FIELD OF THE INVENTION

The present invention relates to a method for isolation or purification of immunoglobuhns from various raw materials and solid phase matrices therefor

BACKGROUND OF THE INVENTION

Immunoglobuhns - or antibodies - constitute a very important class of protems which are present m various body fluids of mammals, birds and fish functioning as protective agents of the animal against substances, bacteria and virus challenging the animal Immunoglobuhns are typically present in animal blood, milk, and saliva as well as other body fluids and secretions

The biological activity, which the immunoglobuhns possess, is today exploited in a range of different applications in the human and veterinary diagnostic health care and therapeutic sector

Diagnostics

.Antibodies have for many years been applied as an important analytic tool in connection with detection and quantification of a large variety of substances of relevance m the diagnosis of diseases and are increasingly important in areas such as quabty control of food products, environmental control, drugs of abuse, and monitoring and control of industrial processes

For these purposes, the desired antibodies can be produced by hyper-immunisation of suitable host animals, such as rabbits and sheep, or, alternatively, by producing monoclonal antibodies in hybridoma cell cultures

Following the primary production of the antibodies in either a host animal or m cell culture, the antibody is typically isolated from the bulk of other substances in the raw material by some sort of isolation process. This is necessary in order to avoid interference from these other substances with the antibody activity m the analytical application

Health care and therapeutic applications

Passive immunisation by intramuscularly injection of immunoglobulin concentrates is a well- known application for temporary protection against infectious diseases, which is typically applied when people are travelling from one part of the world to the other The success of this

kind of treatment on humans is now being followed up in the veterinary field where passive immunisation of new born cattle, horses, pigs and chickens are being apphed and developed to enhance the survival rate of these animals durmg their first weeks of hve An important issue in this field is of course the cost of such a treatment, which to a high degree depends on the cost of producmg the immunoglobulin concentrate

Isolates of animal immunoglobuhns, e g from bovine milk, are also under investigation as an oral health care or even therapeutic product to avoid or treat gastrointestinal infections, e g in AIDS patients For such applications both the degree of purity of the product as well as the cost is of major importance

A more sophisticated apphcation of antibodies for therapeutic use is based on so called "drug- targeting" where very potent drugs are covalently linked to antibodies with specific binding affinities towards specific cells the human organism e g cancer colls This technique ensures that the drug is concentrated on the diseased cells giving maximal effect of the drug without the severe side-effects that frequently occurs when using chemotherapy For such purposes the antibodies have to be very carefully controlled and of high purity, and the typical way of performing the primary production are either by producmg monoclonal antibodies in hybridoma cell culture or by fermenting genetically engineered bacteria, e g E coli

Isolation of immunoglobulins

Ul the above mentioned applications of immunoglobuhns requires some sort of isolation of the antibody from the crude raw material, but each kind of apphcation has its own very varying demands with respect to the final purity and allowable cost of the antibody product

Generally, there exists a very broad range of different methods available for isolation of immunoglobuhns giving a very broad range of final purities, yields and cost of the product

Traditional methods for isolation of immunoglobuhns are based on selective reversible precipitation of the protein fraction comprising the immunoglobuhns while leaving other groups of protems in solution Typical precipitation agents being ethanol, polyethylene glycol, lyotropic (anti-chaotropic) salts such as ammonium sulfate and potassium phosphate, and capryhc acid

Typically, these precipitation methods are giving very impure products while at the same time being time consuming and laborious Furthermore, the addition of the precipitating agent to the raw material makes it difficult to use the supernatant for other purposes and creates a disposal problem This is particularly relevant when speaking of large scale purification of immunoglobuhns from, e g , whey and plasma

Ion exchange chromatography is another well known method of protem fractionation frequently used for isolation of immunoglobuhns However, this method is not generally apphcable because of the restraints in ionic strength and pH necessary to ensure efficient binding of the antibody together with the varying lsoelectπc points of different immunoglobuhns

Protein A and Protem G affinity chromatography are very popular and widespread methods for isolation and purification of immunoglobuhns, particularly for isolation of monoclonal antibodies, mainly due to the ease of use and the high purity obtained being popular it is however recognised that Protem A and Protem G poses several problems to the user among which are very high cost, variable binding efficiency of different monoclonal antibodies (particularly mouse IgGi), leakage of Protem A/Protein G into the product, and low stability of the matrix in typical cleaning solutions, e g 1 M sodium hydroxide Each of these drawbacks have its specific conse¬ quence in the individual apphcation, ranging from insignificant to very serious and prohibitive consequences

Hydrophobic chromatography is also a method widely described for isolation of immunoglobuhns, e g in "Apphcation Note 210, BioProcess Media" published by Pharmacia LKB Biotechnology, 1991 In this reference a state of the art product "Phenyl Sepharose High Performance" is described for the purpose of purifying monoclonal antibodies from cell culture supematants As with other hydrophobic matrices employed so far it is necessary to add lyotropic salts to the raw material to make the immunoglobulin bmd efficiently The bound antibody is released from the matrix by lowermg the concentration of lyotropic salt in a continuous or stepwise gradient It is recommended to combine the hydrophobic chromatography with a further step if highly pure product is the object

The disadvantage of this procedure is the necessity to add lyotropic salt to the raw material as this gives a disposal problem and thereby increased cost to the large scale user For other raw materials than cell culture supematants such as whey, plasma, and egg yolk the addition of lyotropic salts to the raw materials would in many instances be prohibitive m large scale applications as the salt would prevent any economically feasible use of the immunoglobulin depleted raw material in combination with the problem of disposing several thousand litres of waste

Thiophihc adsorption chromatography was introduced by J Porath in 1985 (J Porath et al,

FEBS Letters, vol 185, p.306, 1985) as a new chromatographic adsorption principle for isolation of immunoglobuhns In this paper, it is described how divinyl sulfone activated agarose coupled with various hgands comprising a free mercap o- roup show specific binding of immunoglobuhns in the presence of 0 5 M potassium sulfate, I e a lyotropic salt It was postulated that the sulfone

group, from the vinyl sulfone spacer, and the resulting thio-ether m the ligand was a structural necessity to obtain the described specificity and capacity for binding of antibodies It was however later shown that the thio-ether could be replaced by nitrogen or oxygen if the ligand further comprised an aromatic radical (K L Knudsen et al, Analytical Biochemistry, vol 201, p 170, 1992)

Although the matrices described for thiophihc chromatography generally show good performance, they also have a major disadvantage in that it is needed to add lyotropic salts to the raw material to ensure efficient binding of the immunoglobulin, which is a problem for the reasons discussed above

Other thiophihc hgands coupled to epoxy activated agarose have been disclosed in (J Porath et al , Makromol Chem , Makromol Symp , vol 17, p 359, 1988) and (A Schwarz et al , Journal of Chromatography B, vol 664, pp 83-88, 1995), e g 2-mercaptopvrιdme, 2-mercaptopyπmιdιne, and 2-mercaptothιazohne However all these affmity matrices still have inadequate affmity constants to ensure an efficient binding of the antibody without added lyotropic salts

Binding and isolation of proteins and other bwmolecules

WO 96/00735 and WO 96/09116 disclose resms (matrices) for purifying protems and peptides which resms are characterised by the fact that they contain lonizable hgands and/or functionalities which are uncharged at the pH of binding the target protem or peptide, thereby facilitating hydrophobic interactions, and charged at the pH of desorption, thereby disrupting the established hydrophobic mteraction between the resm and the target protem or peptido WO 96/00735 mentions the possibility of coupling 2-mercapto-benzιmιdazole to epoxy-activated Sepharose 6 B The actual ligand concentration is not disclosed, however the coupling was performed with an epoxy-activated Sepharose wherein the content of epoxy-groups is disclosed to be m the range of 1 02- 1 28 mmol/g dry matter

WO 92/16292 discloses a number of different hgands coupled to divmyl sulfone activated agarose and the use of the resulting solid phase matrices for thiophihc adsorption of protems, preferably immunoglobuhns Specifically is mentioned solid phase matrices compπsmg 4-amιno-benzoιc acid as a ligand on a divmyl sulfone activated agarose The adsorption of protems, preferably immunoglobuhns in WO 92/16292, is performed at high concentrations of lvotropic salts l e with an ionic strength of on or above 2 25

BRIEF DESCRIPTION OF THE INVENTION

It has now surprisingly been found that several types of aromatic or heteroaromatic substances linked to a sohd phase matrix can be utilised in a novel method for the isolation and/or purification of immunoglobuhns of different kinds from widely different raw materials with high efficiency and with special advantages with respect to the use of little or no salts, especially lyotropic salts, in the binding process and with respect to the ability to bind a wide range of immunoglobulins Furthermore, these matrices have special advantages with respect to stabihty in NaOH which is especially relevant when the sohd phase matrices are to be regenerated after use

Thus, an object of the present invention is to provide a method for the isolation of immunoglobuhns from a solution containing one or more immunoglobuhns, compnsmg the following operations

a) contactmg a solution containing one or more immunoglobulins and havmg a pH in the range of 2 0 to 10 0 and a total salt content corresponding to a ionic strength of at the most 2 0 with a solid phase matrix of the general formula

M-SP1-L,

wherein M designates the matrix backbone, SP1 designates a spacer, and L designates a ligand compnsmg a mono- or bicychc optionally substituted aromatic or heteroaromatic moiety,

whereby at least a part of the immunoglobuhns becomes bound to the sohd phase matrix,

b) separating the sohd phase matrix having immunoglobuhns bound thereto from the solution

c) optionally washing the sohd phase matrix, and

d) contactmg the solid phase matrix with an eluenl in order to hberate the one or more immunoglobulins from the sohd phase matrix,

with the first proviso that at least two of the criteria (a), (b), and (c) are fulfilled

(a) the sohd phase matrix has a bmdmg efficiency of at least 50% when tested at a pH m the range of 2 0 to 10 0 m the "Standard Immunoglobulin Binding Test" described herein, or

(b) the sohd phase matrix has an average bmdmg efficiency of at least 60% for all of the immunoglobuhns tested in the "Monoclonal Antibody Array Binding Test" when tested at a pH in the range of 2 0 to 10 0, or

(c) the stability of the sohd phase matrix in 1 M NaOH is so that incubation of the matrix in 1 M NaOH in the dark at room temperature for 7 days reduces the binding efficiency at a pH m the range of one pH unit lower than the bmdmg maximum pH value to one pH unit higher than the bmdmg maximum pH value, as determined in the "Standard Immunoglobulin Binding Test" described herem, with less than 25 % compared to a corresponding untreated matrix and

with the second proviso that the molecular weight of the ligand -L is at the most 500 Dalton

The present invention furthermore a solid phase matrix, compnsmg a functionahsed matrix backbone carrying a plurality of functional groups of the following formula

M-SP1-L

wherein M designates the matrix backbone, SP1 designates a spacer, and L designates a hgand compnsmg a mono- or bicychc optionally substituted aromatic or heteroaromatic moiety,

and wherein at least two of the criteria (a), (b), and (c) are fulfilled

(a) the sohd phase matrix has a bmdmg efficiency of at least 50% when tested at a pH m the range of 2 0 to 10 0 m the "Standard Immunoglobulin Binding Test" described herein, or

(b) the sohd phase matrix has a binding efficiency of at least 40% for all of the immunoglobuhns tested m the "Monoclonal Antibody Array Binding Test" when tested at a pH in the range of 2 0 to 10 0, or

(c) the stability of the sohd phase matrix m 1 M NaOH is so that mcubation of the matrix in 1 M NaOH in the dark at room temperature for 7 days reduces the binding efficiency at a pH in the range of one pH unit lower than the bmdmg maximum pH value to one pH unit higher than the bmdmg maximum pH value, as determmed in the "Standard Immunoglobulin Binding Test" described herein, with less than 25 % compared to a corresponding untreated matrix,

with the first proviso that the molecular weight of the ligand -L is at the most 500 Dalton, and with the second proviso that when M is agarose and SPl is derived from vinyl sulfone then L is not 4-amιnobenzoιc acid

which is especially suited for use in the method according to the invention

It has furthermore been found that the matrices mentioned above, wherein the aromatic or heteroaromatic moiety is carrying an acidic group, optionally via a spacer SP2, are equally suited for the isolation and purification of proteins without the need to add lyotropic salts to the protem containing solution (the raw material) and without the need to use large amounts of organic solvents for elution of the bound proteins from the matrix

Thus, the present invention also provides a solid phase matrix, comprising a functionahsed matrix backbone carrying a plurality of functional groups of the following formula

M-SP1-X-A-SP2-ACID

wherein M designates the matrix backbone, SPl designates a spacer, X designates -0-, - S-, or -NH-, A designates a mono- or bicychc optionally substituted aromatic or heteroaromatic moiety, SP2 designates an optional spacer, and ACID designates an acidic group,

with the first proviso that the molecular weight of the ligand -L is at the most 500 Dalton, and with the second proviso that when M is agarose and SPl is derived from vinyl sulfone then L is not 4-amιnobenzoιc acid,

and a method for the isolation of protems from a solution containing one or more of protems, compnsmg the following operations

a) contacting a solution containing one or more proteins havmg a pH m the range of 1 0 to 6 0 and a total salt content corresponding to a ionic strength of at the most 2 0 with a sohd phase matrix according to any of the claims 44-57, whereby at least a part of the protems becomes bound to the sohd phase matrix,

b) separatmg the sohd phase matrix havmg proteins bound thereto from the solution,

c) optionally washmg the sohd phase matrix, and

d) contacting the sohd phase matrix with an eluent in order to liberate one or more of the proteins from the sohd phase matrix, wherein the eluent used comprises less than 10% (v/v) of organic solvents.

DETAILED DESCRIPTION OF THE INVENTION

Isolation of immunoglobulins

In general, the method for isolation of immunoglobuhns may be divided into several steps:

(a) Equilibration of the solid phase matrix

(b) Contacting the solid phase with immunoglobulin solution

(c) Washing the solid phase

(d) Separation of the sohd phase from the solution (d) Elution of the bound immunoglobulin

(e) Regeneration of the solid phase matrix

It may however depend on the specific application whether all steps are performed each time or at all. Thus, the only mandatory steps are the contacting, separation, and the elution steps, while the equilibration, washing, and regeneration steps may or may not be performed according to the specific requirements relevant to the actual application. The type of the separation step depends on the actual set-up (see below).

Equilibration

Before contacting the sohd phase matrix with the immunoglobulin containing solution it is preferred to ensured that both the matrix and the solution are in a condition resulting in the wanted binding of immunoglobulin. In this respect, it may therefore be necessary to adjust parameters such as pH, ionic strength, and temperature and in some instances the addition of substances of different kind to promote binding of immunoglobulins and/or to prevent binding of impurities.

Thus, it is an optional step to perform an equilibration of the sohd phase matrix by washing it with a solution (e.g. a buffer for adjusting pH, ionic strength, etc., or for the introduction of a detergent) bringing the necessary characteristics to the sohd phase.

Contactmg

When the sohd phase matrix is m the form of particles of either spherical or irregular form the contacting of a solution contammg one or more immunoglobuhns may be performed either in a packed bed column or in a fluidised/expanded bed column containing the sohd phase matrix It may also be performed in a simple batch operation where the solid phase matrix is mixed with the solution for a certain time to allow bmdmg of the immunoglobuhn (s)

Whenever the solid phase matrix is in the form of permeable or semi-permeable membranes or sheets the contacting is generally performed by pumpmg/forcmg the immunoglobuhn contammg solution across the surface and or through a porous structure of the membrane or sheet to ensure that the immunoglobuhns are coming in close contact with the hgands immobilised on the surface and/or in the porous structures

Further guidelines for this process are given in "Purification Tools for Monoclonal .Antibodies", Gagnon, P , Validated Biosystems, 1996)

Washmg

After contactmg the sohd phase matrix with the immunoglobuhn contammg solution there is optionally performed a washmg procedure to remove unbound or loosely bound substances such as other protems, hpids, nucleic acids or other impurities from the matrix However in some cases where very high purity of the immunoglobuhn is not critical the washmg procedure may be omitted sav g a process-step as well as washmg solution

In other cases where very high purity of the immunoglobuhn is needed there may be employed several different washmg procedures with different washmg buffers before elution is commenced

The washmg buffers employed will depend on the nature of the chromatographic adsorbent and the hgand binding the immunoglobuhns The washmg buffer should not disturb the bmdmg of the immunoglobuhn to the adsorbent i e pH, salt concentration and other additives should be adjusted so that only the unwanted impurities are removed either by simple substitution of the solution contammg impurities and present m and around the adsorbent with the washmg buffer - or in combination herewith also releasmg impurities bound to the adsorbent The releasmg of impurities bound to the matrix may be accomphshed either by changing pH and/or ionic strength or by adding a substance to the washmg buffer which mteracts competitively with either the adsorbent or the impurity, and thereby displacing the impurity from the adsorbent

The washing (operation (c) in the method according to the invention) is preferably performed in order to remove remainders from the solution containing the immunoglobuhns, and in order to remove other type of biomolecules.

Elution

Elution of the bound immunoglobuhn is generally performed by contacting the sohd phase matrix comprising the bound immunoglobuhn with a solution that releases the immunoglobuhn from the ligand on the matrix. The immunoglobuhn is hereby released into the solution and can be washed out of the matrix. The solution employed to release the immunoglobuhn should generally have different characteristics than what was used for binding of the immunoglobulin e.g. the solution may have a different pH, a different ionic strength, a different temperature and/or it may comprise organic solvents (preferably only small amounts), detergents, chaotropes or other denaturing reagents. Combinations of changes in these different parameters are also generally employed.

Elution may also be performed by applying a solution gradually changing the conditions from binding to non-binding conditions, a procedure which typically is phrased gradient elution.

Once the immunoglobulin have been released from the sohd phase matrix into the eluting solution it may be recovered from this by different optional means known per se. In the most simple case the immunoglobulin may be used directly without any changes but in many instances some sort of concentrating procedure would be preferred e.g. ultra-filtration, freeze- drying or precipitation (e.g. salting out). The immunoglobulin solution may also very well be purified further in a further processing step of optional character.

Regeneration

The sohd phase matrix may optionally by cleaned i.e. regenerated after elution of the immunoglobulin. This procedure is typically performed regularly to minimise the building up of impurities fouling up the surface of the sohd phase and/or to sterilise the matrix to avoid contamination of the product with microorganisms proliferating and escaping from the sohd phase and the equipment used during the process. Popular ways of performing such a regeneration step is to wash the sohd phase matrix with solutions able to clean the matrix and/or kill microorganisms. Typical solutions for these purposes would be, e.g., 0.1-1.0 M sodium hydroxide; solutions of peracids or hydrogen peroxide; denaturants such as guanidinium hydrochloride; solutions comprising active chlorine such as hypochlorite solutions, organic solvents such as ethanol; detergents etc. .An especially preferred method for this purpose is to

use 0 1- 1 0 M sodium hydroxide due to the very high efficiency, low cost, ease of neutralization with hydrochloric acid and lack of waste problems

In a preferred embodiment of the present mvention the method includes (I) equilibration (optional step), (u) contactmg, (ni) washmg (optional step), (IV) separation, (v) elution, and (vi) regeneration, where cycle of steps (ι)-(v) are repeated one or several times before regeneration, and were the sohd phase matrix is reused after regeneration

The conditions employed m both the binding, washing and elution step(s) may be very decisive for the resulting bmding efficiency, yield and purity of the immunoglobuhn Different sohd phase matrices according to the invention may need different binding, washing and elution conditions to ensure an optimal result Likewise the nature of the raw material will have a very significant impact on the conditions chosen for that particular isolation procedure e g very dilute solutions of monoclonal antibodies in hybridoma cell culture supematants (typically 10- 100 μg/ml) behave differently than the same type of antibodies present m more concentrated solutions such as ascites fluids (1-5 mg/ml) and immunoglobuhns present in, e g , whey (1-2 mg/ml) need other conditions than immunoglobuhns from plasma/serum (5-20 mg/ml) etc

Also the composition i e the contents of different types of impurities may vary significantly between different raw materials, e g , egg yolk has a very different composition as compared to hybridoma coll culture supematants

As mentioned above it is generally possible to add different substances to the immunoglobu n contammg solution as to enhance the bmdmg of antibodies to the sohd phase matrix

In a particular embodiment, the present mvention relates to methods for the isolation of immunoglobuhns and sohd phase matrices therefor yielding an isolated immunoglobuhn of a purity of at least 10 % such as at least 30%, preferably at least 50% such as at least 70%, more preferably at least 80% such as 90%, m particular at least 99%

As mentioned above, it is beheved that the bmdmg efficiency maximum pH value for the solid phase matrices is in the range of 2 0 to 10 0, most likely m the range of 3 0 to 9 0 It is therefore most relevant to conduct the isolation process near that maximum (which of course may vary for different combinations of immunoglobu ns/sohd phase matices Thus, the pH of the solution contammg the immunoglobuhns (or protems in general) is preferably in the range of 2 0 to 10, such as m the range of 3 0 to 90 However, depending on the ligand type and the matrix backbone, the pH range is preferably 3 0 to 7 0 or 60 to 90

It is beheved that, when the hgand is of the type -X-A-SP2-ACID, then should the pH of the solution contammg the immunoglobuhns be in the range of 2 0 to 6 0, preferably in the range of 2 5 to 5 5 such as in the range of 3 0 to 5 5, or in the range of 4 0 to 5 5, corresponding to an expected bmdmg efficiency maximum for that specific type of matrix

With respect to contactmg operation (a) above, it has been found that it is not necessary to add excessive amounts of lyotropic salt in order for the immunoglobuhns to bind to the matrix Thus, the total salt content, including e g NaCI, of the solution contammg the immunoglobuhns need only be so that it corresponds to a ionic strength of at the most 2 0, preferably m the range of 0 05 to 2 0, such as 0 05 to 1 4, especially m the range of 0 05 to 1 0 As an alternative requirement, the concentration of lyotropic salt as such should be as low as possible, thus it has been shown that it is possible to operate with a solution contammg immunoglobuhns where the concentration of lyotropic salts is at the most 0 4 M, preferably at the most 0 3 M, m particular at the most 0 2 M, such as at the most 0 1 M

Examples of lyotropic salts are given m "Purification Tools for Monoclonal jAntibodies", Gagnon, P , Validated Biosystems, 1996), where the Hofmeister series of lyotropic ions are presented

With respect to the concentration of immunoglobuhns m the solution it is beheved that the solid phase matrices can operate for a very large range concentration range, thus, it is beheved that the sohd phase matrices operate equally efficient for concentration of immunoglobuhns m the solution contammg the immunoglobuhns m the range of 0 001 to 0 2, preferably 0 01 to 0 1 mg/ml, as m hybridoma cell culture supematants, m the range of 0 2 to 2 0 mg/ml as in milk and whey, in the range of 5 0 to 20 mg/ml as for different animal sear and plasma, and even m the range of 20-80 mg/ml as for colostrum

It has been found that the present mvention is especially suitable for solutions compnsmg in the range of 0 1 to 30 mg immunoglobuhns per gram of sohd phase matrix, such as m the range of 0 2 to 2 or m the range of 5 0 to 25 mg per gram of sohd phase matrix

Thus, the solution contammg the immunoglobuhns may be artificially as well as biologically solution of immunoglobuhns such as crude fermentation broths, mammalian cell cultures such as hybridoma cell cultures, fermentation broths from cultures of genetically engineered microorganisms such as E coh, ascites fluids such as mouse and rat ascites fluid, milk, whey, blood, plasma and serum from man, mouse, rat, cow, pig, rabbit, goat, gumea pig, and donkey, and egg yolk such as chicken egg yolk

Furthermore, it has been shown (see the examples) that special advantages with respect to purity may be obtained when the solution contammg the immunoglobuhns comprises a

negatively charged detergent Without bemg bound to any theory it is beheved that the detergent suppresses the adherence of other biomolecules to the matrix Examples of such detergent are octyl sulfate, bromphenol blue, octane sulfonate, sodium laurylsarcosinate, and hexane sulfonate

.Also, in the washmg step (operation (c) of the method according to the mvention) it is, probably for the same reasons, advantageous to use an negatively charged detergent The detergent may be used alone or in combination with an buffer, e g a lyotropic salt buffer Use of lyotropic salts m the washmg step (small volume) represents only a minor waste product problem compared with usmg lyotropic salts m the bmdmg processes (operation (a)) (in that the bmding process includes the use of large volumes is most cases)

Also, the excellent properties of the sohd phase matrices for use in the method according to the mvention may be expressed even without the use of organic solvents in the eluation step (operation (d)), thus, preferably, the eluent used comprises less than 10% (v/v) more preferably less than 5%, of organic solvents Most preferably no organic solvents are used at all

.Alternatively, as has been show example 14, a larger amount of non-toxic solvents, e g propylene glycol, may be used, e g up to 40% propylene glycol

The contactmg step (operation (a)) as well as the following step, I e separation, washmg, and eluation, may be performed m various way The physical measures selected are often guided by the scale and whether the process has to be repeated The sohd phase matrices according to the mvention may be used m almost any of the set-ups used for development and for industrial purposes Thus, the sohd phase matrix may be contacted with the solution contammg the immunoglobuhns, e g m a stirred batch process, in a packed bed chromatographic column process, and m a fluidised bed process Further guidelines are given m "Purification Tools for Monoclonal Antibodies", Gagnon, P , Vahdated Biosystems, 1996)

Other necessary measures for performing the isolation of immunoglobuhns accordmg to the mvention follow conventional methodologies

The present mvention provides a method for the isolation and purification of immunoglobuhns from a large variety of raw materials havmg different concentrations of immunoglobuhns, typically rangmg from about 10 μg/ml m hybridoma cell culture supematants and about 1-2 mg/ml in milk and whey to about 5-20 mg/ml in different animal sera/plasma, and up to 50-60 mg/ml m colostrum The nature and relative concentration of different impurities potentially interfering with the bmd g and isolation of immunoglobuhns are also varying to a great extent between the different immunoglobuhn sources

For some applications of immunoglobu ns it is of high important that the immunoglobu ns are extremely pure, e g having a purity of more than 99 % This is particularly true whenever the immunoglobuhn is to be used as a therapeutic, but is also necessary for other applications In the diagnostic field the degree of purity needed may depend on a number of factors such as whether the antibody is to be used un-deπvatised in which case there may not be required a high degree of purity, I e less than 50 %, or whether the antibody has to be labelled with a signal molecule such as an enzyme, e g horseradish peroxidase, m which case the antibody often is required to be at least 80% pure or more For other applications the need for purity may differ correspondingly It seems however to be a general demand that the purity of the immunoglobu n is at least 10 % on a dry matter basis to enable a proper use of the product However, the present mvention provides, as it should be clear, guidelines for solvmg these problems

Solid phase matrices

As described above, the method accordmg to the mvention includes the use of a sohd phase matrix, where the sohd phase matrix comprises a functionahsed matrix backbone carrymg a plurality of functional groups of the following formula

M-SP1-L

wherein M designates the matrix backbone, SPl designates a spacer, and L designates a ligand compnsmg an mono- or bicychc optionally substituted aromatic or heteroaromatic moiety which has to fulfil certain criteria

It should be noted that the present mvention also relates to these sohd phase matrices as such Thus, the definitions below relate to the method according to the invention as well as to the sohd phase matrices accordmg to the mvention

The solid phase matrix may comprise, as the matrix backbone, any natural or synthetic and organic or inorganic material known per se to be apphcable m sohd phase separation of protems and other biomolecules, e g natural or synthetic polysacchaπdes such as agar-agar and agaroses, celluloses, cellulose ethers such as hydroxypropyl cellulose, carboxymethyl celluose, starches, gums such as guar gum, and gum arable, gum ghatti, gum tragacanth, locust bean gum, xanthan gum, pectms, mucins, dextrans, chitms, chitosans, alginates, carrageenans, hepaπns, gelatms, synthetic polymers such as polyamides such as polyacrylamides and polymethacrylamides, polyimides, polyesters, polyethers, polymeric vinyl compounds such as polyv ylalcohols and polystyrenes, polyalkenes, inorganic materials such as sihcious materials

such as silicon dioxide including amorphous silica and quartz; silicas; metal silicates, controlled pore glasses and ceramics; metal oxides and sulfides, or combinations of these natural or synthetic and organic or inorganic materials.

The matrix backbone is preferably selected from agar-agar, agaroses, celluloses, cellulose ethers such as hydroxypropyl cellulose, carboxy ethyl cellulose, polyamides such as poly(meth)acryl- amides, polyvinylalcohols, silicas, and controlled pore glasses.

Especially interesting sohd phase materials as matrix backbones are e.g. agar or agarose beads such as Sepharose and Superose beads from Pharmacia Biotech, Sweden and Biogel A from

Biorad, USA; dextran based beads such as Sephadex, Pharmacia Biotech; cellulose based beads and membranes such as Perloza cellulose from Secheza, Czechoslovakia; composite beads such as Sephacryl and Superdex, Pharmacia Biotech; beads of synthetic organic polymers such as Fractogel from Toso-Haas, USA; POROS media from Perceptive Biosystems, USA, Bio-Rex, Bio- Gel P and Macro Prep from Biorad, HEMA and Separon from TESSEK and Hyper D and Trisacryl media from BioSepra, USA, Enzacryl and Azlactone, 3M, USA; beads of siliceous materials such as controlled pore glass, PROSEP, from Bioprocesing, England and Spherocil, BioSepra; and coated silica composites in the form of beads or membranes such as ACTI-DISK, ACTI-MOD and CycloSep from Arbor Technologies, USA.

Typically, the sohd phase matrix backbone, as well as the resulting functionahsed sohd phase matrix, may, e.g., be in the form of irregular particles or spherical beads, membranes or sheets, moulded surfaces, or sticks. The sohd phase material may further be fully or partly permeable or completely impermeable to proteins. In a particularly interesting embodiment of the present invention, the matrix is in the form of irregular or spherical beads with sizes in the range of 1- 10000 μm, preferably 10-1000 μm; such as 10-60 μm for high performance apphcations and such as 50-500 μm, preferably 50-300 μm, for preparative purposes.

A particular interesting form of matrix is a density controlled matrix in the form of a conglomerate comprising density controlling particles. These conglomerates, which are especially applicable in large scale operations for fluidised or expanded bed chromatography as well as different batch-wise chromatography techniques in non-packed columns, e.g. simple batch adsorption in stirred tanks, are described in the WO 92/00799, which is hereby incorporated by reference.

The hgands L may be attached to the sohd phase material by any type of covalent bond known per se to be applicable for this purpose, either by a direct chemical reaction between the ligand and the sohd phase material or by a preceding activation of the solid phase material or of the ligand with a suitable reagent known per se making it possible to link the matrix backbone and

1G the hgand Examples of such suitable activatmg reagents are epichlorohydrm, epibromohydrm allyl-glycidylether, bis-epoxides such as butanedioldiglycidylether, halogen-substituted ahphatic compounds such as di-chloro-propanol, divmyl sulfone, carbonyldnmidazole, aldehydes such as glutanc dialdehyde, quinones, cyanogen bromide, periodates such as sodium-meta-peπodate, carbodnmides, chloro-tπazines such as cyanunc chloride, sulfonyl chlorides such as tosyl chlorides and tresyl chlorides, N-hydroxy succinimides, 2-fluoro- l-methylpyrιdmιum toluene-4- sulfonates, oxazolones, maleimides, pyπdyl disulfides, and hydrazides Among these, the activating reagents leavmg a spacer group SPl different from a smgle bond, e g epichlorohydrm, epibromohydrm, allyl-glycidylether, bis-epoxides, halogen -substituted ahphatic compounds, divmyl sulfone, aldehydes, quinones, cyanogen bromide chloro-triazines, oxazolones, maleimides, pyridyl disulfides, and hydrazides, are preferred

Especially interesting activatmg reagents are beheved to be epoxy-compounds such as epichlorohydrm, allyl-glycidylether and butanedioldiglycidylether

In certam mstances the activatmg reagent may even constitute a part of the functionahty contributmg to the binding of immunoglobuhns to the solid phase matrix e g m cases where divmyl sulfone is used as the activatmg reagent In other cases the activating reagent is released from the matrix durmg reaction of the functional group with the matrix This is the case when carbodumidazoles and carbodnmides are used

The above mentioned possibilities makes it relevant to defme the presence of an optional spacer SPl linking the matrix M and the hgand L In the present context the spacer SPl is to be considered as the part of the activating reagent which forms the link between the matrix and the hgand Thus, the spacer SPl corresponds to the activatmg reagents and the couphng reactions mvolved In some cases, e g when usmg carbodnmides, the activatmg reagent forms an activated form of the matrix or of the hgand reagent After couphng no parts of the activatmg reagent is left between the hgand and the matrix, and, thus, SPl is simply a smgle bond

In other cases the spacer SP 1 is an mtegral part of the functional group effectmg the bmdmg characteristics, I e the hgand, and this will be especially significant if the spacer SPl comprises functionally active sites or substituents such as thiols, amines, acidic groups, sulfone groups, nitro groups, hydroxy groups, nitnle groups or other groups able to mteract through hydrogen bonding electrostatic bonding or repulsion, charge transfer or the like

In still other cases the spacer SPl may comprise an aromatic or heteroaromatic rmg which plays a significant role for the bmdmg characteristics of the sohd phase matrix This would for example be the case if qumones or chlorotπazines where used as activation agents for the sohd phase matrix or the hgand

Preferably, the spacer SPl is a smgle bond or a biradical derived from an activatmg reagent selected from epichlorohydrm, allyl-glycidylether, bis-epoxides such as butanedioldiglycidylether, halogen-substituted ahphatic compounds such as l,3-dιchloropropan-2-ol, aldehydes such as glutaric dialdehyde, divmyl sulfone, qumones, cyanogen bromide, chloro-tπazmes such as cyanunc chloride, 2-fluoro- l-methylpyπdιnιum toluene-4-sulfonates, maleimides, oxazolones, and hydrazides

Preferably the spacer SPl is selected from short chain ahphatic biradicals, e g of the formula - CH ₂ -CH(OH)-CH ₂ - (derived from epichlorohydrm), -(CH2)3-0-CH ₂ -CH(OH)-CH„- (derived from allyl-glycidylether) or -CH ₂ -CH(OH)-CH ₂ -0-(CH ₂ )4-0-CH2-CH(OH)-CH ₂ - (derived from butane¬ dioldiglycidylether, or a single bond

Due to the risk of leakage of material (e g the hgand and/or the spacer) from a sohd phase matrix mto the eluted product (e g the unmunoglobuhn) the molecular weight of the hgand (or the hgand + the optional spacer) is advantageously chosen as low as possible A major drawback of usmg protem A, protem G, synthetic peptides and other relatively high molecular weight hgands (e g dyes) is that it may be difficult or even impossible to separate any released ligand (optionally including the spacer) from the eluted immunoglobuhn due to the small difference between the respective molecular weights and the natural tendency of the components to bmd to each other This may have a detrimental effect in those cases where the immunoglobuhn is to be used as a therapeutic agent causmg allergic chock or other serious symptoms m the patient The smaller the molecular weight of the hgand (including its spacer) the more efficient can any leaked hgand be separated from the immunoglobuhn product Another significant advantage of havmg the smallest possible molecular weight of the hgand (or the hgand-spacer arm conjugate) is that any leaked material, which may not have been separated from the immunoglobuhn prior to injection/ingestion in the patient will elucidate a minimum of antigenicity the lower the molecular weight and therefore m general be relatively more acceptable to the organism than higher molecular weight hgands

It is therefore, preferred that the hgand L has a molecular weight below 500 Dalton, preferably below 400 Dalton, more preferably below 300 Dalton, such as below 250 Dalton, or even below 200 Dalton

With respect to the hgand-spacer arm conjugate (-SP1-L), it is preferred that the molecular weight is below 500 Dalton, more preferably below 400 Dalton, such as below 300 Dalton, or even below 250 Dalton

Accordmg to the mvention, the matrix comprises hgands which either alone or m combination with a spacer SPl (and even the matrix backbone) make it possible to bind immunoglobuhns thereto It is found that a crucial part of the hgand is a mono- or bicychc aromatic or heteroaromatic moiety which may carry one or more substituents, one of which preferably bemg a substituent compnsmg an acidic moiety

The term "mono- or bicychc" is mtended to mean that the core part of the moiety in question is consisting of one ring or two fused rmgs, e g as in benzene and naphthalene respectively, and, thus, not to hgands comprising two separate rings as in biphenyl

It has been found that the structure of the aromatic or heteroaromatic part of the ligand, L may cover a very wide spectrum of different structures optionally having one or more substituents on the aromatic or heteroaromatic rιng(s) However, it seems to be rather decisive which substituents are present on, e g , a benzene rmg as to whether the hgand wdl bmd the immunoglobuhn (s) efficiently, which is the object of the present mvention, or whether the bindmg is only moderately or low

Even though the hgands are named here and m the following usmg the nomenclature corresponding to the individual and discrete chemical compound, from which they are derived, it should be understood that the actual hgand L is a radical of such a compound

However, based on our preliminary findings, it is especially preferred to employ matrices compnsmg aromatic or heteroaromatic groups (radicals) of the following types as functional groups benzoic acids such as 2-amιnobenzoιc acids, 3-ammobenzoιc acids, 4-amιnobenzoιc acids, 2-mercaptobenzoιc acids, 4-ammo-2-chlorobenzoιc acid, 2-amιno-5-chlorobenzoιc acid, 2-amιno-4- chlorobenzoic acid, 4-ammosahcyhc acids, 5-amιnosahcyhc acids, 3,4-dιamιnobenzoιc acids, 3,5- diammobenzoic acid, 5-amιnoιsophthahc acid, 4-amιnophthahc acid, cmnamic acids such as hydroxy-cmnamic acids, mcotmic acids such as 2-mercaptonιcotmιc acids, naphthoic acids such as 2-hydroxy-l -naphthoic acid, qumohnes such as 2-mercaptoqumohne, tetrazolacetic acids such as 5-mercapto-l -tetrazolacetic acid, thiadiazols such as 2-mercapto-5-methyl-l,3,4-thιadιazol, benzimidazols such as 2-amιno-benzιmιdazol, 2-mercaptobenzιmιdazol, and 2-mercapto-5-nιtro- benzimidazol, benzothiazols such as 2-ammobenzothιazol, 2-amιno-6-mtrobenzothιazol, 2- mercaptobenzothiazol and 2-mercapto-6-ethoxybenzothιazol, benzoxazols such as 2- mercaptobenzoxazol, thiophenols such as thiophenol and 2-amιnothiophenol, 2-(4- amιnophenylthιo)acetιc acid, aromatic or heteroaromatic sulfonic acids and phosphonic acids such as l-ammo-2-naphthol-4-sulfonιc acid and phenols such as 2-ammo-4-nιtrophenol It should be noted that the case where M is agarose, S l is derived from vinyl sulfone, and L is 4-amιno- benzoic acid is specifically disclaimed m relation to the sohd phase matrices according to the mvention, cf WO 92/16292

The detailed structure of the hgand seems to determine important functional characteristics relevant for the isolation of immunoglobulins from different sources. Thus, different hgands comprising remote or closely related aromatic structures seems to result in significant changes in the binding strength, binding selectivity, binding capacity and overall yield of immunoglobulin when applied in the isolation of antibodies from different raw materials.

For binding of immunoglobulins at near neutral pH (about pH 5 to pH 9) it is preferred to use a hgand comprising radicals derived from a benzene ring fused with a heteroaromatic ring system, e.g. a hgand selected from benzimidazoles such as 2-mercapto-benzimidazol and 2-mercapto-5- nitro-benzimidazol; benzothiazols such as 2-amino-6-nitrobenzothiazol, 2-mercaptobenzothiazol and 2-mercapto-6-ethoxybenzothiazol; benzoxazols such as 2-mercaptobenzoxazol. Not belonging to the former group of ligands but also preferred for the binding of immunoglobuhns at near neutral pH are ligands chosen from the group of thiophenols such as thiophenol and 2- aminothiophenol.

Thus, as it is clear from the above and the results shown herein, the hgand L is preferably selected from radicals having the following formula

-X-A-SUB

wherein X designates -0-, -S-, or -NH-, A designates an aromatic or heteroaromatic ring or ring system, and SUB designates one or more substituents.

It is understood that X is an integral part of the hgand in that the aromatic or heteroaromatic compound which forms the hgand part of the sohd phase matrix after reaction with an activated matrix backbone, must include a hydroxy group (X is -0-), a mercapto group (X is -S-) or an amino group (X is -NH-) directly attached the aromatic or heteroaromatic moiety. Examples of such compounds are 3-hydroxy-cinnamic acid, 2-mercapto-benzoic acid, and 2-amino-benzoic acid. It should be understood that if the aromatic or heteroaromatic compound comprises, e.g., a hydroxy group as well as an amino group, the resulting sohd phase matrix may comprise a mixture of hgand being attached to the linker through the amino group and through the hydroxy group, respectively.

The aromatic radicals are preferably selected from benzene radicals and naphthalene radicals.

The aromatic radical is preferably a benzene radical such as phenyl, 1,2-phenylene, 1,3- phenylene, 1,4-phenylene, 1,2,3-benzenetriyl, 1,2,4-bcnzenetriyl, 1,3,5-benzenetriyl, 1,2,3,4- benzenetetrayl, 1,2,3,5-benzenetetrayl, 1 ,2,4,5-benzenetetrayl, and 1 ,2,3,4,5-benzenepentayl.

The heteroaromatic radicals are preferably selected from monocychc heteroaromatic radicals such as thiophene, furan, pyran, pyrrole lmidazole, pyrazole, isothiazole, isoxazole, pyπdine, pyrazine, pyrmudine, and pyridazme radicals, and bicychc heteroaromatic radicals such as mdole, purine, quinohne, benzofuran, benzimidazole, benzothiazole, and benzoxazole radicals

The heteroaromatic radical is preferably selected from radicals of pyndme, benzimidazole, benzothiazole and benzoxazole

A preferred group of hgands for high purity immunoglobuhn isolates is chosen among amino- benzoic acids hke 2-amιno-benzoιc acid 2-mercapto-benzoιc acid 3-ammobenzoιc acid 4-amιno- benzoic acid, 4-amιno-2-chlorobenzoιc acid, 2-amιno-5-chlorobenzoιc acid, 2-amιno-4- chlorobenzoic acid, 4-amιnosahcyhc acids, 5-ammosahcyhc acids, 3,4-dιammobenzoιc acids, 3,5- diaminobenzoic acid, 5-ammoιsophthahc acid, 4-amιnophthahc acid

Another preferred group of hgands giving a high degree of purity of the isolated immunoglobuhn is the group of cmnamic acids such as 2-hydroxy-cιnnamιc acids, 3-hydroxy-cιnnamιc acid and 4- hydroxy-cmnamic acid

Still another preferred group of hgands for isolation of high purity immunoglobu ns are derived from the group of heteroaromatic compounds compnsmg a carboxylic acid and an ammo group as substituents such as 2-ammo-nιcotιnιc acid, 2-mercapto-nιcotιnιc acid 6-ammo-nιcotιnιc acid and 2-amιno-4-hydroxypyπmιdιne-carboxyhc acid

Agarose matrix backbones and spacers derived from epoxy compounds are especially relevant m combination with these preferred groups of hgands

With respect to the substituents on the aromatic or heteroaromatic moiety, SUB preferably comprises at least one acidic group

In a particularly interesting embodiment of the present invention, SUB comprises at least one substituent of the following formula

-SP2-ACID

wherein SP2 designate an optional second spacer and ACID designates an acidic group

In the present context the term "acidic group" is mtended to mean groups havmg a pKa-value of less than about 6 0, such as a carboxyhc acid group (-COOH), sulfonic acid group (-SO2OH)

sulfuiic acid group (-S(O)OH), phosphinic acid group (-PH(0)(OH)), phosphonic acid monoester groups (-P(0)(OH)(OR)), and phosphonic acids group (-P(0)(OH)2) The pKa-value of the acidic group should preferably be in the range of 1 0 to 6 0

The acidic group is preferably selected from carboxylic acid, sulfonic acid, and phosphonic acid

The group SP2 is selected from Ci β-alkylene, and C ₂ 6-alkenylene, or SP2 designates a single bond Examples of relevant biradicals are methylene, ethylene, propylene, propenylene, ISO- propylene, and cyclohexylene Preferably, SP2 designates methylene ethenylene, or a smgle bond

In one embodiment of the present invention SUB designates one group SP2-ACID In this case, SP2 is preferably a smgle bond

SUB may, however, designate a substituent -SP2-ACID as well as one or more further substιtuent(s) independently selected from hydroxy, ammo, cyano, mono- and dι(Cι β- alkyl)amιno, halogen such as lodo, bromo, chloro, and fluoro, sulfanyl, nitro, Ci 6-alkylcarboxy, and ammocarboxy, mono- and dι(Cι 6-alkyl)amιnocarboxy, carboxy sulfono, sulfonamide, phosphonic ester with Ci β-alkyl, optionally substituted Ci 12-alkyl, optionally substituted C ₂ 12- alkenyl optionally substituted Ci ι_-alkynyl, and optionally substituted Ci 12-alkoxy, thioester, or the substituent is an oxygen atom which together with two valences of a carbon atom of the aromatic or heteroaromatic moiety form an oxo group Furthermore, SUB may designate a further group -SP2-ACID as defmed above It should be understood that the substituents defmed for SUB correspond to the optional substituents for L

In another preferred embodiment, SUB designates a substituent -SP2-ACID as well as one or more further substιtuent(s) independently selected from hydroxy, ammo, cyano, halogen, sulfanyl, nitro, optionally substituted Ci 6-alkyl methyl, ethyl, propyl, butyl, isobutyl, and cyclohexyl, optionally substituted C26-alkenyl, optionally substituted C2 β-alkynyl, optionally substituted Ci β-alkoxy, carboxy, and sulfono, or the substituent is an oxygen atom which together with two valences of a carbon atom of the aromatic or heteroaromatic moiety form an oxo group J*USO m this case, SP2 preferably designates methylene, ethenylene, or a smgle bond, preferably a smgle bond

In the present context, the term "Ci 12-alky ^] " is mtended to mean alkyl groups with 1- 12 carbon atoms which may be straight or branched or cyclic such as methyl, ethyl, propyl, isopropyl, butyl, isobutyl, tert-butyl, pentyl, hexyl, octyl, dodecyl, cyclopentyl, cyclohexyl, decahnyl, etc

The term "optionally substituted Ci 12-alkyl" is intended to mean a Ci 12-alkyl group which may be substituted with one or more, preferably 1-3, groups selected from carboxy , protected carboxy such as a carboxy ester, e g Ci β-alkoxycarbonyl, aminocarbonyl, mono- and dι(Cι β-alkyl)- aminocarbonyl, amino-Ci 6-alkyl-amιnocarbonyl, mono- and dι(Cι h-alky])amιno-Cι 6-alkyl- ammocarbonyl, ammo, mono- and dι(Cι 6-alkyl)amιno, Ci β-alkylcarbonylammo, hydroxy. protected hydroxy such as acyloxy, e g Ci β-alkanoyloxy, sulfono, Ci β-alkylsulfonyloxy, nitro phenyl, phenyl-Ci e-alkyl, halogen, nitnlo, and mercapto

Examples of substituted Ci 12-alkyl groups are carboxy-Ci 12-alkyl (e g carboxymethyl and carboxyethyl), protected carboxy-Ci 12-alkyl such as estenfied carboxy-Ci β-alkyl (e g Ci β-alkoxy- carbonyl-Ci 12-alkyl such as methoxycarbonylmethyl, ethoxycarbonylmethyl, and methoxycarbonylethyl), ammocarbonyl-Ci 12-alkyl (e g ammocarbonylethyl, aminocarbonylethyl and ammocarbonylpropyl), Ci β-alkylammocarbonyl-Ci 12-alkyl (e g methylammocarbonvlmethyl and ethylaminocarbonylmethyl) ammo-Ci (,-alkyl-amιnocarbonyl-Cι 12-alkyl (e g aminomethylaminocarbonylmethyl and ammoethylammocarbonylmethyl), mono- or dι(Cι β- alkyl)ammo-Cι β-alkylaminocarbonyl-Ci 12-alkyl (e g dimethylammomethylammocarbonylmethyl and dimethylammoethylammocarbonylmethyl), mono- or dι(Cι G-alkyl)ammo-Cι 12-alkyl (e g di- methylammomethyl and dimethylammoethyl), hydroxy-Ct 12-alkyl (e g hydroxymethyl and hy- droxyethyl), protected hydroxy-Ci 12-alkyl such as acyloxy-Ci 12-alkyl (e g Ci β-alkanoyloxy-Ct 12- alkyl such as acetyloxyethyl, acetyloxypropyl, acetyloxybutyl, acetyloxypentyl, propionyloxy- methyl, butyryloxymethyl, and hexanoyloxymethyl)

In the present context, the term "C2 12-alkenyl" is intended to mean mono-, di- or polyunsaturated alkyl groups with 2-12 carbon atoms which may be straight or branched or cyclic in which the double bond(s) may be present anywhere in the chain or the rιng(s), for example vmyl, 1-propenyl, 2-propenyl, hexenyl, decenyl, 1 ,3-heptadιenyl cyclohexenyl etc Some of the substituents exist both m a cis and a trans configuration The scope of this mvention comprises both the cis and trans forms

In the present context, the term "C2 12-alkynyl" is mtended to mean a straight or branched alkyl group with 2-12 carbon atoms and incorporating one or more triple bond(s), e g ethynyl, 1- propynyl, 2-propynyl, 2-butynyl, 1,6-heptachynyl, etc

In the expressions "optionally substituted C2 12-alkenyl" and "optionally substituted C2 12- alkynyl", the term "optionally substituted" is mtended to mean that the moiety may be substituted one or more times, preferably 1-3 times, with one of the groups defmed above for "optionally substituted Ci 12-alkyl"

The term "optionally substituted Ci 12-alkoxy" designates, as in traditional chemical nomenclature, an optionally substituted Ci 12-alkyl-oxy group, which may be substituted one or more times, preferably 1-3 times, with the substituents indicated for "optionally substituted alkyl" described above

The terms "Ci e-alkyl", "C26-alkenyl", "C∑e-alkynyl", and "Ci e-alkoxy" reflect the shorter analogues of the "Ci 12-alkyl", "C2 12-alkenyI", "C2 12-alkynyl" and "Ci 12-alkoxy" groups

The terms "Ci β-alkylene" and "C ₂ 6-alkenylene" is intended to mean biradicals of the groups defmed for "Ci β-alkyl" and "C2 β-alkenyl", respectively

The present invention should not be bound to any specific theory, however, it is envisaged that the special electronic configuration of the aromatic or heteroaromatic moiety m combmation with one or more heteroatoms, which may be located in the heteroaromatic rmg system or as a substituent thereon, is mvolved in the specific binding of immunoglobu ns, as well as the bmdmg of other protems

Thus, m an mteresting embodiment of the present invention the hgand comprises at least one nitrogen, sulfur or phosphorous atom, e g as a ring atom or as a substituent on the (hetero)aromatιc rmg, such as an ammo or nitro group or a sulfonic acid group or a phosphonic acid group

An especially interesting combmation of substituents seem to be any combmation of at least one amino or mercapto group with at least one acidic group selected from carboxyhc acids, sulfonic acids, and phosphonic acids

It is envisaged that a combmation of two or more of the hgands t} e defmed here on the same matrix backbone may lead to certam to certain advantages with respect to high binding efficiency and or high purity of the immunoglobuhn However, in an important embodiment of the present mvention, all of the functional groups on the sohd phase matrix are substantially identical

It may also be found to enhance bmdmg efficiency and purity of the product by coupling the hgand to a matrix already compnsmg negatively or positively charged moieties such as positively charged ammo-groups or negatively charged carboxyhc acid, sulfonic acid or phosphonic acid groups

The ligand concentration may also be of major significance for the functional characteristics of a matrix according to the invention e g a hgand may show a high degree of selective bmdmg of

immunoglobuhns at one hgand concentration, while an increase m the ligand concentration results in a decrease m the bmding selectivity As is well-known to a person skilled in the art too high hgand concentrations may lead to strong bmdmg of unwant d impurities by mechanism of multiple bmding pomts, because the hgands are too closely spaced on the sohd phase backbone If the hgand concentration is kept low the hgands will be spaced with larger distances and therefore not course the bmdmg of impurities by bmdmg at multiple sites jAnother negative effect of too high hgand concentration is the risk of bmding the wanted protem e g the immunoglobuhn by multiple bmding sites Such a multiple bmding may lead to difficulties in releasing the protem e g the immunoglobuhn with an appropriate elution buffer In some mstances it may even be necessary to utilise strongly denaturmg conditions and/organic solvents for release of the product from such to highly substituted sohd phase matrices - with loss of biological activity as a consequence

Ligand concentration of solid phase matrices may be disclosed in several different ways One way of describmg the hgand concentration is to disclose the amount of ligand present per gram of dry matter (e g in μmol/g dry matter) This is the result obtained directly if for example the hgand concentration is measured by elemental analysis of dried (e g freeze-dπed) samples of the sohd phase matrix The hgand concentration may, however, also be disclosed as the amount of hgand present on one ml wet and sedimented solid phase matrix (also often referred to as one ml packed bed matrix) This is a figure which is easily calculated from a determination based on dried sohd phase matrix (e g μmol g dry matter), if the dry matter content of the hydrated sohd phase matrix has been determmed at the same time (l e gram of dry matter/ml wet sedimented sohd phase matrix) Still another way of disclosing the hgand concentration is as the amount of hgand present m one gram of wet but suction drained matrix This figure is agam easily calculated from a determination based on dry matter, if the sohd phase dry matter content per gram of wet, but suction drained matrix has been determmed at the same time

Thus, the hgand concentration of the sohd phase matrices of the invention is preferably m the range of 10-990 μmol/g dry matter of sohd phase matrix, such as 100-990 μmol/g, more preferably 200-980 μmol g, m particular 250-975 μmol/g, or the hgand concentration the sohd phase matrices of the mvention is preferably m the range of 1- 145 μmol/ml of hydrated, sedimented sohd phase matrix, such as 10-120 μmol/ml, more preferably 15-100 μmol/ml, in particular 20-80 μmol/ml, or the hgand concentration the sohd phase matrices of the mvention is preferably in the range of 1- 130 μmol g wet, but suction dra ed sohd phase matrix such as 10-110 μmol/gram, more preferably 20-100 μmol/g, in particular 20-90 μmol/gram

It is, as should already be clear from the above, the aim of the present mvention to provide sohd phase matrices having a high bmdmg efficiency

Thus, the sohd phase matrices, which are useful within the scope of the present mvention must fulfil two of three criteria (a), (b), and (c) (see above), e g criteria (a) and (b), criteria (a) and (c), or criteria (b) and (c) Preferably all three criteria are fulfilled

With respect to criterion (a), it is highly desirably in combination with the other criteria set forth herem or as an alternative thereto, that the sohd phase matrix has a bmding efficiency of at least 50% when tested at a pH m the range of 2 0 to 10 0 m the "Standard Immunoglobulin Binding Test" described herem It is envisaged that the bmding efficiency maximum (which can be estimated quite accurately, within half a pH unit, by testing the bmdmg efficiency over an pH range usmg, e g , increments of 0 5 pH units) of most of the matrices accordmg to the invention is m the range of 3 0 to 9 0, e g in the range 3 0 to 7 0 or in the range of 6 0 to 90 dependmg on the nature of the hgand

It has been found that the bindmg efficiency at pH 4 5 and pH 7 0 is especially relevant when performing a general evaluation of a sohd phase matrix for isolation of immunoglobuhns, thus, m a preferred embodiment, the present invention relates to a sohd phase matrix havmg a bmding efficiency of at least 50% at pH 4 5 or pH 7 0, the "Standard Immunoglobuhn Binding Test" described herem

Thus, m a particularly mterestmg embodiment of the present mvention, the sohd phase matrix has a bmdmg efficiency of at least 50%, preferably at least 60%, more preferably at least 70%, in particular at least 80%, such as at least 90%, in the "Standard Immunoglobulin Binding Test" described herem, at least one pH-value of the solvent m the range of pH 1 0 to pH 11 0, m particular m the range of pH 3 0 to pH 9 0, and more particularly at pH 4 5 or 7 0

Furthermore, it is also the aim of the present mvention to provide sohd phase matrices which are able to bmd a vide range of immunoglobuhns, so that the end user can rely on one sohd phase matrix stead of an number of products which has to be tested individually for each clone of immunoglobuhns

Thus, with respect to the criterion (b), the sohd phase matrix preferably has an average bmdmg efficiency of at least 50%, such as at least 60%, preferably at least 70%, especially at least 80%, in particular at least 90%, for the immunoglobuhns tested m the "Monoclonal Antibody Array Binding Test" when tested at a pH in the range of 2 0 to 10 0, such as in the range of 3 0 to 9 0, e g in the range of 3 0 to 7 0 or in the range of 6 0 to 9 0 Typically, the bmding efficiency is

determmed at two pH values, e at pH 4 5 and pH 7 0, and the optimum is then found by varying the pH value in mcrements of 0 5 around the one of the two pH values giving the most promismg bmding efficiency

The functional stabihty of the matrix, which is mterestmg and important with respect to lower risk of leachmg and the possibility of regeneration, may be influenced by the chemical structure of the hgand, I e the stabdity to harsh regeneration conditions such as 1 M sodium hydroxide is dependent on the hgand structure, as well as the matrix backbone and any spacer moiety

Therefore, with respect to criterion (c), it is a preferred that the stabihty (see example 8) of the solid phase matrix ui l NaOH is so that incubation of the matrix in 1 M NaOH in the dark at room temperature for 7 days reduces the bmding efficiency at a pH in the range of one pH unit lower than the bmdmg maximum pH to one pH unit higher than the binding maximum pH value as determmed accordmg to the "Standard Immunoglobuhn Binding test" described herem, with less than 50%, preferably less than 25%, compared to a corresponding untreated matrix Preferably the reduction is less than 15%, such as at less than 10% m particular less than 5%

It has been found that sohd phase matrices compnsmg a functionahsed matrix backbone carrymg a plurahty of functional groups of the following formula

M-SP1-X-A-SP2-ACID

wherein designates the matrix backbone, SPl designates a spacer, X designates -0-, -S- or -NII-, A designates a mono- or bicychc optionally substituted aromatic or heteroaromatic moiety, SP2 designates an optional spacer, and ACID designates an acidic group,

with the first proviso that the molecular weight of the hgand -L is at the most 500 Dalton,

are novel m themselves (specifically disclaimmg 4-ammobenzoιc acid disclosed in K L Knudsen et al, .Analytical Biochemistry, vol 201, p 170, 1992 and WO 92/16292, which has been used for the isolation of immunoglobuhns m combmation with lyotropic salts, as a ligand), and that they are well suited for the isolation and/or purification of immunoglobuhns as well as for the isolation and or purification of protems and other biomolecules m general

It has furthermore been found that the above-mentioned sohd phase matrices compnsmg functional groups of the formula M-SP1-X-A-SP2-ACID, are equally apphcable for pH-dependent reversible bmding of protems and other biomolecules

Thus, the present invention also provides a method for the isolation of protems from a solution contammg one or more of protems, compnsmg the following operations

a) contactmg a solution contammg one or more proteins havmg a pH in the range of 1 0 to 6 0 and a total salt content corresponding to a ionic strength of at the most 2 0 with a sohd phase of the formula M-SP1 -X-A-SP2-ACID, whereby at least a part of the protems becomes bound to the solid phase matrix,

b) separatmg the solid phase matrix havmg protems bound thereto from the solution,

c) optionally washing the sohd phase matrix, and

d) contacting the solid phase matrix with an eluent m order to hberate one or more of the proteins from the solid phase matrix wherein the eluent used comprises less than 10% (v/v) of organic solvents

The pH of the solution contammg the protems is preferably m the range of 1 0 to 6 0, such as 2 0 to 6 0, especially in the range of 3 0 to 5 5, such as 4 0 to 5 0, and the pH of the eluent is m the range of 6 0 to 11, preferably m the range of 6 0 to 9 0

As in the method for the isolation of the immunoglobuhns, the total salt content of the solution contammg the proteins preferably corresponds to a ionic strength of at the most 2 0, such as m the range of 0 05 to 2 0, m particular in the range of 0 05 to 1 4, especially m the range of 0 05 to 1 0, and or the concentration of lyotropic salts preferably is at the most 0 4 M, such as at the most 0 3 M, m particular at the most 0 2 M, especially at the most 0 1 M Furthermore, as above it is advantageous to use a negatively charged detergent m the contactmg step (operation (a)) and or in the washmg step (operation (c)) Preferably, the washmg step (operation (c)) preferably implies the use of an morganic or organic salt buffer compnsmg a negatively charged detergent

The method for the isolation of protems and other biomolecules may be employed for a number of proteins, examples of which are proteases such as pro-enzymes, trypsins, chymotrypsins, subtilism, pepsin, plasmmogen, papam, renm, thrombm, and elastase, hpases, glucosidases, xylanases, lect s, albumins, protems from fermentations broths, protein from milk and whey protems from blood, plasma, and serum, protems from fish waste, protems from slaughterhouse waste such as organ and tissue extracts, e g alkaline phosphatase from bovmc intestines, and protems from vegetable extracts such as potato, tomato, coconut, e g horse radish peroxidase

Synthesis of solid phase matrices

Generally a solid phase matrix may be derivatised so as to comprise covalently linked hgands according to the invention be methods know per se, e.g., activation of the sohd phase matrix with a suitable reagent known per se followed by coupling of the ligand to the activated matrix, optionally incorporating a spacer SPl between the hgand and the matrix by coupling the spacer to the activated matrix first followed by couphng the hgand to the spacer via a suitable condensation reagent or even a second activation of the spacer followed by couphng of the ligand.

The sequence and choice of reagents may depend on the actual hgand to be coupled and the sohd phase material to be derivatised with consideration to, e.g., the content of reactive groups such as hydroxyl, amino, mercapto, and silanols etc. In some cases it may be preferable to activate or derivatise the hgand instead of the sohd phase matrix followed by a reaction of the derivatised hgand with the sohd phase matrix.

Thus, in a preferred method for of synthesising a sohd phase matrix according to the invention, the sohd phase matrix is first reacted with a reagent able to react with the sohd phase matrix and thereby activate it for further reaction with the ligand, optionally washing away the activation reagent followed by a reaction of the activated sohd phase matrix with a solution comprising the hgand and optionally followed by washing the sohd phase matrix comprising the covalently immobilised ligand with one or more suitable solutions cleaning the matrix for surplus reactants.

In some cases in may be possible to combine the activation and the coupling of the hgand by mixing the two reagents and let the reactions take place in parallel. This is a great advantage as it saves costs and time as well as minimising the volume of waste water. Thus, the activation and the coupling step is preferably performed in one combined step.

Furthermore, it is a significant advantage if the activation and/or the coupling reaction can be performed without the need to add organic solvents to the reaction medium. These organic solvents are often used to solubilise the reactive reagents or to ensure that hydrolysis of reactive species are kept at a minimum. However, the use of organic solvents adds to the cost and risk of the process because of the risk of explosions, the risk of health damage, the waste problems and the relatively high cost of the solvents themselves. Thus, the activation and or the couphng procedure is preferably performed without the addition of any organic solvent to the reaction medium.

EXAMPLES

The mvention is illustrated by the following examples 1- 15

1. Derivatisation of solid phases

1A) Epichlorohydrm activation of agarose beads

Activation of agarose beads from Hispanagar

"High " level of activation

Approximately 1000 ml of a 1 1 suspension of agarose beads m water (Hispanagar, 6 % agarose beads, particle size 100-140 μm) was washed with demmerahsed water on a smtered glass funnel followed by suction draining for one mmute 700 gram of wet, but drained agarose beads were weighed mto a mixture of 560 ml water and 70 ml 32,5 % w/v sodium hydroxide This suspension were then added 90 ml epichlorohydrm (.ALDRICH cat no E 105-5) followed by gentle stirring with a paddle at room temperature (20-25 °C) for 6 hours The agarose beads were then washed on a suction filter with approx 20 htres of water and finally suspended m water The activated agarose beads were found to be stable in this suspension for several weeks when stored at 4 °C

The concentration of active epoxy groups on the activated agarose beads were determmed by thiosulfate titration as described m Porath, J , Laas, T Janson, J -C Journal of Chromatography, vol 103, pp 49-69, 1975 and Sundberg, L , & Porath, J , Journal of Chromatography, vol 90, pp 87-98, 1974 The results from this titration mdicated that the activated beads had a concentration of 70 μmol epoxy-groups per gram of wet, but suction drained beads, corresponding to 972 μmol/g dry matter, or 54 μmol /ml wet sedimented beads (aqueous solution)

"Low" level of activation

For production of a matrix with a lower content of active epoxy groups the same procedure as described above was followed with the exemption that the reaction mixture consisted of 200 g wet, but suction drained agarose beads, 160 ml water, 20 ml 2 M sodium hydroxide and 11 5 ml epichlorohydrm

Thiosulfate titration mdicated the presence of 21 μmol epoxy groups per gram wet, but suction drained beads, corresponding to 292 μmol/g dry matter, or 16 μmol/ml wet sedimented beads (aqueous solution)

Activation of agarose beads from Pharmacia and Biorad:

The same activation procedure as described above were employed for the activation of agarose beads from Pharmacia (Sepharose 4B and Sepharose 6B) and Biorad (Biogel A-5m Gel, particle size 38-75 μm and Biogel A- 15m Gel particle size 75-150 μ ).

Titration of active epoxy groups on these sohd phases gave the following results:

μmol epoxy groups per gram drained beads:

Sepharose 4B: 40

Sepharose 6B: 52

Biogel A 5m Gel: 65

Biogel A15m Gel: 46

IB) Epichlorohydrin activation of Fractogel

Fractogel TSK HW-55 (F), particle size 32-63 μm, from MERCK (cat.no.: 14981) and Fractogel TSK HW-65 (F), particle size 32-63 μm, MERCK (cat.no.: 14984) were activated with epichlorohydrin with the same procedure as described above for agarose beads. The resulting concentration of active epoxy groups on these solid phases were 98 and 53 μmol g of drained beads respectively.

1C) Butanedioldiglycidyl ether activation of agarose beads

100 gram 6 % agarose beads from Hispanagar was washed with water on a sintered glass funnel and drained by suction for one minute. The beads were then suspended in 75 ml 0.6 M NaOH and hereafter added 75 ml 1,4-butanediol diglycidyl ether. Gentle stirring with a paddle was performed at room temperature for 18 hours whereafter the matrix was washed with water (approx. 3 litre).

Thiosulfate titration of the amount of epoxy groups incorporated into the matrix gave a content of 55 μmol/g suction drained matrix.

ID) Divinyl sulfone activation of agarose beads.

Activation of agarose beads from Hispanagar:

Approximately 1400 ml of a 1:1 suspension of agarose beads in water OHispanagar, 6 % agarose beads, particle size 100-140 μm) was washed with deminerahsed water on a sintered glass funnel followed by suction draining for one minute. 700 gram of wet, but drained, agarose beads were weighed into 350 ml 0.5 M potassium phosphate buffer pH 11.5. 35 ml divmyl sulfone was added and the resulting suspension was paddle stirred at room temperature for 2 hours. The matrix was then transferred to a sintered glass funnel and washed with 20 htres of water, 5 htres of 30 % ethanol in water and finally 5 htres of water. The resulting activated matrix was determined to have a content of 45 μmol active vinyl groups per gram suction drained beads as determined by the thiosulfate titration method.

IE) Coupling of ligands to activated matrices

General coupling procedure:

All couplings of different hgands to the activated matrices mentioned in example 1 A-D were performed accordmg to the following general procedure:

1) The activated beads were washed on a suction filter with 2-3 volumes of deminerahsed water. The beads were drained by slight suction on a sintered glass funnel and 20 g of wet, but drained gel were weighed into a 100 ml plastic bottle with screw cap.

2) 1 g of hgand was dissolved in 20 ml of water and titrated to pH 10.5-11.0 with 2 M sodium hydroxide (for some hgands with low solubility the pH was adjusted to pH 11.5-12.5). The resulting solution was mixed with the activated matrix. The gel was incubated with the solution by gentle mixing on a roller mixer for 18 hours at room temperature.

3) The gel was then washed with 2 htres of water.

In those instances where the hgand had poor solubility in water, a 50 % ethanol solution was employed for dissolution instead followed by titration to pH 10.5-11.0 with 2 M sodium hydroxide. At the same instances the final washing with water was substituted with one washing step of 1 litre 50 % ethanol followed by another washing step with 1 litre of water.

When divinyl sulfone activated agarose was used for coupling the pH of the couphng mixture was adjusted to pH 11.5 instead of 12.6.

Whenever possible the concentration of coupled ligand on the matrices was determmed by elementary analysis of Carbon, Hydrogen, Nitrogen, Oxygen and Sulfur In some mstances it was furthermore possible to determmed the amount of coupled hgand by acid-base titration of characteristic functional groups on the coupled ligand

Coupling of ligands to epoxy-activated 6 % agarose beads

The following chemical substances (ligands) were coupled to epichlorohydrin activated 6 % agarose beads (Example 1-A) (Hispanagar, particle size 100-140 μm) according to the above given general coupling procedure

2-hydroxybenzoιc acid, 3-hydroxybenzoιc acid, 4-hydroxybenzoιc acid, 2,5-dιhydroxybenzoιc acid, 2-hydroxycmnamιc acid 3-hydroxycmnamιc acid, 4-hydroxycmnamιc acid 3,5-dιnιtrosahcyhc acid, 2-hydroxy-3-methoxybenzoιc acid, 3-hydroxy-4-methoxybenzoιc acid, 2-hydroxy-5-met- hoxybenzoic acid, 4-hydroxy-3-methoxybenzoιc acid, 3,5-dιmethoxy-4-hydroxybenzoιc acid, 2- ammo-4,5-dιmethoxybenzoιc acid, 5-sulfosahcyhc acid 5-chlorosahcyhc acid, 4-hydroxy-3,5-dι- mtrobenzoic acid, 2-amιnobenzoιc acid, 3-amιnobenzoιc acid, 4-amιnobenzoιc acid, 2-ammo-3,5 duodobenzoic acid, 2-mercaptobenzoιc acid, 2-mercaptonιcotιnιc acid, anιhne-2-sulfonιc acid, 2- pyndylhydroxymethanesulfonic acid, 4-acetamιdophenol, 5-mercapto-l -tetrazolacetic acid, 1- hydroxy-2-naphthoιc acid, 3-hydroxy-2-naphhtoιc acid, 2-hydroxy- l -naphthoic acid, 2,3-pyrιdme- di-carboxyhc acid, 4-pyrιdylthιoacetιc acid, 2-pyrιmιdylthιoacetιc acid, 2-mercaptochmohne lmidazole, 2-mercaptoιmιdazole, 2-mercapto-l-methyhmιdazole, 3-mercapto-l,2,4-tπazole, 5- mercapto-1-methyltetrazole, 2-mercaptothιazohne, 2-mercapto-5-methyl-l,3,4-thιadιazole, 2,5-dι- mercapto-l,3,4-thιadιazole, benzimidazole, 2-hydroxybenzιmιdazole, 2-ammobenzιmιdazole 2- mercaptobenzimidazole, 2-mercapto-5-nιtrobenzιmιdazole, benzothiazole, 2-amιnobenzothιazole, 2-ammo-6-mtro-benzothιazole, 2-ammo-6-ethoxybenzothιazole, 2-mercaptobenzthιazole, 6- ethoxy-2-mercaptobenzothιazole, 6-amιno-2,5-dιhydroιmιdazo(2, l-b)benzothιazole, 2- mercaptobenzoxazole, 2-(2-hydroxyphenyl)benzoxazole, phenol, 2-chlorophenol, 3-chlorophenol, 4-chlorophenol, 2,4,6-tπmethylphenol, 2,3,5-tπmethylphenol, 4-methoxyphenol, 2,6- dimethoxyphenol, 3,4,5-tnmethoxyphenol, thiophenol, 4-chlorothιophenol, 2-amιnothιophenol benzyl mercaptan, 4-methoxybenzyl mercaptan, 4-methylthιo-m-cresol, aniline, 2,4-dιmethylam- hne, 3,5-dιmethoxyanιlme, 3,4,5-trιmethoxyanιhne, 2-methylmercaptoanιhne, 4-methylmercap- toanihne, 2,4,6-trι-methyl-m-phenylendιammo, 2,3-dιcyanhydrochmone, 2-phenylphenol, 4- phenylphenol, 4-benzyloxyphenol, 4,4-dιammophenylsulfone, 2-hydroxypyrιdme, 2,3-dι- hydroxypyπdine 2,6-dιhydroxypyndιne, 2-hydroxy-5-nιtropyπdιne, 3-cyano-4,6-dιmethyl-2- hydroxypyndine, 4-hydroxy-2-mercaptopyrιdιne, 2-mercaptopyπchne, 2-amιnoρyπdιne, 4-amιno- 2-chlorobenzoιc acid, 3-amιno-4-chlorobenzoιc acid, 2-ammo-5-chlorobenzoιc acid, 2-ammo-4- chlorobenzoic acid, 2-amιno-5-nιtrobenzoιc acid, 4-amιnosahcyhc acid, 5-amιnosahcylιc acid 3 4-

diaminobenzoic acid, 3,5-dιamιnobenzoιc acid, 4-ammomethylbenzoιc acid, 5-ammoιsophthahc acid, 4-ammophthahc acid, 4-ammohιppunc acid, 3-amιno- l,2,4-trιazole-δ-carboxyhc acid, 1- am o-2-naphthol-4-sulfonιc acid, 2-(4-amιnophenylthιo)acetιc acid, 2-amιno-4-nιtrophenol, 4- ammophenylacetic acid, 1-ammocyclohancarboxyhc acid, 2-amιnobenzylalcohol

The hgand concentration as determined by elementary analysis on freeze-dπed samples on the respective matrices generally all were m the range between 50 to 70 μmol of ligand per gram of wet, but suction drained matrix

IF) Coupling of ligands to various solid phases

The following other sohd phases epoxy activated agarose beads from Hispanagar, activated to a low level (Example 1-A), epoxy activated agarose beads from Pharmacia and Biorad .(Example 1- A), epoxy activated Fractogel from Merck (Example 1-B), butanedioldiglycidyl ether activated agarose beads from Hispanagar (Example 1-C), and divmyl sulfone activated agarose beads from Hispanagar (Example 1-D) were each coupled with 2-mercapto-benzoιc acid (2-MBA), 4-amιno- benzoic acid (4-.ABA), and 2-mercapto-benzιmιdazole (2-MBI)

The general coupling procedure described above in Example 1-E was followed durmg all couplmgs

The obtained ligand concentrations were determined by elemental analysis on freeze-dried samples of the respective matrices and is calculated and given as μmol of hgand per gram of wet, but suction dried matrix (one gram of wet, but suction dried matrix corresponds to approx 1 1 - 1 3 ml sedimented beads, while the dry matter content may vary considerably more between the different type of beads

Ligand concentration for the synthesised solid phase matrices. (stated as μmol/g wet, but suction drained beads)

Epoxy activated agarose beads from Hispanagar, activated to a low level Example 1-A)

2-MBA 4-ABA 2-MBI

18 20 20

Epoxy activated agarose beads from Pharmacia and Biorad Example 1-A).

Matrix \ Ligand 2-MBA 4-ABA 2-MBI

Sepharose 4B 37 40 38

Sepharose 6B 51 47 50

Biogel A 5m 59 60 58

Biogel A 15m 44 43 41

Epoxy activated Fractogel from Merck (Example 1-B):

Matrix \ Ligand 2-MBA 4- ABA 2-MBI

Fractogel TSK HW-55 96 92 98

Fractogel TSK HW-65 53 51 53

Butanedioldiglycidyl ether activated agarose beads from Hispanagar (Example 1-C):

2-MBA 4-ABA 2-MBI

51 48 44

Divinyl sulfone activated agarose beads from Hispanagar (Example 1-D):

2-MBA 4-ABA 2-MBI

45 45 42

2. Standard Immunoglobulin binding test

For the purpose of testing all the different sohd phase materials synthesised according to example 1 a standardised test, which can be reproduced any time, has been devised. The test is designed to determine the immunoglobulin binding efficiency of the different matrices under standardised conditions with respect to composition and pH of the raw material.

To ensure maximal relevancy of the test for isolation of monoclonal antibodies from dilute cell culture supematants we have simulated the conditions used for culturing hybridoma cells by mixing a typical cell culture media with fetal calf serum and added purified mouse

immunoglobuhn to this "artificial culture supernatant " All reagents are standard reagents and commercially available

Definition of "artificial culture supernatant "

For 250 ml solution

236 5 ml cell culture growth medium, DMEM (Imperial, UK, cat no 7-385-14) 12 5 ml fetal calf serum (Life Technologies, Denmark, cat no 10106-060) 1,0 ml purified polyclonal murine IgG (Sigma, USA, cat no 1-8765, 10 mg/ml) 0 244 g sodium azide (Sigma, USA cat no S-2002),

resulting in a solution contammg 40 μg murine IgG/ml, 5 % fetal calf serum, and 15 uiM sodium azide, and havmg a pH of approx 8 0

This solution was shown to be stable at 4 °C for several weeks without any deterioration of the immunoglobuhns

Standard procedure

1) Approximately 100 mg of the matrix to be tested is washed with 10 ml demmerahsed water on a smtered glass funnel followed by suction draining for 60 seconds 100 mg of wet (drained) sohd phase matrix is weighed mto a 3 0 ml test tube and 2 50 ml "artificial culture supernatant" havmg the pH value at which the matrix is to be tested is added The test tube is closed with a stopper, and the suspension is mcubated on a roller mixer for 2 hours at room temperature (20- 25 °C) The test tube is then centrifuged for 5 min at 2000 RPM m order to sediment the matrix The supernatant is then isolated from the sohd phase matrix by pipetting mto a separate test tube, avoiding the carry-over of any matrix particles Following this a determmation of the concentration of non-bound immunoglobuhn m the supernatant is performed by smgle radial immunodiffusion (as described m D Catty and C Raykundaha "Antibodies - a practical approach" Vol I, pp 137-168, 1988) usmg rabbit anti mouse immunoglobuhns as the precipitatmg antibody (DAKO, Denmark, cat no Z109)

The percentage of mouse unmunoglobuhn bound to the matnx is then calculated accordmg to the following formula

Percentage bound = (1 -(cone supernatant/cone startmg material)) x 100%

The precision of this method is better than +/- 5 %

2A) Screening for high immunoglobulin binding efficiency

The above described standard procedure for testmg the bmdmg efficiency was used for testmg a broad range of different solid phase matrices based on epichlorohydrm activated 6 % agarose beads from Hispanagar and synthesised according to example 1 A and 1 E

The results of the bmdmg test performed at pH 4 5 and pH 7 0 respectively is presented in the Table I below

Table I

Ligand Capacity at Capacity at pH 4 5 pH 7 0

2-hydroxybenzoιc acid 0 0

3-hydroxybenzoιc acid 0 30

4-hydroxybenzoιc acid 0 0

2,5-dιhydroxybenzoιc acid 60 0

2-hydroxycmnamιc acid 20 0

3-hydroxycιnnamιc acid 80 0

4-hydroxycmnamιc acid 40 0

3,5-dmιtrosahcyhc acid 0 0

2-hydroxy-3-methoxybenzoιc acid 0 0

3-hydroxy-4-methoxybenzoιc acid 40 0

2-hydroxy-5-methoxybenzoic acid 0 0

4-hydroxy-3-methoxybenzoιc acid 0 0

3,5-dιmethoxy-4-hydroxybenzoιc acid 0 30

2-amιno-4,5-dιmethoxybenzoιc acid 20 0

5-sulfosahcyhc acid 0 0

5-chlorosahcyhc acid 0 0

4-hydroxy-3,5-dιnιtrobenzoιc acid 0 0

2-ammobenzoιc acid 80 0

Ligand Capacity at Capacity at pH 4 5 pH 7 0

3-ammobenzoιc acid 100 0

4-ammobenzoιc acid 90 0

2-ammo-3,5-duodobenzoιc acid 0 0

2-mercaptobenzoιc acid 100 0

2-mercaptonιcotmιc acid 100 0 anιlme-2-sulfomc acid 0 0

2-pyrιdylhydroxymethansulfonιc acid 0 0

4-acetamιdophenol 0 0

5-mercapto- l-tetrazole acetic acid 70 0

1 -hydroxy -2- naphthoic acid 0 0

3-hydroxy-2-naphthoιc acid 0 0

2-hydroxy- l -naphthoic acid 60 0 pyrιdιne-2,3-dιcarboxybc acid 0 0

4-pyrιdylthιoacetιc acid 0 0

2-pyrιmιdylthιoacetιc acid 0 0

2-mercaptochmohne 80 60 lmidazole 0 0

2-mercaptoιmιdazole 0 0

2-mercapto- 1 -methyhmidazole 20 0

3-mercapto- 1 ,2,4-trιazole 0 0

5-mercapto- l-methyltetrazole 0 0

2-mercaptothιazohne 20 0

2-mercapto-5-methyl- l,3,4-thιadιazole 0 20

2,5-dιmercapto- 1 ,3,4-thιadιazole 100 20 benzimidazole 0 0

Ligand Capacity at Capacity at pH 4 5 pH 7 0

2-hydroxybenzιmιdazole 0 0

2-amιnobenzιmιdazole 40 20

2-mercaptobenzιmιdazole 70 70

2-mercapto-5-nιtrobenzιmιdazole 80 90 benzothiazole 0 0

2-ammobenzothιazole 20 0

2-amιno-6-nιtro-benzothιazole 80 60

2- ammo-6- ethoxy-benzothiazole 0 0

2 -m ercap tobenzoth mzole 70 60

6-ethoxy-2-mercaptobenzothιazole 20 40

6-ammo-2,5-dιhydroιmιdazo(2,l-b)benzothιazole 0 20

2-mercaptobenzoxazole 80 60

2-(2-hydroxyphenyl)benzoxazole 0 0 phenol 0 0

2-chlorophenol 0 0

3-chlorophenol 0 0

4-chlorophenol 0 20

2,4,6-tnmethylphenol 20 0

2,3,5-trιmethylphenol 20 20

2,6-dιmethoxyphenol 0 0

3,4,5-tnmethoxyphenol 0 0 thiophenol 70 60

4-chlorothιophenol 100 70

2-amιnothιophenol 70 50 benzyl mercaptan 0 0

Ligand Capacity at Capacity at pH 4 5 pH 7 0 aniline 20 20

2 4-dιmethylanιhne 0 0

3,4,5-trιmethoxyanιime 0 0

2-methylmercaρtoanuιne 60 0

2,4,6-trι-methyl-m-phenylendιamιne 20 0

2 3-dιcyanhydrochmone 20 0

2-phenylphenol 0 0

4-phenylphenol 20 20

4-benzyloxyphenoI 0 0

1,4-dιammophenylsulfone 20 0

2-h ydroxypy πdm e 0 0 ,3-cbhydroxypyrιdιne 20 0 -hydroxy-2-mercaptopyπdme 60 40 -ammo-2-chlorobenzoιc acid 0 40 -ammo-4-chlorobenzoιc acid 0 0 -ammo-5-chlorobenzoιc acid 80 0 -amιno-4-chlorobenzoιc acid 40 0 -amιno-5-nιtrobenzoιc acid 0 0 -ammosahcyhc acid 80 20 -ammosahcyhc acid 80 30 ,4-dιamιnobenzoιc acid 80 0 ,5-dιammobenzoιc acid 60 0 -amιnomethylbenzoιc acid 0 0 -ammoιsophthahc acid 60 20 -amιnophthalιc acid 60 20

Ligand Capacity at Capacity at pH 4 5 pH 7 0

4-ammohιppurιc acid 0 20

3-amιno-l,2,4-trιazol-5-carboxyhc acid 0 20 l-ammo-2-naphthol-4-sulfonιc acid 80 20

2-(4-amιnophenylthιo)acetιc acid 80 0

2-ammo-4-nιtrophenol 80 20

4-ammophenylacetιc acid 0 0

1-ammocyclohexancarboxyhc acid (reference) 0 0

2-amιnobenzylalcohol 20 0

As can be seen from the table some hgands do not bind the immunoglobuhn at all while others show very efficient bmdmg in the range of 80-100 % and still other hgands show intermediate binding efficiencies in the range of 30-60 %

As can be seen from the result from 1-amιnocyclohexancarboxyhc acid (reference), an aromatic or heteroaromatic moiety seems to be required for efficient bindmg

IF) Coupling of ligands to other solid phases

The following other sohd phases epoxy activated agarose beads from Hispanagar, activated to a low level (Example 1-A), epoxy activated agarose beads from Pharmacia and Biorad OExample 1- A), epoxy activated Fractogel from Merck (Example 1-B), butanedioldiglycidyl ether activated agarose beads from Hispanagar (Example 1-C), and divmyl sulfone activated agarose beads from Hispanagar OExample 1-D) were each coupled with 2-mercapto-benzoιc acid (2-MBA), 4-amrno- benzoic acid (4- ABA), and 2-mercapto-benzιmιdazole (2-MBI)

The general couphng procedure described above in Example 1-E was followed durmg all couplmgs

The obtained ligand concentrations were determmed by elemental analysis on freeze-dπed samples of the respective matrices and is calculated and given as μmol of hgand per gram of wet, but suction dried matrix (one gram of wet, but suction dried matrix corresponds to approx 1 1- 1 3 ml sedimented beads, while the dry matter content may vary considerably more between the different type of beads

Ligand concentration for the synthesised solid phase matrices: (stated as μmol per gram wet, but suction drained beads)

Epoxy activated agarose beads from Hispanagar, activated to a low level (Example 1-A):

2-MBA 4- ABA 2-MBI

18 20 20

Epoxy activated agarose beads from Pharmacia and Biorad (Example 1-A).

Matrix \ Ligand 2-MBA 4-ABA 2-MBI

Sepharose 4B 37 40 38

Sepharose 6B 51 47 50

Biogel A 5m 59 60 58

Biogel A 15m 44 43 41

Epoxy activated Fractogel from Merck (Example 1-B):

Matrix \ Ligand 2-MBA 4-ABA 2-MBI

Fractogel TSK HW-55 96 92 98

Fractogel TSK HW-65 53 51 53

Butanedioldiglycidyl ether activated agarose beads from Hispanagar (Example 1-C):

2-MBA 4-ABA 2-MBI

51 48 44

Divinyl sulfone activated agarose beads from Hispanagar (Example 1-D):

2-MBA 4-ABA 2-MBI

45 45 42

3. Monoclonal antibody array binding test

The following example illustrates the differences in binding efficiency between prior art sohd phase matrices and sohd phase matrices accordmg to the mvention for immunoglobuhn purification

For the comparative study 7 different cell hnes capable of producmg 7 different monoclonal antibodies were acquired from the American Type Culture Collection (ATCC) and propagated according to a standard procedure as described below Hereafter the bmdmg efficiency ol each monoclonal antibody was tested with each of the sohd phases protem A agarose (prior art matrix) Avidchrom (prior art matrix) and epoxy-lmked 2-mercapto-benzoιc acid agarose 4- ammo-benzoic acid agarose and 2-mercapto-benzιmιdazole agarose

The study was designed to determme the antibody binding efficiency durmg batch mcubation of the 5 different solid phases with culture supematants from the 7 different commercially available cell hnes

Monoclonal antibodies Cell lines The following seven cell hnes available from the .American Type Culture Collection were included m the standardised set-up

Cultures The monoclonal antibody culture supematants used in the study were produced by culture of the correspondmg mouse and rat hybridoma cells m a medium contammg fetal calf serum (RPMI-X, Medicult, Denmark cat no 20230500 + 5 % fetal calf serum Imperial, United Kingdom, cat no 83041) The methodology used for culturmg the five cell lines is well estabhshed in the prior art and described in G Brown and N R Ling "Antibodies - a practical

approach" Vol I, pp.81-104, 1988). After 3 weeks of culture the cells were removed by centrifugation and the supernatant filtered to remove any remaining particles. The concentration of monoclonal antibody in the five different culture supematants were determined by single radial immunodiffusion (as described in D. Catty and C Raykundalia "Antibodies - a practical approach " Vol I, pp.137-168, 1988) using rabbit anti mouse immunoglobulins and rabbit anti rat immunoglobulins as the precipitating antibodies (DAKO, Denmark, cat.no. :Z 109 and Z147) and found to be in the range of 30 to 60 μg/ml for all clones. Hereafter the content of monoclonal antibody in each culture supernatant was standardised by dilution to reach a final concentration of 30 μg/ml. To ensure similar conditions for all the supematants the dilution was performed with culture medium including 5 % fetal calf serum.

Solid phases: Protein A agarose from Repligen Corporation, USA, cat.no. :IPA-300. lot no.: RN 2917; Avidchrom from Unisyn Technologies, USA, cat. no.: 3100-0025, lot no.: 96-0404-1; 2-mercapto-benzoic acid agarose, 4-amino-benzoic acid agarose and 2-mercapto-benzimidazole agarose were based on epichlorohydrin activated 6 % agarose beads from Hispanagar, Spain and synthesised as described in example 1 A and 1 E. The hgand concentrations were measured by elemental analysis and found to be 65, 69 and 69 μmoles/gram wet, but drained matrix respectively (corresponding to 903, 958 and 958 μmoles/g dry matter as measured by elemental analysis on freeze dried samples).

The five different solid phase matrices were tested for their monoclonal antibody binding efficiency by incubating them with the 7 different monoclonal antibody supematants (standardised at 30 μg antibody/ml) according to the following procedure:

Standard procedure for the "Monoclonal antibody array binding test ":

Approximately 100 mg of the matrix to be tested is washed with 10 ml deminerahsed water on a sintered glass funnel followed by suction draining for 60 seconds. 100 mg of wet (drained) sohd phase matrix is weighed into a 3.0 ml test tube and 4.0 ml monoclonal antibody culture supernatant adjusted to the pH value at which the matrix is to be tested is added. The test tube is closed with a stopper, and the suspension is incubated on a roller mixer for 2 hours at room temperature (20-25 °C). The test tube is then centrifuged for 5 min. at 2000 RPM in order to sediment the matrix. The supernatant is then isolated from the sohd phase matrix by pipetting into a separate test tube, avoiding the carry-over of any matrix particles. Following this a determination of the concentration of non-bound immunoglobulin in the supernatant is performed by single radial immunodiffusion.

The percentage of monoclonal antibody bound to the matrix is then calculated according to the following formula:

Percentage bound = (1- (cone, in supernatant/ 30 μg/ml) ) x 100 %

The precision of this method is better than +/• 5 %.

pH adjustments to culture supematants for the different solid phases:

Protein A agarose: The monoclonal antibody culture supematants were adjusted to pH 8.2 by the addition TRIS/HC1 to a final TRIS concentration of 0.05 M.

Avidchrom: The monoclonal antibody culture supematants were adjusted to pH 7.4 by addition of potassium hydrogen phosphate/HCl to a final phosphate concentration of 0.05 M.

2-mercapto-benzoic acid and 4-amino-benzoic acid agarose: The monoclonal antibody culture supematants were adjusted to pH 4.5 by addition of acetic acid/sodium hydroxide to a final acetic acid concentration of 0.05 M.

2-mercapto-benzimidazole: The monoclonal antibody culture supematants were adjusted to pH 7.0 by addition of potassium hydrogen phosphate/HCl to a final phosphate concentration of 0.05 M.

Binding efficiency %

2-MBA: 2-mercapto- enzoic aci agarose (epic oro y rin) 4-ABA: 4-amino-benzoic acid agarose (epichlorohydrm) 2-MBI: 2-mercapto-benzimidazole agarose (epichlorohydrin)

As can be seen from the table the sohd phase matrices according to the mvention i.e. 2-mercapto- benzoic acid agarose, 4-amino-benzoic acid agarose and 2-mercapto-benzimidazole agarose exhibits a very constant high binding efficiency with the different clones (typically in the range of 50-100 % binding), while the prior art sohd phase matrices, protem A agarose and Avidchrom, exhibits much more varying binding efficiency (in the range from 0-100 % binding). The average binding efficiency has been calculated for each adsorbent and it is also from these data seen that the prior art adsorbents with average binding efficiencies of 52 and 53 % are significantly less efficient than the adsorbents according to the invention which have average binding efficiencies in the range from 75-95%.

4. 2-Mercaptobenzoic acid as the ligand

Isolation of immunoglobulins under different binding and washing conditions

As is indicated from the results in table I 2-mercaptobenzoιc acid seems to be a very mterestmg hgand for isolation and purification of monoclonal antibodies from dilute culture supematants Further studies of this solid phase matrix employmg the "artificial culture supernatant" as described in example 2 was therefore performed with the aim of establishing the optimal bmding and washing conditions so as to achieve the maximal bmdmg capacity as well as yield and purity of the antibody in the eluate

2-mercapto-benzoιc acid agarose was based on epichlorohydrm activated 6 % agarose beads from Hispanagar and synthesised as described m example 1 A and 1 E The hgand concentration was measured by two different methods and found to be 65 μmol /g wet but drained matrix as determined by elementary analysis and 60 μmol /g as determmed by acid-base titration of the immobilised benzoic acid part of the ligand

Generally the experiments were performed according to the following procedure

1) A small ahquot of 2-mercaptobenzoιc acid agarose was washed with water (all water unless otherwise stated had the quahty of Milh Q water) on a smtered glass funnel by gentle suction followed by draining of the interstitial water by hght suction for one minute

2) 0 4 gram of wet, but dramed matrix was then weighed into a test tube followed by the addition of 10 ml "artificial culture supernatant" havmg a specific pH-value for that particular experiment With or without any further additives the suspension was hereafter mcubated on a roller mixer for two hours at room temperature to ensure efficient bmd g of the im¬ munoglobuhn

3) Following mcubation the matrix was transferred to a column with a 5 mm mner diameter dramed for excess "artificial culture medium" and washed accordmg to a scheme specific for the particular experiment Washmg was performed by adding 4 x 4 ml washmg buffer to the column and collecting the run-through from the column m one fraction

4) The final elution of bound immunoglobulin was performed with a specific elution buffer by addition of 4 x 2 5 ml buffer to the column and collecting the eluate m one fraction No pumps were employed m the experiments, all columns were run by gravity (at an approximate flow rate of 0 5- 1 0 ml/mm)

5) Analyses were performed to determme the relative distribution of immunoglobuhn between the non-bound fraction in the supernatant after bmdmg, the washmg fractιon(s) and the eluate This was done by smgle radial immunodiffusion (as described m D Catty and C Raykundaha "Antibodies - a practical approach" Vol I, pp 137-168, 1988) using rabbit anti mouse immunoglobuhns as the precipitatmg antibody (DAKO AS, Denmark, Cat no Z 109)

The binding capacity was then calculated from the amount of non-bound unmunoglobuhn present in the supernatant and expressed as a percentage of the total amount added to the matrix in the raw material

The yield was calculated as the percentage of the added immunoglobuhn found in the eluate fraction (I e a yield of 100 % is equal to the presence of 1 mg IgG in the eluate)

The purity of the eluted immunoglobuhn was analysed by SDS-PAGE (sodium dodecyl sulfate polyacrylamide electrophoresis) under reducmg conditions followed by stammg of the protem bands with coomassie brilliant blue (Precast gel 4-20 % tris-glycme, 1 mm cat no EC6025, running 1 hour at 30 mA, tris-glycme SDS runnmg buffer cat.no LC2675, tris-glycme sample buffer cat no LC2676, coomassie stammg kit LC6025 all chemicals from Novex, USA)

The degree of purity as expressed in percent of the total protem contents was determmed by scanning and image processmg of the coomassie sta ed and dried polyacrylamide gel For this purpose we employed the CRE.AM system available from Kem-En-Tec AS, Denmark (cat no 6010 + 6050)

4A) The effect of performing binding at different pH-values

The following experiment was performed to estabhsh the pH-range in which the 2-mercapto- benzoic acid matrix would bmd immunoglobuhns efficiently from the "artificial culture supernatant" As was shown in Table I example 2, this matrix bmds 100 % at pH 4 5 and 0 % at pH 7.0 In this experiment the bmding efficiency, yield and purity of the eluate is determined when bmding is performed m the pH range 3 0-6 5 In all mstances the washing buffer used was 10 mM citric acid buffer adjusted to the same pH as the binding pH with 1 M sodium hydroxide The elution buffer used was in all mstances 005 M bone acid/NaOH + 0 5 M sodium chloride pH 8 6

Results

pi I of binding percent binding Yield (%) Purity (%)

3 0 100 90 < 5

3 5 100 95 < 5

4 0 100 100 < 5

4 5 100 100 < 5

5 0 95 95 5

5 5 40 40 10

6 0 0 0 -

6 5 0 0 .

As can bo seen from the table efficient bmding is achieved at pH-values below 6 0 reachmg 100% at pH 4 5 At the same time there is an mdication that a relatively higher purity may be obtained if the bmdmg step is performed at a higher pH than 4 5

4B) The effect of different washing procedures/pH in washing buffer

A series of tests were performed with the aim of optimising the purity of the eluate while mamta mg the yield at a high level For this purpose a range of different washing procedures were tested All tests were performed with pH 4 5 as the pH of bindmg and all eluates were performed with 0 05 M boric acid/NaOH + 0 5 M NaCl pH 8 6

Results

Washing buffer punt y (%) yield (%)

10 mM citric acid/NaOH pH 4 5 < 5 100 10 mM citric acid/NaOH pH 5 5 5 95 10 mM citric acid/NaOH pH 6 5 15 80 20 mM TRIS/HC1 pH 7 5 20 70 20 mM TRIS/HC1 pH 8 5 20 60

As can be seen from the table the purity of the eluate may be increased by washmg with a higher pH, but an increase m pH above pH 5 5 decreases the yield significantly

4C) The effect of different washing procedures/Lyotropic salts at high pH

Experiments were performed as described in 3 B except that a series of washmg buffers containing different lyotropic salts at pH 8.0 were tested for their abihty to improve the purity of the eluate without significantly decreasing the yield.

Results:

Washing buffer: purity (%) yield (%) 0.7 M ammonium sulfate/NaOH pH 8.0 ND < 10 0.9 M ammonium sulfate/NaOH pH 8.0 25 30

1.0 M ammonium sulfate/NaOH pH 8.0 25 80

1.1 M ammonium sulfate/NaOH pH 8.0 20 95 1.3 M ammonium sulfate/NaOH pH 8.0 20 95 0.8 M potassium phosphate pH 8.0 20 95

1.0 M potassium phosphate pH 8.0 15 95

0.9 M sodium sulfate/NaOH

+ 0.05 M sodium bicarbonate pH 8.0 20 95

1.0 M sodium sulfate/NaOH + 0.05 M sodium bicarbonate pH 8.0 20 95

1.0 M sodium chloride

+ 0.05 M potassium phosphate pH 8.0 ND 0

2.0 M sodium chloride

+ 0.05 M potassium phosphate pH 8.0 ND 0 4.0 M sodium chloride

+ 0.05 M potassium phosphate pH 8.0 20 80

The results indicate that the presence of lyotropic salts in the washing buffer combined with a higher pH than the binding pH may increase the purity of the eluate significantly without decreasing the yield. It is also evident that a certain concentration of the lyotropic salt is necessary to obtain this result. Too low concentrations results in loss of immunoglobuhn in the washing fraction, resulting in very low yields. .As can be seen the necessary concentration is dependent on the nature of the lyotropic salt, e.g. ammonium sulfate which is considered a strongly lyotropic salts according to the Hofmeister series (see Gagnon cited herein) need only to have a concentration of about 1.0-1.1 M to ensure a high yield in the eluate, while sodium chloride, which is considered a poor weakly lyotropic salt according to the Hofmeister series, needs to have a concentration of about 4 M before the yield is increasing to an acceptable level.

4D) The effect of different washing procedures/different additives

The effect of adding detergents and other additives to the washing buffer was investigated in tests performed as described above (example 4 B and 4 C)

Results:

Washing buffer: purity (%) yield (%)

0.01 M citric acid/NaOH pH 6.5 + 3 mg/ml octyl sulfate 50 80

0.01 M citric acid/NaOH pH 5.8

+ 0.05 mg/ml bromophenol blue 70 90

1.0 M ammonium sulfate/NaOH pH 7.5

+ 10 mg/ml octane sulfonic acid 80 80 1.0 M ammonium sulfate/NaOH pH 8.0 + 5 mg/ml sodium laurylsarcosinate 60 80 1.0 M ammonium sulfate/NaOH pH 8.0 + 5mg/ml octane sulfonic acid + 5 mg/ml sodium laurylsarcocinate 80 70 0.9 M potassium phosphate pH 9.2 + 5 mg/ml octane sulfonic acid 80 80 0.9 M potassium phosphate pH 9.2 + 5 mg/ml hexane sulfonic acid 60 90 1.0 M ammonium sulfate/NaOH pH 8.0 + 5 mg/ml tween 20 25 90

1.0 M ammonium sulfate/NaOH pH 8.0 + 5 mg/ml pluronic F68 25 80

The results from these experiments clearly indicate the positive effect on the purity of the eluate obtained by washing the matrix with buffers containing a negatively charged detergent (e.g. octane sulfonic acid, hexane sulfonic acid, octyl sulfate and sodium laurylsarcosinate), while the addition of uncharged detergents such as Tween 20 and pluronic F-68 seems to have little or no effect on the purity of the eluted immunoglobulin. Likewise it is shown that bromophenol blue, which is known to have a high affinity for binding to albumin (an unwanted impurity) also has a significant effect on the purity without compromising the yield of product. Furthermore the obtained effect seems to be independent of whether the washing buffer comprises high concentrations of lyotropic salts or not as well as the choice of lyotropic salt used, if present.

4E) The effect of different additives during binding

The following experiments were performed to mvestigate the effect on purity and yield of the addition of different detergents and other chemical substances to the "artificial culture supernatant" durmg the mcubation with the 2-mercapto-benzoιc acid agarose For all tests the pH of bmdmg was pH 5 0, the washmg buffer used was 1 1 M ammonium sulfate/NaOH pH 8 0 and the elution buffer was 0 05 M boric acid/NaOH + 0 5 M sodium chloride pi I 8 6 The experiments were otherwise performed as described in the general procedure above

Results

Substance added purity (%) yield (%)

None 25 95

5 mg/ml Tween 20 30 80

10 mg/ml benzoic acid 25 50 5 mg/ml 1 -octyl- 2-pyrrohdone 20 80

5 mg/ml N-octanoyl-N-methylglucamme 20 80

1 mg/ml lauryl sulfobetame 20 80 5 mg/ml lauryl sulfobetame ND 0 5 mg/ml suberic acid 25 80 5 mg/ml sebacic acid 25 80

5 mg/ml octane sulfonic acid 25 90 5 mg/ml caproic acid 60 90 5 mg/ml capryhc acid 70 80

0 5 mg/ml sodium laurylsarcosinate 70 90 1 0 mg/ml sodium laurylsarcosinate 85 90

2 0 mg/ml sodium laurylsarcosinate 90 70

1 mg/ml bromophenol blue 80 90

The results indicate that the addition of certain negatively charged detergents (or amphophihc substances) to the "artificial culture supernatant" prior to the mcubation with 2-mercapto- benzoic acid agarose has a significant influence on the final purity of the eluate This is for example the case for substances such as caproic and capryhc acid as well as lauryl sarcosinate, while other negatively charged substances such as benzoic acid, lauryl sulfobetame, suberic acid, sebacic acid and octane sulfonic acid seems to have very httle effect m the concentrations tested It is also noted that the neutral detergents Tween 20 and the positive detergent 1-octyl-N- methylglucamine seems to have no effect either

4F) The effect of different washing buffers in combination with the addition of sodium laurylsarcosinate to the raw material

The following example demonstrates the effect of combining the addition of a negatively charged detergent to the raw material with a series of different washing buffer compositions In all experiments there is added 1 mg/ml sodium lauryl sarcosinate to the "artificial culture supernatant" prior to mixmg with the 2-mercapto-benzoιc acid agarose, pH of binding were pH 5 0 and the elution buffer were in all cases 0 05 M boric acid/NaOH H 0 5 M sodium chloride pH 8 6 Otherwise the general procedure described above was followed

Results

Washmg buffer purity (%) yield (%) water 45 90

0 001 M sodium citrate pH 6 0 45 90 0 001 M sodium citrate pH 6 5 50 90 0 001 M sodium citrate pH 7 0 50 90 0 001 M potassium phosphate pH 7 5 50 90

0 001 M sodium citrate pH 6 5 + 5 % monopropylene glycol 50 80 0 001 M sodium citrate pH 6 5 + 20 % monopropylene glycol 45 95

1 0 M ammonium sulfate/NaOH pH 7 5 60 85 1 0 M ammonium sulfate/NaOH pH 7 0 60 90 0 9 M ammonium sulfate/NaOH pH 7 0 60 75

5. 4-Amino-benzoic acid as the ligand

Isolation of monoclonal antibodies under different conditions

4-amιno-benzoιc acid is another aromatic acid accordmg to the invention that seems to be very mterestmg for use in monoclonal antibody purification (table I example 2) The following tests demonstrates the influence of different bmdmg and washmg conditions on the performance of 4- ammo-benzoic acid agarose based on 6 % agarose from Hispanagar and synthesised accordmg to example 1 A and 1 E The matrix used was analysed by elemental analysis and determined to have a content of 69 μmol 4-ammo-benzoιc acid groups per ml wet, but dramed matrix

The tests were performed as described m the general procedure in example 4

5A) The effect of performing binding at different pH-values

The following experiment was performed to estabhsh the pH-range m which the 4-ammo-benzoιc acid matrix would bmd immunoglobulins efficiently from the "artificial culture supernatant" As was shown m Table I example 2, this matrix bmds 90 % at pH 4 5 and 0 % at pH 7 0 In this experiment the bmding efficiency, yield and purity of the eluate is determmed when bmdmg is performed in the pH-range 4 0-6 5 In all mstances the washing buffer used was 10 mM citric acid buffer adjusted to the same pH as the binding pH with 1 M sodium hydroxide The elution buffer used was in all instances 0 05 M boric acid/NaOH + 0 5 M sodium chloride pH 8 6

pH of binding percent bmdmg Yield (%) Purity (%)

4 0 100 90 10

4 5 90 95 10

5 0 60 55 20

5 5 20 ND ND

6 0 0 ND ND

6 5 0 ND ND

As can be seen from the table efficient binding is achieved at pH-values below 5 5 reachmg 90 % at pH 4 5 At the same time there is an indication that a relatively higher purity may be obtamed if the binding step is performed at a higher pH than 4 5

5B) The effect of different washing procedures/pH in washing buffer

A series of tests were performed with the aim of optimising the purity of the eluate while maintaining the yield at a high level For this purpose a range of different washmg procedures were tested .All tests were performed with pH 4 5 as the pH of bmding and all eluates were performed with 0 05 M boric acid/NaOH + 0.5 M NaCl pH 8 6

Results.

Washmg buffer purity (%) yield (%)

10 mM citric acid/NaOH pH 4 5 10 90

10 mM citric acid/NaOH pH 5.5 25 90

10 M citric acid NaOH pH 6 0 60 80

10 mM citric acid/NaOH pH 6 5 75 55

As can be seen from the table the purity of the eluate may be mcreased by washmg with a higher pH, but an mcrease m pH above pH 6 0 decreases the yield significantly

6. 2-Mercapto-nicotinic acid

Isolation of monoclonal antibodies under different conditions

2-mercapto-nicotinic acid is another aromatic acid according to the invention that seems to be very interesting for use in monoclonal antibody purification (table I, example 2). The following tests demonstrates the influence of different bmding and washing conditions on the performance of 2-mercapto-nicotinic acid agarose based on epichlorohydrin activated 6 % agarose from Hispanagar and synthesised according to example 1 A and 1 E. The matrix used was analysed by elemental analysis and determined to have a content of 63 μmol 2-mercapto-nicotinic acid groups per ml wet, but drained matrix.

The tests were performed as described in the general procedure in example 4.

In these two tests the effect of varying binding pH on yield and purity of the resulting eluate was investigated while keeping washmg and elution conditions constant. In both instances the washing buffer was 1.1 M ammonium sulfate/NaOH pH 8.0 + 5 mg/ml octyl sulfate and the elution buffer was 0.05 M boric acid/NaOH pH 8.6 + 0.5 M sodium chloride.

The "artificial culture supernatant" was adjusted to pH 4.5 and 5.0 with 1 M hydrochloric acid respectively and no further additions were made.

Results: Binding pH Yield, % Purity, %

4.5 85 80

5.0 75 95

As can be seen this matrix provides an excellent yield of immunoglobuhn in the eluate at both binding pH-values while the purity of the eluted immunoglobulin is significantly increased by raising the binding pH from pH 4.5 to pH 5.0.

Effect of adding sodium lauryl sarcosinate to the raw material

In the following tests the effect of adding different amounts of sodium lauryl sarcosinate to the "artificial culture supernatant" at two different binding pH-values is investigated. In all tests the washing buffer used was 1.1 M ammonium sulfate/NaOH pH 7.5 and the elution buffer was 0.05 M boric acid NaOH pH 8.6 + 0.5 M sodium chloride.

Prior to mixing with the solid phase matrix the "artificial culture supernatant" was added sodium lauryl sarcosmate to three different concentrations and then adjusted to pH 4 5 and 5 0 with 1 M hydrochloric acid respectively

Results

Binding at pH 4 5

Concentration of SLS, mg/ml aid, % Purity, %

0 5 90 50

1 0 90 65

1 5 65 85

Bmding at pH 5 0 Concentration of SLS mg/ml Yield, % Purity, %

0 5 90 50

1 0 90 85 1 5 40 >95

SLS = Sodium Lauryl Sarcosmate

7. 2-Mercapto-benzimidazoIe

Isolation of monoclonal antibodies

.As IS mdicated m Table I, 2-mercaptobenzιmιdazole represents another very mterestmg group of hgands (the benzimidazoles, benzoxazoles and benzothiazoles) for unmunoglobuhn isolation The following example illustrates the apphcation of this hgand for binding and isolation of immunoglobuhns from the "artificial culture supernatant" described in example 2

2-mercapto-benzιmιdazole agarose was based on epichlorohydrm activated 6 % agarose beads from Hispanagar and synthesised as described m example 1 A and 1 E The hgand concentration was measured by elemental analysis and found to be 69 μmol /g wet, but dramed matrix

In the following tests the yield and purity obtamed by mcubation of the 2-mercaptobenzιmιdazole agarose with "artificial culture supernatant" contammg different concentrations of added polyvmyl pyrrohdone is determmed by following the general procedure described in example 4 The pH of bmding was adjusted to pH 7 5 with hydrochloric acid, the washmg buffer was 0 01 M potassium phosphate + 0 5 M sodium chloride pH 7 5 and the elution buffer was 0 01 M citric acid/NaOH pH 3 5

Results

Concentration of PVP, mg/ml Yield, % Purity, %

0 0 95 25

0 5 80 70

1 0 70 80

2 0 40 90

4 0 5 ND

PVP polyvinyl pyrrohdone

The results indicate that 2-mercapto-benzιmιdazole agarose is able to bind almost all of the applied monoclonal antibody (l e giving a yield of 95%) and at the same time give an eluate which is substantially purified The purity can even be increased by adding substances such as polyvinyl pyrrohdone

8. Stability at high pH

2-mercapto-benzιmιdazole was coupled to epichlorohydrm activated 6 % agarose beads (Hispanagar) prepared as described in example 1 A as well as to divmyl sulfone activated 6 % agarose beads (Hispanagar) prepared accordmg to example 1 D Both couphng procedures were according to example 1 E

The contents of 2 mercapto-benzimidazole of the two matrices were determined by elemental analysis and found to be 69 μmol /g wet (dramed) matrix and 42 μmol /ml wet (dramed) matrix respectively

Both matrices were tested for their stabdity towards mcubation with 1 M sodium hydroxide by following the procedure described below

Standard stability test

I) Approximately 1000 mg of the matrix to be tested is washed with 100 ml demmerahsed water on a smtered glass funnel followed by suction draining for 60 seconds 500 mg of wet (dramed) sohd phase matrix is weighed mto a 10 0 ml test tube labelled "NaOH" and 9 0 ml 1 M sodium hydroxide is added followed by mixing gently for 1 mm Another 500 mg of wet (dramed) sohd phase matrix is weighed mto a 10 ml test tube labelled "Water" and 9 0 ml water is added followed by gentle mixing for 1 mm

The test tubes are closed tightly with stoppers stored dark at room temperature (20-25 °C) for 7 days

The matrices are then washed separately with 200 ml water on a sintered glass funnel followed by suction draining for 60 seconds

II) Each of the sohd phases matrices are tested in the "Standard Immunoglobulin Binding Test" defined in Example 2

The stabuity of the sohd phase matrix towards 1 M sodium hydroxide is then calculated and expressed as a percentage compared to the control which has only been incubated in water according to the followmg formula

Stabihty =

(percentage bound of NaOH treated matrix / percentage bound of control) x 100%

Results

Sohd phase matrix Stabuity, %

2-mercaplo-benzιmιdazole-epιchlorohydrιn-agarose 98 2-mercapto-benzιmιdazole-dιvmyl sulfone-agarose 0

The results indicate that matrices produced with divmyl sulfone activated agarose have poor stability m 1M NaOH, whereas epoxy activated agarose gives stable sohd phase matrices It is furthermore demonstrated that 2-mercapto-benzιmιdazole is a stable hgand in it self

9. Isolation of polyclonal antibodies from different species

2-mercaptobenzoιc acid as the ligand

The followmg example illustrates the bmdmg efficiency of 2-mercaptobenzoιc acid agarose towards polyclonal antibodies from different species, as well as yield and purity of the antibody in the eluate For the study sera from 5 different species were used

Polyclonal antibodies: The polyclonal antibodies used originated from normal sera from the followmg species goat, horse, rabbit, swme and human The sera were obtained from freshly drawn blood by mild centnfugation after coagulation for 24 hours at room temperature

Solid phase matrix 2-mercaptobenzoιc acid agarose was based on epoxy activated 6% agarose beads from Hispanagar and synthesised as described m example 1A and IE The hgand

concentration was measured by elemental analysis and found to be 65 μmol /gram wet, but dramed matrix

The sohd phase matrix was tested for it's polyclonal antibody bmding efficiency in a column accordmg to the followmg procedure

1) The matrix was washed with water on a smtered glass funnel and finally dramed 1 gram of wet, but dramed sohd phase matrix was weighed mto a small column (mner diameter of 5 mm) The matrix was washed with 5 ml of buffer (10 mM sodium citrate pH 5 0 ) 1 ml of the sample (adjusted to pH 5 0 with 1 M hydrochloric acid) was apphed to the column The column was washed with 20 ml of washmg buffer I (1 1 M ammonium sulfate pH 8 0 contammg 5 mg/ml sodium 1-octanesulfonate) The column was washed with 5 ml of washing buffer II (1 1 M ammonium sulfate pH 8 0) The matrix was eluted with 10 ml of elution buffer (0 05 M boric acid/NaOH + 0 5 M sodium chloride pH 8 6) No pumps were employed in the experiments, all columns were run by gravity (at an approximate flow rate of 0 5- 1 0 ml/mm)

2) Analyses were performed to determine the relative distribution of immunoglobuhn between the non-bound fraction in the run-through after bmdmg, the washing fractιon(s) and the eluate This was done by smgle radial immunodiffusion (as described in D Catty and C Raykundaha ".Antibodies - a practical approach" Vol I, pp 137-168, 1988) using species specific anti- immunoglobuhns as the precipitating antibodies

The binding capacity was then calculated from the amount of non-bound immunoglobu n present m the run-through and expressed as a percentage of the total amount added to the matrix m the raw material

The yield was calculated as the percentage of the added immunoglobu n found m the eluate fraction

The purity of the eluted immunoglobu n was analysed by SDS-PAGE (sodium dodecyl sulfate polyacrylamide electrophoresis) under reducmg conditions followed by staining of the protem bands with coomassie brilliant blue (Precast gel 4-20 % tris-glycme, 1 mm cat no EC6025, runnmg 1 hour at 30 mA, tris-glycme SDS runnmg buffer cat no LC2675, tris-glycme sample buffer cat no LC2676, coomassie stammg kit LC6025 all chemicals from Novex, USA)

The degree of purity as expressed m percent of the total protein contents was determmed by scanning and image processmg of the coomassie stained and dried polyacrylamide gel For this purpose we employed the CREAM system available from Kem-En-Tec AS, Denmark (cat no 6010 + 6050)

Results

Serum binding capacity Yield (%) Purity (%)

(%)

Goat 60 50 60

Swme 60 60 70

Rabbit 80 60 90

Horse 60 40 80

Human 70 60 75

10. Isolation of IgG from bovine serum

2 mercaptobenzimidazol as the ligand

The followmg example illustrates that it is possible to isolate and purify IgG from bovine serum with 2-mercaptobenzιmιdazol as the hgand

Bovine serum The bov e serum used was normal serum The serum was obtained from freshly drawn blood by mild centπfugation after coagulation for 24 hours at room temperature

Solid phase matrix 2-mercaptobenzιmιdazol agarose was based on epoxy activated 6 % agarose beads from Hispanagar and synthesised as described m example 1A and IE The hgand concentration was measured by elemental analysis and found to be 69 μmol /g wet, but dramed matrix

The sohd phase matrix was tested for it's polyclonal antibody bmding efficiency in a column according to the following procedure

1) The matrix was washed with water on a smtered glass funnel and finally dramed 2 gram of wet, but dramed sohd phase matrix is weighed mto a small column (mner diameter of 5 mm) The matrix was washed with 5 ml of buffer (10 mM sodium citrate pH 7 0 ) 2 ml of bovine serum was apphed to the column The column was washed with 10 ml of washing buffer (10 mM sodium citrate, 0 25 M NaCl pH 7 0) The matrix was eluted with 20 ml of elution buffer (10 mM sodium citrate pH 3 0) The flow rate was 1 0 ml/mm

2) Analyses were performed to determine the relative distribution of immunoglobuhn between the non-bound fraction in the run-through after binding, the washing fraction(s) and the eluate. This was done by single radial immunodiffusion (as described in D. Catty and C Raykundalia "Antibodies - a practical approach" Vol I, pp.137-168, 1988) using rabbit anti cow δ immunoglobuhns (D>AKO, Denmark cat.no.: Z247) as the precipitating antibody

The binding capacity was then calculated from the amount of non-bound immunoglobulin present in the run-through and expressed as a percentage of the total amount added to the matrix in the raw material. 0

The yield and purity was determmed as described in example 4.

Results

Binding capacity 85 %

Purity 80 %

Yield 85 %

5 11. Isolation of immunoglobulins from egg yolk

2-mercaptobenzimidazol as the ligand

The following example illustrates that it is possible to isolate immunoglobulins from egg yolk with 2-mercaptobenzimidazol as the hgand.

Egg yolk: Egg yolks (from normal chicken eggs) were diluted 1: 1 with 0.25 M NaCl. The sample was centrifuged in 20 minutes at 10.000 rpm.

Solid phase matrix: 2-mercaptobenzimidazol agarose was based on epoxy activated 6 % agarose beads from Hispanagar and synthesised as described in example 1A and IE. The hgand concentration was measured by elemental analysis and found to be 69 μmol /gram wet, but drained matrix.

The sohd phase matrix was tested for it's efficiency to bind immunoglobulins from egg yolk in a 0 column according to the following procedure:

1) The matrix was washed with water on a sintered glass funnel and finally drained. 2 gram of wet, but drained sohd phase matrix is weighed into a small column (inner diameter of 5 mm). The matrix was washed with 10 ml of buffer. (10 mM KH2PO4 6.1 ). 4 ml of the sample was

apphed to the column The column was washed with 10 ml washmg buffer (10 mM KH2PO4 6 I) The matrix was eluted with 16 ml of elution buffer (10 mM sodium citrate pH 3 5)

2) Analyses were performed to determme the relative distribution of immunoglobuhn between the non -bound fraction m the run-through after bmdmg, the washmg fractιon(s) and the eluate This was done by smgle radial immunodiffusion (as described m D Catty and C Raykundaha "Antibodies - a practical approach" Vol I, pp 137-168, 1988) usmg rabbit anti chicken IgG (Sigma, USA cat no C-6409) as the precipitatmg antibody

The binding capacity was then calculated from the amount of non-bound immunoglobuhn present in the run-through and expressed as a percentage of the total amount added to the matrix in the raw material

The yield and purity was determmed as described in example 4

Results

Bmding capacity 80 %

Purity 60 %

Yield 80 %

12. Depletion of IgG and haemoglobin from fetal calf serum

The followmg example illustrates the efficiency of some of the sohd phases accordmg to the mvention to deplete IgG and haemoglobm from fetal calf serum The study was designed to determme the bmdmg efficiency durmg batch mcubation with 14 different sohd phases

Fetal calf serum" The fetal calf serum was obtamed from freshly drawn blood by mild centnfugation after coagulation for 24 hours at room temperature

Solid phase The followmg sohd phases were used 2-mercaptobenzιmιdazol agarose, thiophenol agarose, 4-chlorothιophenol agarose, 2-ammothιophenol agarose, 4-methylmercaptoanιhne agarose, 2- mercapto-5-mtrobenzιmιdazole agarose, benzylmercaptan agarose, 2-chlorophenol agarose, 3- chlorophenol agarose, 4-chlorophenol agarose, 2-mercaptobenzoxazol agarose, 2-mercaptopyndme agarose, 2,5-dιmercapto-l,3,4,-thιadιazol agarose, 6-ethoxy-2-mercaptobenzothιazol agarose j l these agaroses were based on epoxy activated 6 % agarose beads from Hispanagar and synthesised as

described in example 1A and IE. The hgand concentration was measured for all matrices by elemental analysis and found to be in the range of 60-70 μmol /g wet, but drained matrix.

The sohd phase matrices were tested for their efficiency to deplete IgG and haemoglobin from fetal calf serum according to the following procedure:

1) The matrix was washed with 0.25 M NaCl on a sintered glass funnel and finally drained. 0.5 gram of wet, but drained sohd phase matrix is weighed into a test tube and added 5 ml of fetal calf serum. The suspension was then incubated on a roller mixer for two hours at room temperature.

2) After incubation the test tube was centrifuged for 5 min. at 2000 RPM to sediment the matrix and a sample of the supernatant was taken out for determination of the amount of IgG left in the serum. This was done by single radial immunodiffusion (as described in D. Catty and C Raykundalia "Antibodies - a practical approach" Vol I, pp.137-168, 1988) using rabbit anti cow immunoglobuhns (DAKO, Denmark, cat.no.: Z247) as the precipitating antibody.

3) The concentration of haemoglobin was measured spectrophotometrical at 414 nm. The percentage of haemoglobin left unbound in the serum is calculated as:

0AbS414 nm, absorbed fetal calf serum/AbS414 nm, fetal calf serum)x 100%

Results

Sample % haemoglobin % IgG left left in serum in serum

2-mercaptobenzimidazol agarose 60 30

Thiophenol agarose 60 50

4-chlorothiophenol agarose 40 40

2-aminothiophenol agarose 60 30

4-methylmercaptoaniline agarose 90 60

2-mercapto-5-nitrobenzimidazole agarose 30 30

Benzylmercaptan agarose 70 60

2-chIorophenol agarose 80 60

3-chlorophenol agarose 70 50

4-chlorophenol agarose 75 50

2-mercaptobenzoxazol agarose 75 50

2-mercaptopyridine agarose 75 60

2,5-dimercapto-l,3,4,-thiadiazol agarose 10 40

6-ethoxy-2-mercaptobenzothiazol agarose 60 60

13. Isolation of trypsinogen and chymotrypsinogen from bovine pancreas with 2- mercapto-benzoic acid agarose

The following example illustrates the use of 2-mercapto-benzoic acid agarose as a suitable matrix for isolation and purification of proteases, e.g. trypsin and chymotrypsin from bovine pancreas.

Pancreas extract: The two proteases were isolated as the proenzymes trypsinogen and chymotryp¬ sinogen from a bovine pancreas extract produced by extraction with sulfuric acid as described in M. Laskowski, Methods in Enzymology, vol. II, pp 9-10, 1955. After extraction the suspension was adjusted to pH 2.5 by addition of 2 M sodium hydroxide and clarified by filtration and centnfugation for 30 min. at 4000 RPM. Just prior to purification the extract was adjusted to pH 4.5 with 2 M NaOH and centrifugated at 4000 rpm for 5 minutes the supernatant was collected.

Solid phase matrix 2-mercapto-benzoιc acid agarose was based on epichlorohydrm activated 6 % agarose beads from Hispanagar and synthesised as descnbed in example 1 A and 1 E The hgand concentration was measured by two different methods and found to be 65 μmol /g wet, but dramed matrix as determmed by elementary analysis and 60 μmol /g as determmed by acid-base titration ol the immobilised benzoic acid part of the hgand

The sohd phase matrix was tested for the efficiency to bind trypsin and chymotrypsin accordmg to the followmg procedure

1) The matrix was washed with water on a smtered glass funnel and finally dramed 2 5 gram of wet, but dramed solid phase matrix is weighed mto a column (mner diameter of 5 mm) The matrix was washed with 10 ml buffer (10 mM sodium citrate pH 4 5) δO ml of the extract was applied to the column The matrix was washed with 15 ml of washmg buffer (10 mM sodium citrate pH 4 5) The matrix was eluted with 10 ml elution buffer (50 mM boric acid 0 5 M NaCl pH 8 7)

2) The purity of the eluate was analysed by SDS-PAGE (sodium dodecyl sulfate polyacrylamide electrophoresis) under reducmg conditions followed by stammg of the protem bands with coomassie brilliant blue (Precast gel 4-20 % tns-glycme, 1 mm cat no EC6025 runnmg 1 hour at 30 mA, tris- glycme SDS runnmg buffer cat no LC2675, tns-glycme sample buffer cat no LC2676 coomassie stammg kit LC6025 all chemicals from Novex, USA)

Results

Total amount of protem m eluate 65 mg

Trypsin m eluate 35 %

Chymotrypsin in eluate 20 %

As can be seen from these results it is surprisingly found that this type of hgand 1 e aromatic hgands compnsmg an acidic group accordmg to the mvention, here represented by 2-mercapto-benzoιc acid as the specific hgand, are able to bmd very efficiently protems such as proteases at relatively low pH values and at relatively high ionic strength (I e approx 025 m ionic strength)

14. Purification of immunoglobulins from horse serum

The followmg example illustrates the use of 2-mercapto-benzιmιdazole coupled to agarose beads for purification of immunoglobuhns from horse serum It further illustrates the use of different elution conditions with this type of matrix

Horse serum: The horse serum was obtamed from freshly drawn blood by mild centrifugation after coagulation for 24 hours at room temperature.

Solid phase matrix: 2-mercapto-benziidazole agarose was produced as described in example 1A and 1 E. The hgand concentration was determined to be 69 μmol/g wet but suction drained matrix.

Procedure:

1) The matrix was washed with water on a sintered glass funnel and finally drained. 2 g of wet, but drained 2-mercapto-benzimidazole agarose is weighed into a small column (inner diameter of 5 mm). The matrix was washed with 5 ml of 10 mM potassium phosphate, pH 7.0. 2 ml of horse serum adjusted to pH 7.0 with 0.1 M HC1 was apphed to the column. The column was washed with 10 ml of washing buffer (10 mM potassium phosphate, 0.1 M NaCl, pH 7.0). The matrix was eluted with 20 ml of elution buffer (see below). The flow rate was 1.0 ml min.

This procedure was followed in three identical experiments except for the use of three different elution buffers:

Elution buffer A = 20 mM sodium citrate pH 3.0

Elution buffer B = 50 mM ethanol amineΗCl pH 11.0

Elution buffer C = 10 M potassium phosphate pH 7.0 + 30 % v/v 1,2-propane diol

2) Analyses were performed to determine the relative distribution of immunoglobulin between the non-bound fraction in the run-through after bmding, the washing fraction(s) and the eluate. This was done by single radial immunodiffusion (as described in D. Catty and C Raykundalia "Antibodies - a practical approach" Vol I, pp.137-168, 1988) using rabbit anti horse immunoglobulin G (Sigma, USA, cat.no.: H-9015) as the precipitating antibody.

The binding capacity was then calculated from the amount of non -bound immunoglobuhn present in the run-through and expressed as a percentage of the total amount added to the matrix in the raw material. The yield and purity was determined as described in example 4.

Results:

Elution buffer A Elution buffer B Elution buffer C

Binding capacity % 85 85 85

Yield % 80 85 80

Purity % 90 90 85

The results mdicated that 2-mercapto-benzιmιdazole agarose is an efficient sohd phase matrix for purification of horse immunoglobuhns and that elution may be performed with either weakly acidic or weakly basic buffers or alternatively with a neutral buffer compnsmg a non-toxic organic solvent such as 1,2-propanedιol without compromising yield and purity of the eluted immunoglobuhn

15. Bovine serum albumin binding efficiency by different solid phase matrices

The followmg example illustrates the efficiency of different sohd phase matrices in a standard bmding assay for bovme serum albumm

Solid phases A selected range of sohd phase matrices were produced on the basis of epichlorohydrm activated agarose beads from Hisapanagar as described in example 1 A and 1 E The hgands tested are listed in the table below

Bovme serum albumin solution pH 4 0 (BSA pH 4 0) Purified bovme serum albumin (Biofac AS Denmark) was dissolved to a fmal concentration of 10 mg/ml m 20 mM sodium citrate pH 4 0 + 0 2 M sodium chloride

Bovine serum albumin solution pH 70 (BSA pH 70) Purified bovme serum albumin (Biofac AS, Denmark) was dissolved to a fmal concentration of 10 mg/ml m 20 mM sodium citrate pH 7 0 + 0 2 M sodium chloride

Procedure

Standard albumin binding assay

The sohd phase matrices were washed with 10 volumes of deminerahsed water on a vacuum suction filter and dramed by gentle suction for 1 mm Two samples of 1 0 g suction dramed matrix were then weighed mto two 10 ml test tubes followed by the addition of 6 0 ml of BSA pH 4 0 to the first test tube and 6 0 ml BSA pH 7 0 to the second test tube Two 1 0 g samples of non -derivatised suction dramed plain agarose beads from Hispanagar were also added 6 0 ml of the two BSA solutions as negative controls The test tubes were then close with a stopper and the suspension mcubated on a roller mixer for 2 hours at room temperature (20-25 °C) The test tube was then centrifuged for 5 mm at 2000 RPM m order to sediment the matrix The supematants were then isolated from the sohd phase matrix by pipettmg mto a separate test tubes, avoiding the carry-over of any matrix particles and filtered through a small non-adsorbmg 0 2 μm filter O ilhpore USA) Followmg this a determmation of the concentration of non-bound BSA in the

supernatant is performed by measuring the optical density (OD) at 280 nm on a spectrophotometer.

The amount of BSA bound to the matrices were then calculated according to the following formula:

mg BSA bound per g suction drained matrix = (l-(OD of test supematant/OD of control)) x 60

The precision of this method is better than +/- 5 %.

Results:

The table gives the amount of BSA bound in mg per gram wet, but suction drained matrix as a function of the coupled hgand and the pH of the BSA solution.

Ligand coupled to sohd phase matrix BSA pH 4.0 BSA pH 7.0

3-hydroxy-benzoic acid 22 0

4-hydroxy-benzoic acid 5 0

3,5-dihydroxy-benzoic acid 36 58

2,4-dihydroxy-benzoic acid 41 0

2-hydroxy- l-naphthalic acid 58 0

3-amino-benzoic acid 56 0

2-amino-benzoic acid 51 0

4-amino-benzoic acid 59 0

3,4-di-amino-benzoic acid 9 0

5-amino-iso-phthalic acid 17 51 l-amino-2-naphthol-4-sulfonic acid 21 52 p-coumaric acid 26 0

2-mercapto-benzoic acid 59 0

2-mercapto-nicotinic acid 30 0

5-mercapto- 1 -tetrazol-acetic acid 26 0

3-amino-l,2,4-triazol-5-carboxylic acid 6 30

2,5-di-mercapto-l,3,4-thiadiazol 0 20

Ligand coupled to solid phase matrix BSA pH 4.0 BSA pH 7.0

2-amino-6-nitro-benzothiazol 28 46

2-mercaptobenzthiazol 42 45

Sulfa-thiazol 28 0

Sulfa-methizol 20 0

2-amino-pyridin 0 0

2-mercapto-pyridin 8 26

2-hydroxy-pyridin 0 19

2-mercapto-5-nitro-benzimidazol 58 47