Slide 1 Hello, this is Ferdinando Almeida. I'm here today to talk about results from three experiments that we conducted to determine the effects of heat damage on the nutritional composition and on the amino acid digestibility of canola meal, sunflower meal, and cottonseed meal fed to pigs. Slide 2 This is the outline of this podcast. We'll start with an introduction; then I'll move on to the objectives of the research; discuss some of the materials and methods; and then we'll look into the results from the three experiments; overall summary; and then I'll leave you with some implications of this research. Slide 3 Many feed ingredients in the swine industry are processed, particularly oilseed meals; they are heat processed primarily to decrease the concentration of antinutritional factors such as trypsin inhibitors. But also, heat process is used to desolventize these meals. And then, in the desolventizing process, depending on the equipment that is used, we'll have varying temperatures, pressure, moisture, and also the time of exposure of these ingredients to heat will vary. Slide 4 Some of the consequences of heat process is the formation of the Maillard reactions. And it has been shown that Maillard reactions, they decrease not only amino acid concentration but also the amino acid digestibility in feed ingredients. And as I mentioned before, because there are different equipments being used in different processing facilities, we will see some variations in feed ingredient composition within the same plant, and also among different plants. Slide 5 Now, this diagram gives a brief explanation of the Maillard reactions. And here, on the top left, we have first a carbohydrate. And we have circled there the carbonyl group of that carbohydrate, which will likely react with the amino group of amino acids or proteins. And in this case here, we have lysine, and the epsilon amino group of lysine is very susceptible to this reaction under heat, pressure, and moisture. So the reaction is divided into initial, intermediate, and advanced stages. In the initial stages, we have the formation of the Schiff bases, which then, as the reaction progresses, form the Amadori compounds. These Amadori compounds, they will go through a series of other reactions called the Strecker degradation, which will then yield to the formation of pre-melanoidins and then melanoidins at the more advanced stages of the reactions. The effects of these reactions, as I said before, because amino acids are involved in it, will decrease the amino acid concentration and also the digestibility because once this reaction occurs, it is irreversible and therefore the protein or amino acids affected are no longer available to the animal. Slide 6 Now, one of the problems that we have is that if we have a protein source which has been heat damaged, in the case of oilseed meals for example, we have two kinds of lysine in that heat damaged material. We have what we call reactive lysine, which is the lysine that has a free epsilon amino group and has not participated in the Maillard reactions, but we also have blocked lysine. And this blocked lysine is the lysine that participated in the Maillard reaction, and therefore is not available to the animal. However, when we take this material to do a standard amino acid analysis, we have to acid hydrolyze the protein before we can analyze for amino acids. And then, under the acid hydrolysis steps, what we have here is reactive lysine, which is the same reactive lysine that we had before the acid hydrolysis, but we also have what we call regenerated lysine. Regenerated lysine is the blocked lysine which is converted back to this form of lysine. But under the curve in the chromatogram, we have both reactive lysine and regenerated lysine accounting for total lysine, which is the value that we use in diet formulation. So what we have here is an overestimation of the concentration of reactive lysine in feed ingredients that have been heat damaged. Slide 7 The Maillard reactions are also involved with other amino acids. For example, we can have pre-melanoidins, cysteine, and arginine reacting, and therefore reducing the digestibility of cysteine and arginine. There are also arginine and lysine interactions under such circumstances, and under the Maillard reactions we also have the formation of crosslinks of other amino acids. So, the proteins, they become more aggregated overall, and therefore digestive enzymes are not as efficient in digesting these proteins and that's why we have an overall reduction in the amino acid digestibility if heat damage has occurred. Slide 8 Some ways we can assess the degree of heat damage in feed ingredients have been proposed, and these include the measurement of reactive lysine, which has been studied using different methods. For example, some research has been conducted using the fluorodinitrobenzene method. Others have used the homoarginine. Others, the furosine procedure. And also, some research has been conducted using the sodium borohydride procedure. Because most of the research in this area has been conducted using the furosine procedure, this is what we're going to focus on for these three experiments. And so we're going to determine reactive lysine using the furosine procedure. Slide 9 Now, this diagram shows how the furosine method works. We have a heat damaged protein. In that protein, we have the Amadori compounds, which were formed in the initial stages of the Maillard reaction, and we also have reactive lysine. Now, the Amadori compounds, under acid hydrolysis, they yield furosine, pyridosine, and regenerated lysine as I previously explained. It is believed that these three compounds are formed in ratios of 32% for furosine, 28% for pyridosine, and 40% for regenerated lysine. Therefore, if we can measure the amount of furosine, we can then back calculate the amount of regenerated lysine using these ratios. And by doing so, we can then use the total lysine that was analyzed and subtract the concentration of regenerated lysine and therefore, we can calculate the concentration of reactive lysine. Slide 10 Others have also used other measurements of the degree of heat damage in feed ingredients. For example, Gonzalez-Vega and coauthors have conducted a study with soybean meal. And in this study, they had a control soybean meal, and then they had two other batches of soybean meal that were autoclaved for 15 minutes or 30 minutes. And we can see here in the pictures, we see the control is light color, and then as they autoclaved for 15 or 30 minutes, we can see that soybean meal becomes much darker. And this is a result of the formation of those advanced Maillard reaction products, such as melanoidins. What they observed here, also, is that the crude protein concentration does not change as we heat damage this feed ingredient. However, the lysine concentration goes down from 3.05% in the control soybean meal to 2.69% in soybean meal that was autoclaved for 30 minutes. For this reason, we can then calculate the lysine:crude protein ratio, and that value will also decrease as the ingredient is heat damaged. So we can see here the control soybean meal, with a lysine:crude protein ratio of 6.29, and the soybean meal autoclaved for 30 minutes with a lysine:crude protein ratio of 5.57. They also determined the concentration of furosine in this soybean meal, and as we can see here, the concentration of furosine increased from 0.015% in the control soybean meal all the way up to 0.026% in soybean meal that was autoclaved for 30 minutes. Although this has been shown for soybean meal, it has not been shown for canola meal, sunflower meal, or cottonseed meal. Slide 11 So, the objectives of these experiments were to evaluate the effects of heat damage on the nutritional composition and also on the standardized ileal digestibility, or SID, of amino acids in canola meal, sunflower meal, and cottonseed meal. And also we wanted to determine prediction equations for the SID lysine using the furosine concentration in these three feed ingredients. Slide 12 This table summarizes the materials and methods. On the first row, we have the ingredients that we used, so canola meal (abbreviated as CM), sunflower meal, cottonseed meal. And then we autoclaved these ingredients at 130 Celsius with different times. The design was a replicated 5x5 Latin square, so we had 10 observations for each treatment. And the initial body weight of the pig for the first experiment with canola meal was 26.5 kg, for the second experiment with sunflower meal was 23.1 kg, and for the third experiment with cottonseed meal was 35.0 kg. Slide 13 Each period consisted of seven days. The first five days were adaptation to the diets, and followed by two days of eight-hour collection of ileal digesta. And the pigs were surgically equipped with a T-cannula in the distal ileum, and they were placed in crates such as this one on the right, with fully slatted floor and a feeder and a nipple drinker. Slide 14 The way we processed the ingredients, we had a common batch of each ingredient, which was then divided into four batches. So we had one batch that was not autoclaved, and three other batches that went into the autoclave for 130 Celsius. And, as I said before, the time was different for each batch. And so, at the end we had five diets mixed: four diets containing each batch of ingredient, so, one that was not autoclaved and three others that were autoclaved, and also a nitrogen-free diet that was used to determine the basal endogenous losses of amino acids and crude protein. Slide 15 For the statistical analysis, we used the Mixed procedure of SAS. Diet was included in the model as the fixed effect, pig and period were the random effects, and we used the LSMeans to calculate the means. We then conducted some contrasts to determine the linear and quadratic effects of increasing levels of heat treatment. Slide 16 Now, let's move on to the results. Slide 17 And for the first experiment, which was again conducted with canola meal: Slide 18 This graph shows the chemical composition of canola meal. We have concentration on the Y-axis in percent, and then on the X-axis we have here in this case, ADF, NDF, and lignin. Canola meal that was not autoclaved is in the red bar, canola meal that was autoclaved for 20 minutes is the yellow bar, canola meal autoclaved for 30 minutes is the blue bar, and canola meal autoclaved for 45 minutes is the green bar. So as we can see here, starting with ADF, as autoclaving time increases from 0 to 45 minutes, we see that there is an increase in the ADF concentration from 20% to 31%. We observed a similar pattern for the NDF concentration, which was increased from 33% in canola meal that was not autoclaved to 47% in canola meal autoclaved for 45 minutes. And also, we see an increase in the lignin concentration from 8% in canola meal that was not autoclaved to 17% in canola meal autoclaved for 45 minutes. And the reason for this is because, during the Maillard reactions, we have the formation of what is called artifact lignin. So the increases we see here for ADF and NDF are likely a result of this artifact lignin, which is then increasing the concentration of each of these components. Slide 19 This next graph shows the concentration of acid detergent insoluble nitrogen, and we have a similar patter for the colors here. So we see that, as we increase the time of autoclaving, there is an increase in the concentration of ADIN from 0.4% in canola meal not autoclaved to 1.8% in canola meal autoclaved for 45 minutes. Slide 20 Now, as I said before, we determined the furosine concentration. And as we can see here in this graph, the concentration of furosine increases from 0.016% in canola meal that was not autoclaved to 0.033% in canola meal autoclaved for 20 or 30 minutes. However, if we further autoclave that for 45 minutes, we see a slight decrease in the concentration of furosine from 0.033 to 0.025. And the reason for this is because the more we autoclave this canola meal, the likelihood of the Maillard reaction to progress to the more advanced stages increases. And as the reaction progressed to the advanced stages, that means we have less of a substrate -- in this case of the Amadori compounds -- which originate the furosine, and therefore we see a decrease in the concentration of this component. Slide 21 Now continuing with the chemical composition of canola meal, we have on the X-axis total lysine, the lysine:crude protein ratio, reducing sugars, and also reactive lysine. And as we can see here the concentration of total lysine, lysine:crude protein ration, the reducing sugars and reactive lysine, they all decrease as we autoclave canola meal. Exactly what has been observed previously for soybean meal, and this is exactly what we would expect happening in canola meal as well. One thing that I want to point out here is that the reactive lysine concentration is less than the concentration of total lysine, and this is what we expected, and this is what has been observed for other feed ingredients as well. Slide 22 Now, looking into the standardized ileal digestibility of amino acids, I'm just showing the results for the first four limiting amino acids. And we can see here that the SID of lysine, methionine, threonine, and tryptophan, they all reduced as we autoclaved canola meal for 20, 30, or 45 minutes. This is a quadratic response. And we can see also that lysine is the amino acid that is most affected by heat damage. It decreases from 68% in canola meal that was not autoclaved all the way down to 21% when canola meal is autoclaved for 45 minutes. Slide 23 We then calculated the concentration of reactive lysine, and tried to use those concentrations to predict the concentration of SID lysine in percent. And you can see here that there is a pretty good correlation with an r-square of 0.82. And that indicates that the reactive lysine may be used as a predictor of heat damage in canola meal. Slide 24 Now, let's move on to the second experiment, with sunflower meal. Slide 25 And I have similar graphs as for the canola meal experiment, so we have first sunflower meal that was not autoclaved in the red bars; sunflower meal autoclaved for 20 minutes in the yellow bar; sunflower meal autoclaved for 40 minutes in the blue bar, and sunflower meal autoclaved for 60 minutes in the green bar. We can see here, again, that there is a slight increase in the concentration of ADF as we autoclave sunflower meal. We observed a similar pattern for the concentration of NDF, going from 32% in sunflower meal that was not autoclaved to 43% in sunflower meal autoclaved for 60 minutes. And we see a slight increase in the concentration of lignin from 5.6 in the non-autoclaved sunflower meal to 6.7% in sunflower meal autoclaved for 60 seconds. Slide 26 The concentration of ADIN was also increased from 0.22% in sunflower meal that was not autoclaved to 0.28% in sunflower meal autoclaved for 60 minutes. Slide 27 The furosine concentration was not changed from sunflower meal that was not autoclaved, it was 0.013, to the sunflower meal that was autoclaved for 20 minutes at 0.012%. However, as we autoclaved sunflower meal for 40 or 60 minutes, we see an increase in the concentration of furosine. Slide 28 In this graph, we have concentrations of total lysine, lysine:crude protein ratio, reducing sugars, and reactive lysine. And, just as we observed for canola meal, we observed a similar pattern in sunflower meal, where we have a reduction in the concentration of total lysine, lysine:crude protein ratio, reducing sugars, and reactive lysine as a result of heat damage. Slide 29 This graph shows the standardized ileal digestibility of lysine, methionine, threonine, and tryptophan, and in the case of sunflower meal, there was a linear effect of increasing time of autoclaving. And we can see that the digestibility of these amino acids decreased as we increased the time of autoclaving. Slide 30 We also conducted a regression analysis to see how well we can predict the SID lysine from the concentration of reactive lysine. And again, we observed a pretty good correlation with an r-square of 0.83. Slide 31 Now, the third and last experiment, conducted with cottonseed meal. Slide 32 In this experiment, we observed that the concentration of ADF did not change as a result of increasing time of autoclaving. However, we observed a slight increase in the concentration of NDF from 26% in cottonseed meal that was not autoclaved to 30% in cottonseed meal that was autoclaved for 60 minutes. And there was also a slight increase in the concentration of lignin from 5.5% in the non-autoclaved cottonseed meal to 6.7% in cottonseed meal autoclaved for 60 minutes. Slide 33 Again, showing the results for the acid detergent insoluble nitrogen, which was increased from 0.2% in cottonseed meal that was not autoclaved to 0.27% in cottonseed meal autoclaved for 60 minutes. Slide 34 The concentration of furosine was increased from 0.027 in cottonseed meal that was not autoclaved to 0.040 in cottonseed meal that was autoclaved for 15 or 35 minutes, and that number was then reduced to 0.03 in cottonseed meal autoclaved for 60 minutes. So, in this case, we see a pattern that is similar to what we observed for canola meal in the first experiment. Slide 35 For the concentrations of total lysine, lysine:crude protein ratio, reducing sugars, and reactive lysine, there was a decrease in these concentrations as we heat damaged cottonseed meal. Slide 36 And, for the SID of amino acids, we also observed a quadratic reduction in the digestibility of lysine, methionine, threonine, and tryptophan. Slide 37 When we used the concentration of reactive lysine to predict the concentration of SID lysine in cottonseed meal, we observed a correlation of 0.64, which is not as high as we observed for the previous two experiments with canola meal and sunflower meal, which suggests that reactive lysine may not be the best predictor for the degree of heat damage in cottonseed meal. Slide 38 For the overall summary: Slide 39 We have in this table a summary of the main findings in these three experiments. So, as we can see here, under the nutrient composition, we see that there is an increase in the fiber components for canola meal, there's an increase in the fiber components concentration for sunflower meal, and also an increase in the concentration of ADIN for cottonseed meal as we autoclaved or as we heat damaged each of these meals. And we also observed that there is a reduction in the concentrations of reducing sugars, lysine (and this is total lysine), reactive lysine, and also the lysine:crude protein ratio. And this is very much standard across all three feed ingredients evaluated. For the SID of amino acids, we see that it always decreases as we increase the time of autoclaving or as we increase the degree of heat damage. However, this can be sometimes a quadratic reduction, in the case of canola meal and cottonseed meal, and in other cases it can be a linear reduction, as we observed for sunflower meal. Slide 40 So the implications of this research is that our data provides evidence of the negative effects of heat damage in commonly fed feed ingredients. The protein quality of canola meal, sunflower meal, and cottonseed meal can be assessed by using the concentrations of lysine, lysine:crude protein ratio, and also reducing sugars, and in some cases, fiber components. Slide 41 From our data, however, it seems that the lysine:crude protein ratio is the most reliable measurement of the degree of heat damage. We looked at NRC values for canola meal, sunflower meal, and cottonseed meal, and those values are close to the values of the lysine:crude protein ratio that we observed in the present data. So we suggest that if one is using canola meal with a lysine:crude protein ratio less than 5.2, or sunflower meal with a lysine:crude protein ratio of less than 3.4, or cottonseed meal with a lysine:crude protein ratio of less than 3.9, we should be concerned about heat damage in these feed ingredients, as values less than these may indicate that these ingredients have been heat damaged. Slide 42 With that, I would like to acknowledge Evonik Industries for funding this research. Slide 43 And thank you for listening to this podcast. I encourage you to visit our website at nutrition.ansci.illinois.edu to find out more about other feed ingredients.