Cole. C.G.B. and Roberts J.J. Department of Food Science. University of Pretoria. Pretoria. 0002. RSA.

Originally published in: THE SA JOURNAL OF FOOD SCIENCE AND NUTRITION. 8(4) 1996. 139-143.


Gelatine (Type B) was manufactured from the hides of bovines of known age at slaughter, after 4 weeks of liming at 22°C. The extractability of the hide, gelatine colour and fluorescence were measured. The 335/385 nm fluorescence indicated the presence of the pentosidine cross-link which was correlated with the gelatine colour and collagen extractability (p > 0.95). Type A gelatine from calf skin exhibited an additional excitation maximum at 295 nm due to the alkali labile pyridinoline collagen cross-link thus explaining the extractability phenomena observed.


Prior study of gelatine fluorescence appears to have been limited to the investigation of Thornton (1966) in which home-made equipment using filters was used to demonstrate gelatine sol and gel fluorescence depolarisation. Recently, there have been many reports on the fluorescence of collagen which is the precursor of gelatine. Tanaka et al. (1988) reported an increase in fluorescence with time, for rat tail collagen incubated in the presence of ribose at 35°C for up to two weeks. They showed that a similar increase in collagen fluorescence occurred as a result of animal age. In addition, when soluble collagen was incubated with ribose, dimers and trimers of the collagen -chain were formed which confirmed that the fluorescence was associated with cross-linking. Sell and Monnier (1989) isolated the "pentosidine" collagen cross-link from human collagen (dura mater). They correlated the increase in quantity of fluorescence with age and elucidated the structure of the cross-link by synthesis. This involved heating equimolar quantities of lysine, arginine and ribose at pH 7.2 and 80°C for one hour. They claimed that similar treatment of a mixture of lysine, arginine and glucose did not result in the formation of a fluorescent product. Hence, there was a "problem" in accounting for the source of the ribose used in the formation of the collagen cross-link. Monnier et al. (1990) used the 335/385 nm fluorescent collagen "pentosidine" cross-link as the basis of a theory that senescence was the result of the Maillard reaction between functional proteins and ribose. The resulting impaired functionality of the protein was regarded as the well documented manifestations of age. In their review "The Maillard reaction in vivo", Dyer et al. (1991) state that "since pentosidine contains only 5 carbons from the sugar component, its formation from glucose must involve the loss of a carbon atom either from glucose itself or from a later intermediate in the reaction. ...While its (pentosidine's) origin from ribose vs. glucose and other sugars may be uncertain.....". This indicated that there was evidence (from Baynes et al. 1990) for the formation of pentosidine from glucose. In the study of Uchiyama et al. (1991) human articular cartilage was treated with collagenase. An increase in 335/385 nm fluorescence of the digests was shown to correlate linearly with increasing age. Furthermore, they demonstrated that the 295/395 nm fluorescence was due to the pyridinoline collagen cross-link and that the amount of this cross-link was largely invariant with age. In addition the 335/385 nm fluorophore was shown to be identical with "pentosidine" of Sell and Monnier (1989). Eyre (1988) gave the details of the methodology for the isolation of hydroxylysylpyridinoline (HP) and lysylpyridinolin (LP) cross-links of collagen which were fluorescent and it was stated that these cross-links could be detected in polyacrylamide electrophoresis gels using an excitation wavelength of 330 nm and an emission wavelength of 395 nm. However, pyridinoline cross-links were not found in skin collagen (Yamauchi et al. 1988).

Based on the observation that gelatine colour also increased with animal age (loc sit) it appeared reasonable to hypothesize that gelatine colour could be the result of the Maillard reaction (in vivo). From the report by Odetti et al. (1994) it appeared that fluorometry was an accepted technique for studying pentosidine's relationship to ageing phenomena. Hence, if gelatine exhibited fluorescence at 335/385 nm and the fluorescence increased with colour and animal age, it would follow from the work of Uchiyama et al. (1991) that the colour could be attributable to the formation of pentosidine collagen cross-links.


All fluorescence determinations were conducted in duplicate and reagent grade chemicals were used, unless otherwise indicated.

The salted hides of animals of known date of birth and date of slaughter were supplied by the Irene Animal Production Institute, Pretoria, RSA.

Gelatine production.

The whole hide was cut into approximately 100 x 100mm pieces. A quantity (5-7 kg) was taken and washed free of salt overnight using a stainless steel (13 RPM) tumbler fitted with a continuous supply of water (through one axel) and a perforated plate in the door for drainage. The washed hide was placed in 20 kg of conditioning liquor containing of sodium sulphide (2g/l) and calcium hydroxide (40g/l). (Commercial chemicals were used). The hide was conditioned for 4 weeks at 22°C ± 2°C. The conditioning liquor was stirred every second or third day in order to prevent localised reductions in pH. After stirring the temperature was recorded. After conditioning the hide was again washed overnight in the tumbler and then acidulated using 5 x 20 l lots of sulphurous acid solution (0.1M) (commercial) over 4 days followed by soaking in tap water for 1 day. The material was then converted to gelatine by sequential 5 hour extractions in water (at 45°C, 50°C, 55°C, respectively) and finally by boiling (93°C) for 7 hours. After each extraction the gelatine solution was separated from the solid residue using a colander. The volume of each separated gelatine solution was determined. An aliquot of the solution was then filtered through Whatman 541 filter paper after which the concentration was determined gravimetrically in duplicate by drying 10 ml of solution at 105°C for 40 to 48 hours. The weight of the dry film was multiplied by 11.4286 to obtain the concentration (% gelatine containing 12.5% moisture) of the original solution. From the volume and gelatine concentration, the total amount of gelatine extracted was calculated. The gelatine produced in each extraction was expressed as the % of the total amount of gelatine recovered from a particular sample of hide. This was a measure of the "extractability" of the hide. Statistical correlations between animal age, raw material extractability, gelatine colour and fluorescence intensity were calculated using the linear regression facility of the Quattro spreadsheet (Borland International, 4585 Scotts Valley Drive, Scotts Valley, CA 95066, USA) .

The bulk of each gelatine extraction liquor (excluding the boil solutions) was filtered through paper pulp and then vacuum evaporated at 40 to 42°C using a rising film glass evaporator, to about 10% concentration, before refiltration through paper pulp. The concentrated gelatine liquor was then treated with a 5% ammonia solution as well as a 5% hydrogen peroxide solution until the liquor exhibited an approximate 10 ppm excess of peroxide and a pH of 5.0 to 5.5 (using Merck indicator strips). The solution was then set in a refrigerator, cut into slices and dried overnight in a current of air to about a 10% moisture content. The sheets of dry gelatine were ground (using a domestic coffee grinder) to a powder to facilitate analysis using British Standard 757:1975 methods.

The colour and clarity of the gelatine were determined by in-house methods:

a) Gelatine solution (60 ml at 40°C) from the Bloom strength determination (6.67%) was diluted to 100 ml using 40°C distilled water and then compared, using 100 ml Nessler tubes, to 100 ml of a 4% solution of a standard gelatine (with an ascribed colour of 8.0). Solution was poured out of the darker Nessler tube until a match was obtained. The colour of the unknown gelatine was then calculated in accordance with Beer's Law. For example, if 60 ml of unknown matched 100 ml of standard (colour 8) then the colour of the unknown was 100x8/60 = 13.3. This method was used to determine the colour of all the gelatines produced, including the colour of the gelatine in the boil liquor. Using the quantity of gelatine in each extraction and its colour the overall or composite colour of the gelatine from each lot of hide was calculated.

b) Gelatine solution clarity was determined by filling a Turbidimeter (ICM Turbidimeter. ICM. 163 S.W. Freeman, Hillsboro. OR 97123. USA) cuvette (20 ml) with molten Bloom sample (6.67% gelatine) and reading the turbidity in NTU. The meter was calibrated daily against a 40 NTU standard supplied by the instrument manufacturer.

Gelatine fluorescence was determined using 1% solutions of gelatine made by weighing aliquots (0.20 g) of gelatine powder into 50 ml clear glass bottles. Distilled water (20 ml) was added by pipette. Screw caps were applied to the bottles and the gelatine was allowed to swell for 20 min. The bottles were then placed in a 40°C waterbath and swirled several times over 30 min. to achieve complete solution of the gelatine. The sample fluorescence intensity was determined using a Schimadzu RF-5000 (Schimadzu Corporation, 3, Kandu - Nishikicho 1 Chome, Chiyoda-ku, Tokyo 101, Japan.) spectrofluorophotometer in transmission mode and a 1 cm2 quartz cuvette . The excitation wavelengths were scanned using a fixed emission wavelength and the emission wavelengths were scanned using a fixed excitation wavelength. Results were recorded by means of a thermal printer. Fluorescence was monitored also at 335/385 nm for the pentosidine collagen cross-link and the 295/395 nm for the pyridinoline collagen cross-link with results being recorded from the cathode-ray tube (CRT).


Figure 1. Fluorescence emission spectrum of pale
                Type B gelatine from 10 month old calf
                skin (YSA/2). Excitation at 335 nm..
                Colour = 4.4.

Figure 2    Fluorescence emission spectrum of dark
                Type B gelatine from 12 year old Inguni
                cow skin (IOB/1). Excitation at 335 nm.
                Colour = 16.0

Figure 3.     Fluorescence excitation spectrun of Type B
                    gelatine. Emission at 395 nm.

Figure 4.    Excitation spectrum of Type A calfskin
                 gelatine. Emission at 385 nm.

In Figures 1 and 2 the first peaks at 360 nm are due to Rayleigh scatter (Munck 1989). Figure 1 shows the fluorescence emission spectrum (FES) of Type B Gelatine from a 10 month-old animal (i.e. alkali treated collagen). It was found to be limited to an excitation wavelength peak at 335 nm with emission maxima at 385 and 410 nm. Figure 2 shows the FES of gelatine from a 12 year old animal. The marked increase in fluorescence intensity of the gelatine from the older animal was noted. Figure 3 shows the excitation scan and confirms that there was only one peak at 335 nm (for alkali process gelatines).

Type A gelatine was made from calf (less than 6 months of age) skin using the acid conditioning process (Reich et al. 1962) usually used for the manufacture of gelatine from pigskin. This gelatine exhibited 335/385 nm fluorescence and it also had a 295 nm fluorescence excitation maximum as shown in Figure 4.

Table 1. Colour and fluorescence data for gelatines derived from bovines of known age.






in Davis Units.

FLUORESCENCE. 385 nm Emission Intensity - Excitation @ 335 nm.
45°C 50°C 55°C 60°C 70°C 3 RUNS 1ST 2ND 3RD 1ST 2ND 3RD
YSA 10 4.3 35 26 20 - - 83 3.6 4.4 4.8 38 48 68
CT4 18 6.4 23 28 28 - - 76 5.6 5.6 5.6 104 133 142
5Y4 58 7.2 20 21 - 29 - 70 5.6 6.8 8.4 104 126 149
ST24 144 15.0 10 13 - 26 - 49 8.9 12.3 13.3 133 168 214
IOE 143 14.9 - 11 - 23 37 72 9.4 12.3 16.0 175 207 -
CTO4 156 16.4 10 11 14 - - 35 10.0 9.4 10.0 187 223 239
3Y4A 40 5.2 28 30 - - - - 6.6 6.0 - 102 138 -
6Y4A 78 7.9 15 21 - 34 - 70 8.4 7.6 8.9 141 165 174

Samples had equal conditioning - 4 weeks, except for YSA which had 2 weeks.
- Not applicable.

The extractability, 335/385 fluorescence and colour data for gelatines from bovines with ages ranging from 10 to 156 months is shown in Table 1. It should be noted that an alkali conditioning time of 4 weeks was used for all the gelatines except those derived from the 10 month old animal where a conditioning time of 2 weeks was used. The following linear correlations were obtained:

1. Animal age & Overall colour.
    r = 0.976, df = 6, p = 0.9995

2. Animal age and 1st extraction colour.
    r = 0.928, df = 6, p = 0.9990

3. Animal age and 45°C extractability.
    r = -0.902, df = 5, p = 0.995

4. Animal age and 1st extraction fluorescence.
    r = 0.895, df = 6, p = 0.9975

5. 1st (45°C) Extractability & 1st Colour.
    r = -0.910, df = 5, p = 0.9975

6. 1st extraction colour & 1st extraction fluorescence.
    r = 0.957, df = 6, p = 0.9995

7. 2nd (50°C) Extractability & Colour.
    r = -0.852, df = 6, p = 0.995

8. 2nd extraction Colour & Fluorescence.
    r = 0.771, df = 6, p = 0.975

The above correlations show that the most significant linear correlation was (1) between animal age and overall colour. There were also significant correlations between gelatine 1st and 2nd extractability and colour, as well as between 1st and 2nd extraction gelatine colour and pentosidine fluorescence intensity. Therefore, the Maillard cross-linking, as determined by fluorescence intensity, affected both hide extractability and gelatine colour (a not unexpected observation). Furthermore, it implies that the colour moiety was covalently linked to the protein and was probably not a removable contaminant.

The drop in correlation coefficients from 1st extraction to 2nd extraction could indicate that the Maillard cross-link is not the only contributor to gelatine extractability and colour. The implication is that there could be other contributing factor(s) responsible for these phenomena. In addition, the less significant correlation between animal age and first extraction colour and fluorescence confirmed that the amount of Maillard reaction was not simply a function of animal age nor was gelatine colour purely a function of Maillard reaction. (For example, first extraction colour could vary with the time of conditioning as well as animal age).

From correlation 3 above it was evident that as animals became older the collagen extractability decreased. In addition, from the approximately 290 nm fluorescence in Figures 4 it would appear that bovine hide gelatine does in fact contain a detectable amount of pyridinoline cross-link. This cross-link was evidently alkali labile as found by Eyre et al. (1984) as this fluorescence was not seen in Figure 3. Furthermore, the 295/395 nm fluorescence was not found with any gelatine from alkali treated hides. Hence, if it was accepted that the pyridinoline cross-link does exist in bovine hide, that the amount is constant (as shown by Uchiyama et al. 1991) and that the cross-link is alkali labile, then the extractability phenomena as shown in Table 1 could be explained as follows:

Although it is recognised that there are a number of nonfluorescent collagen cross-links, it was significant that young 10 month-old animal hide collagen exhibited very high extractability after a very moderate (2 weeks) liming presumably due to the destruction of predominating pyridinoline cross-links. As the animal age increased so the pyridinoline cross-link became of lesser significance compared to the alkali stable (pentosidine) cross-links. The result was that liming had a progressively smaller effect on the extractability of the hide.


This investigation showed that bovine hide gelatine exhibited fluorescence due to both the pyridinoline (295/395 nm) and the pentosidine (335/385 nm) collagen cross-links. In most cases, however, the pyridinoline fluorescence was not observed because this raw material had been treated with alkali prior to conversion to gelatine.

The strong correlation (r = 0.95) between the first extraction gelatine colour and the first extraction pentosidine 335/385 nm fluorescence showed that, at least in part, the colour of bovine hide gelatine was attributable to the in vivo Maillard reaction. The superior correlation (r = 0.976) between overall colour and senescence indicated that virtually all the colour of gelatine was related to animal age but that there was at least one more chromophore unrelated to pentosidine fluorescence which needs to be identified or characterised.

The response of bovine hide to liming can be explained by the presence of alkali labile pyridinoline cross-links which become less and less important as additional alkali stable (pentosidine) cross-links are formed with senescence.


British Standard 757. 1975. (BS 757:1975). Methods for sampling and testing gelatine. British Standards Institution. London.

Dyer, D.G., Blackledge, J.A., Katz, B.M., Hull, C.J., Adkisson, H.D., Thorpe, S.R., Lyons, T.J., Baynes, J.W. 1991. The Maillard reaction in vivo. Zeitschrift für Ernährungswissenschaft 30(1): 29-45.

Eyre, R.E., Paz, M.A., Gallup, P.M. 1984. Cross-linking in collagen and elastin. Annual Review of Biochemistry 53: 717-748.

Eyre, D. 1987. Collagen cross-linking Amino Acids. Methods in Enzymology 144: 115-139.

Monnier, V.M., Sell, D.R., Miyata, S., Nagaraj, R.H. 1990. The Maillard reaction as a basis for a theory of ageing. In The Maillard Reaction in Food Processing, Human Nutrition and Physiology. Finot, P.A., Aeschbacher, H.K., Hurrell, R.F., Liardon, R. (Ed.). p. 393-414. Advances in Life Sciences. Birkhauser Verlag Basel.

Munck, L., de Francisco.A. 1989. Fluorescence analysis in foods. p. 37. Longman Scientific & Technical and John Wiley & Sons Inc. New York.

Odetti, P., Pronzato, M.A., Noberasco, G., Cosso, L., Traverso, N., Cottalasso, D., and Marinari, U.M. 1994. Relationships between glycation and oxidation related fluorescence in rat collagen during ageing. An in vivo and in vitro Study. Laboratory Investigation 70(1): 61-67.

Reich, G., Walther., S., and Stather, F. 1962. Uber einige gesetzmäßigkeiten bei der herstellung von gelatine aus schweinshaut nacht dem sauren aufschlußverfahren.[On some of the laws governing the production of gelatine from pigskin according to the acid conditioning process.] Deutsche Lederinstitut, Frieberg/SA. 18: 15-23.

Sell, D.R. and Monnier, V.M. 1989. Structure elucidation of a senescence cross-link from human extracellular matrix. Journal of Biological Chemistry 264: 21597-21602.

Tanaka, S., Avigad, G., Eikenberry, E.F., Brodsky, B. 1988.Isolation and partial characterization of collagen chains dimerized by sugar-derived cross-links. Journal of Biological Chemistry 263: 17650-17657.

Thornton, A.C.R. 1966. Depolarisation of fluorescence studies on gelatin sols and gels. Gelatine and Glue Research Association Research Report A.33. April 1966.

Uchiyama, A., Ohishi, T., Takahashi, M., Kushida, K., Inoue, T., Jugie, M., Horiuchi, K. 1991. Fluorophores from ageing human articular cartilage. Journal of Biochemistry (Tokyo) 110: 714-718.

Yamauchi, M., Woodley, D.T., Mechanic, G.L. 1988. Ageing and cross-linking of skin collagen. Biochemical and Biophysical Research Communications. 152: 898-903.


Submitted in partial fulfilment of the requirements for the degree of Ph.D (Food Science). University of Pretoria.

Thanks to Leiner Davis Gelatin (South Africa) for their support.