WO1990011291A1

WO1990011291A1 - Monitoring reactions in solid-phase peptide synthesis by conductivity measurements

Info

Publication number: WO1990011291A1
Application number: PCT/GB1990/000422
Authority: WO
Inventors: Claus Schafer Nielsen
Original assignee: Biotech Instruments Limited
Priority date: 1989-03-21
Filing date: 1990-03-20
Publication date: 1990-10-04

Abstract

The invention provides a method of monitoring reactions in solid phase peptide synthesis, wherein the electrical conductivity of the reaction solution is monitored. Conductivity measurements provide an indication of the progress of reactions to completion, and may then be used to provide feed-back control in automatic or semi-automatic synthesis systems. The method may be used for syntheses in batch mode or in continuous flow systems. It is particularly suitable for use with syntheses based on the FMOC protection strategy. The conductivity monitored may be due to the production of ion pairs in the reaction solution during a coupling step. These ion pairs may be formed between a growing peptide chain and an acid catalyst used to catalyse the coupling step, or between ester groups released from an incoming amino acid during coupling and a proton acceptor present in the reaction solution. Washing and deprotection steps in a synthesis can also be monitored in accordance with the invention. The invention additionally provides apparatus suitable for use in the above described method.

Description

Title: Monitoring reactions in solid-phase peptide synthesi by conductivity measurements

DESCRIPTION

Field of Invention

This invention concerns peptide synthesis.

Background to the Invention

Peptides consist of linear chains of amino acids (each comprising an amino group, a carboxyl group, a hydrogen atom, and a distinctive side chain, all bonded to. a carbon atom), linked by peptide bonds between the amino group and the carboxyl group of adjacent amino acids. Techniques are known for the chemical synthesis of peptides by the sequential addition of desired amino acids, which form peptide bonds by a condensation reaction, resulting in a growing peptide chain.

Solid phase methods of peptide synthesis have been devised, in which the carboxyl terminus of the growing peptide chain is anchored to a solid support, and a desired sequence of amino acids added in stepwise manner to the amino terminus at the other end of the growing peptide chain. Briefly, the carboxyl terminus of a first, fully protected amino acid is attached to a support, generally of polymeric material but typically a polystyrene or a polyamide-based resin. The alpha-amino group is protected only temporarily, the protection being removed at each addition cycle. The side chain protection, blocking the reactive groups of each amino acid, is permanent and is only removed at the end of the synthesis. Side chain protection is usually by means of benzyl esters and ethers.

The stepwise synthesis cycle starts with the removal of the alpha-amino group protection (deprotection). After washing and neutralisation, the next amino acid, with a similarly protected alpha-amino group, is added in the presence of an activation agent. After this coupling reaction is complete, excess reagents are removed by washing. The procedure is repeated until the desired sequence of amino acids has been produced. At the end of a synthesis, all protecting groups are removed, and the peptide is cleaved from the solid phase support.

Such techniques are well known and are described in greater detail in, e.g. the work "Synthetic Peptides in Biotechnology" published in 1988 by Alan R. Liss, Inc. Solid phase peptide synthesis methods also lend themselves well to automation, as is described in, e.g. the article "Automation of Peptide Synthesis" by Raymond Newton and John E. Fox in the work mentioned above. Automatic and semi-automatic peptide synthesiser equipment is commercially available and is widely used.

For successful synthesis, all the reactions involved must go as far as possible to completion: any incomplete reactions lead to a serious build up of side products and deletion peptides which can only be removed at^' the end of the synthesis. It is accordingly important to be able to monitor the completeness of the reactions involved. especially since information thus obtained on the course of the various reactions can also be used to reduce the duration of coupling, deprotection and washing cycles in order to speed up the overall synthesis.

The pivotal reaction to monitor during a synthesis is the coupling reaction between the growing peptide chain and the next amino acid added to it. Near completeness of these acylation reactions during solid phase peptide synthesis by stepwise methods, such as that of Merrifield (8) and its later developments (6,9,10), is essential to ensure a high overall yield of the final peptides. The time course of the coupling reaction in each step is essentially unpredictable, being dependent on the sequence of the peptide, the type of amino acid added and the activation method used.

The ideal monitoring technique will offer continuous, non interactive monitoring of coupling, deprotection and washing steps by means of a simple set-up. It should not introduce additional reagent cycles or other time consuming procedures and it should occur in real time in order to allow instant, preferably automatic, control of the synthesis. In addition, the ideal monitoring method should be based on measurements of unreacted groups during coupling, as this gives the most sensitive indication of the completeness of the coupling reaction.

Traditional monitoring is performed by chemical analysis of small samples withdrawn from the support resin at various intervals during the coupling reaction (11). These methods include semiquantitative and quantitative versions for quantitating free amines with the ninhydrin reaction (12,13) or by fluorescamine (14). Whilst being universally applicable and accurate, chemical analysis on resin samples is difficult to incorporate in a fully automated system and, if incorporated (15), serves only to document the end point incorporations achieved after coupling or even after the whole synthesis has been completed. Moreover, this type of chemical analysis is inherently time consuming, destructive and discontinuous. Complex plumbing problems are involved thus increasing expense, and valuable sample is generally consumed in the process.

Titration methods have also been used for monitoring peptide syntheses. These take place in a non-destructive way on the entire batch of support resin and require no sampling. Titration of amino groups by picric acid followed spectrofotometrically (16), by perchloric acid followed potentiometrically (17) or by other means (see discussion in (16)) determines the quantity of unreacted amino groups present during coupling reactions or of free amino groups present after deprotection. Titration methods can thus give an accurate picture of the reactions taking place during a synthesis and even allow for feedback operations. They do, however, share the drawback of introducing additional reagent cycles to the peptide synthesis, some of which may be harmful to the peptide chain or the protecting groups employed.

These drawbacks are partly eliminated in a recently introduced technique (18), in which unreacted amino groups on the resin are coupled with dimethoxytrityl chloride (DMT) added to the resin after elution of the activated amino acid. Spectrofotometric determination of the trityl compound eluted in dilute trichloroacetic acid is used as a measure of the incompleteness of the coupling reaction. This technique leaves the resin-bound peptide in a still protected state such that repetition of the coupling, or other corrections, can be made in case of incomplete reaction. However, the technique cannot be used to make real time measurements of the coupling reactions, and it necessitates the use of additional solvent and reaction cycles.

Recently, taking advantage of the fact that chromophores are constituent parts of peptide synthesis procedures, other approaches have been developed using direct spectrofotometric methods readily adaptable to automated systems. FMOC protecting groups (19), used to protect the alpha-amino groups of the growing peptide chain, can be measured in the liquid phase during deprotection steps and during couplings (20,21). This, however, gives only limited and inaccurate information about the time course of couplings (it measures small changes in a large amount of circulating surplus of FMOC-protected amino acid, yielding a not very easily interpreted set of data), while the deprotection step, in contrast, is monitored accurately.

In an alternative method, the yellow colour produced when the carboxyl-activating group dihydroxybenzotriazole (Dhbt) (3) is deprotonized by unreacted amino groups on the resin can be measured and used in real time monitoring of coupling reactions (4,22). This elegant method is applicable to coupling reactions employing Dhbt-esters and can be used with other ester types as well if free Dhbt is added to the coupling suspension. The method, however, requires technically complicated optical monitors for measurements directly on the resin and requires the use of resins with high transparency or with well defined light reflection properties (22). Furthermore, the technique puts some restraints on the chemistry used, since reactions where tertiary amine is added as a catalyst (e.g. during esterification of the first amino acid to the linker (4) for the purposes of protection) cannot be monitored in this way. The technique is further of little use in the monitoring of washing steps.

Thus, whilst a wide number of approaches to the problem of monitoring peptide syntheses have been attempted, few have proved entirely satisfactory. Techniques based on measurements of light absorption by reactants have been the most popular of the non-destructive methods, but even these require the use of expensive optical systems, whilst the high concentrations of chemicals involved in the synthesis make accurate measurements difficult.

The present invention therefore aims to provide an alternative approach to monitoring reactions in solid phase peptide syntheses.

Summary of the Invention

According to the present invention there is provided a method of monitoring reactions in solid phase peptide synthesis, comprising monitoring the electrical conductivity of the reaction solution.

Acidic and basic species are involved in the reactions occuring in solid phase peptide synthesis, so by monitoring the conductivity of the reaction solution an indication can be obtained of the rate of reaction for both removal of protecting groups and coupling of incoming amino acid groups. From this kinetic data, predictions can be made as to the expected extent of reaction, and synthesis conditions can be modified as appropriate if required.

The invention also provides a method of solid phase peptide synthesis, characterised by monitoring the electrical conductivity of the reaction solution.

The invention is suitable for use during peptide synthesis in a batch mode, but is also applicable to continuous flow methods of solid phase peptide synthesis.

The invention is primarily applicable for use with the FMOC protection strategy of peptide synthesis, which is currently regarded as the most successful peptide synthesis method and an example of which is described below, but can also be used in other solid phase peptide syntheses involving acidic and basic species, such as the BOC method. Both of these techniques are described in the article by Newton and Fox referred to above. A number of variants and modifications of the FMOC and BOC methods have been devised, using different catalysts, linking groups etc., as are known to those skilled in the art, and the invention is not intended to be limited to any particular method of this type.

Preferably, it is the change in conductivity of the reaction solution due to the production of ion pairs in the solution during a coupling step which is monitored during the synthesis.

In a first version of the invention, the coupling step is acid catalysed so as to allow formation of ion pairs between the deprotected growing peptide chain and the acid catalyst.

The catalyst conveniently used is hydroxybenzotriazole (HObt). It is thought that this catalyst, acting as a proton donor, forms an ion pair with the deprotected alpha-amino group of the growing peptide chain (i.e. an NH + group is created at the free end of the chain). The formation of these ion pairs will cause an initial increase in conductivity of the reaction solution on addition of the HObt catalyst. As the incoming amino acid reacts with the growing peptide chain during the coupling step, the ion pairs are removed as the alph-amino groups become involved in coupling. Thus, conductivity will fall again during coupling, and these changes in conductivity can be used to monitor the progress of the coupling reaction.

In this first version of the invention, the conductivity is conveniently measured by use of two spaced apart electrodes located in a reaction vessel for synthesis reaction solutions, and by applying an AC voltage across the electrodes. By amplifying and rectifying the AC voltage resulting from current flowing through a reaction solution, a measure of the electrical conductivity of the solution can be obtained, with conductivity being proportional to the AC signal value.

Thus, the present invention also provides apparatus for the solid phase synthesis of peptides, characterised in that it comprises a reaction vessel having two spaced apart electrodes located therein; means for applying an AC voltage across the electrodes; means for amplifying and rectifying the AC voltage resulting from current flowing through a reaction solution in the vessel; and means for displaying the resulting AC signal value.

The electrode geometry is not critical, and electrodes can be fitted to any suitable reaction vessel, possibly forming part of an automatic or semi-automatic solid phase peptide synthesiser.

The amplification and rectification circuitry may be of conventional construction, as will be well known to those skilled in the art.

The display means may comprise any convenient form of display such as a numerical display, a visual display unit or a graphical display, as will also be well known to those skilled in the art.

In one typical embodiment two platinum flashed, tungsten electrodes are located in a reaction vessel with a spacing of 10mm, and an electrode voltage of 0.2 volts AC is applied, operating at 1 KHz.

In a second version of the invention, the ion pairs are formed when the incoming amino acid to be coupled with the peptide chain being sythesised is in the form of an amino acid ester, since the ester group will then be released in its free acid form during coupling. In the presence of a suitable proton acceptor, an ion pair will be formed between the released acid group and the proton acceptor, resulting in an increase in the electrical conductivity of the reaction solution due to the presence of the ion pairs.

A suitable proton acceptor might conveniently be a tertiary amine such as diisopropylethyl amine (DIEA) . However, the proton acceptor should not itself be acylated in the presence of the incoming amino acid. DIEA fulfils this criterion and also, usefully, reacts only slowly with the base labile FMOC group often used to protect the alpha-amino group of a growing peptide chain.

The incoming amino acid is preferably in the form of a pentafluorophenol ester, however, the method of the present invention is equally applicable for monitoring reactions involving dihydroxybenzotriazole esters or symmetric anhydrides.

The measurements made during the monitoring of a synthesis may be fed back to a computer or other data processor, and data obtained from the measurements may be used to control the duration of subsequent synthesis steps. Thus, the method of the present invention is particularly useful in computer-controlled, or other automatic or semi-automatic, peptide syntheses, since measurements can be carried out on the system in real time and then used to provide control feed-back so as to maximise the efficiency of a synthesis.

According to the second version of the present invention, conductivity is conveniently measured by means of two spaced apart electrodes located in a reaction vessel for synthesis reaction solutions, an AC voltage then being applied across the electrodes. Again, by amplifying and rectifying the AC voltage resulting from current flow through a reaction solution, a measure of the conductivity of the solution can be obtained, conductivity being proportional to the AC signal value.

The exact electrode geometry and reaction vessel shape are not critical, however, the reaction vessel preferably takes the form of a flow-through chromatographic column having an upper and a lower electrode, one at each of its ends.

The upper electrode is preferably a ϋ-shaped rod inserted in the upper end of the reaction vessel, such that reactants injected into the reaction vessel are guided into the ϋ-shaped cavity provided by the electrode and onto the surface of a solid phase support resin contained in the reaction vessel. In this way, reactants can be delivered onto the reaction surface without "splashing". The upper electrode is conveniently made of electroplated stainless steel.

The lower electrode preferably comprises an end piece located in the lower end of the reaction vessel, which end piece is made of PTFE (polytetrafluoroethylene) containing 30% graphite.

During conductivity measurement carried out in accordance with the present invention, a quantity of 3-hydroxy 4- oxodihydrobenzotriazol (Dhbt) is preferably included in the reaction solvent used during deprotection steps, so as to increase conductivity of the reaction solution and hence aid accurate measurement.

The invention can thus provide a novel, simple, non¬ destructive method of detecting the rate of reactions for both the removal of protecting groups and the subsequent coupling of the incoming amino acid. The method is relatively insensitive to reactant concentration; capable of operating on any scale of synthesis; very cheap; and capable of providing accurate kinetic data, which enables a prediction to be made as to the expected extent of reaction. Moreover, adaptation of the reaction vessel to accommodate this new method is relatively simple.

The invention will be further described, by way of illustration, with reference to the accompanying drawings in which:

Figure 1 illustrates the reactions involved in the solid phase synthesis of peptides using an example of FMOC chemistry;

Figure 2 is a schematic illustration of automatic solid phase peptide synthesiser apparatus in accordance with the present invention, for use in a method of the first version of the present invention;

Figure 3 illustrates the reaction vessel of the apparatus of Figure 2;

Figure 4 illustrates the control circuitry of a conductivity sensor for monitoring conductivity of reaction solutions in the reaction vessel of the apparatus of Figure 2;

Figures 5, 6 and 7 are graphs of typical outputs from the conductivity sensor of the apparatus of Figure 2;

Figure 8 illustrates a reaction vessel for use in a method of the second version of the present invention;

Figure 9 is a graph illustrating the change in conductivity with concentration of solutions of FMOC- glycine, pentafluorophenol free acid and Dhbt in a DIEA/DMF mixture;

Figure 10 illustrates the correlation found between conductivity in a reaction solution during a peptide synthesis and the number of FMOC groups inserted on the support resin during coupling reactions;

Figure 11 illustrates the baseline drift in conductivity of solutions of free pentafluorophenol acid; and

Figure 12 shows the results of a continuous conductivity recording made in accordance with the second version of the present invention.

Detailed Description of the Drawings

Referring firstly to Figure 1 , the reactions involved in a conventional solid phase peptide synthesis (in this case, a batch synthesis rather than continuous flow, although the chemistry involved in these two cases is essentially the same) using the FMOC protection strategy are illustrated.

Using this strategy, FMOC (9-fluorenylmethoxycarbonyl-), the structure of which is illustrated in Figure 1 , is used to protect the alpha-amino group of the growing peptide chain, the caryboxyl terminus of which is attached to a polymeric support, e.g. functionalised polystyrene or polyamide beads, as is shown at (1). The beads will typically be poured into a reaction vessel, and will undergo the following reactions.

After washing with dimethyl formamide (DMF) and drying to remove solvent, the FMOC protecting group is removed by reaction with a base, e.g. 20% piperidine in DMF. This is the deprotection step, shown at (2). The product of deprotection is compound (3), which is a weak base.

The next amino acid to be linked to the growing peptide chain (in this case AA_) is generally added as the activated o-pentafluorophenyl (OPFP) ester (4), in the presence of a suitable catalyst. This catalyst may be a weak acid such as HObt, in which case ion pairs will be formed between the weak acid catalyst and the weak base (3). This causes an increase in the electrical conductivity of the reaction solution which will then fall off again as the coupling reaction proceeds, which changes in conductivity can be used to monitor the course of the coupling reaction in accordance with the first version of the present invention.

The coupling reaction involves acylation of the peptide chain by the incoming amino acid to form compound (5). The ester group of the incoming acid is released in its free acid form. This free acid, in the presence of a proton acceptor, forms ion pairs, the presence of which affects conductivity of the reaction solution and can be used to monitor the coupling reaction, in accordance with the second version of the present invention (see below).

Thus, two alternative methods for monitoring conductivity during a synthesis emerge, one in which the coupling step is acid catalysed (the first version) and one in which a proton acceptor is present to form ion pairs with released ester groups (the second version).

Following the coupling reaction, the deprotection and further coupling cycle begins again, with subsequent amino acids AA ....A being added one at each cycle. When the desired peptide chain is complete, washing with aqueous acid will remove it from the support resin to produce the end product (6).

The FMOC technique may of course be modified or varied in a number of ways. In particular, the incoming amino acid may be added in forms other than the OPFP ester.

Figure 2 illustrates schematically the overall layout of automatic solid phase peptide synthesiser apparatus for use in carrying out a batch synthesis such as that described in Figure 1. The apparatus comprises a reaction vessel 7 and sources 8 of reagents, including different amino acids (represented as AA. , AA_, etc.), linked by tubes and valves arranged in such a way that specified quantities of specified reagents can be delivered to the reaction vessel 7 in a specified sequence by a positive nitrogen gas pressure delivery system under control of control means (not shown) in conventional manner. On-off valve 9 is positioned in the supply tubes between sources 8 and the reaction vessel. The control means also controls delivery of nitrogen gas to the reaction vessel for agitating the contents and for removal of reagents from the reaction vessel at appropriate stages.

Reaction vessel 7 is illustrated in further detail in Figure 3. The vessel is generally of conventional construction, and comprises a generally conical glass vessel with a sintered glass frit 10 positioned at its lower end and a ground glass Quickfit (Trade Mark) joint 11 at the top. This construction allows the polymeric support material, usually polystyrene or polyamide based resin beads, to which the growing peptide is attached, to be poured into the reaction chamber. A connector 12, having a male ground glass joint 13, and also fitted with a sintered glass frit 14 to contain the polymeric support material within the reaction vessel, fits into the top of the vessel and in use is attached to the vessel by standard laboratory coiled spring connectors. Connector 12 includes an upper inlet tube 15 and vessel 7 includes a lower outlet tube 16, both of which tubes are modified to take standard Omnifit screw-on connectors, enabling the vessel to be connected into the apparatus of Figure 2.

In accordance with the invention, vessel 7 has a pair of platinum flashed, tungsten electrodes 17 and 18 cemented into the sides of the vessel, the inner ends of the electrodes being spaced apart by a distance of 10mm.

Electrodes 17 and 18 are designed to measure the conductivity of reaction solutions contained in the vessel 7, and are linked to a conductivity sensor having circuitry as shown in Figure 4.

Briefly, an AC voltage of 0.2 volts operating at 1 KHz is applied across the electrodes, the AC voltage resulting from current flowing through reaction solution in the vessel 7 is amplified and rectified, and the resulting signal strength is measured and displayed on display means (not shown). Signal strength is a measure of conductivity, i.e. the greater the signal the higher the conductivity of the solution.

The variation of conductivity with time is conveniently obtained as a graphical representation such as is illustrated in Figures 5, 6 and 7, e.g. on a printer associated with the conductivity sensor, but other representations of conductivity can also be obtained, as will be apparent to those skilled in the art.

The illustrated apparatus is used in generally conventional manner for batch synthesis of peptides.

When the reaction scheme illustrated in Figure 1 is employed, use of the apparatus in accordance with the first version of the present invention is as follows.

The base of the peptide chain (i.e. amino acid AA..), attached to its polymeric support material in reaction vessel 7, is washed with DMF and dried to remove solvent.

The FMOC protecting group is removed by reaction with 20% piperidine in DMF, to produce compound (3). The rate of removal of the FMOC group varies according to the conformation of the peptide at the particular stage in the synthesis. Slow removal indicates probable folding or insolubility of the growing peptide chain. If the rate is excessively slow, then the reaction time can be extended or the reaction repeated to ensure completion. The peptide is then ready for the addition of the next amino

The normal method is to add the amino acid (e.g. AA_) as an active OPFP-ester with hydroxybenzotriazole (HObt) as catalyst. The HObt is a weak acid and this protonates the exposed peptide amino group, causing an increase in conductivity of the reaction solution as the acid-amino group ion pairs are formed (during this reaction, the support resin behaves essentially as an ion exchange resin). As the protonated amino group reacts with the incoming amino acid the conductivity of the reaction solution falls again. As the rate of fall of signal from the conductivity sensor is directly proportional to the rate of formation of the peptide bond, this can be used to calculate the half life of the reaction. From the half life the time required for the reaction to be completed to the required specification (e.g. greater than 99.5%) can be calculated. After coupling, the peptide and resin are washed to remove excess reagents and are then ready for the next addition cycle.

The process is repeated until the desired sequence of amino acids has been produced. All protecting groups are then removed and the product peptide is finally cleaved from the solid support by treatment with an aqueous acid.

The FMOC technique may be modified or varied in a number of ways. For example catalysts such as dicyclohexyl- carbodiimide (DCC) may be used in place of HObt, and amino acids may be added in forms other than OPFP-esters.

By using the half life of the reaction, it is not necessary to attempt to measure when, say, 99.5% of the reaction is over: this involves trying to monitor a very small change, +/-0.1%, in a large number, 99.5%, and so is very prone to error. Instead, by calculating the half life, a prediction can be made early on in the reaction as to the probable time required for completion of the reaction to the required specification, so this approach is not dependent on trying to measure small differences in large numbers.

This method lends itself very readily to the automation of the entire peptide synthesis cycle. Fitted to an automatic peptide synthesiser, the conductivity sensor can be easily monitored by control means; the required kinetic data can then be calculated and appropriate action taken. Problems such as slow reactions are countered by extending reaction times. Insolubility problems, often shown by a flattened reaction curve, could be overcome by changing solvents. Correct delivery of reagents can also be confirmed.

The method also offers an advantage over many other monitoring techniques in that no additional chemical ingredients are involved in the monitoring process; the HObt is present as a catalyst anyway.

Typical outputs from the sensor in its present form are shown in Figures 5, 6 and 7.

A typical output from the conductivity sensor shows several distinct regions, as illustrated in Figure 5. Region (a) indicates the start of the coupling cycle when there is little or no output from the sensor. Region (b) is where piperidine has been added, and the sensor output rises as the FMOC group is removed. The level of the response at (b) indicates the quantity of growing peptide present. The unbound products of this reaction are then removed by solvent washing - region (c). Since the released FMOC group forms weak carbonates in solution, its release causes an increase in conductivity, whereas its removal by solvent washing causes conductivity to fall again.

A second quantity of piperidine is then added at (d): this ensures complete removal of the FMOC groups. All the unbound products of these reactions are removed at (e) with solvent washes: this results in a further decrease in monitor output until the next amino acid derivative to be coupled is added at (f). This results in a large increase in conductivity, which falls off during (g) as the coupling reaction proceeds until completion at (h). The slope of the region at (g) seems to be proportional to the rate of the coupling reaction.

It will be clear from Figure 5 that the method of the present invention can be used to monitor the progress of deprotection as well as coupling reactions.

Figures 6 and 7 help to illustrate the use of the conductivity sensor and also some of the typical problems encountered in solid phase peptide synthesis. The reference lettes (a) to (g) refer to regions of the graphs corresponding to those shown in Figure 5.

Figure 6 shows the case where the coupling reaction is proceeding slowly and shows how the sensor can be used to follow the coupling reaction and extend the reaction time if necessary.

Figure 7 shows the case where removal of the FMOC group proceeds more slowly than normal and indicates how the deprotection step can be automatically extended following the output of the conductivity sensor. We have discovered that this seems to give an indication of steric hindrance occuring in the growing peptide chain and therefore potentially can be used to gain further information about possible mechanisms involved in the coupling reaction.

Cases in which no HObt or equivalent catalyst has been present have not resulted in the characteristic reaction curve of Figures 5 to 7. This supports the explanation given above of the way in which HObt affects conductivity.

The reaction vessel shown in Figure 8 is for use in a the present invention. The vessel is designed as a flow- through chromatographic column having a 12mm I.D. glass tube 19 equipped with end pieces 20 made of PTFE containing 30% graphite. This material is in itself electrically conducting and thus serves as an electrode in the lower end of the reactor. The upper electrode 21 is a U-profiled rod made of electroplated stainless steel. This electrode is inserted in the centre of the upper end of the reactor so that reactants injected into the reactor are guided into the ϋ-shaped cavity of the electrode to the surface of the support resin without "splashing" of reactants.

The reaction vessel is also equipped with a filter 22 and electrode jacks 23.

A typical peptide synthesis might be carried out in the reaction vessel shown in Figure 8 and monitored as follows.

The monitoring method is based on continuous measurement of electrical conductivity of reaction solutions in the reaction vessel. It is shown that there is a close correspondence between the degree of acylation of the growing peptide chain (as determined from the number of FMOC groups released during deprotection) and the conductivity profile obtained during coupling of incoming amino acids to the growing peptide chain. Measurements taken are fed back to a computer so as to provide data for software control of the duration of the acylation, deprotection and washing steps involved in the synthesis. The method is described for use with pentaf luorophenol- esters (see references (1) and (2)), but is equally applicable to dihydroxybenzotriazole esters (see (3) and (4)) and symmetric anhydrides using the FMOC-polyamide strategy (5,6,7) in a continuous flow set-up with DMF as the general solvent.

a) Materials used

Pentafluorophenyl (Pfp) esters of FMOC amino acids and the acid labile resin (4-hydroxymethyl phenoxyacetic acid on polydimethylacrylamide in kieselguhr, Pepsyn KA) were obtained from Milligen, Bedford, MA. FMOC-Ser and FMOC-Thr were delivered as dihydroxybenzotriazole (Dhbt) esters from the same manufacturer. Dhbt (free acid, analytical grade) was purchased from Fluka, Switzerland. Dimethylformamide (DMF); pentafluorophenol (Pfp-OH) ester (free acid); piperidine and 4- ( dimethyl Jaminopyridine (DMAP), all of reagent grade, were purchased from Merck, Darrmstadt. N,N-diisopropylethylamine (DIEA) "peptide synthesis grade" was from Applied Biosystems, Foster City, CA. The syntheses was peroformed on an EASY-PEP computer controlled continuous flow peptide synthesizer from Kem- En-Tec, Copenhagen. A model 525 conduct imeter from CRISON, Barcelona, was used for recording electrical conductivity in the reaction vessels and the voltage output was interfaced to the computer by means of one of the 12-bit A/D converters in the synthesizer.

b) Method

The reaction vessel of Figure 8 is incorporated into a standard continuous flow set-up for solid phase peptide synthesis. Between 4 and 5 ml of pre-swollen resin containing approximately 50 umoles of linker per ml is used as the solid phase support for the synthesis. Just prior to injection into the reaction vessel, a 3-fold molar excess of FMOC-amino acid ester (approx. 100 mM) is dissolved in 1.5 times the resin volume of 15 mM DIEA in DMF. The injection is carried out with a pump speed of 2.5 ml per min and no recirculation of the ester is performed after injection of the entire volume into the reaction vessel. Excess reactant solution is employed to ensure an even distribution of the amino acid ester throughout the resin. During coupling of the first amino acid to the linker, 20 mM 4,4 dimethylaminopyridine (DMAP) is included in the reaction solvent. Deprotection is carried out in 20% piperidine, 5 mM Dhbt in DMF for 10 minutes. Inclusion of Dhbt in the deprotection solvent is not essential, but serves to increase its conductivity, thus making it easier to monitor the loading of piperidine into the reaction vessel.

In this experiment, determination of the amount of amino groups acylated during the couplings was made by absorbance measurement at 280 nm of the FMOC-groups released from the resin during deprotection in 20% piperidine in DMF. The resin was washed thoroughly in DMF before the deprotection step in order to remove the unreacted amino acid and the released ester groups that would otherwise interfere with the readings at 280 nm. This determination was then compared with the results of monitoring by conductivity measurements in accordance with the present invention.

c) Results Monitoring of the synthesis by electrical conductivity is based on the generation of ionized molecules during the reactions. Ideally, the amount of ions generated should be in proportion to the amount of reactant that has been converted in the reaction. In acylations by amino acid esters, the ester group is released as the free acid during the acylation reaction: thus, in the presence of a suitable proton acceptor, an ion pair will be formed, resulting in an increase in the electrical conductivity of the reaction solution.

During the coupling reaction, the proton acceptor should not itself be acylated. A number of tertiary amines fulfil this criterion, and for our experiment we chose DIEA. This amine is often used as a catalyst in acylation reactions with symmetrical anhydrides (11), and it has been reported to react only very slowly with the base labile FMOC group used for protection of the amino functionality (T ,.: 10.1 hours for FMOC-Val-OH in 50% DIEA in DMF, (23)). DMAP has been reported to cause partial racemization of some BOC-amino acids (24), but to our knowledge no such effect has been reported with DIEA used with FMOC-amino acids. However, due to the theoretical risk of racemization, the concentration of the base should be kept as low as possible.

Most solid phase peptide syntheses employ resins with less than 100 umoles of linker per mL of resin. Accordingly, the concentration of the tertiary amine can be kept below this value for the present purpose.

The results of these experiments are illustrated in Figures 9 to 12. Figure 9 shows the conductivity of 15 mM DIEA in DMF as a function of the concentration of Pentafluorophenol (Pfp- OH), Dhbt (free acids) and FMOC-protected Glycine. The conductivity recorded with 100 mM Pfp-OH, Dhbt or FMOC- Glycine in DMF without any DIEA present was less than 10 uS/cm. With 15 mM DIEA a continuous non linear increase in conductivity was seen up to a concentration of at least 100 mM of these acids, with Pfp-OH giving the highest values. With 15 mM DIEA, the deviation from linearity of the conductivity as a function of the concentration of free acid was moderate up to about 50 mM of the acid. The deviation from linearity can be incorporated in computer software used to control the experiment, and good corrections can be made up to at least 100 mM of free acid, which is well above the levels encountered in most reactions. At lower concentrations of DIEA an increasing deviation from linearity was seen, making interpretation of conductivity values at high concentrations of free acid more uncertain. The concentration of DIEA was set to 15 mM in all subsequent experiments.

To test the correlation between the increase in conductivity during acylation and the progression of the reaction, aliquots of the resin (containing Glycine (Gly) as the first amino acid) were incubated for 90 min at room temperature in 15 mM DIEA in DMF containing increasing amounts of FMOC-Val-OPfp. After incubation, the resin samples were rapidly filtered and the filtrate from each sample was collected for conductivity measurement. The resin samples were then washed in DMF and incubated in 20% piperidine in DMF for 10 minutes before a new wash and collection of the filtrate for determination of the FMOC- groups released during deprotection. As can be seen from Figure 10, a good correlation was found between the progress of the acylation reaction, as indicated by the conductivity measurements, and that deduced from measurement of the FMOC-groups released during deprotection (as determined by absorbance measurements at 280 nm) .

In other experiments we observed a baseline drift when measuring the conductivity at time intervals in a solution of FMOC-amino acid-Opfp ester in DIEA/DMF with no resin present. Since in the same solutions no drift was observed after reaction of the ester with amino groups on an added resin, we carried out an experiment to see if the baseline drift was inhibited by the pentafluorophenol groups released during the reaction. To aliquots containing FMOC-Val-Opfp in 15 mM DIEA in DMF were added increasing amounts of Pfp-OH dissolved in the same solvent. The final concentration of FMOC-Val-Opfp was 100 mM and the concentration of Pfp-OH added ranged from 0 to 16 mM . The conductivity in each of the aliquots was measured immediately after preparation and again after incubation for 90 minutes at room temperature.

Figure 11 shows the baseline drift (expressed as uS/cm x min) calculated after subtraction of the initial measurements from those made after 90 min. As is seen from Figure 11 , the drift decreases as a function of the concentration of free Pfp-OH in the system (curve 24). In a similar experiment where DMAP rather than DIEA was used as a solvent, in the same concentration, a significantly higher baseline drift was observed (curve 25). However, with both amines the drift rate was low compared to the rate of the signal recorded during the initial phase of the acylation reactions (about 50 uS/cm.min). In the later stages of the acylation reactions, the concentration of Pfp-OH released during the reaction was typically in the range of from 25 to 100 mM, which is more than adequate to quench the baseline drift in DIEA/DMF solutions.

A monitoration of all steps in a typical peptide synthesis is shown in Figure 12. The synthesis in this case was of the tetrapeptide Gly-Glu-Leu-Ile, and conductivity is shown in arbitrary units (A.U.). The acylation steps were terminated by the computer program when the rate of change in the monitor signal was less than 2% per hour. The rate was calculated as the averaged rate over the 50 latest subsequent measuring points recorded at 4-second intervals. In a similar manner, the washing steps after the acylation and the deprotection reactions were terminated when the rate of change was less than.10% per hour.

The monitor profiles of Figure 12 show nonidentical reaction rates for individual acylation steps. The recorded slow rate of the isoleucine coupling (region 26) is in accordance with the common experience that this amino acid couples relatively slowly. Similarly, the esterification of the first amino acid (Gly) to the linker (region 27) proceeded at a moderate rate, ais would be expected for this type of reaction. The latter reaction (and that involving isoleucine) did not reach the rate threshold of 2% per hour and was terminated instead when reaching an arbitrary time limit of 3 hours defined for the duration of any reaction.

In Figure 12, peaks 28 indicate signals recorded during deprotection in 20% v/v piperidine in DMF. Deprotection is seen to cause an increase in conductivity, which is probably as a result of the formation of ionizable species during deprotection. The fact that this occurs means that conductivity measurements can be used to monitor and control both the coupling and the deprotection steps.

d) Discussion

The use of such conductivity measurements as these for monitoring coupling reactions in solid phase peptide synthesis is seen to be a convenient alternative to conventional monitoring techniques. Adaptation of the reaction vessel to the technique is simple, and in the design presented here, one of the electrodes may even be used to obtain a splash free delivery of solvent to the reaction vessel.

In order to obtain a steady monitor signal, the reagents were not recirculated through the reaction vessel during the reaction. To ensure a complete loading of the reagent vessel, it was therefore necessary to inject an excess volume of reagent. The excess volume (together with the ester groups released in the initial phase of the reaction) was eluted from the resin and lost to the waste channel. In some monitor profiles a transient increase in the signal was recorded during loading of the activated amino acid into the vessel (not shown). This 'spike* is caused by ester groups that are released during the very first part of the coupling reaction and are washed out of the reaction vessel during loading. If the distance between the upper and the lower electrode in the reaction vessel is small, the resulting spike in the monitor readings can be integrated to obtain a measure of the amount of ester groups released during this initial phase.

The omission of a recirculation procedure has the advantage that the pump of the synthesizer is made available for other tasks, e.g. for injection of reagents into another reactor.

With 10 - 50 umoles functional groups per mL resin, a 2-3 fold molar excess of FMOC-amino acid ester can easily be dissolved in e.g. 1.5 times the resin volume of DMF. Recirculation in order to adapt a large reagent volume therefore seems unwarranted.

With the set-up used, it would be a straightforward task to incorporate a reference cell in the flow line so that the background conductivity of the dissolved ester could be subtracted from the readings. The use of such a reference cell is not a prerequisite for determination of reaction rates but may be useful in determination of the amount of ester groups released. We have found that FMOC- amino acid esters from some manufacturers contain varying amounts of material that contribute to the electrical conductivity-of the solution when the esters are dissolved in DMF. Therefore, if the absolute amount of such esters released during a reaction is to be determined, it is necessary to perform a subtraction of the background signal. Background subtraction further makes the recording independent of the absolute amount of activated amino acid used.

The difference in conductivity observed when titrating DIEA with Pfp, Dhbt or FMOC-amino acid in DMF (see Figure 9) makes it necessary to inform the computer control system about the nature of the amino acid derivative used in each step. Again, this is only relevant if the absolute amount of ester/amino acid released during the reaction is to be determined. For precise work, it may even be necessary to make calibration curves for each ester type and for all amino acids used as symmetrical anhydrides.

The above described technique allows for determination of both the reaction rate and the amount of ester consumed. Both recordings are generated by the released ester groups rather than by the unreacted amino groups on the resin. While this is less than ideal with respect to precision, it is believed that the simultaneous determination of both the rate and the amount provides a satisfactory basis for the decision of whether or not a reaction is completed. As an additional benefit, the technique is highly suited for monitoring deprotection and washing procedures. This defines precisely when to stop washing and, in our experience, may cut solvent consumption, compared to arbitrarily defined washings, by about 50%.

As noted above, the most precise monitoring methods are those monitoring unreacted amino groups. It will be clear that modification of the technique of the present invention to provide a record of the unreacted amino groups is theoretically possible. References

1. Kisfaludy, L., Schon, I., (1983) Preparation and Applications of Pentafluorophenyl Esters of 9- Fluorenylmethoxycarbonyl Amino Acids for Peptide Synthesis, Synthesis, 325-327.

2. Atherton, E., Sheppard, R.C., (1985) Solid Phase Peptide Synthesis using N-alpha-Fluorenylmethoxy- carbonylamino Acid Pentafluorophenyl Esters, J. Chem Soc, Chem. Comm. 165-166.

3. Konig, W. , Geiger, R. , (1970) Eine Neue Methode zur Synthese von Peptiden: Aktivierung der Carboxylgruppe rait Dicyclohexylcarbodiimid und 3-Hydroxy-4-oxo-3,4-dihydro-

1 ,2,3-benzotriazin, Chem. Ber. 103, 2034-2040.

4. Atherton, E., Cameron, L., Meldal, M., Sheppard, R.C., (1986) Self-indicating Activated Esters for use in Solid Phase Peptide Synthesis. Fluorenylmethoxycarbonylamino Acid Derivatives of 3-Hydroxy-4-oxodihydrobenzotriazine. J. Chem. Soc, Chem. Comm., 1763-1765.

5. Atherton, E., Sheppard, R.C., (1974), An Investigation into the use of Polar Supports based on Poly (dimethylacrylamide), "Peptides", Wolman, Y., ed., J. Wiley and Sons, N.Y., 123-137.

6. Atherton, E., Gait, M.J, Sheppard, R.C., Williams, B.J., (1979) The Polyamide Method of Solid Phase Peptide and Oligonucleotide Synthesis, Bioorganic Chem. 8,351- 370. 7. Arshady, R. , Atherton, E. , Clive, D.L.J., Sheppard, R.C., (1981) Peptide Synthesis. Part 1. Preparation and use of Polar Supports based on Poly(dimethylacrylamide). J. Chem. Soc, Perkin Trans. I., 529-537.

8. Merrifield, R.B., (1968) Automatic Synthesis of Proteins, Sci. Amer., 218, 56-62, 67-72.

9. Barany, G., Merrifield, R.B., (1980), in "The Peptides", Gross, E., Meienhofer, J., eds., Acad. Press, N.Y., 1-284.

10. Barany, G., Kneib-Cordonnier, N., Mullen, D.G., (1987), Solid-Phase Peptide Synthesis: a Silver Anniversary Report, Int. J. Pept. Prot. Res. 30, 705-739.

11. Stewart, J. and Young, J.D., (1984) "Solid Phase Peptide Synthesis", 2nd.ed., Pierce Chemical Company.

12. Kaiser, E., Colescott, R.L., Bossinger, CD., Cook, P.I., (1970), Color Test for Detection of Free Terminal Amino Groups in the Solid Phase Synthesis of Peptides, Anal. Biochem. 34, 594-598.

13. Sarin, V.K., Kent, S.B.H., Tarn, J.P., Merrifield, R.B., (1981) Quantitative Monitoring of Solid-Phase Peptide Synthesis by the Ninhydrin Reaction. Anal. Biochem. 117,147-157.

14. Felix., A.M., Jimenez, M.H., (1973) Rapid Fluorimetric Detection for Completeness in Solid Phase Coupling Reactions. Anal, Biochem. 52, 377-381.

15. Kent, S.B.H., (1988), Chemical Synthesis of Peptides and Proteins, Ann. Rev. Biochem. 57, 957-989.

16. Hodges, R.S., Merrifield, R.B., (1975) Monitoring of Solid Phase Peptide Synthesis by an Automated Spectrophotometric Picrate Method. Anal. Biochem. 65, 241- 272.

17. Brunfeldt, K., Christensen, T. , Villemoes, P., (1972), Automatic Monitoring of Solid Phase Peptide Synthesis of a Decapeptide, FEBS letters 22, 238-244.

18. Horn, M. , Novak, C. , (1988), A Monitoring and Control Chemistry for Solid Phase Peptide Synthesis, IBF, April 1988, 30-37.

19. Carpino, L.A. , Han, G.Y., (1972), The 9-Fluorenyl- methoxycarbonyl Amino-Protecting Group, J. Org. Chem. 37, 3404-3405.

20. Meienhofer, J., Waki, M., Heimer, E.P., Lambros, T.J., Makofske, R.C., Chang, CD., (1979), Solid Phase Synthesis without Repetitive Acidolysis, Int. J. Pept. Prot. Res. 13, 35-42.

21. Dryland, A., Sheppard, R.C., (1986), Peptide Synthesis Part 8. A System for Solid-Phase Synthesis Under Low Pressure Continuous Flow Conditions, J. Chem. Soc, Perkin Trans I, 125-137.

22. Cameron, L., Meldal, M. , Sheppard, R.C., (1987) Feedback Control in Organic Synthesis. A System for Solid Phase Peptide Synthesis with True Automation. J. Chem. Soc, Chem. Comm., 270-272. 23. Atherton, E., Logan, C.J., Sheppard, R.C., (1981) Peptide Synthesis. Part 2. Procedures for Solid Phase Synthesis using N-alpha-Fluorenylmethoxycarbonylamino Acids on Polyamide Supports. Synthesis of Substance P and Acyl Carrier Protein 65-74 Decapeptide. J. Chem. Soc, Perkin Trans. I, 538-546.

24. Atherton, E., Benoiton, N.L., Brown, E., Sheppard, R.C., Williams, B.J., (1981) Racemization of Activated, ϋrethane-protected Amino-acids by p-dimethylaminopyridine. Significance in Solid Phase Peptide Synthesis. J. Chem. Soc. Chem. Comm., 336-337.

Claims

1. A method of monitoring reactions in solid phase peptide synthesis, comprising monitoring the electrical conductivity of the reaction solution.

2. A method of solid phase peptide synthesis, characterised by monitoring the electrical conductivity of the reaction solution.

3. A method according to Claim 1 or Claim 2, wherein the peptide synthesis is carried out using the FMOC protection strategy.

4. A method according to any one of Claims 1 to 3, wherein the peptide synthesis is carried out in a batch mode.

5. A method according to any one of Claims 1 to 3, wherein the peptide synthesis is carried out using a continuous flow method.

6. A method according to any one of the preceding claims, wherein the change in conductivity of the reaction solution due to the production of ion pairs in the solution during a coupling step is monitored.

7. A method according to Claim 6, wherein coupling of a growing peptide chain with an incoming amino acid is acid catalysed so as to allow formation of ion pairs between the growing peptide chain and an acid used as a catalyst.

8. A method according to Claim 7, wherein the acid used as a catalyst is hydroxybenzotriazole (HObt).

9. A method according to Claim 6, wherein an incoming amino acid to be coupled with the peptide chain being synthesised is in the form of an amino acid ester, and the ion pairs produced in the reaction solution are formed between ester groups released during coupling and a suitable proton acceptor present in the reaction solution.

10. A method according to Claim 9, wherein the proton acceptor is a tertiary amine which is not itself acylated in the presence of the incoming amino acid.

11. A method according to Claim 10, wherein the proton acceptor is diisopropylethyl amine (DIEA) .

12. A method according to any one of Claims 9 to 11, wherein the incoming amino acid is in the form of a pentafluorophenol ester or a pentafluorophenyl ester.

13. A method according to any one of the preceding claims, wherein a quantity of 3-hydroxy 4- oxodihydrobenzotriazol (Dhbt) is included in the reaction solvent used during deprotection steps.

14. A method according to any one of the preceding claims, wherein conductivity is monitored by means of two spaced apart electrodes located in a reaction vessel for synthesis reaction solutions, and wherein an AC voltage is applied across the electrodes; the AC voltage resulting from current flowing through the reaction solution is amplified and rectified; and a measure of the conductivity of the solution is obtained therefrom.

15. A method according to Claim 14, wherein the voltage applied across the electrodes is 0.2 volts AC, operating at 1 kHz.

16. A method according to any one of the preceding claims, wherein conductivity measurements obtained by the method are processed and used to provide control of or to modify subsequent stages of the synthesis.

17. Apparatus for solid phase synthesis of peptides, characterised in that it comprises a reaction vessel having two spaced apart electrodes located therein; means for applying an AC voltage across the electrodes; means for amplifying and rectifying the AC voltage resulting from current flowing through a reaction solution in the vessel; and means for displaying the resulting AC signal.

18. Apparatus according to Claim 17 and for use in the method of any one of Claims 1 to 4 or 6 to 16, wherein the electrodes are platinum flashed, tungsten electrodes.

19. A reaction vessel for use in the method of any one of Claims 1 to 3 or 5 to 16, comprising a flow-through chromatographic column having an upper and a lower electrode, one at each of the ends of the column.

20. A reaction vessel according to Claim 19, wherein the upper electrode comprises a ϋ-shaped rod inserted into the upper end of the reaction vessel, such that reactants injected into the reaction vessel are guided into the ϋ- shaped cavity provided by the upper electrode and onto the surface of a solid phase support resin contained in the reaction vessel.