A sensible technologist needs a fine balance between determination and boldness on the one hand, and modesty and caution on the other. Thus, I have always thought that it must be possible to predict whether or not beef will be tough or tender from a single, rapid, non-destructive measurement made on the intact carcass soon after slaughter. Early attempts failed, such as the Armour Tenderometer, but then so did all the first airplanes until the Wright brothers found the answer, not by luck or from basic theoretical knowledge, but by applied research. On the other hand of the balance, modesty and caution are required, otherwise one is deprived of two of the key requirements for applied research - knowing how much one depends on others, and impartiality.

Connective tissue is one cause of toughness in beef, and fluorescence is a way of detecting it. Fluorescence is a very interesting phenomenon in which a substance absorbs light energy, and re-emits it at a higher wavelength. Thus, when an ultraviolet (UV) black-light in a trendy club or bar excites the fluorescence of a cotton shirt, the cotton fibres re-emit the UV light at a longer wavelength - blue-white. Connective tissue collagen does the same thing, and we can detect it using a fibre-optic probe.


The importance of connective tissue in beef toughness is not universally acknowledged by meat scientists. Every butcher knows that connective tissues are responsible for major differences in tenderness between cuts of meat, as in the contrast of stewing beef with prime beef steak, and for differences in tenderness between beef from young and old cattle. But what about customers who complain about toughness in a prime roast or steak? Those who dispute the importance of connective tissue argue that, despite the wealth of biochemical information published on connective tissue in meat, little has been published that correlates collagen biochemistry with taste panel responses to commercial beef. Scientific journals seldom allow negative findings to be reported, because failure to find an effect does not necessarily constitute a proof of nonexistence. Thus, we do not know how many research groups have searched for such relationships and failed.

Thus, given these uncertainties, some meat scientists have argued that connective tissues are gelatinized by appropriate cooking methods so that any toughness in a prime steak can only originate from the state of the myofibrils or sarcomeres. And they have concluded that connective tissues do not play a major role in commercial problems with beef toughness. I have always favoured the traditional butcher's view that connective tissues are important, even when comparing prime steaks between prime carcasses, and this is what causes me to believe that rapid detection of connective tissue must be possible in the intact carcass. However, there is no doubt whatsoever that sarcomere length and integrity, plus the state of the cytoskeleton, are very important in explaining meat toughness in the real world, as in the notorious toughness of cold-shortened meat and beef consumed before aging. There is no doubt that connective tissue toughness in beef is only one of several possible causes of toughness, but what I wished to learn was how much, and how important in typical beef.


There are two things that most concern us about connective tissue in beef: how much, and how strong. As cattle get older, their connective tissues get stronger and stronger, and more resistant to cooking. Despite some initial uncertainties in the 1980s, pyridinoline now is widely recognised as a cooking-resistant cross-link between three tropocollagen molecules. In the medical field, the presence pyridinoline in the urine is used as a marker of pathological collagen degradation.

Differences in the degree of crosslinking may occur between different muscles of the same carcass, and between the same muscle in different species. For example, collagen from the longissimus dorsi is less cross-linked than collagen from the semimembranosus, and collagen from the longissimus dorsi of a pork carcass is less cross-linked than collagen from the bovine longissimus dorsi. Nutritional factors such as high-carbohydrate diet, fructose instead of glucose in the diet, low protein, and pre-slaughter feed restriction may reduce the proportion of stable crosslinks. Non-enzymatic glycosylation (a reaction between lysine and reducing sugars) may be involved in the interaction between diet and collagen strength. In general, the turnover rate of collagen is accelerated in cattle fed a high energy diet. The rate of collagen turnover in skeletal muscle may be about 10% per day and the turnover time for collagen may be inversely proportional to collagen fibril diameter.


It is common knowledge among those interested in the microscopic study of animal cells that both collagen and elastin fibres are fluorescent, emitting blue-white light when excited with UV. With excitation using UV light at 335 nm and 370 nm, and measurement at 385 nm and 440 nm, respectively, the fluorescence of collagen increases exponentially with age and is a reliable marker for biological age. The cause of this age-related increase in fluorescence is thought to be associated with cross-linking and polymerization from glycosylation by reducing sugars. Elastin contains crosslinking amino acids which are fluorescent.

During the 1980s, as biochemists developed collagen fluorescence as a method for measuring biological age, I decided to investigate the fluorescence of collagen and elastin in meat. Microscope spectrofluorometry takes a while to master, especially if building and programming one's own apparatus as I do, but eventually some relatively simple techniques were established, then these were adapted for on-line application using fibre-optics. As a development strategy this offered two important advantages: firstly, working under the microscope enabled the exact type of tissue generating a fluorescence spectrum to be determined, and secondly, many of the techniques of microscope photometry were readily adaptable to working with optical fibres.

Although microscopy is useful for identifying the microstructural sources of fluorescence in a meat product, the fluorescence emission spectrum of a standard tissue such as bovine tendon may vary considerably in shape when measured by microscopy, or at the semi-micro level using a fibre-optic meat probe. Unfortunately, fluorescence measurements are difficult to standardize and they tend to reflect the apparatus with which they are recorded as much as the nature of the sample. Eventually, however, it proved possible to obtain similar fluorometry spectra with both microscopy and meat probes.

There are two different methods to obtain information about the connective tissue content of meat using fluorometry. One method is to obtain stereological information on the distribution of collagen in meat, taking a broad-band measurement of the overall fluorescence emission through a small window of a moving probe as an indicator of collagen distribution. The other method is to measure the fluorescence emission spectrum of the meat through a large stationary window, using the shape of the spectrum to obtain information on both the amount and biochemical type of the collagen in the field. The latter method depends on the fact that biochemical Types I and III collagen have different fluorescence emission spectra. Maximum excitation for most types of collagen in meat is at 375 nm, although 370 nm may give the best separation of Types I and III collagen using a ratio of emission at 440 to 510 nm. Thus, for practical purposes, excitation may be obtained with the 365 nm peak of a mercury lamp. But Type I fibres emit a pre-quenching spectrum for longer than Type III fibres, probably because Type I fibres have a larger diameter than Type III fibres. The relative intensities of pre- and post-quenching emission spectra may be quantified by taking a ratio (such as 440:510 nm). In this case, the fluorescence of elastin is added to the pre-quenching peak around 440 nm that is associated with Type I collagen fibres. This peak is an indicator for connective tissues that may cause meat toughness. On the other hand, the fluorescence of the endomysium is combined with the fluorescence of Type III collagen fibres around adipose cells. Both of these contribute to the post-quenching peak at 510 nm which is regarded as the background fluorescence of meat with low connective tissue toughness. This is a fortunate coincidence that enables fluorescence to be used to predict connective tissue toughness in meat.


It is certainly difficult, and perhaps impossible, to name any invention that does not use parts or ideas of earlier inventions, even if the application has been altered in the new invention. Thus, one might trace back the whole of our modern technology to one hungry, cold butcher sitting in the mouth of a cave, contemplating on his front lawn herds of prime steak clad in warm skin coats. At his feet were some sharp, flint flakes, and the rest is history, even if unwritten. Thus, the fluorescence probe for connective tissue was based on a modification of an earlier invention, the fat-depth probe. The principle of the fat-depth probe is to make a series of measurements through a small probe-window relative to depth in the meat. Thus, having found fat depth, and knowing carcass weight, it is possible to predict meat yield. But, instead of searching for a discrete boundary between fat and muscle, the connective tissue probe was modified to pass through muscle detecting connective tissue fluorescence in relation to meat depth.

For UV light, it was not possible to mount the light source and detector within the shaft of the probe, as is possible with the small diodes used in fat-depth probes. Instead, a single optical fibre was used to create a window on the shaft of the probe, then, back near the main controller, the source and detector light paths were separated by a dichroic mirror. At wavelengths less than 400 nm, the light from the mercury arc was reflected from the dichroic mirror into the optical fibre, passing along the fibre and out of the probe window. Fluorescent connective tissues at the probe window emitted light that was collected by the optical fibre and passed back towards the controller. Instead of being reflected into the illuminator, most of the fluorescence, at wavelengths over 400 nm, passed through the dichroic mirror to the photomultiplier. Filters were required in front of the arc lamp to remove some of the heat and most of the light over 400 nm, while a high-pass filter in front of the photomultiplier helped prevent any stray light below 400 nm reaching the photomultiplier.


Many of our customers rate beef tenderness very highly as a key attribute of beef quality that needs to be improved and made more consistent. This is especially important if we are battling against bad publicity that beef might be a health risk. Everything in life carries a risk to benefit ratio. Flying to a warm beach for a short winter holiday can do wonders for your enjoyment of life, but not if the airplane crashes. A pint of beer can dissolve many worries, but not if you get drunk, drive a car, or become an alcoholic. Thus, if we wish our customers to keep eating beef against bad publicity, we must make the beef tasty and tender, otherwise it is all risk and no benefit.

Having searched the scientific literature fairly thoroughly (although one can never be sure one has read absolutely everything), I was unable to find any hard evidence that sporadic toughness in good quality beef was caused by connective tissue. Certainly, I cannot fault my colleagues who, faced with similar experience, had concluded that beef toughness must have some cause other than connective tissue, such as a rapid growth rate reducing degradative enzyme activity.

But, staying loyal to the traditional butcher's view of the importance of connective tissue, and armed with my first probe, I tested three age groups of Canada Grade A carcasses, assembled on production records and dentition: at approximately 12 months (59 carcasses at 293 ± 37 kg), 17 months (54 at 344 ± 43 kg), and 24 months (54 at 352 ñ 46 kg). In rib-eye and round steaks, age-related effects in connective tissue distribution were detectable with the meat probe. From 12 to 17 months the incidence of fluorescence peaks per centimetre increased, probably as layers of perimysium developed and became detectable. From 17 to 24 months, the incidence of fluorescence peaks per centimetre decreased, probably as layers of perimysium were pushed apart by muscle fibre growth. Fluorescence peaks became wider from 17 to 24 months, probably as layers of perimysium grew thicker.

A search was made for relationships between probe signals and taste panel evaluations of meat toughness (chewiness). At 12, 17 and 24 months, respectively, absolute values for the strongest simple correlations of chewiness with the probe signal were r = .32 (P < .01), .32 (P < .01), and .47 (P < .0005), for semitendinosus, and .31 (P < .01), .29 (P < .05), and .18 (NS), for longissimus dorsi. The corresponding multiple correlations derived from stepwise regressions of indices were R = .64, .61, and .86 for semitendinosus, and .63, .47 and .65 for longissimus dorsi. Here it was: the first evidence that rapid, non-destructive measurements of connective tissue fluorescence in beef could predict taste panel responses!

Obviously, as stressed earlier in this article, connective tissue is not the sole cause of variation in tenderness in top quality beef, but it does have an effect, exactly where one might anticipate, in the older animals, and in cuts of meat with an intermediate level of connective tissue toughness. Nobody is suggesting that this technology in its present form should be used for statutory grading, but it would be foolish to ignore the commercial possibilities that it offers. Thus, for a plant that hand-picks beef for a premium treatment such as aging vacuum packed primal cuts, this technology could help considerably in avoiding any meat with excessively high connective tissue levels. Thus, taking the 20 animals that the taste panel found had the toughest meat (all ages pooled) and comparing them with the remainder, the tough carcasses had a higher mean peak intensity than the others (0.027 ± 0.013 versus 0.020 ± 0.010, respectively, P < 0.005). Bearing in mind that a measurement takes only a couple of seconds in a meat cooler and is virtually non-destructive, this technology has some potential as an on-line method to predict meat toughness originating from connective tissue.

Where beef is a by-product of the dairy industry, the commercial justification for using fluorescence probe technology to sort beef is strong, because everyone agrees that connective tissues may make a substantial contribution to the toughness of beef from old cows. In a recent trial of Danish beef, taste panel tested samples of longissimus dorsi and semitendinosus muscles from 20 beef carcasses covering a wide chronological age were tested. In longissimus dorsi, the number of fluorescence peaks per centimetre was correlated with toughness (r = 0.73, P < 0.0005), whereas the half-width of peaks was the best single predictor for semitendinosus (r = 0.66, P < 0.005). Using stepwise regression of a number of features of the signal stronger prediction equations are possible (R = 0.95 for meat toughness). Correlations of taste with the fluorescence signal also were detected (r = 83, P < 0.0005 for peaks per centimetre), although they might have been secondary to a correlation of taste with toughness (r = 61, P < 0.005). However, semitendinosus juiciness was correlated with the height of the largest fluorescence peak in a transect (r = 0.78, P < 0.0005) even though no significant correlations of semitendinosus juiciness with toughness were detectable (r = 0.06). Thus, from this preliminary study it appears that the UV probe works as well on a wide age range of carcasses as it does on a narrow age range. Another concern is that differences in the degree of aging might fool the probe, but this did not happen in a recent trial funded by a large Canadian retail chain interested in comparing beef from different sources (Canadian, versus US, versus Australian).


I have no idea whether the connective tissue probe really is going to improve our quality control and grading of beef to give our customers a consistent, tender product. If I was a self-employed inventor, I might well have a try, but I am not. Thus, the Ontario Cattlemen's Association and the Canadian Meat Council are the owners and current developers of this technology (I donated the rights to the technology to them). They have produced their first commercial prototype and, by all accounts so far, it is working well enough.. The basic problem seems to be, however, that nobody wants to know if they have tough beef. Thus, in Canada at least, lack of vertical integration is more of a problem than the technological problem of detecting tough beef. Right now, I have a promising method for sarcomere length which might be combined in a single probe to detect both short sarcomeres and connective tissue fluorescence - but why bother? Also, I have shown that the probe can be reduced in size down to a 16-gauge hypodermic that could be used on live animals, but nobody wants to use it for genetic selection. Only when there is direct vertical integration of the beef industry will producers have a serious interest in producing tender beef. Until then, keep chewing! Ever wondered why beef consumption keeps going down and down?