streptomycin in 95% air/5% C0₂. Islets were used 2days after isolation and culture.

Islets (150 islets) were mixed with Mal-PEG-lipid (50μΙ., 50 mg/mL in PBS) and incubated for 20min at room temperature with gentle mixing. The islet suspension was put on ice for 5min. The suspension was added into 8-arm PEG-SH solution (50μΙ., 5, 7.5, 10 mg/mL, in PBS, pH7.4) with vigorous mixing. The mixture was left for 15min at room temperature. After washed with BSA/PBS by centrifugation twice, 4-arm PEG-Mal solution (50μΙ., 50 mg/mL, in PBS, pH7.4) was added to the islets and incubated for lOmin on ice with gentle mixing. After washed with BSA/PBS twice, the polymer membrane-encapsulated islets were prepared. And then, again, those islets were mixed with Mal-PEG-lipid (50μΕ, 50 mg/mL in PBS) and incubated for 20min at room temperature with gentle mixing. The islet suspension was put on ice for 5min. The islet suspension was added into 8-arm PEG-SH solution (50μΙ., 5, 7.5, 10 mg/mL, in PBS, pH7.4) with vigorous mixing. The mixture was left for 20min at room temperature. After washed with BSA/PBS by centrifugation twice, 4-arm PEG-Mal solution (50μΙ.,

50mg/mL, in PBS) was added to those islets and incubated for lOmin on ice with gentle mixing. After washed with BSA/PBS twice, the polymer membrane-encapsulated islets were prepared.

Alexa488-apyrase-SH (20μΙ., 3mg/mL in PBS) was added to those islets and incubated for 5min on ice. After washed with BSA/PBS twice, Alexa488-apyrase immobilized polymer membrane-encapsulated islets were prepared.

All samples were analyzed by confocal laser scanning microscopy.

Encapsulation of erythrocyte with polymer membrane using alginate-SH

Alginate having SH groups (alginate-SH) was prepared as follows. Sodium Alginate obtained from brown algae was purchased from Sigma- Aldrich. All other reagents and solvents were purchased from Sigma-Aldrich (Sweden) and used as received. Dialysis membranes were purchased from Spectra Por-6 (MWCO 3500). The NMR experiments (δ scale; J values are in Hz) were carried out on Jeol JNM-ECP Series FT NMR system at a magnetic field strength of 9.4 T, operating at 400 MHz for 1H.

Synthesis of thiol derivative of Alginic acid was performed following a modified reported procedure. Briefly, sodium alginate (500 mg, 2.84 mmol of disaccharide repeat units) was dissolved in 250 ml de-ionized water at room temperature. To it 3,3'-dithiobis(propanoic hydrazide; DTPH) (67.6 mg, 0.284 mmol) synthesized from 3,3'-dithiobis(propanoic acid) was added followed by HOBt (434.5 mg, 2.84 mmol). Thereafter, pH of the resultant solution was adjusted to 5.0 using 1 mM NaOH solution and solid (l-ethyl-3-(3- dimethylaminopropyl) carbodiimide or EDC (108.9 mg, 0.568 mmol) was added and stirred overnight. The solution was loaded into a dialysis bag (Spectra Por-6, MWCO 3500) and dialyzed against dilute HC1 (pH=3.5) containing 0.1 M NaCl (2x2L, 48 h), then dialyzed against deionized water (2><2L, 24 h). The solution was lyophilized and 590 mg disulphide alginate derivative (~9 % modification) was obtained.

200 mg of disulphide alginate derivative was dissolved in 100 ml de-ionized water and the pH was adjusted to 9.0. Thereafter, dithiothreitol (DTT) (180 mg, 1.136 mmol) was added and reaction mixture was stirred overnight at room temperature. The solution was loaded into a dialysis bag (Spectra Por-6, MWCO 3500) and dialyzed against dilute HC1 (pH=3.5) containing 0.1 M NaCl (2x2L, 48 h), followed by dialysis against deionized water (2><2L, 24 h). The solution was lyophilized and 190 mg of alginate-SH (-18 % modification) was obtained.

Rabbit erythrocyte (5x 10⁵ cells) were mixed with Mal-PEG-lipid (50μΙ_^, 50 mg/mL in PBS) and incubated for 30min at room temperature with gentle mixing. The islet suspension was put on ice for lOmin. The suspension was added into alginate-SH solution (50μΕ, 10 mg/mL, in PBS, pH7.4) with vigorous mixing. The mixture was left for 20min at room temperature. After washed with BS A/PBS by centrifugation twice, 4-arm PEG- Mai solution (50μΕ, 50 mg/mL, in PBS, pH7.4) was added to the erythrocyte and incubated for lOmin on ice with gentle mixing. After washed with BS A/PBS twice, the polymer membrane-encapsulated erythrocytes were prepared. And then, again, those erythrocytes were mixed with Mal-PEG-lipid (50μΕ, 50 mg/mL in PBS) and incubated for 20min at room temperature with gentle mixing. The erythrocyte suspension was put on ice for 5min. The suspension was added into alginate-SH solution (50μΕ, 10 mg/mL, in PBS, pH7.4) with vigorous mixing. The mixture was left for 20min at room temperature. After washed with BSA/PBS by centrifugation twice, 4-arm PEG-Mal solution (50μΕ,

50mg/mL, in PBS) was added to those erythrocytes and incubated for lOmin on ice with gentle mixing. After washed with BSA/PBS twice, the polymer membrane-encapsulated erythrocytes were prepared.

For lysis assay, cystein solution (250 μg/mL in PBS) was mixed with those erythrocytes for 30 sec and then washed with BSA/PBS by centrifugation twice.

Alexa488-apyrase-SH (20μΕ, 3mg/mL in PBS) was added to those erythrocytes and incubated for 5min on ice. After washed with BSA/PBS twice, Alexa488-apyrase immobilized polymer membrane-encapsulated erythrocytes were prepared.

All samples were analyzed by confocal laser scanning microscopy.

Encapsulation of pancreatic islets with polymer membrane using alginate-SH

Islets (150 islets) were mixed with Mal-PEG-lipid (50μί, 50 mg/mL in PBS) and incubated for 20min at room temperature with gentle mixing. The islet suspension was put on ice for 5min. The suspension was added into alginate-SH solution (50μΕ, 10 mg/mL, in PBS, pH7.4) with vigorous mixing. The mixture was left for 15min at room temperature. After washed with BSA/PBS by centrifugation twice, 4-arm PEG-Mal solution (50μΕ, 50 mg/mL, in PBS, pH7.4) was added to the islets and incubated for lOmin on ice with gentle mixing. After washed with BSA/PBS twice, the polymer membrane-encapsulated islets were prepared. And then, again, those islets were mixed with Mal-PEG-lipid (50μΙ., 50 mg/mL in PBS) and incubated for 20min at room temperature with gentle mixing. The islet suspension was put on ice for 5min. The islet suspension was added into alginate-SH solution (50μΕ, 10 mg/mL, in PBS, pH7.4) with vigorous mixing. The mixture was left for 20min at room temperature. After washed with BSA/PBS by centrifugation twice, 4-arm PEG-Mal solution (50μΕ, 50mg/mL, in PBS) was added to those islets and incubated for lOmin on ice with gentle mixing. After washed with BSA/PBS twice, the polymer membrane-encapsulated islets were prepared.

Alexa488-apyrase-SH (20μΕ, 3mg/mL in PBS) was added to those islets and incubated for 5min on ice. After washed with BSA/PBS twice, Alexa488-apyrase immobilized polymer membrane-encapsulated islets were prepared.

All samples were analyzed by confocal laser scanning microscopy.

Insulin release from islets in response to a glucose challenge using a dynamic perfusion system

Islet function was tested in a dynamic perfusion system. Islets are cultured for 1 day and 7 days after encapsulation with polymer membrane by above-mentioned method. As a control, islets without any treatment were used. For each group, 10 islets were perfused with 1.67 mM glucose (0 to 24 min), then with 16.7 mM glucose (24 to 60 min), and again 1.67 mM glucose (60 to 96 min). The concentration of mouse insulin was analyzed using an enzyme-linked immunosorbent assay (EIA) kit (Mercodia, Uppsala, Sweden).

Membrane permeability test

10 mg/mL glucose solution (in PBS, Sigma- Aldrich, Inc), 3.5 mg/mL human insulin solution (in PBS, Actrapid Penfill, Bagsvaerd, Novo Nordisk), 10 mg/mL human serum albumin (HSA, CSL Behring, King of Prussia, Germany) solution (in PBS), and 10 mg/mL IgG solution (in PBS, Berigloblin, CSL Behring) were prepared. And Mal-PEG- lipid (50μί, 50 mg/mL), 4-arm PEG-Mal solution (50μί, 50mg/mL), and 8-arm PEG-SH solution (50μΕ, 25 mg/mL) were dissolved into each solution. All three samples were mixed and left at room temperature for 30min. After washed with PBS (4°C) once, the gel was immersed into 2mL PBS (37°C) and shaken at 37oC. For human insulin, HSA, and IgG, 25 μΐ_^ of the supernatant was taken with time for the measurement of absorbance at 280 nm and 25 μΐ. of PBS was added in replacement. For glucose, the supernatant was measured by glucose assay kti (Sigma- Aldrich, Inc).

Human serum was mixed with human insulin (3.5 mg/mL). Mal-PEG-lipid (80μΙ., 50 mg/mL), 4-arm PEG-Mal solution (δθμί, 50mg/mL), and 8-arm PEG-SH solution (δθμί, 25mg/mL) were dissolved into the human serum. All three samples were mixed and left at room temperature for 30min. After washed with PBS (4°C) once, the gel was immersed into 3mL PBS (37°C) and shaken at 37°C. The supernatant (100 μΐ.) was taken with time and 100 μΐ. of PBS was added in replacement. The concentration of insulin was analyzed using EIA kit (Mercodia). And the concentration of C3, IgG and IgM was analyzed using IMMAGE Immunochemistry Systems (Beckman Coulter, Inc. Brea, CA, USA). The limit of detection for C3, IgG and IgM was 0.0583, 0.333 and 0.0417 g/L.

Lysis assay

Hemolytic assays were performed using rabbit erythrocytes. As described above, polymer membrane-encapsulated erythrocyte (in which the procedure was repeated second times to form polymer membrane) and PEG-erythrocyte (single PEG layer) were prepared. As a control, erythrocyte was used without any modification. All samples were washing with Mg²⁺/EGTA buffer (8 mM EGTA, 2 mM MgS0₄, 1 g/L gelatin in veronal-buffered saline, pH 7.5) on ice three times. A suspension of 1% erythrocytes (50 μΕ) was mixed with 1/8- diluted human serum in Mg²⁺/EGTA buffer (100 μΐ,) and shaken for 1 h at 37°C. The supernatant was then collected by centrifugation, and the absorbance of the supernatants at 405 nm was measured to calculate the percentage of erythrocyte lysed.

Blood experiment

The slide-chamber model was used to evaluate the biocompatibility of the polymer membrane on substrate in human whole blood. The chambers and blood-collection materials were coated with heparin according to the manufacturer's protocol (Corline System AB, Uppsala, Sweden). Whole blood from seven healthy donors was collected into heparin-coated tubes.

Glass slides were washed with 2-propanol (Sigma-Aldrich, Inc.) and pure water, then immersed in 5% 3-aminopropyl triethoxysilane (APTES, Sigma-Aldrich, Inc.) in toluene solution for 1 h at RT. After washes with ethanol and pure water, they were dried in vacuo at 80°C overnight, resulting in APTES-coated glass slides. HS-PEG-Mal was dissolved in dichloromethane (2 mg/mL) and added to the APTES-coated glass slides, which were incubated for 4 days at RT with gentle shaking and then sequentially washed with dichloromethane, ethanol, and pure water.

The Mal-PEG-modified surfaces were exposed to Mal-PEG-lipid (ΙΟΟμΙ., 50 mg/mL in PBS) and incubated for lOmin at room temperature. And 8-arm PEG-SH solution (ΙΟΟμΙ., 5 mg/mL in PBS) was added with vigorous mixing. The mixture was left for 30min at room temperature. After washed with PBS twice, 4-arm PEG-Mal solution (ΙΟΟμΕ, 50 mg/mL, in PBS, pH7.4) was added to the surface and incubated for lOmin at room temperature with gentle mixing. After washed with PBS twice, the polymer membrane- covered surface was prepared. And then, again, the surface was exposed to Mal-PEG-lipid (ΙΟΟμΕ, 50 mg/mL in PBS) and incubated for lOmin at room temperature. And 8-arm PEG-SH solution (ΙΟΟμΕ, 5 mg/mL in PBS) was added with vigorous mixing. The mixture was left for 30min at room temperature. After washed with PBS twice, 4-arm PEG-Mal solution (ΙΟΟμί, 50 mg/mL, in PBS, pH7.4) was added to the surface and incubated for lOmin at room temperature with gentle mixing. After washed with PBS twice, the polymer membrane-covered surface was prepared.

Whole blood was added to each well of the slide chambers, which were covered with the polymer membrane-covered surface. As a control, Mal-PEG-modified surfaces was used after reacted with cystein (lOmg/mL in PBS) (single PEG layer). The chamber was rotated vertically at 22 rpm for 30 min in a 37°C water bath. Then 1.2 mL of blood was collected from each well and mixed with EDTA-K3 solution at a final concentration of 10 mM, and the platelet concentration was analyzed in a Coulter AcT 5diff® hematology analyzer (Coulter Corporation, Miami, FL, USA). The blood was then centrifuged at 3000g- for 25 min at 4°C, and plasma was collected and stored at -70°C until further analysis by sandwich EIAs. for C3a, sC5b-9, and thrombin-antithrombin complexes (TAT). C3a, C5a, sC5b-9, and TAT in plasma were measured by conventional sandwich EIAs. For soluble C3a, plasma was diluted 1 :500 to 1 :6000 in working buffer (PBS containing 0.05% Tween 20, 10 mg/mL BSA, and 10 mM EDTA). As previously reported, C3a was captured by anti-human C3a mAb 4SD17.3 and detected by biotinylated polyclonal rabbit anti-C3a antibody and HRP-conjugated streptavidin. Zymosan-activated serum, calibrated against purified C3a, served as a standard. Values were expressed as ng/mL. C5a was analyzed with a commercial kit (HyCult Biotechnology, Uden, The Netherlands) according to their protocol. Samples were diluted 1 :5-1 :25, and values were expressed as ng/mL.

For sC5b-9, plasma was diluted 1 :2-l :50 in working buffer. sC5b-9 was captured by anti- human C5b-9 mAb aEl 1 (Diatec Monoclonals AS, Oslo, Norway) and detected with anti- human C5 polyclonal rabbit antibody (Dako) and HRP-conjugated anti-rabbit IgG (Dako). Zymosan-activated serum containing 6xl0⁴ AU/mL served as a standard. Values were expressed as AU/mL. For TAT, plasma was diluted 1 :20 in normal citrate-phosphate- dextrose plasma. TAT was captured by anti-human thrombin mAb and detected by HRP- coupled anti-human antithrombin mAb (Enzyme Research Laboratories, South Bend, IN, USA). A standard prepared by diluting pooled human serum in normal citrate-phosphate- dextrose plasma was used. Values were expressed as μg/L.

Claims

1. A cross-linked polymer for surface modification of a cell or cell aggregate, said cross- linked polymer comprising micelles of a maleimide-conjugated amphiphilic first polymer, and a second polymer having multiple thiol groups.

2. The cross-linked polymer according to claim 1, further comprising a third polymer having multiple maleimide groups.

3. The cross-linked polymer according to claim 1 or 2, wherein the maleimide- conjugated amphiphilic first polymer comprises a hydrophobic phospholipid part, a hydrophilic part selected from poly(ethylene glycol) and alginate, and a maleimide group cojugated to said hydrophilic part.

4. The cross-linked polymer according to any of claims 1-3, wherein said second

polymer is thiol alginate or poly(ethylene glycol) having multiple thiol groups; such as 8-arm PEG-SH.

5. The cross-linked polymer according to any of claims 2-4, wherein said third polymer is 4-arm PEG-Mal.

6. The cross-linked polymer according to any of claims 1-5, further comprising said maleimide-conjugated amphiphilic first polymer inserted in a cell surface membrane and cross-linked to said micelles through said second polymer having multiple thiol groups.

7. The cross-linked polymer according to any of claims 1-6, further comprising one or more biologically active substances, such as for modulating immune response, bound to maleimide and/or thiol groups of said cross-linked polymer.

8. The cross-linked polymer according to any of claims 1-7, having a molecular cut-off between 190 kDa and 900 kDa.

9. A cell or cell aggregate encapsulated with a cross-linked polymer according to any of claims 6-8.

10. The cell or cell aggregate according to claim 9, wherein said cell or cell aggregate is selected from pancreatic islet cells, pancreatic islets, endothelial cells, erythrocytes, primary cells, stem cells (e.g. mesenchymal stem cells (MSC), embryonic stem (ES) cells, induced pluripotent stem (iPS) cells) or differentiated cells from stem cells.

11. Method for preparing a cross-linked polymer for surface modification of a cell or cell aggregate, comprising the steps

i. bringing said cell or cell aggregate in contact with an aqueous solution of a maleimide-conjugated amphiphilic first polymer in a concentration above the critical micelle concentration of said first polymer;

ii. allowing said first polymer to insert into the cell surface of said cell or cell aggregate and form an equilibrium state of distribution between micellar polymer and cell-bound polymer;

iii. adding a second polymer having multiple thiol groups; and

iv. allowing said first and said second polymer to react to form cross-linkage between said first and said second polymer.

12. The method according to claim 11, further comprising the steps

v. adding a a third polymer having multiple maleimide groups;

vi. allowing said second and said third polymers to react to form cross-linkage between said second and said third polymer; and optionally

vii. repeating steps i-vi a desired number of times.

13. The method according to claim 12, further comprising binding one or more

biologically active substances, such as for modulating immune response, to maleimide and/or thiol groups of said cross-linked polymer.

14. The method according to any of claims 10-13, wherein said cross-linked polymer is a cross-linked polymer according to any of claims 3-5.

15. The method according to any of claims 10-14, wherein said cell or cell aggregate is selected from pancreatic cells, pancreatic islets, endothelial cells, erythrocytes

16. A kit of parts for preparing the cross-linked polymer according to any of claims 1-9 or performing the method according to any of claims 10-15, comprising said first, second and third polymer, and optionally instructions for use, suitable buffers and/or suitable reaction vials.

17. The cell according to claim 9 or 10, for use in medicine.

18. A pancreatic islet cell according to claim 9, for use in a method for treatment or

prevention of Type 1 diabetes and/or Type 2 diabetes.