Enzyme studies

The vulnerability of collagen to protease attack.
It is increasingly common for proteolyic enzymes to be incorporated in the beamhouse process steps, particularly in the liming/unhairing, as indicated in Section 3. The advantages are clear, but there is a problem associated with the desired reactions on the hair and in opening up, which is attack on the substrate itself. This effect can be investigated by observing the solubilisation of the collagen as an increasing concentration of hydroxyproline in solution. However, such an approach has the disadvantage of requiring separate determinations for every enzyme used in the beamhouse. What would be better is a determination of the principles of the reaction.

Collagen is resistant to attack by general protease's, because of its high degree of structure, which can only be fitted into the reactive site of specific collagenases. But, if the collagen structure loses that high degree of structure, it becomes vulnerable to proteolysis. One way by which collagen can deform is by breaking the peptide links of the oc-chains. This reaction is clearly possible during liming, so the question is: at what point during liming does the collagen sustain sufficient damage to allow degradation by general proteases?

A sensitive measurement of the intactness of the collagen backbone is the effect of liming on the racemization of aspartic acid during the hydrolysis of asparagine. The mechanism can be either direct hydrolysis by hydroxyl catalysis or by intramolecular catalysis, illustrated in Fig. 5.

Using molecular dynamics, it has been estimated that the rate of asparagine deamidation by intramolecular catalysis at 370C would be 10,000 times slower in an extended a·chain than in a random Coal's. Therefore, any disruption to the normal collagen structure would be reflected in the amount of racemized aspartic acid produced during liming.

From Fig. 6, the percentage of D-aspartic acid in the limed samples initially increases Z~d then decreases; this may be explained by all of the non structural, non collagenous aspartic acid changing its molecular conformation, while passing into the liming solution within a period of 18 hrs. The rate of racemization is low during the early part of the liming process, increasing only after 24 hrs.

Some racemization by acid hydrolysis always occurs during the processing of the specimens for analysis, and this explains why, in Fig. 6, there is a slightly positive measurement for the untreated cdllagen, at time zero. Also, there is some accumulation of racemized amino acids in slowly metabolized tissues, during the lifetime of the animal; this phenomenon is exemplified by red blood cells.

From the literature, deamination of amide sidechains by the intramolecular mechanism follows first order kinetics: in synthetic polypeptide the reaction displays a rate constant of 1.0 x 10" sec'' at 250C, but the observed rate in collagen decreases rapidly below the melting temperature, calculated to be over 1,000 times slower. This can be compared to the observed rate of ammonia production during liming of 8.7 x 10 secs at 250C, an order of magnitude faster than the rate observed in the synthetic polypepide. Therefore, the measurements indicate that the hydrolysis originates from the intermolecular reaction.

These observations mean that conventional liming processes do not cause major damage to the high order of collagen structure during the first 24 hrs. If there was significant conformational damage to the ct-chains, then the formation of succinimide rings within the peptide chains would be easier and therefore faster. That kind of destabilization is only obtained by prolonged liming. This means that the collagen triple helix remains effectively intact for up to 24 hrs, but then starts to denature significantly. The implication for enzyme assisted opening up and liming is that their application should be restricted to the earlier part of the process; if enzymic action is allowed to continue to the end of normal liming or beyond, there is a likelihood of enzymolysis damage, which can rapidly adversely affect leather quality and performance.

Monitoring the site of enzyme activity.
The effects of proteolytic enzymes on hides and skins during beamhouse processing can be extremely powerful, to the extent that considerable damage can be caused if the processes are not monitored and controlled. This is not simple, because there is currently no quick and easy way to follow the reaction. The amount of solubilised protein and collagen can be analysed by total Kjeldahl nitrogen and hydroxyproline respectively and proteolytic enzyme penetration through pelt cross section can be monitored by the gelatine film test". In each case, the procedure is lengthy, too long in terms of the process steps to which they apply, and the gelatine test can be unreliable.

A established way of measuring dissolved protein concentration is the biuret test, which utilises basic copper(II) sulphate tb produce a purple colour reaction. The optical density is measured at 580nm and comparison with a calibration graph yields the concentration. The test is accurate and correlates well with the more conventional protein analyses. This is illustrated in Fig. 7, by some data from bating studies22, in which the total protein content in solution was measured by the biuret test and plotted against the hydroxyproline content from solubilised collagen.

It is not unusual for the biuret test to work poorly, because of the instability of the copper salt,which tends to precipitate by hydrolysis. However, during this study, it was found that consistently reliable results could be obtained by the following procedure:

to 2ml of bate liquor,
add Iml of 5% copper(II) sulphate solution,
add Iml of 30% sodium hydroxide solution,
make up to 10ml accurately, then measure optical density at 580nm.

This procedure can be used in a more novel way. It was argued that, since the test visualises dissolved protein and peptide fragments, it would also visualise where protein breakdown had occurred, in the presence of intact collagen. In this way, the colour reaction would identify sites of enzyme activity, because protein including collagen would be degraded. The developed procedure is as follows:

to a cut edge of pelt, apply a drop of 5%- copper(II) sulphate solution,
immediately add a drop of 30% sodium hydroxide solution,
observe the colour reaction.

Where the pelt proteins are intact, because there is no enzyme activity, the cross section will be stained blue. But where there is or has been enzyme action, the cross section will be stained purple. The colour difference can be subtle, because of the blue background, but it is discernible. The test can be applied to limed or bated or pickled pelt, because the pH of the reaction is determined by the sodium hydroxide. Although the test is not quantitative, it is a useful way of monitoring the degree of penetration of proteolytic enzymes through pelt. Because the effects of these enzymes on leather properties depend on where in the hide or skin they operate, it is important for the tanner to know how the reaction is progressing, so it can be controlled, by changing the rate or by stopping it. An additional application is for detecting bacterial activity in rawstock, typically related to delay in curing after flay or to inadequate curing. Here, grain damage correlates with the extent of bacterial growth from the flesh side towards the grain.

Organic tanning
In considering organic based tanning, the challenge is to find the limits to metal free tanning. There are many ways to tan using organic agents, but the candidates for new processes are likely to be based either on the familiar polyphenolic vegetable tannins or on synthetic polymeric agents. Semi metal tanning is a model for organic tanning processes; Lere, the reaction is based on hydrolysable vegetable tannins, in which the multiple interactions between the polyphenol and the protein are enhanced by crosslinking the polyphenol molecules with metal salts. What is required is an equivalent process in which the crosslinker is organic. Reactions are already known in which condensed tannins can be polymerised by alhehydes, but it is only recently that this type of chemistry has been shown to have applicability to tanning technology.

Polyphenolic organic tanning
Recent NMR solution studies of catechin crosslinking by oxazolidine have shown that reaction occurs at the A ring, at the C-6 and C-8 positions, but there is no reaction at the catechol C ring. The powerful interaction between oxazolidine and the resorcinol moiety of the flavone ring system is evidence of polymerisation being at the heart of the new tanning chemistry, rather than the reaction depending on crosslinking the polyflavone to collagen. This is further illustrated by the tanning effect of model polyphenols crosslinked with oxazolidine, shown in Table 8.

From Table 8, it is apparent that the polymerisation reaction is disposition of the hydroxyl groups: 1,2 and 1,4 arrangements are inactive in this context, but the 1,3 arrangement is highly reactive. This explains the reactivity of the catechin flavonoid A ring and may also explain the difference in reactivity between quebracho and mimosa in this reaction.

The reactivity of the flavonoid ring system provides a new range of tannages. it is clear that the currently available condensed vegetable tannin extracts can form the basis of novel processes. However, it is equally clear that, in order to avoid a return to traditional vegetable character leathers, a 'higher tech.' approach to the chemistry is required, in which the principles of polyphenol reactions are exploited, rather than the plant extracts themselves.

Synthetic organic tanning
The goal of a synthetic organic tanning method that confers all the properties currently provided by chromium(III) offers the potential for making improvements in processing and leather performance. Such a process would eliminate part of the present environmental impact of leather making and facilitate biological, end of pipe effluent treatment.

It has been argued that shrinkage temperature depends on the degree of disorder in the transition state and on the bond energy change in the formation of the transition complex. If the entropy decreases and/or the enthalpy increases, the shrinkage temperature rises. Those parameters depend on the order and structure imposed on the collagen by the tanning reaction. It has also been argued that most chemicals that can interact with collagen have an effect on the entropy of the transition and this is why there is a plethora of tannages capable of raising the shrinkage temperature by differing amounts, up to about 85"C. It is only when there is a powerful enthalpic contribution to structure making that higher hydrothermal stability is achieved.

There are two contributions to a crosslink, to define the rigidity or stiffness of the structure making effect: the interaction between the crosslink and the collagen and the rigidity of the crosslink itself. In other words, the crosslink must be short and non labile. The ability of the bonding with collagen, the degree of reversibility, is determined by the chemistry; a strong interaction is obtained by covalent reaction or by multiple weaker interactions, such as hydrogen bonding. If the crosslink is formed by a second interaction, linking species directly bound to collagen, there is a third factor in determining hydrothermal stability, the lability of this bonding.

In the light of this theory, it is possible to explain both traditional tannages and new chemistries. Chromium(III) reacts covalently, with a rigid crosslink, causing stabilization by making a large cooperating unit, but other metal tanning salts such as aluminium(III) or zirconium(IV) react electrostatically, producing a labile bond, with consequently low hydrothermal stability. But, if these ionic metals are bound to collagen by chelating sites, the ability is reduced and higher shrinkage temperature is achieved27. Indeed, if unipoint bound chromium(III) is crosslinked by a highly rigid moiety, such as mellitate (benzene-1,2,4,5- tetracarboxylate), then the collagen is also stabilised by imposed structure; therefore, such leather is stabilised by both mechanisms, so the shrinkage temperature is raised to very high values. Moreover, the new theory sheds some light on the variability within the chrome tanning reaction. Shrinkage temperature increase is dominated by the pH changes during tanning", with the result that higher shrinkage temperature is achieved with low basicity chrome and reactive collagen, than by tanning with high basicity, astringent chrome. That means, the reaction is more effective when smaller chrome species are used and the astringency is controlled by the reactivity of the substrate, i.e. when more structure is imposed by shorter crosslinks.

Aldehydes react covalently with collagen, but the crosslinks are polymeric and hence non rigid. Vegetable tannins, syntans and resins may have the geometry to interact with collagen in a non labile manner, but, if the interaction is only hydrogen bonding at both ends of the crosslink, this constitutes a labile crosslink. In these cases, the interaction with collagen must be specific and it is only when those species are crosslinked by a short species that the crosslink becomes stiffened, so the overall interaction constitutes structure making. Retanning hydrolysable tannins with aluminium(III) produces a powerful multiple complex, because of the high affinity of aluminium(III) for phenolic hydroxyl groups. Conversely, chromium(III) has lower affinity for those reaction sites, so the interaction is typically weaker and the result is lower shrinkage temperatures in semi chrome leathers.

The aldehydic crosslinking reactions of condensed tannins depend on the individual tannin, which determines the primary interaction with collagen; then the relative amenities of the crosslinker for the polyphenol and itself determine the rigidity of the crosslinks. Here the reaction is aided by the reaction mechanism for collagen and condensed polyphenols, which is a mixture of hydrogen bonding via the phenolic hydroxyls and covalent bonding'". The degree to which covalent bonding occurs depends on the particular condensed tannin, which would account, at least in part, for variability in the effect the tannin has in this high stability application. The synthetic tannages also depend on similar criteria. Some examples of the model are given in Table 9.

From the proposed theory, it is possible to draw up the rules governing high hydrothermal stability, which then allows targeting of appropriate chemistries.

1. High energy interaction between collagen and the primary organic compound polymer/macromolecule) : (a) covalent, (b) multiple hydrogen bonding, (c) preferably partial covalency.
2. Maximise the magnitude of the primary interaction, consistent with minimising the size of the molecular species.
3. Increase the degree of structure, by fixing the organic compound with immobilising, covalent cross links.
4. Crosslinking compounds should have higher affinity for the organic compounds than for themselves i.e. they should react as monomeric species, rather than as polymeric chains.

Discussion
The new technologies presented in this paper demonstrate the value of the scientific approach to process development. It is no longer enough merely to optimise current processing technology or to import technologies and materials from other industries into the modem leather industry. Such an approach has its place, as technology transfer, but it is not the basis for a development programme for the future. It is necessary to look beyond short term expediency, to undertake fundamental studies of all aspects of leather production, in order to achieve the breakthroughs needed to take leather into the next millennium and retain its favoured position as the material of choice for many applications.

This is a view shared by Rutland: 'research innovation - whether performed in the tannery, by a supplier or at an independent laboratory - represents the lifeblood of an industry, ours or any other. The day we cease to recognise the potential benefit of continuing research effort and fail to provide a suitable forum for this effort will be a sad day indeed for the tanning industry.

Acknowledgement

This paper contains contributions from collaborations between the BSLT research group and the following scientists: Prof. Christine Evans of the University of Westminster, UK, Dr. Matthew Collins and Ms. Emma Waite of the University of Newcastle upon Tyne, UK, Dr. Richard Hancock of Royal Holloway College, University of London, UK, Dr. Ioannis Ioannidis, of El. Re. De., Athens, Greece, Prof. Bi Shi of Sichuan Union University, Chengdu, P R China.

The BSLT research group includes: Ph.D. students, Graham Lampard, Murat Tozan, Syed Muhammad, Ms. Ozlem Menderes, MSc. students, Hossam el Baba, Simon Phillips, Colleagues at University College Northampton, Prof. Geof. Attenburrow, Mr. DanteDavighi, Mr. Paul Richardson.

The author thanks them all for their contributions to the ongoing programme of fundamental studies in leather science.  

Back