AFFINITY ADSORBENTS FOR I MUNOGLOBULINS
Field of the Invention
This invention relates to compounds and their use as affinity ligands. Background to the Invention Immunoglobulins are glycoproteins consisting of Y-shaped building block having two identical light chains (of weight -25 kDa) and two identical heavy chains (of weight ~50 kDa). Each chain is composed of constant and variable regions which are divided into individual domains. Five distinct classes of immunoglobulin (corresponding to heavy chain isotypes: a (IgA), d (IgD), e (IgE), m (IgM) and g (IgG)) are recognised in higher mammals, the molecules differing in size, charge, biological properties, amino acid composition and carbohydrate content. The light chains may take the form of one of two isotypes, termed kappa and lambda.
Immunoglobulins are widely used in diagnostic, therapeutic, preparative and analytical applications. Interest in such proteins stems from developments in hybridoma technology, alone or combined with genetic engineering, and also in the use of polyclonal preparations such as human IGIV. There is also an increasing interest in the use of immunoglobulin classes, subclasses and fragments, since they may differ both chemically and functionally from the whole polyclonal preparation or the intact molecule. For example, Fab fragments offer a number of advantages over the intact protein for imaging and therapy, since they exhibit rapid pharmacokinetics and reduced non-specific binding that is often associated with the glycosylated Fc portion of the immunoglobulin.
A concern in the administration of immunoglobulins in vivo is the presence of contaminants in the preparation. Such contaminants include DNA, viruses, pyrogens and leachates from separation media. High purity immunoglobulin preparations must comply with GMP procedures and strict FDA guidelines. The requirement for high purity coupled with the need for more effective biopharmaceuticals and research tools has led to the development of many new purification techniques, especially those based on affinity interactions.
The most commonly used affinity adsorbents for immunoglobulin purification are immobilised bacterial surface proteins. These natural proteins
typically interact with the Fc portion of IgG (e.g. Proteins A and G) or immunoglobulin light chains (e.g. Protein L). Protein L, obtainable from Peptostreptococcυs magnus, is often used because it binds regardless of heavy chain class. However, natural protein affinity ligands are generally poor adsorbents due to difficulties in immobilisation, high cost, leakage and poor stability.
Other purification techniques such as affinity chromatography using protein A or immunoaffinity chromatography have been used, but since these methodologies also involve the use of natural proteins, they suffer from the disadvantages described above.
There exists the need for ligands which are able to mimic the selectivity and affinity of natural protein ligands but which are also stable, inexpensive and readily ligated. Summary of the Invention
The present invention is based on the discovery of a particular class of compounds which addresses the problem described supra.
A first aspect of the invention is the use of a compound of formula (I)
wherein
R1 and R2 are the same or different and are each optionally substituted alkyl or aryl; and
R3 is a solid support optionally attached via a spacer; for the affinity binding of an immunoglobulin or fragment thereof. Preferably, the affinity of the compound for the immunoglobulin or fragment is at least 50% of that of Protein L.
A second aspect of the invention is a compound of formula (I), wherein any substitutent on R1 or R2 is selected from aryl, alkyl, OH, NH2, COOH and CONH2, and R3 may alternatively be a functional atom or group, optionally attached via a spacer, which is capable of reaction with a solid support. Another aspect of the invention is the use of a compound of the invention as an affinity ligand.
Compounds of the invention may be used as affinity ligands for immunoglobulin separation, isolation, characterisation, identification, quantification and purification. In particular, they may act as affinity ligands for immunoglobulins and fragments thereof, e.g. Fabs. The compounds may be able to mimic the affinity and selectivity of natural ligands such as Protein L, without suffering from the limitations of such ligands. Furthermore, the compounds may bind both kappa and lambda light chains, and may act as affinity ligands for a broader range of IgG types (e.g. mouse, bovine and human IgG) relative to Protein L. Description of the Invention
The term "alkyl" as used herein refers to a straight or branched chain alkyl moiety having from one to six carbon atoms, and includes, for example, methyl, ethyl, propyl, isopropyl, butyl, tert-butyl, pentyl, hexyl and the like. "C 6 alkyl" has the same meaning.
The term "aryl" as used herein refers to aromatic ring systems comprising six to ten ring atoms, and polycyclic ring systems having two or more cyclic rings at least one of which is aromatic. This term includes, for example, phenyl and naphthyl. The term "halogen" as used herein refers to F, Cl, Br or I.
Certain compounds and combinations of substituents are preferred; in particular, see the subclaims.
The groups -NHR1 and -NHR2 may be obtained by substitution of a triazine ring using any suitable amine compound. Preferably, -NHR1 and -NHR2 derive from amines such as alanine, 1 ,5-diaminopentane, tyramine, -xylylenediamine, phenethylamine, isoamylamine, 4-aminobutanoic acid,
4-aminobenzamide, 1 -amino-2-propanol, 2-methylbutylamine or 4-aminobutyramide,
With regard to formula (I), R1 and/or R2 is preferably alkyl, -alkyl-C(0)OH, -alkyl-NH2, -alkyl-OH, -aryl-C(O)NH2, -alkyl-aryl, -alkyl-aryl-OH or -alkyl-aryl- alkyl-NH2. More preferably, R1 and/or R2 is 1-carboxyethyl, 5-aminopentyl, (4- hydroxyphenyl)methyl, (4-aminomethyl-phenyl)methyl, 2-phenylethyl, 3-methylbutyl, 3-carboxypropyl, 4-amidophenyl, 2-hydroxypropyl, 2-carboxyethyl, 2-methylbutyl or 3-amidopropyl. R3 is preferably halogen (e.g. Cl) or a solid substrate optionally attached via a spacer. More particularly, preferred compounds of the invention include those wherein R1 is 4-amidophenyl and R2 is 3-carboxypropyl; R1 is 2-phenylethyl and R2 is 2-hydroxypropyl; or R1 is (4-hydroxyphenyl)methyl and R2 is 3-amidopropyl. In each case, R3 is preferably halogen (e.g. Cl) or a solid substrate.
Compounds of the invention may be chiral. They may be in the form of a single enantiomer or diastereomer, or a racemate.
A compound of the invention may be in a protected amino, protected hydroxy or protected carboxy form. The terms "protected amino", "protected hydroxy" and "protected carboxy" as used herein refer to amino, hydroxy and carboxy groups which are protected in a manner familiar to those skilled in the art. For example, an amino group can be protected by a benzyloxycarbonyl, tert- butoxycarbonyl, acetyl or like group, or in the form of a phthalimido or like group. A carboxyl group can be protected in the form of a readily cleavable ester such as the methyl, ethyl, benzyl or tert-butyl ester. A hydroxy group can be protected by an alkyl or like group. Some compounds of the formula may exist in the form of solvates, for example hydrates, which also fall within the scope of the present invention.
Compounds of the invention may be in the form of salts, for example, addition salts of inorganic or organic acids. Salts may also be formed with inorganic bases. Salts may be prepared by reacting the compound with a suitable acid or base in a conventional manner.
R3 may be any suitable substrate known in the art, preferably agarose. The substrate may be in any suitable form, for example beads, and may be
linked to the ligand via a spacer. The presence of a spacer may hold the ligand away from the support and reduce steric hindrance. Preferably, a spacer has a length of about 1 to about 10 atoms, more preferably about 6 atoms. The spacer may be attached using any suitable method known in the art. Preferably, the substrate or spacer comprises an amine group which can substitute at the R3 position of formula (I).
A compound of the invention may be prepared by any suitable method known in the art and/or by the following processes:
Scheme 1
Scheme 4
Scheme 3 depicts the synthesis of an intermediate compound, suitable for use in Scheme 4. In Scheme 4, the dashed line represents a direct coupling process, the solid lines representing a semi-solution process. R, and R
2 represent -NHR
1 and -NHR
2 of formula (I) respectively. It will be understood that the processes detailed above are solely for the purpose of illustrating the invention and should not be construed as limiting. A process utilising similar or analogous reagents and/or conditions known to one skilled in the art may also be used to obtain a compound of the invention.
Any mixtures of final products or intermediates obtained can be separated on the basis of the physico-chemical differences of the constituents, in a known manner, into the pure final products or intermediates, for example by chromatography, distillation, fractional crystallisation, or by the formation of a salt if appropriate or possible under the circumstances.
The following Examples illustrate the invention. In the Examples, all reagents had a purity of at least 98%. Solvents were pro-analysis and obtained from Sigma-Aldrich, Fisher, Merck or Acros. Human
IgG (395% pure), bovine IgG and mouse IgG (reagent grade) were obtained from
Sigma-Aldrich. Human Fab (398% pure) was obtained from Cappel, ICN. Human
F(ab')2 and human Fc ( 95% pure) were obtained from Calbiochem. Protein L (immobilised on agarose) was obtained from Pierce. Anti-human IgG (Fab and
Fc specific), anti-human lambda light chains (both alkaline phosphatase conjugates) and anti-human kappa light chains (horseradish peroxidase conjugate) were purchased from Sigma-Aldrich.
Silica gel 60 (0.040-0.063mm) was obtained from Merck. TLC assays were performed using Polygram SilG/UV 254 silica plates (obtained from Macherey-Nagel) and Silica gel 60 F254 (obtained from Merck).
1H and 13C NMR spectroscopy was perfomed using a Jeol JNM Lambda LA400 FT NMR Spectrometer. Other spectrometric readings were performed using either a HITACHI U-2000 or Shimadzu UV-160A spectrometer.
I
8
Example 1 : 4-r4-f4-Carbamoyl-phenylamino)-6-chloro-ri.3.51triazin-2- ylaminolbutyric acid (B in Scheme 1) 4-(4,6-Dichloro-[1,3,5]triazin-2-ylamino)benzamide (A)
3.68g of cyanuric chloride (20mmol) was dissolved in acetone (90ml) and ice water (20ml) at 0°C. To this, a mixture of 2.72g of 4-aminobenzamide (20mmol) dissolved in acetone (30ml) and water (60ml), and 1.68g of NaHC03 (20mmol) in water (30ml) was added dropwise. The reaction mixture was stirred for two hours at 0°C. The reaction was monitored by TLC (solvent system: ethyl acetate/methanol 95:5) and stopped when no cyanuric chloride was detected. The resultant yellowish solid product was filtered off, washed with hot water and heptane and dried in vacuo over solid P2O5 (phosphorus pentoxide). Yield: 90% (5.15g, 18.1mmol). R 0.6 (EtOAc/MeOH 95:5). H-NMR (400MHz, [D6]DMSO, 25°C): d 7.34 (s, 1 H, NH), d 7.65, 7.67 (d, 2H, ArH), d 7.87, 7.89 (d, 2H, ArH), d 7.92 (s, 1H, NH), d 11.32 (s, 1H, NH). 13C-NMR (500MHz, [D6]DMSO, 25°C): d 120.16, 128.42 (ArC), d 129.63, 140.11 (ArC, quaternary), d 166.88 (CONH2), d 166.18, 167.69, 169.41 (Ctriazine). MS (El, CONCEPT) calculated for C10H7CI2N5O: 283.00, found 283.00. MS (ESI, Q-tof) calculated for C10H7CI2N5O (M+H)+: 284.0, found 284.0. Melting point found: m.p. >250°C. 4-[4-(4-Carbamoyl-phenylamino)-6-chloro-[1,3,5]triazin-2-ylamino]butyric acid (B)
To a solution of 1.98g of dichloride (A) (7mmol) in acetone (100ml) and water (15ml), a mixture of 0.72g of 4-aminobutyric acid (7mmol) in water (30ml) and 0.58g of NaHC03 (7mmol) in water (30ml) was added. The reaction was carried out at 45-50°C with constant stirring for 24h, monitored by TLC (solvent system: ethyl acetate/methanol 95:5) and stopped when no 4-aminobutyric acid was detected by ninhydrin coloration test. The white precipitate formed was filtered off, washed with water and dried in vacuo over solid P205. The white solid was dissolved in an aqueous solution K2CO35%(w/v) and washed four times with ethylacetate. The aqueous phase was neutralised with HCI (5M) and the resultant white precipitate filtered, washed with water and dried in vacuo over solid P205. Yield: 36% (0.87g, 2.5mmol). R, 0.4 (EtOAc/MeOH 95:5). 1H-NMR (400MHz, [D6]DMSO, 25°C): d 1.71 -1.82 (m, 2H, NHCH2CH2CH2COOH), d 2.25-
2.31 (m, 2H, NHCH2CH2CH2COOH), d 3.26-3.29 (t, 2H, NHCtf2CH2CH2COOH), d 7.20 (s, 1H, NH), d 7.76-7.84 (m, 4H, ArH and 1 H, NH), d 8.16, 8.24 (s, 2H, - CONH2), d 10.11, 10.23 (s, 1H, -COOH). 13C-NMR (400MHz, [D6]DMSO, 25°C): d 24.31, 24.55, 31.38 (aliphatic CH2), d 119.47, 128.32 (ArC), d 128.61, 128.67 (ArC quaternary), d 142.04 (CONH2), d 165.80 (COOH), d 168.30, 168.36, 174.65 (Ctriazine). MS (LSIMS, CONCEPT) calculated for C14H15CIN603(M+H)+: 351.09, found 351.0. MS (ESI, CONCEPT) calculated for C14H15CIN603 (M+Na)+: 373.09, found 373.1. Melting point found: 218-219°C. Example 2: 1 -(4-Chloro-6-phenethylamino-f1.3.5.triazin-2-ylamino.-propan- 2-ol (D in Scheme 2)
(4,6-Dichloro-[1 ,3,5]triazin-2-yl)-phenethylamine (C)
3.68g of cyanuric chloride (20mmol) was dissolved in acetone (90ml) and ice water (20ml) at 0°C. To this, 2.5ml of phenethylamine (20mmol) in acetone (30ml) and 1.68g of NaHC03 (20mmol) in water (40ml) were added dropwise. The reaction mixture was stirred for two hours at 0°C, monitored by TLC (solvent system: ethyl acetate/hexane 1:1) and stopped when no cyanuric chloride was detected. The resultant white solid product was filtered off, washed with water and dried in vacuo over solid P2O5. Yield: 68% (3.65g, 13.6mmol). R 0.7 (EtOAc/Hexane 1:1). 1H-NMR (400MHz, [D6]DMSO, 25°C): d2.79, 2.81, 2.83 (t, 2H, NHCH2CH2C6H5), d 3.48, 3.50, 3.51 , 3.53 (q, 2H, NHCH2), d 7.17-7.29 (m, 5H, ArH), d 9.19, 9.20, 9.22 (t, 1H, NHCH2CH2C6H5). 13C-NMR (500MHz, [D6]DMSO, 25°C): d 34.15, 42.29 (aliphatic CH2), d 126.37, 128.42, 128.88 (ArC), d 138.74 (ArC quaternary), d 165.29, 168.46, 169.47 (Ctriazine). MS (El, CONCEPT) calculated for C^H^C^N.,: 268.03, found 268.03. MS (ESI, Q-tof) calculated for C11H10CI2N4(M+H)+: 269.03, found269.0. Melting pointfound: 134- 135°C (as suggested by Suter and Zutter, 1965). 1 -(4-Chloro-6-phenethylamino-[1 ,3,5]triazin-2-ylamino)-propan-2-ol (D)
To a solution of 1.87g of dichloride (C) (7mmol) in acetone (35ml), a mixture of 0.54ml of 1-amino-2-propanol (7mmol) in water (5ml) and 0.58g of NaHC03 (7mmol) in water (10ml), was added. The reaction was carried out at 45-50°C with constant stirring for 4h, monitored by TLC (solvent system: ethyl acetate/hexane 1:1) and stopped when no 1-amino-2-propanol was detected by
ninhydrin. The resultant white solid product was filtered off and washed with water. A precipitate was formed in the filtrate and added to the filtrate. The white solid product was dried in vacuo over solid P205 and further purified by column chromatography (silica, solvent system EtOAc/Heptane 1:1 to separate the contaminants and acetone for elution of purified product). The solvent was evaporated and the white solid dried in vacuo over solid P205. Yield: 83% (1.8g, 5.8mmol). R, 0.2 (EtOAc/Hexane 1 :1). 1H-NMR (400MHz, [D6]DMSO, 25°C): d 1.01 , 1.03, 1.04 (t, 3H, C/_/3), d 2.78, 2.79, 2.81 (t, 2H, NHCH2CH2C6H5), d 3.11- 3.22 (m, 2H, NHCH2CH(COCH3)), d 3.41-3.46 (m, 2H, NHCH2 CH2), d 3.71-3.80 (m, 1 H, CH2CHOH) d 4.63, 4.64, 4.66, 4.68 (q, 1 H, NHCH2CH2C6H5), d 7.16-7.29 (m, 5H, ArH), d 7.54-7.90 (m, 2H, CH2CHOH and NHCH2CH(COCH3)).13C-NMR (400MHz, [D6]DMSO, 25°C): d 21.40 (CH3), d 35.18, 42.17, 48.14 (aliphatic CH2), d 65.27 (CHOH), d 126.62, 128.80, 129.09 (ArC), d 139.64 (ArC quaternary), d 165.61, 165.86, 168.04 (Ctriazine). MS (El, CONCEPT) calculated for C14H18CIN50: 307.12, found 307.1. MS (ESI, Q-tof) calculated for C14H18CIN50 (M+Na)+: 330.12, found 330.1. Melting point found: 188-189°C. Example 3: 4-r2-(4.6-Dichloro-ri.3.5]triazin-2-ylamino)ethyn-phenol (H in Scheme 3) (4-te/ -Butoxy-phenyl)-acetonitrile (E) 3.25g of 4-hydroxybenzonitrile (27mmol) was dissolved in dried ether
(15ml) under N2 atmosphere. To this mixture, 17ml of tert- butyltrichloroacetimidate (104mmol), in dried ether (10ml) and a catalytic amount of BF3.Et20 (475ml) were successively added dropwise at 0°C. The mixture was stirred at room temperature for 24h and monitored by TLC (solvent system: dichloromethane). At the end of the reaction, solid NaHC03 was added and stirring continued for 30min. Cyclohexane (30ml) was added and the mixture filtered. The filtrate was washed 3xNaOH 10% (w/v), 1x brine and dried over MgSO4. The solvent was evaporated, a brown syrup obtained and left drying in vacuo over solid P2O5. A column chromatography was performed (silica, solvent system: dichloromethane) affording a yellow syrup (E). Yield: 32% (1.63g, 8.6mmol). R, 0.6 (dichloromethane). 1H-NMR (400MHz, [D6] CDCI3, 25°C): d 1.31 -1.37 (m, 9H, -OtyCH s), d 3.69 (s, 2H, -CH2CN), d 7.00-7.04 (m, 2H, ArH),
d 7.23-7.26 (m, 2H, ArH). 13C-NMR (400MHz, [D6] CDCI3, 25°C): d 22.97 (- CH2CN), d 28.77 (C(CH3)3), d 78.81 (C(CH3)3), d 118.06 (CN), d 124.50, 124.61 (ArC), d 128.50, 155.65 (ArC quaternary). 2-(4-tert-Butoxy-phenyl)-ethylamine (F) 0.47g of LiAIH4 (12.4mmol,) was suspended in dried ether at 0°C under
N2 atmosphere.1.57g of (4-fetf-Butoxy-phenyl)-acetonitrile (E) (8.3mmol) in dried ether was added dropwise. The reaction was left for 5h at 40°C. The control was made by TLC (solvent system: dichloromethane/methanol 9:1). When the reaction was complete, water (0.27ml) was added dropwise, followed by NaOH 10%(w/v) (0.27ml) and then water (0.54ml). The reaction mixture was filtered and the solid washed with ether (3x). The filtrate was recovered, washed with NaOH 10%(w/v) and brine and then extracted 3xKHS04 1 N. The aqueous phase was washed once with ether and then solid NaHC03 added until pH»9. The basic layer was extracted 3x with ether. The resultant organic layer was washed with brine, dried over MgS04, filtered and concentrated under vacuum. A yellowish oil was obtained and left drying in vacuo over solid P205. Yield: 29% (2.4mmol, 0.5g,). R 0.25 (CH2CI2/MeOH 9:1 ). 1H-NMR (400MHz, [D6] CDCI3, 25°C): d 1.27- 1.33 (m, 9H, -OC^/^), d 2.69, 2.71 , 2.73 (t, 2H, CH2), d 2.93, 2.95, 2.97 (t, 2H, NCCH2), d 6.91 , 6.93 (d, 2H, ArH), d 7.07, 7.09 (d, 2H, ArH), d 7.26, 7.27 (s, 2H, NH2) . 13C-NMR (400MHz, [D6] CDCI3, 25°C): d 28.83 (C(CH3)3), d 39.44 (CH2), d 43.61 (NCCH2), d 78.17 (C(CH3)3), d 124.19, 129.11 (ArC), d 134.63, 153.62 (ArC quaternary). p^-fe/ -Butoxy-pheny -ethyll-t^β-dichloro-tl.S.δltriazin^-ylJ-amine iG) 0.46g of cyanuric chloride (2.5mmol) was dissolved in acetone (5ml) and ice water (5ml) at 0°C. To this, 0.5g of 2-(4-tetf-Butoxy-phenyl)-ethylamine (F) (2.5mmol) in acetone (5ml) and 0.23g of NaHCO3 (2.75mmol) in water (5ml) were added dropwise. The reaction took 2h and was monitored by TLC (solvent system: EtOAc/Heptane 1 :1 ). The resultant yellowish solid product was filtered off, washed with water and dried in vacuo over solid P2O5. A column chromatography was performed (silica, solvent system: EtO Ac/Heptane 2:8). The contaminant-free samples were pooled and concentrated under vacuum, affording a white solid (G). The samples containing the product and small
amount of impurities were also pooled, concentrated under vacuum and recrystallized in heptane giving a white solid (G). Total yield: 50% (0.42g, 1.25mmol). R,0.3 (EtOAc/Heptane 2:8). 1H-NMR (400MHz, [D6] CDCI3, 25°C): d 1.27-1.35 (m, 9H, -OC(CH3,3), d 2.82, 2.84, 2.85 (t, 2H, CH2), d 3.69, 3.71 , 3.73, 3.74 (q, 2H, NCCH2), d 5.98 (s, 1 H, NH), d 6.89-6.92 (m, 2H, ArH), d 7.05, 7.07 (d, 2H, ArH) . 13C-NMR (400MHz, [D6] CDCI3, 25°C): d 28.81 (C(CH3)3), d 34.41 (CH2), d 42.59 (NCCH2), d 78.47 (C(CH3)3), d 124.44, 129.37 (ArC), d 132.23, 154.32 (ArC quaternary), d 165.75, 169.75, 170.99 © triazine). 4-[2-(4,6-Dichloro-[1,3,5]triazin-2-ylamino)ethyl]-phenol (H) To0.17g of[2-(4-tetf-Butoxy-phenyl)-ethyl]-(4,6-dichloro-[1 ,3,5]triazin-2- yl)-amine (G) (O.δmmol) dissolved in dichloromethane (2ml), 0.23ml of TFA were added (5eq., 2.98mmol). The reaction mixture was stirred for 2h at room temperature and monitored by TLC (solvent system: EtOAc/Heptane 1:1). To recover the yellow precipitate, NaHCO310% (w/v) was added, at0°C, until pH»7. Dichloromethane was evaporated under vacuum and the aqueous mixture containing the product filtered. A column chromatography was performed (silica, solvent system: EtOAc/Heptane 1 :1 ) affording a white solid (H) dried in vacuo over solid P205. Yield: 70% (0.12g, 0.42mmol). Rf0.4 (EtOAc/Heptane 1:1 ). 1H- NMR (400MHz, [D6] DMSO, 25°C): d 2.67, 2.69, 2.70 (t, 2H, CH2), d 3.40, 3.42, 3.44, 3.45 (q, 2H, HNCH2), d 6.64, 6.67 (d, 2H, ArH), d 6.98, 7.00 (d, 2H, ArH), d 9.13-9.18 (m, 2H, NH and OH). 13C-NMR (400MHz, [D6] DMSO, 25°C): d 33.25 (CH2), d 42.55 (HNCH2), d 115.10 (ArC), d 128.57 (ArC, quaternary), d 129.55 (ArC), d 155.77 (ArC quaternary), d 165.12, 168.33, 169.34 ©triazine). MS (ESI, QUATTRO) calculated for C14H18CIN50: 284.02, found 283.9. Melting point found: 168-170°C.
Example 4: Coupling of compounds B and D to aminated agarose
To 1g of moisty aminated agarose (»24mmoI/g), was added a solution containing 42mg of B (5 molar equivalent, 0.12mmol) and 12mg of NaHC03 (6 molar equivalent, 0.15mmol) in 5ml of 50%(v/v) DMF:H20. A solution of 32mg of D (5 molar equivalent, 0.12mmol) and 2mg of NaHC03 in 5ml of 50%(v/v) DMF:H20 was prepared and mixed with 1 g of moisty aminated agarose (»24mmol/g). The coupling reactions were carried out at 85°C for 72h. At the end
of the reaction the agarose beads were sequentialy washed with DMF:water (1 : 1 ; 1:0, 1:1, 0:1). The resins were then packed in 1ml disposable columns and washed 2x with regeneration buffer followed by water, and stored in a solution of ethanol 20% (v/v) at 0-4°C. Ligands B and D are also referred to herein as "5/9" and "8/7" respectively. Example 5: Alternative, semi-solution coupling process
Aminated agarose (1g, 24mmol/g) or aminated agarose with a -OCH2CH(OH)CH2NH2 spacer (1g, 20mmol/g) was added to a solution of DMF:H2050%(v/v) (5ml) containing 2 molar equivalent (48mmol or 40mmol) of compounds A, C or H and 2 molar equivalent (48mmol or 40 mmol) of NaHC03. The reaction was carried out at 30°C for 24h. At the end of the reaction the agarose beads were sequentially washed with DMF:water (1:1; 1:0, 1:1, 0:1).
For the A- and C-substituted resins, the second chloride (R2 in Scheme 3) was displaced by reaction with solutions containing, respectively, 4- aminobutanoic acid and NaHCO3 (0.12mmol/0.1 mmol, 5eq. of each) or 1 -amino-
2-propanol and NaHCO3 (5 molar equivalent of each) in water (5ml), at 85°C for
72h. This resulted in the formation of B- and D-substituted resins respectively.
The H-substituted resin (1g) was reacted at 85°C for 72h with different compounds in order to replace the second chloride. In each case, a 5 molar equivalent of an aqueous solution (5ml) of the one of the following reagents was used:
L-alanine ("3/1");
1,5-diaminopentane ("3/2"); 4-aminobutanoic acid ("3/7");
1-amino-2-propanol ("3/9"); β-alanine (with 5 molar equivalent of NaHC03) ("3/10");
2-methylbutylamine ("3/11"); and
4-aminobutyramide (with 10 molar equivalent of NaHCO3) ("3/12"). [The term alongside each reagent refers to the corresponding resulting ligand. These ligands are referred to collectively as "3/x"]
At the end of the reaction the agarose beads were washed with distilled water, packed in 1ml disposable columns and washed 2x with regeneration buffer followed by water, and stored in a solution of ethanol 20% (v/v) at 0-4°C. Example 6: Determination of densities of ligands 8/7 and 5/9
The ligand densities of the resins of Examples 4 and 5 were determined by solubilisation of the immobilised ligands. Immobilised ligands (30mg of moist gel) were hydrolysed in HCI 5M (0.3ml) at 60°C for 10min. On cooling, ethanol (3.7ml) was added to the hydrolysed ligand and the absorbance read at the characteristic wavelength calculated for each ligand (8/7 - e (296nm)=26.3E3 l(mol.cm); 5/9 - e (263nm)=3.7E3 l(mol.cm); and 3/x - e (280nm)=3.47E3 l(mol.cm)) against a solution of agarose submitted to the same treatment. The extinction coefficient, e, for each ligand was determined by constructing a standard curve with the measurements of the absorbance read at the characteristic wavelength for different free ligand concentration solutions.
The densities of ligands 8/7 and 5/9 are shown in Table 1.
Example 7: Assessment of binding to human IgG and its fragments, by affinity chromatography
The derivatised resins of Example 5 were assessed for their affinity for human Fab and Fc fragments using affinity chromatography. The affinities were compared with that of Protein L.
8/7-, 5/9- and 3/x-derivatised resins were packed into 4ml columns (0.8x
6cm) to a final volume of about 0.5ml (0.5g of gel). The resulting matrices were washed with 2x3ml regeneration buffer (NaOH 0.1 M in 30% isopropanol (v/v)) and then distilled water, to bring the pH value to neutral. Immobilised Protein L
was also packed as described above. The resins were equilibrated with 10ml of equilibration buffer (PBS, 10mM phosphate, 150mM NaCI, pH7.4). Protein (human IgG, hFab, hF(ab')2 or hFc) was reconstituted to 0.5mg/ml in equilibration buffer and the absorbance at 280nm (A280nm) measured. Protein solution (1 ml) was loaded on to each column. The columns were washed with equilibration buffer until the absorbance of the samples at 280nm was less than or equal to 0.005. Bound protein was eluted using elution buffer (Glycine-HCI 0.1M, pH 2.0) and neutralized by addition of 90ml of 1 M Tris.HCI, pH9. After elution, the columns were washed with regeneration buffer, followed by distilled water, and stored at 0-4° C in 20%(v/v) ethanol. Immobilised Protein L was regenerated using elution buffer and stored in an aqueous sodium azide solution (0.02% (w/v)).
The results are shown in Figures 1 and 2. It is evident that the ligands showed a higher affinity for human Fab over Fc. The introduction of a spacer resulted in an increase in the amount of Fab bound but also decreased specificity for this fragment. The exceptions to this were ligands 8/7, 5/9 and 3/12, which had greater yields and maintained their specificity. The affinities of these particular ligands (bound via a spacer) are summarised in Figure 3. Example 8: Assessment of the affinity and selectivity of ligands 8/7, 5/9 and 3/12 to kappa and lambda light chains, and IgG from different sources
Immobilised ligands 8/7, 5/9, 3/12 and (for comparison) Protein L were packed into 4ml columns (0.8x 6cm) to a final volume of about 0.2ml. The resulting matrices were washed with 1 ml regeneration buffer and then distilled water, to bring the pH value to neutral. The resins were then equilibrated with 10ml of equilibration buffer 0.2ml of a solution containing either human myeloma IgG! kappa or lambda light chain was loaded on to each column. The columns were washed with a total of 17 column volumes of equilibration buffer (2x500ml; 4x250ml; 3x500ml) and fractions collected. Bound protein was eluted using elution buffer and two fractions collected (1x650ml; 1x500ml) to which 65 and 50ml of Tris.HCI (1M, pH9.0) were added, respectively. After elution, the columns were washed with regeneration buffer, followed by distilled water and stored at 0-4° C in 20%(v/v) ethanol.
Quantitative ELISA
The amount of each protein collected from was determined using a quantitative ELISA.
Antibody anti-human kappa light chain was diluted 1 : 1 ,000 in PBS-Tween and anti-human lambda light chain diluted 1 :6,500 in PBS-Tween. The substrate solution for the kappa chain was an oPD solution (1.85mM oPD, 5mM NaH2P04,
2mM citric acid and 0.04% H202) (100ml) left incubating for 5min in the absence of light. The reaction was stopped by the addition of 2M H2S04 (50ml) and the absorbance read at 490nm. The substrate solution (100 ml) for the lambda chain was 1 mg/ml of pNPD in substrate buffer-0.1M diethanolamine HCI, at ph 9.8.
Calibration curves to correlate hlgG., kappa (mg/ml) with A4g0nm were constructed using hlgG, kappa standard solutions from 5-0.5ng/ml. The reactions were stopped by the addition of 2M sulphuric acid (50ml). The absorbance was read at 490nm. Calibration curves to correlate hlgG, lambda (mg/ml) with A405nm were constructed using hlgG., lambda standard solutions from 10-1 ng/rril.
As Figure 4 shows, the ligands of the invention bind with high affinity to human IgG from both kappa and lambda light chains. Protein L, effectively only binds to kappa light chains.
Assessment of affinity for IgG isolated from different sources by affinity chromatography
The procedure described in Example 7 was followed except that 0.5mg/ml solutions of human, bovine or mouse IgG were loaded on to each column.
The results are shown in Figure 5. The ligands of the invention bind have a high affinity for bovine, mouse and human IgG. As previous work has already shown, Protein L does not bind bovine IgG.