Syntheses of isocyanates via amines and carbonyl fluoride

Article abstract of DOI:10.1016/j.jfluchem.2015.05.007

Isocyanates are widely used in many different areas, but the most common synthesis route-phosgene route cannot fit the more and more rigorous restriction of safety and environment. Here, a facile synthesis method of isocyanates via amines and carbonyl fluoride is proven feasibly by expanding its applications to the syntheses of nine different isocyanates. And two differences with the phosgene route are proposed. The reaction could occur under milder conditions and afford isocyanates in good yields, especially for the isocyanates containing electron withdrawing groups. It is appealing for industrial application.

Full text of DOI:10.1016/j.jfluchem.2015.05.007

Animal Feed Science and Technology 205 (2015) 10–22
Contents lists available at ScienceDirect
Animal Feed Science and Technology
journal homepage: www.elsevier.com/locate/anifeedsci
Effects of ruminally protected methionine and/or
phenylalanine on performance of high producing Holstein
cows fed rations with very high levels of canola meal
N. Swanepoela,b,∗, P.H. Robinsonb, L.J. Erasmusa
a
Department of Animal and Wildlife Sciences, University of Pretoria, 0001, South Africa
Department of Animal Science, University of California, Davis, CA 95616, USA
b
a r t i c l e i n f o
a b s t r a c t
Article history:
Canola meal is the second largest protein feed in the Northern latitudes and inclusion levels
in dairy rations are expected to increase due to projected large increases in production of
canola seed in Canada. However, a recent study (Swanepoel et al., 2014) showed that even
though higher inclusions of canola meal (CM) had a positive effect on production when CM
directly substituted for high protein corn based dried distillers grains (DDG), that there was
an optimum point at 120–135 g/kg of diet dry matter (DM) after which animal performance
seemed to decline. Only the amino acids (AA), methionine (Met), phenylalanine (Phe) and
leucine (Leu) could have limited production based upon plasma AA concentrations at the
highest CM inclusion level. Our objective was to determine if either Met or Phe, or both,
was limiting performance of early lactation dairy cows fed a ration containing 180 g/kg
of diet DM as CM, by supplementing a calculated target of 7.5 g of intestinally absorbable
Phe/cow/day and/or 8.0 g of intestinally absorbable Met/cow/day in ruminally protected
(RP) forms to four pens of ∼320 early lactation cows/pen in a 4 × 4 Latin square with 28
days experimental periods. Dry matter intake was not affected (avg.: 27.6 0.4 kg/day)
by feeding either of the RP AA, or the combination. Phenylalanine supplementation alone
had no effect on milk production or composition, and body condition score (BCS) change
compared to Control. Supplemental Met alone modestly increased (P<0.01) milk protein
and fat content, while decreasing (P<0.01) milk lactose content and yield, but with no impact
on BCS change compared to Control. Combination Met and Phe supplementation decreased
milk and lactose yields, as well as lactose content (P<0.01), while increasing milk protein
content and the BCS change (P<0.01). Urine volume (avg.: 16.7 0.31 L/day) and flow of
microbial protein (MCP) from the rumen (avg.: 2092 52.7 g CP/day) were not affected by
any treatment. Plasma Met levels increased (P<0.01) with both Met treatments and plasma
tryptophan (Trp) levels decreased (P<0.01) with both Phe treatments. However, plasma Phe
levels did not change with any treatment. Results are interpreted to suggest that delivery of
Met with RP Met feeding was higher than animal requirements and caused an oversupply
of Met. Addition of Phe to the Met supplementation changed the way energy was utilized
by the cows, redirecting energy liberated by Met from milk components toward BCS gain.
Received 15 October 2014
Received in revised form 1 April 2015
Accepted 2 April 2015
Keywords:
Spot urine purine
Estimated microbial flow
Plasma amino acids
Protein feeding
Abbreviations: AA, amino acid; ADF, acid detergent fiber; ADICP, AD insoluble CP; ADIN, acid detergent insoluble N; AL, allantoin; aNDF, amylase-treated
NDF; aNDFom, aNDF free of residual ash; BCS, body condition score; BCAA, branched-chain AA; BW, body weight; CM, canola meal; CP, crude protein;
CR, creatinine; DC305, DairyComp 305 management system; DDG, dried distillers grains; DHIA, Dairy Herd Improvement Association; DIM, days in milk;
DM, dry matter; MCP, microbial CP; NDF, neutral detergent fiber; NEL, net energy for lactation; OM, organic matter; PD, purine derivatives; RDP, rumen
degradable CP; RP, rumen protected; SCC, somatic cell count; SG, specific gravity; TMR, total mixed ration; TP, true protein.
∗
Corresponding author at: Department of Animal and Wildlife Sciences, University of Pretoria, 0001, South Africa. Tel.: +27 79 891 8920.
E-mail address: nanswanepoel@gmail.com (N. Swanepoel).
http://dx.doi.org/10.1016/j.anifeedsci.2015.04.002
0377-8401/© 2015 Elsevier B.V. All rights reserved.

N. Swanepoel et al. / Animal Feed Science and Technology 205 (2015) 10–22
11
It remains unclear if Phe was limiting in the Control ration or if RP Phe was not fed at
high enough levels to have a measurable response on production. However, it is clear that
AA limitations, requirements and production responses are governed by much more than
plasma AA levels. Results further suggest that AA are bioactive metabolites to the extent
that they can change animal performance, even when they are not “limiting” per se, and that
their supplementation to practical dairy cattle diets should be approached with extreme
caution for this reason.
© 2015 Elsevier B.V. All rights reserved.
1. Introduction
Canola meal (CM) is the second largest protein feed in the Northern latitudes and inclusion levels in dairy rations are
expected to increase due to projected increased production of canola seed in Canada (Growing Great, 2015). The goal of the
Canola Council of Canada with this program was to produce 15 million tons of canola seed annually by 2015 but, with record
breaking seed production in 2013; annual production currently stands at 18 million tons. Vegetable oil demand worldwide
is expected to rise 60% in the next decade (Canola Council of Canada Annual Report, 2013) and, with the increased crushing
capacity in Canada, this will have a cascading effect resulting in increased amounts of CM produced and used in North
American dairy rations.
A recent study (Swanepoel et al., 2014) showed that higher inclusions of CM in lactating dairy cow rations had a positive
effect on production when CM directly substituted for high protein dried corn distillers grains (DDG), but that there was an
optimum point at 120–135 g/kg of diet dry matter (DM) after which animal performance started to decline. This agrees with
other studies comparing CM to DDG which reported that higher proportions of CM, included at up to 120 g/kg DM, tended to
have higher milk and protein yields (Mulrooney et al., 2009). It was clear, however, in Swanepoel et al. (2014) that the high
rumen degradable protein (RDP) content of CM, and resultant high rumen ammonia levels, did not limit microbial protein
(MCP) production when CM was included in the diet at 200 g/kg DM, suggesting that it may have been the availability of
absorbable amino acids (AA), and/or specific AA(s), that limited productive performance of the cows. In Swanepoel et al.
(2014), only methionine (Met), phenylalanine (Phe) and leucine (Leu), could have limited production, based upon their
declining plasma AA concentrations as the CM inclusion level in the diet increased.
Our objective was to determine if either Met or Phe, or both, was limiting performance of early lactation dairy cows fed
a ration containing CM as the sole supplementary crude protein (CP) source, by supplementing Met and/or Phe in ruminally
protected (RP) forms.
2. Materials and methods
The experimental design used four pens of ∼320 early lactation cows/pen in a 4 × 4 Latin square with 28 days experimental
periods, utilizing the William’s experimental design (Williams, 1949) to balance for potential carryover of treatment effects.
The study took place during winter from 28 December 2012 to 18 April 2013 with temperature ranging between −3.6 and
28.3 ◦C and humidity between 22.7 and 100.0%. All cows were cared for relative to applicable laws of the state of California
and the USA, and were consistent with requirements for “The care and use of animals for scientific purposes”, as per the
South African National Standard (SANS 10386-2008).
2.1. Farm and management
The same commercial dairy farm (located in Hanford, CA) used in Swanepoel et al. (2014) was selected for this study.
Every week cows were randomly allocated to one of four early lactation pens from a single fresh pen and, once confirmed
pregnant, cows were moved from these pens to mid lactation pens. At the start of the 1st period, treatments were randomly
allocated to each of the four early lactation pens and rotated after each 28 days experimental period consistent with a
William’s design.
2.2. Diets
Mixing of the total mixed rations (TMR) and all other farm practices were as outlined in Swanepoel et al. (2014). All
four of the pens were fed the same base TMR based on alfalfa hay, whole crop winter wheat and corn silages, and corn
grain, with a premix containing dry ingredients (i.e., almond hulls, fuzzy and cracked pima cottonseed, wheat straw, liquid
molasses, mineral premix, CM), with CM inclusion in all TMR targeted at 180 g/kg of total ration DM (Table 2). Cows were fed
ad libitum to achieve ∼3% refusals on an as fed basis, with each pen receiving a total of ∼16,000 kg of as-mixed TMR/day in
two feedings. Cows were fed one full 11,000 kg load of TMR (which contained the RP AA) at the 1st feeding, between 04:30
and 07:30 h, to a clean bunk as bunks were cleared of all residual feed while the cows were at morning milking. A second
∼5000 kg load of TMR was fed at the 2nd feeding between 11:00 and 12:30 h and weights for each load of TMR fed were

12
N. Swanepoel et al. / Animal Feed Science and Technology 205 (2015) 10–22
recorded on record sheets at the time of feeding and used together with daily refusals to calculate DM intake/cow/pen. The
“TMR Tracker” system (Digi-Star LLC, Fort Atkinson, WI, USA) kept a record of the actual ingredient profiles of each load of
TMR mixed.
2.3. The rumen protected AA products
The RP Met product (Smartamine M; Adisseo USA, Inc., Alpharetta, GA, USA) contains 750 g/kg d,l-methionine with a
250 g/kg fat encapsulation (stearic acid) and a pH sensitive intestinal release. Using a variety of methods, the average rumen
stability of the Met in Smartamine M has been estimated and assumed to be between 750 and 800 g/kg (Schwab, 1995,
2007; Rulquin and Kowalczyk, 2003; Chen et al., 2011; Osorio et al., 2013). Specific gravity (SG) is reported by Adisseo
to be 0.70 g/cm3 (http://www.sfm.state.or.us/cr2k subdb/MSDS/SMARTAMINE.PDF). Since Smartamine M’s coating can be
damaged and its integrity compromised by physical impact, cutting and abrasion, a physical inspection of the individual
product beadlets were conducted twice a week, after the RP Met has been mixed into the TMR and delivered to the feedbunks,
to ensure acceptable product delivery. The number of beadlets which were destroyed or physically changed during mixing
and feeding was negligible.
The RP Phe product was manufactured by QualiTech, Inc. (Chaska, MN, USA) according to the same specifications as their
RP Lys product described in Sakkers et al. (2013) except that the Phe product did not contain the Co-EDTA marker. The RP Phe
contained 600 g/kg Phe combined with 400 g/kg fat as a matrix after which the pellets were sprayed with another coating of
the same fat matrix. The fatty acid profile of the fat matrix, as reported in Wrinkle et al. (2012), primarily contained rumen
stable C14:1 trans, C16:0 and C18:0 fatty acids. Due to reactivity of Lys with its fat coating, it is difficult to incorporate it into
a RP form with acceptable rumen stability and intestinal release. Therefore QualiTech adapted its coating to increase the
proportion of fat used with its RP Lys product used in Sakkers et al. (2013). Since Phe has not been ruminally protected in the
past, this procedure of fat coating was followed to ensure high protection and post ruminal delivery of the Phe. However,
since Phe is not as reactive as Lys, it is likely that our RP Phe product had higher rumen stability than the 527 g/kg reported
by Sakkers et al. (2013) for the similarly protected Lys product. Therefore, a whole tract stability of 600 g/kg was assumed
for our Phe product. The SG, determined to be slightly higher than 1.207 g/cm3 according to the procedure described by
Swanepoel et al. (2010) using different concentrations of a saline solution, is attributed to the long chain fatty acids used in
the fat coating (Wrinkle et al., 2012).
Therefore, assuming a total duodenal delivery of 580 g/kg Met (i.e., 750 g/kg Met multiplied by 775 g/kg rumen stability)
and 360 g/kg Phe (i.e., 600 g/kg Phe multiplied by 600 g/kg rumen stability) treatments were created by adding either RP Phe
alone (PHE), RP Met alone (MET) or both AA together (M/P) by mixing 20.9 g/cow/day of RP Phe (estimated to deliver 7.5 g of
intestinally absorbable Phe/cow/day) and/or 13.7 g/cow/day of RP Met (estimated to deliver 8.0 g of intestinally absorbable
Met/cow/day) into the base TMR by adding a pre-weighed bag of the RP product(s) to the dry ingredient premix prior to its
addition to the TMR mixer.
2.4. Sample collection, preparation and analytical methods
2.4.1. Total mixed rations and ingredients
Individual feed ingredients and TMR were sampled twice during the last 7 days (i.e., the sampling week) of each of the
four experimental periods. Ten handfuls (of 200 g each) of each TMR were collected according to Robinson and Meyer (2010)
at pre-marked posts with evenly spaced intervals along the bunk-line immediately after feeding and before the cows had
access to it. Ingredient samples from all four periods were pooled (n = 4 samples/ingredient), while TMR samples were pooled
within period and pen (n = 16 TMR samples) for chemical analysis.
All TMR samples, silages and other wet ingredients were weighed, dried at 55 ◦C for 48 h and air equilibrated for 24 h
before being sent for chemical analysis to the UC Davis service laboratory. All samples were ground to pass a 1 mm screen
on a model 4 Wiley Mill (Thomas Scientific, Swedesboro, NJ, USA). Oven DM was determined as the gravimetric loss when
dried at 105 ◦C for 3 h in a forced air oven (NFTA, 2006). Ash determination was based on gravimetric loss by heating samples
to 550 ◦C for at least 3 h (Method 942.05, AOAC, 2005). Total N was determined by the Leco method (Method 990.03, AOAC,
2005). Acid detergent fiber (ADF) was determined as the residue after acid detergent extraction with the ADF residue sub-
sampled for extraction with sulfuric acid to determine lignin (sa) or analyzed for N to determine acid detergent insoluble N
(ADIN) according to Method 973.18 of AOAC (1997). Neutral detergent fiber (NDF) was determined using neutral detergent
and heat (Method 2002.04, AOAC, 2006). Heat-stable amylase was used to remove starch and inactivate enzymes that may
degrade the fiber (aNDF). Results for NDF were also reported on an ash-free basis (aNDFom). Starch was determined as total
glucose minus free glucose multiplied by a factor of 0.9 as described by Smith (1969). Fat was quantified using the Randall
modification of the standard Soxhlet extraction (Method 2003.05, AOAC, 2006).
2.4.2. Animal measurements
A group of ∼195 cows with the lowest days in milk (DIM, i.e., 10–125 DIM) were selected from each pen at the start of the
study (i.e., the cows most likely to complete the study due to their low DIM at the start) and coded in DairyComp 305 (DC305
(Valley Agricultural Software, Tulare, CA, USA) is an electronic herd record system) to be used as the base group of cows
from which all representative subgroups were selected for animal samples (i.e., urine, blood) and measurements (i.e., body

N. Swanepoel et al. / Animal Feed Science and Technology 205 (2015) 10–22
13
condition scores; BCS). Milk production and composition data used all cows which remained eligible (i.e., in their originally
assigned pen) throughout the study, regardless of their DIM at the start of the study. For a cow to remain eligible (i.e., to be
included in any sampling dataset and the resulting statistical analysis), they had to have been in their originally assigned
pen for the entire 16 weeks study, which was checked by examination of daily records of cow pen assignment within DC305.
In addition, no cow was ever physically observed to be in an incorrect pen.
2.4.2.1. Milk production and composition. Milk data were collected on Day 28 of each experimental period by Dairy Herd
Improvement Association (DHIA) personnel. Milk yields were recorded for each cow and milk samples collected by drawing
a small representative sub-sample from the sample collection flask (after a short period of mixing) and preserving it with
a 2-bromo-nitropropane-1, 3-diol for subsequent analytical testing. Fat, true protein, lactose and somatic cell count (SCC)
were determined with the Bentley Combi using optical infrared analysis at the DHIA laboratory in Hanford (CA, USA).
2.4.2.2. Body condition score. A subgroup of ∼140 cows/pen from the base group of ∼195 cows/pen (see Section 2.4.2) were
body scored throughout the study. This was completed by the same trained scorer on the first day of Period 1 and at the
end of each experimental period. The 5 point BCS system of Ferguson et al. (1994) was used and adapted as described in
Swanepoel et al. (2014) to include intermediate points between the ¼ point scores.
2.4.2.3. Urine. On Day 4 of the sampling week of each experimental period, spot urine samples were collected from the
first ∼35 cows which voluntarily urinated during normal morning lockup (for normal health and reproductive checks;
∼50 min/pen/day) and immediately placed in ice. Samples were only retained if the cow was a part of the original base
group of ∼195 cows/pen. The SG of each untreated urine sample was measured using a digital handheld pen refractometer
(Atago USA, Inc., Bellevue, WA, USA). A small quantity of urine (7 ml) was then combined with 1 ml of 100 ml/L sulfuric acid
to reduce the final pH to <2 in order to prevent bacterial destruction of allantoin (AL), diluted with water to a final volume of
35 ml and frozen at −20 ◦C. Urine samples were chemically analyzed for creatinine (CR) and AL as described by Swanepoel
et al. (2014). Two inter-run standard urine samples were used in each analytical run to assess variation among runs. The
average concentration of the inter-run standards over all runs was then used to correct sample concentrations in each run.
2.4.2.4. Blood plasma. A subgroup of 24 cows/pen was selected from the base group of ∼195 cows/pen for blood sampling
with collection and treatment following the same methods as outlined in Swanepoel et al. (2014).
2.5. Calculations
Final oven DM was calculated as air equilibrated DM (i.e., dried at 55 ◦C and air equilibrated for 24 h) multiplied by the
laboratory oven DM (i.e., dried at 105 ◦C) of the air equilibrated sample.
Data backups of the DC305 herd record system were used to determine the number of cows in each pen on each day of
the collection week of each period, and used together with the weights of each load of TMR fed and daily refusals to calculate
DM intake per cow/pen as:
(Total intake(kg as fed/day) − daily refusals)
× (TMR DM proportion).
cows in pen
Milk energy content (MJ/kg) was calculated using a prediction equation from Tyrrell and Reid (1965), summing the
energetic weights of the milk components as:
ꢀ
ꢁ
[(4.163 × fat (g/kg)) + (2.413 × (true protein (g/kg)/0.94)) + (2.16 × Lactose(g/kg))] − 11.72 × 2.204
× 4.184,
1000
with the factor 1000 converting kcal to Mcal, 2.204 converting Mcal/lb to Mcal/kg and 4.184 converting Mcal/kg to MJ/kg.
True protein (TP) was converted to CP assuming 60 g/kg non-protein N in total milk N (Akers, 2002). Milk energy output
(MJ/day) was calculated by multiplying milk energy content (MJ/kg) by daily milk yield (kg/day). Body condition score change
was calculated as the difference between the BCS at the end and at the beginning of each period and BCS change energy
(MJ/day) was calculated as:
ꢂ
ꢃ
BCS change × 300
× 4.184,
28
assuming 1 unit BCS change over 28 days = 300 Mcal energy (Chilliard et al., 1991) with the factor 4.184 converting Mcal/day
to MJ/day. Urine volume (L/day) was calculated using an equation derived from data published by Burgos et al. (2005) as:
332.66 × (((SG − 1) × 1000) − 0.884).
Total daily purine derivative (PD) excretion (mmol/day) is the sum of PD excreted in urine and milk of lactating dairy
cows (Chen and Gomes, 1992), and was estimated assuming allantoin is 0.906 of total PD in urine (obtained from values

14
N. Swanepoel et al. / Animal Feed Science and Technology 205 (2015) 10–22
reported by Vagnoni and Broderick, 1997; Valadares et al., 1999; Gonzalez-Ronquillo et al., 2003; Reynal & Broderick, 2005;
Moorby et al., 2006) and PD excretion in milk is a constant 0.05 of urine PD excretion (Chen and Gomes, 1992).
Microbial purines absorbed from the intestine (X, mmol/day) were calculated using the equation reported by Chen and
Gomes (1992), using a constant assumed body weight (BW) of 673 kg (Swanepoel et al., 2014), as:
(Total daily PD − 0.385 × (BW0.75))
,
0.85
assuming that the endogenous contribution of PD is 0.385 mmol/kg BW0.75 and that 0.85 represents recovery of absorbed
purines as PD in the urine. Microbial CP production (g CP/day) was then estimated as:
ꢄ
ꢅ
(X(mmol/day) × 70)
(0.116 × 0.83 × 1000)
× 6.25,
assuming an N content of 70 mg N/mmol for purines, a ratio of purine N:total N in mixed rumen microbes as 11.6:100, a
microbial purine digestibility of 0.83 and the conversion of microbial N to MCP by the factor 6.25.
A partial net energy (NE) output (MJ/day) balance was calculated by summing the milk, BCS change energy and main-
tenance, with maintenance net energy needs calculated from NRC (2001) and assuming a constant BW of 673 kg, as:
(6730.75 × 0.08) × 4.184.
Net energy for lactation (NEL) density (MJ/kg DM) of the rations were estimated using the biological responses of the
cows, as expressed in the partial NE output, and measured DM intake on a pen basis as:
Net energy output (MJ/day)
.
DM intake (kg/day)
2.6. Statistical analysis
Cows were only included in the statistical analysis if they did not move from their originally assigned pen during the
study, for health or any other reason. Thus the number of cows eligible for statistical analysis of milk production was 608,
and for the BCS dataset it was 348. Outlier analysis (completed blind to treatments by excluding values deemed to be not
biologically possible), excluded 12 cows from the milk production dataset (i.e., 8 cows for a milk fat level > 57 g/kg, 3 cows
for milk yields below 11.5 kg/day and 1 cow for a milk lactose proportion > 85 g/kg), and 5 cows which were removed from
the BCS dataset due to abnormally high BCS changes within an experimental period. A group of 24 (i.e., 6 cows/pen) cows
were randomly selected from the 96 eligible blood cows for plasma AA analysis. A total of 529 urine samples were collected
from 363 cows, as several cows were sampled in more than one period, but the group of 114 urine samples selected for AL
and CR assays only came from the 42 cows which had repeated urine samples between periods.
Animal production, BCS, urine AL, urine CR and plasma AA levels were analyzed using the MIXED procedure of SAS (2000)
for a 4 × 4 Latin square design, with cow as the experimental unit within pen in the random statement and period, pen and
treatment as fixed effects. Dry matter intake (n = 4 pens, calculated on a pen basis with 4 pens/period), TMR components and
ingredients and NE balance (n = 4 pens) used pen as the experimental unit in the GLM option of SAS (2000) with period, pen
and treatment as fixed effects.
Table 1
Chemical analysis of ingredients used in the total mixed rations (g/kg dry matter) fed to the treatment groups.a
Dry matter
Organic matter
Crude protein
ADFb
aNDFomc
aNDFd
ADINe
Alfalfa, hay
Alfalfa, fresh chop
Alfalfa, haylage
906
253
398
973
906
876
919
920
932
359
322
162
897
867
840
930
931
988
954
960
858
893
937
958
174
226
202
44
380
74
216
219
38
267
277
323
233
170
27.0
264
317
440
352
272
173
322
303
316
295
219
77.0
354
418
614
485
423
200
337
333
372
305
239
78.0
367
436
660
523
437
211
1.5
1.9
2.0
1.6
2.8
IR
2.2
2.3
1.1
0.6
0.8
<0.5
Almond, hulls
Canola meal, pellets (solvent)
Corn, flaked grain
Cottonseed, cracked pima
Cottonseed, fuzzy linted
Wheat, straw
Wheat, silage
Corn, silage
Citrus, wet pulp (orange)
71
66
66
IR = insufficient residue from ADF determination for N analysis.
a
n = 4; except alfalfa, fresh chop and alfalfa, haylage = 2.
Acid detergent fiber, expressed inclusive of residual ash.
Neutral detergent fiber assayed with heat stable amylase, expressed exclusive of residual ash.
Neutral detergent fiber assayed with heat stable amylase, expressed inclusive of residual ash.
Acid detergent insoluble nitrogen.
b
c
d
e

N. Swanepoel et al. / Animal Feed Science and Technology 205 (2015) 10–22
15
Reported values are least squares means with differences accepted as significant if P≤0.01 and trends accepted if P≤0.05.
3. Results
3.1. Ration evaluation and intakes
The chemical composition of the ingredients used in the TMR (Table 1) was similar to ingredients listed in NRC (2001).
Only the alfalfa hay had a lower NDF proportion (337 vs. 450 g/kg) and canola pellets a lower CP proportion (380 vs. 420 g/kg).
There was no difference in the chemical profiles of the TMR fed to the four treatment groups (Table 2). At 169 g/kg, the
level of CM was slightly lower than the targeted 180 g/kg, but it did not differ among treatments and was well above the
suggested optimum level of 135 g/kg according to Swanepoel et al. (2014).
The only difference in the ingredient profiles of the treatment diets was inclusion of 13.4 and 20.5 g/cow/day of RP Met and
RP Phe, respectively (which was equal to the targeted 13.7 and 20.9 g/cow/day). Even though alfalfa haylage was substituted
with alfalfa fresh chop in the 3rd and 4th periods, it resulted in no changes in the nutrient profile of the treatment TMR,
which were the same among periods. The TMR met all nutrient requirements of lactating dairy cows producing 45–50 L of
milk/day (NRC, 2001).
3.2. Animal measurements
3.2.1. Intake, milk production and its composition
Dry matter intake (Table 3) was not affected (avg: 27.6 + - 0.40 kg/day) by treatment. Intakes for the Control ration were
higher (27.8 vs. 25.28 kg/day) than for the all CM ration in Swanepoel et al. (2014), even with lower milk production (44.1
vs. 47.4 kg/day).
Table 2
Ingredient profile and chemical composition (g/kg dry matter) of total mixed rations fed to the treatment groups.a
Treatments
Control
Control vs. (P)
PHE
MET
M/P
SEM
PHE
MET
M/P
Ingredient profile, g/kg DMb
Alfalfa, hay
84
83
84
85
4.6
0.96
0.80
0.90
Premix
Almond, hulls
155
51.9
31.3
8.0
155
51.9
31.3
8.0
16.8
169
14.1
1.03
0.00
67.3
58.4
180
164
34.9
154
51.8
31.2
8.0
16.8
169
14.1
0.00
0.66
67.7
57.0
181
164
35.5
154
51.7
31.2
8.0
16.8
169
14.1
1.01
0.66
66.5
57.0
181
164
34.9
1.1
1.00
1.00
0.99
1.00
0.98
1.00
0.94
<0.01
1.00
0.65
0.71
0.48
1.00
0.97
0.85
0.87
0.89
0.93
0.80
0.85
0.72
1.00
<0.01
0.90
0.94
0.79
1.00
0.85
0.69
0.74
0.78
0.85
0.62
0.71
0.46
<0.01
<0.01
0.86
0.95
0.69
0.96
0.98
Cottonseed, cracked pima
Cottonseed, fuzzy linted
Wheat, straw
Mineral, premix
Canola meal, pellets (solvent)
Molasses, liquid
RPP productc
0.43
0.30
0.13
0.09
1.3
0.06
0.014
0.008
2.73
2.58
1.9
16.8
169
14.1
0.00
0.00
67.1
57.2
181
164
34.8
RPM productd
Alfalfa, fresh chop/haylagee
Wheat, silage
Corn, flaked grain
Corn, silage
Citrus, wet pulp (orange)
Nutrient profile, g/kg DMf
Dry matter
5.5
3.03
557
921
160
64
318
305
223
44
546
923
160
63
313
299
216
44
554
921
161
64
317
304
222
45
555
923
159
63
317
302
217
44
9.7
1.0
4.1
1.9
2.8
2.7
2.7
1.0
6.0
0.22
0.28
0.98
0.46
0.16
0.12
0.08
0.58
0.96
0.74
0.98
0.92
0.94
0.80
0.74
0.72
0.68
0.29
0.84
0.28
0.86
0.48
0.70
0.41
0.13
0.88
0.70
Organic matter
Crude protein (CP)
ADICPg
aNDFh
aNDFomi
ADFj
Fat
Starch
172
173
164
175
a
n = 4 pens.
b
Based on average ingredient composition during the sampling week for each pen, each period, as reported by TMR tracker system.
Ruminally protected Phe (QualiTech, Inc., Chaska, MN, USA). Fed at 13.7 g/cow/day to deliver 8 g intestinally absorbable Phe.
Ruminally protected Met (Smartamine M, Adisseo USA, Inc., Alpharetta, GA, USA). Fed at 20.9 g/cow/day to deliver 7.5 g intestinally absorbable Met.
Alfalfa haylage used only in Periods 1 and 2. Alfalfa fresh chop used only in Periods 3 and 4.
c
d
e
f
Total mixed ration samples collected twice during sampling week for each pen, each period (i.e., 32 total samples), samples pooled by period and pen
(n = 2 per period).
g
Acid detergent insoluble CP (g/kg of CP).
Neutral detergent fiber assayed with heat stable amylase, expressed inclusive of residual ash.
Neutral detergent fiber assayed with heat stable amylase, expressed exclusive of residual ash.
Acid detergent fiber, expressed inclusive of residual ash.
h
i
j

16
N. Swanepoel et al. / Animal Feed Science and Technology 205 (2015) 10–22
Table 3
Production performance and body scores for cows fed rations with different ruminally protected amino acids.
Treatment
Control
Control vs. (P)
SEM PHE
M/P vs. (P)
PHE
MET
M/P
27.7
MET
M/P
PHE
MET
n = 4 pens
Dry matter intake (kg/day)
27.8
28.1
28.3
0.40
0.79
0.62
0.87
0.67
0.52
n = 596 cows
Yield (kg/day)
Milk
44.1
1.49
1.32
2.11
43.7
1.48
1.30
2.08
43.5
1.50
1.33
2.07
43.1
1.46
1.31
2.05
0.31
0.10
0.41
0.06
0.06
0.03
0.62
0.41
<0.01
0.03
0.18
0.05
0.17
0.57
0.02
0.15
<0.01
0.03
Fat
0.013
0.008
0.015
True protein
Lactose
<0.01
<0.01
0.29
Components (g/kg)
Fat
True protein
Lactose
Energy density (MJ/kg)
Somatic cell count (‘000)
34.2
30.2
47.8
2.87
127
34.1
30.1
47.7
2.87
130
34.7
30.7
47.5
2.90
109
34.1
30.6
47.6
2.88
98
0.23
0.11
0.06
0.011
0.84
0.44
0.10
0.62
0.85
0.01
<0.01
<0.01
<0.01
0.16
0.67
<0.01
<0.01
0.92
0.82
<0.01
<0.01
0.55
<0.01
0.10
0.08
<0.01
0.38
10.5
0.02
0.01
n = 343 cows
Body condition score (BCS)
BCS change (unit/28 days)
2.65
0.04
2.64
0.04
2.65
0.06
2.66
0.08
0.021
0.011
0.57
0.98
0.69
0.10
0.45
<0.01
0.19
<0.01
0.25
0.25
The PHE treatment had no effect on milk production or composition vs. Control (Table 3), but MET increased milk pro-
tein (30.2 vs. 30.7 g/kg; P<0.01) and fat (34.2 vs. 34.7 g/kg; P=0.01) content, while decreasing milk lactose content (47.8 vs.
47.5 g/kg; P<0.01) and its yield (2.11 vs. 2.07 kg/day; P<0.01). Milk yield tended (P=0.03) to decrease with MET, while M/P
did decrease (P<0.01) milk (44.10 vs. 43.14 kg/day) and lactose (2.11 vs. 2.05 kg/day) yields, as well as milk lactose content
(47.8 vs. 47.6 g/kg; P<0.01), while increasing the milk true protein content (30.2 vs. 30.6 g/kg; P<0.01). The M/P treatment
tended to decrease milk fat yield (P=0.03) and SCC (P=0.02).
Compared to M/P, PHE had a lower milk true protein content (30.6 vs. 30.1 g/kg; P<0.01), higher lactose content (47.6 vs.
47.7; P<0.01) and yield (2.05 vs. 2.08; P=0.02) as well as higher SCC (P=0.01), while tending to a higher milk yield (P=0.05). For
MET, increases (P<0.01) were observed for milk true protein yield (1.33 vs. 1.31 kg/day), milk fat yield (1.50 vs. 1.46 kg/day)
and content (34.7 vs. 34.1 g/kg) and milk energy density (2.90 vs. 2.88 MJ/kg) compared to M/P.
3.2.2. Body condition score
Body condition score (Table 3) was not affected by any treatment and the mean of 2.65 is normal for high producing early
lactation cows. Cows in all treatments gained BCS, which is expected in cows past peak production, but only M/P increased
(P<0.01) the change in BCS vs. Control and PHE (0.08 vs. 0.04 unit change/28 days).
3.2.3. Urine
Urine AL and CR concentrations (Table 4) did not differ among treatments. Control AL concentrations were higher (3812
vs. 3370 mg/L) and CR lower (882 vs. 946 mg/L) than for the all CM ration in Swanepoel et al. (2014). Urine volume (avg.:
16.7 0.31 L/day) also did not differ among treatments. As expected, the calculated flow of MCP from the rumen (avg:
2092 52.7 g CP/day) was also not affected by the treatments. These calculated MCP flow values are higher than the ranges
(763–1959 g CP/day) previously reported in the literature when duodenal samples were collected and MCP flow directly
measured (Khorasani et al., 1993; Robinson et al., 1994, 1996, 1998a,b; Stensig and Robinson, 1997; Timmermans et al., 2000;
Gonzalez-Ronquillo et al., 2003; Moorby et al., 2006), but this is likely due to the higher DM intakes (27.9 vs. 20.7 kg/day)
in our study compared to that literature, which suggests that our MCP flows, estimated using urine AL concentrations, are
Table 4
Urine analysis for cows fed rations with different ruminally protected amino acids.
Treatment
Control
Control vs. (P)
M/P vs. (P)
PHE
MET
M/P
SEM
PHE
MET
M/P
PHE
MET
n = 42 cows
Allantoin (AL, mg/L)
Creatinine (CR, mg/L)
Specific gravity
Urine volume (L/day)
Total PDa excreted (mmol/day)
MCPb yield (g CP/day)
3812
882
1.03
16.4
454
3734
933
1.03
16.6
446
2114
3564
821
1.03
16.8
431
2030
3601
894
1.03
16.8
438
2071
112.5
36.2
0.001
0.31
9.9
0.59
0.20
0.70
0.61
0.56
0.56
0.12
0.17
0.29
0.32
0.10
0.10
0.18
0.78
0.22
0.20
0.26
0.26
0.37
0.35
0.36
0.40
0.55
0.55
0.82
0.11
0.88
0.80
0.59
0.59
2155
52.7
a
Purine derivatives.
Microbial protein.
b

N. Swanepoel et al. / Animal Feed Science and Technology 205 (2015) 10–22
17
Table 5
Free amino acid and ammonia concentrations (␮g/ml) in plasma of cows fed rations with different ruminally protected amino acids.
Treatment
Control
Control vs. (P)
M/P vs. (P)
PHE
MET
M/P
SEM
PHE
MET
M/P
PHE
MET
n = 24 cowsa
Essential amino acids
Threonine
Valine
Methionine
Isoleucine
Leucine
Phenylalanine
Tryptophan
Lysine
14.7
32.9
3.58
15.0
20.2
9.21
16.8
13.2
8.81
15.5
3.80
14.9
32.6
3.71
15.2
20.1
9.49
15.2
13.1
8.67
15.5
3.60
14.6
33.4
4.97
15.8
20.8
9.49
16.4
13.9
8.85
16.6
2.91
14.1
31.5
4.56
14.8
19.7
9.31
15.7
12.9
8.46
15.2
2.95
0.50
0.90
0.167
0.50
0.65
0.252
0.46
0.43
0.285
0.57
0.086
0.73
0.67
0.47
0.74
0.93
0.29
<0.01
0.84
0.58
0.95
0.05
0.81
0.27
0.15
0.19
<0.01
0.48
0.45
0.49
0.25
0.80
0.44
0.66
<0.01
0.38
0.03
0.03
0.07
0.07
0.50
0.23
0.05
0.15
0.02
0.66
0.61
<0.01
0.15
0.34
0.30
0.40
0.12
0.89
0.06
<0.01
0.09
<0.01
0.72
0.40
0.71
0.04
0.65
0.19
0.71
<0.01
Histidine
Arginine
Lys:Met ratio
Non-essential amino acids
Homocystine
Aspartic acid
Tyrosine
Serine
Glutamic acid
Glutamine
Glycine
Alanine
3-Methylhistidine
Ammonia
1.02
1.60
9.92
9.83
7.60
48.0
31.0
23.6
0.71
2.94
1.02
1.56
9.86
9.85
7.43
48.1
30.9
23.8
0.97
1.59
9.98
9.66
7.33
48.9
29.2
24.7
0.98
1.55
9.78
9.48
7.35
47.2
29.8
23.2
0.108
0.083
0.399
0.356
0.216
2.06
1.44
0.91
0.062
0.146
0.95
0.67
0.88
0.96
0.45
0.94
0.95
0.79
0.25
0.48
0.74
0.94
0.89
0.62
0.22
0.70
0.13
0.11
0.07
0.87
0.74
0.58
0.72
0.29
0.26
0.76
0.33
0.62
0.56
0.50
0.78
0.90
0.83
0.27
0.70
0.70
0.36
0.44
0.57
0.97
1.00
0.64
0.61
0.57
0.93
0.49
0.60
0.04
0.21
0.41
0.62
3.06
0.57
2.91
0.67
3.06
a
A randomly selected group of 6 cows/pen/period was used for amino acid analysis as it was decided that additional samples would not change
significance of differences.
biologically sensible. However one study (Reynal and Broderick, 2005), with similar milk production (avg. 42.3 kg/day) and
DM intakes (avg. 25.5 kg/day) to our study, reported MCP flows that were higher (2683 vs. 2092 g CP/day) using urinary
excretion of PD together with N:purine ratios in omasal samples.
3.2.4. Blood plasma
There were no changes in plasma AA levels (Table 5) for MET or M/P vs. Control, except Met which increased with both
(P<0.01), and Trp tended (P=0.04) to decrease with the M/P treatment. In contrast, plasma Trp decreased with PHE vs. Control
(P<0.01), while plasma Phe levels were not impacted by any treatment. Due to higher plasma Met levels, both the MET and
M/P treatments decreased the Lys:Met ratio (P<0.01).
Compared to the M/P treatment, MET tended to increase (P<0.05) valine (Val), Met, Lys, alanine (Ala) and arginine (Arg)
while PHE decreased Met (P<0.01), thereby increasing the Lys:Met ratio from 2.95 to 3.60 (P<0.01).
Since it was not measured, when AA changes are discussed it is assumed that the pool sizes (plasma volumes) remained
constant between treatments.
3.2.5. Partial net energy balance
The partial NE balance (Table 6) for each treatment was calculated to determine where energy was utilized. Compared
to Control, M/P decreased total energy output in milk (126 vs. 124 MJ/day; P<0.01) while increasing energy used for BCS
change (1.7 vs. 3.6 MJ/day; P<0.01). However, there were no other differences in total NE output, or the calculated dietary
NEL densities, between any treatment and the Control. Therefore, the amount of NE expressed by the cows, and in the diets,
Table 6
Partial net energy balance for cows fed rations with different ruminally protected amino acids.a
Treatment
Control
Control vs. (P)
M/P vs. (P)
PHE
125
1.7
171
6.12
MET
126
2.8
173
6.12
M/P
124
3.6
172
6.22
SEM
PHE
MET
M/P
PHE
MET
Milk energy output (MJ/day)
BCSb energy (MJ/day)
126
1.7
172
6.19
0.8
0.48
1.1
0.16
0.98
0.65
0.72
0.73
0.10
0.74
0.73
<0.01
<0.01
0.84
0.19
<0.01
0.81
0.02
0.25
0.59
0.60
Total net energy (MJ/day)
c
NEL (MJ/kg DM)
0.088
0.86
0.60
a
Maintenance energy (MJ/day) calculated using a constant body weight of 673 kg for all treatments.
Body condition score.
Net energy available for lactation. n = 4 pens.
b
c

18
N. Swanepoel et al. / Animal Feed Science and Technology 205 (2015) 10–22
did not change among treatments. However, AA supplementation resulted in energy being partitioned differently, especially
the M/P treatment, which changed the way energy was utilized by either AA treatment alone compared to Control.
The M/P treatment had a lower milk energy output (124 vs. 126 MJ/day; P=0.02) compared to MET but a higher BCS
change energy (3.60 vs. 1.71 MJ/day; P<0.01) compared to PHE.
4. Discussion
In many feeding situations, supplementation of Met has been shown to increase milk protein content (Chen et al., 2011)
ˇ
and/or yield (Cermáková et al., 2012; Osorio et al., 2013) as well as milk fat content (Wang et al., 2010). This has also been
confirmed in a recent systemic review of the literature (Robinson, 2010) and meta-analysis (Patton, 2010), and it agrees with
our results in which milk protein and fat proportions increased modestly with the MET treatment. Patton (2010) suggests
that a slight increase in milk yield can also be expected, and, even though this is consistent with the increases in milk yield
ˇ
reported earlier (Wang et al., 2010; Cermáková et al., 2012; Osorio et al., 2013), it does not agree with the reduction in
milk lactose yield, or the tendency for milk yield to decrease with MET, in our study. A study by Robinson et al. (2000), to
determine effects of a Met oversupply, showed that abomasal Met infusion to increase its intestinal delivery by 34–39%
markedly reduced animal performance. Indeed Robinson et al. (2000) reported a sharp decline in milk and lactose yields,
which is similar to our results in relative terms, although not quantitatively as all changes in our study were modest even
though supplementation with RP Met increased estimated intestinally available Met by ∼38% in our study, suggesting that
an oversupply of Met was possible.
It is generally suggested that Lys and Met should be fed in the ratio of 3:1 (at the intestinal absorptive site) in order to obtain
beneficial responses in milk production and composition (Chalupa and Sniffen, 2006). Since our two Met supplementation
treatments reduced the plasma Lys:Met ratio (which should be reflective of the ratio of absorbable Lys and Met at the
intestinal absorptive site) from 3.80 in the Control ration to 2.91 and 2.95, respectively, this theoretically near perfect ratio
should have resulted in a positive production response. However, this was not the case. Even though the Lys:Met ratio concept
seems clear when looking at the data published in NRC (2001), other studies in the literature do not always agree. For instance,
Chen et al. (2011) reduced this Lys:Met ratio from 3.6 to 3.0, resulting in an increase in milk protein content, but only in a
relatively low protein ration. Supplementation of Met in the peripartal period to decrease the Lys:Met ratio from 3.4 to 2.9
only showed a tendency for milk protein content to increase (Osorio et al., 2013). Rulquin and Delaby (1997) also reported
an increase in milk protein content when the Lys:Met ratio was decreased from 3.8 to 3.0 through Met supplementation,
although their plasma AA analysis showed that the ratio was actually reduced from 4.8 to 2.2. As the predicted Lys:Met ratios
may not reflect the actual ratios, interpretation of the Lys:Met ratio concept in studies outside the NRC (2001) is difficult.
In the study of Robinson et al. (2000), creating an imbalance in the suggested optimum Lys:Met ratio by supplementing Lys
(i.e., changing the ratio from 3.0 to 3.9) had no effect on production, but creating a theoretical imbalance by supplementing
Met (i.e., changing the ratio from 3.0 to 2.3), reduced DM intake, milk production and lactose yields. While restoring the ratio
to 3:1 by supplementing Lys and Met together did not impact production, the negative effects of high Met supplementation
remained, even at the 3.0 Lys:Met ratio. This seems to suggest that Met supplementation to some point results in positive
production responses, regardless of the theoretical Lys:Met ratio, but that production responses became more and more
negative at higher and higher Met supplementation levels, regardless of the Lys:Met ratio. This strongly suggests that it is
the level of Met relative to its requirement that elicits the response rather than its ratio with Lys per se. Indeed this hypothesis
agrees with results of a study where the control ration had the lowest production, even with a calculated ratio close to 3.0,
while positive production responses were seen for both Met and Lys supplementation, even when the supplementation
changed the ratios to 1.3 and 4.6, respectively (Wang et al., 2010).
Robinson et al. (2000) reported that oversupplying both Met and Lys simultaneously changed partitioning of energy, with
their combined infusion improving energetic efficiency of the cows without affecting general animal performance, thereby
increasing the NE density of the ration to prevent a negative energy balance in the cows. This corresponds to our study
where combined supplementation of Phe and Met changed the way energy was utilized vs. when Met was supplemented
alone. While it is clear that Met supplementation alone directed energy from milk lactose to milk fat and protein production,
thereby shifting energy output within milk components, addition of Phe to the Met redirected energy from lactose production
toward BCS gain. As explained by Swanepoel et al. (2010), it is possible that an oversupply of Met resulted in increased fat
synthesis but that supplementation of the limiting AA, in this case Phe, rectified the AA imbalance, thereby restoring the fat
content. This suggests that Phe was limiting in our study, but that its supplementation level was not high enough to allow
for an effect on production. In contrast, it is clear that supplementation of Phe alone had no effect on animal performance,
or at least any response parameters measured in our study, suggesting that Phe only became limiting in the combination
treatment after Met requirements were met, but that a possible oversupply of Met, and undersupply of Phe, in the M/P
treatment did not allow for a production response. Arginine (Arg) metabolism in the liver is mainly directed toward the urea
cycle in order to dispose of AA in excess relative to the limiting AA (Bach et al., 2000). The tendency of Arg to increase in
the plasma for the Met treatment in this study (P=0.06) support the hypothesis that Met was probably limiting and that it’s
supplementation allowed utilization of other AA, reducing the amount of surplus AA catabolized by the liver and therefore
the requirement for Arg in the urea cycle.
Our study was designed to deliver 7.5 g of Phe to the intestine, which is 5.5% higher than the estimated intestinal Phe
delivery levels for the Control ration from Swanepoel et al. (2014). However, plasma Phe levels did not differ significantly

N. Swanepoel et al. / Animal Feed Science and Technology 205 (2015) 10–22
19
between treatments and were only 3.0% higher for PHE vs. Control, possibly suggesting that RP-Phe failed and that little
or no Phe reached the intestinal absorptive site. However this seems unlikely since the manufacturing technology of this
RP-Phe was the same as the RP-Lys product evaluated in Sakkers et al. (2013) which was measured in vivo to have a rumen
escape of 52% of consumed Lys. Haque et al. (2013) demonstrated that the separate infusion of Arg, isoleucine (Ile) and Val
to the duodenum resulted in increased concentrations of Ile and Val but not Arg, showing that plasma AA concentrations
are not always directly associated with its supplementation or removal. In this case the failure of plasma Phe concentra-
tions to increase in response to its duodenal supplementation can be interpreted to suggest that Phe was utilized by the
cow, instead of building up in the plasma, suggesting a more likely hypothesis, which is supported by the data, that Phe
was under-delivered relative to its needs. Indeed it has been suggested that Phe enhances the rate of Trp hydroxylation
(Kaufman, 1971), the first step in catabolizing Trp to synthesize serotonin and, since Trp levels in the plasma of both Phe
supplemented treatments decreased vs. Control, this supports delivery of Phe to the intestinal absorptive site, as well as its
absorption and availability in high enough quantities pre-liver to have elicited this enhanced effect on Trp hydroxylase in the
liver.
Phenylalanine is part of a group of AA which are extensively catabolized by the liver and removal of these AA are directly
correlated to their hepatic inflow (Lapierre et al., 2005; Bach et al., 2000). Thus as more of these AA are absorbed, more will
be removed by the liver (up to 0.49 of portal absorption). Bach et al. (2000) showed that even though supplementation of
the limiting AA reduced extraction and deamination of most AA by the liver, due to their utilization elsewhere, the rate of
catabolism of Phe by the liver remained constant. The liver also has the capability to export AA bound to peptides or proteins
in the blood, rather than free AA, which would not be reflected in the plasma data since blood AA concentrations were
not analyzed in this study due to the difficulties associated with it (Bach et al., 2000). These phenomena could account for
the lack of an increased post-liver plasma Phe level in both Phe treatments which is consistent with Lapierre et al. (2005),
suggesting that the reason for increased removal of AA by the liver is to remove excess AA, thereby equalizing post-liver
supply with both mammary uptake and milk output requirements. This suggests that Phe was not limiting production in
the Control diet and that its supplementation in an RP form was unnecessary.
Another possible explanation for the lack of a production response to supplemental Phe alone could be that when AA
which are usually extracted at levels lower than milk protein requirement (e.g., Phe, Tyr) are supplemented, extraction rates
and efficiencies considerably increase (Guinard and Rulquin, 1994), substituting mammary utilization of peptide-derived
Phe with that of added free Phe. This suggests that supplemented Phe did not provide an additional supply of AA, but rather
replaced peptides as the source of Phe to support milk protein synthesis (Bequette and Backwell, 1997). It is also possible
that, due to the relatively high MCP flow in our study and the large contribution of MCP to total absorbable protein, any
benefit that resulted from supplementation of Phe alone was too small to be measured (Robinson, 2010).
The shift in energy utilization with M/P vs. MET supports the hypothesis that Phe was absorbed into blood and that enough
was delivered to the mammary gland to elicit a response. If the additional free Phe replaced use of peptide-derived Phe in
the mammary gland, it is possible that more Met, which liberated energy from milk fat and lactose, provided the energy
required to incorporate the peptides back into muscle protein, resulting in some of the increased BCS gain. Jaurena et al.
(2005) showed that BCS gain reflects accretion of both fat and muscle mass, which was not the case with Phe supplementation
alone. This is supported by the tendency for a number of AA (i.e., Met, Lys, Arg, Ala, all branched-chain AA (BCAA)) to decrease
in the M/P vs. MET treatments. Since these AA were not incorporated into milk per se, it is possible that they were utilized
to support synthesis of muscle protein, as the increase in BCS change may attest. This is also supported by the decrease in
concentrations of alanine (P=0.04) and Lys (P=0.05), these being the AA which are most abundant in muscle tissue (Bach
et al., 2000).
Responses to AA (mainly Lys and Met) supplementation in general have been inconsistent, relatively small and, in many
cases, not what was expected in a literature which now dates back over 30 years. Some authors have ascribed a lack of
response to other dietary factors and/or other AA being more limiting than those supplemented (Karunanandaa et al., 1994;
Robinson et al., 1998a,b; Liu et al., 2000), although this often seems to be “form” reasoning. It seems more plausible that
supplemented AA are used in biological processes, resulting in effects on animal performance which differ depending on
whether they were actually limiting milk production, or not, with actual limitations being rare. As suggested by Robinson
et al. (1998), Met supplementation consistently enhances milk component yields, albeit to a small degree, beyond that
expected from its role as a limiting AA per se, and seemingly without connection to the expected basal Met flows to the
intestinal absorptive site. The perceived correction of a theoretical AA deficiency through its supplementation consistently
increases its plasma levels while also increasing milk protein content (Rulquin and Pisulewski, 2006; Weeks et al., 2006;
Haque et al., 2012). However, the same cannot be said for calculated AA “imbalances” at the intestinal absorptive site. In most
cases an estimated imbalance of AA manifests as increased milk fat content, or a low ratio of protein:fat in milk (Chamberlain
et al., 1992; Varvikko et al., 1999; Robinson et al., 2000; Cant et al., 2001; Kim et al., 2001; Weeks et al., 2006) possibly through
mechanisms in the mammary gland, such as increased blood flow to maintain milk protein yield, thereby supplying more
milk fat precursors to stimulate milk fat yield (Cant et al., 2001; Weeks et al., 2006). It has been suggested that endocrine
responses to total AA imbalances can override imperfections in AA profiles in order to maintain milk protein yields (Weeks
et al., 2006), although not all of the resulting changes may be deemed positive by dairy farmers. In addition, supplementation
of AA to rations in which they are not limiting are known to negatively affect animal performance through, for instance,
reducing DM intake (Karunanandaa et al., 1994; Robinson et al., 2000; Rulquin and Pisulewski, 2006) and sequestration of
AA in body protein that exacerbate the AA deficiency (Weeks et al., 2006).

20
N. Swanepoel et al. / Animal Feed Science and Technology 205 (2015) 10–22
In some cases, as in our study, supplementing a second AA (or group of AA) may result in the opposite, or a reversal, of the
response from only one AA (Polan et al., 1991; Kim et al., 2001; Wang et al., 2010), or simply a completely different response
as was seen when supplementation of BCAA in addition to Met and Lys, which were deemed limiting, had no additional effect
on milk production but stimulated muscle protein synthesis (Appuhamy et al., 2011). It has been suggested that it is not
blood AA concentrations per se that limits milk protein production, but rather the metabolic machinery which determines
maximum velocity of milk protein production (Cant et al., 2001) or a gastrointestinal event (Bequette et al., 1996), both of
which are under hormonal control.
It may be time to reconsider the “limiting AA” or “broken stick” concept of AA nutrition of lactating dairy cows in favor of
accepting that most AA are bioactive and can change animal performance, even when they are not ‘limiting’ milk production
per se, which is likely the normal situation. Thus recommendations to use RPAA products must consider the potential for
unwanted effects, which could be deemed negative, which are associated with oversupply, or unnecessary supply, of AA to
the intestinal absorptive site relative to animal “needs”.
5. Conclusions
Phenylalanine supplementation alone in an RP form caused no animal response since, even though the results suggest that
it was delivered and absorbed, it likely increased Phe catabolism in the liver since Met was limiting its use for production,
thereby lowering the amount of Phe which reached the mammary gland. In contrast, Met was likely oversupplied with
Met supplementation alone, which showed as increased milk fat and protein proportions. However, supplementation of
the combination of both AA possibly rectified the Met limitation and supplied Phe which became 2nd limiting after the
Met requirements were met. However it seems that we did not supply enough Phe to support a sustained milk production
response. Instead, as a first priority, free Phe may have replaced peptides which were previously mobilized from the muscle
to rectify the Phe limitation, directing those back to muscle protein synthesis with no surplus Phe remaining available to
increase milk production.
Responses to AA supplementation in dairy cows in a research time frame now exceeding 30 years. have been inconsistent
and unpredictable and, although authors provide many reasons to justify the seemingly random (but generally low or no)
responses to AA supplementation, it is clear that AA limitations, requirements and production responses are governed by
much more than their plasma AA levels. Indeed our results strongly suggest that AA should be viewed as bioactive metabolites
to the extent that they can change animal performance characteristics, even when they are not “limiting” per se, and that
their supplementation to practical dairy cattle rations should be approached with extreme caution for this reason.
Conflict of interest
The authors declared that they have no conflict of interest.
Acknowledgements
The authors thank Mr. William van Die and the personnel (i.e., George, Nacho, Pedro, Junior, José) at Cloverdale Dairy for
their interest, assistance and patience during the completion of the study. Efforts of the DHIA staff, especially Kari Furtado,
in Hanford in arranging and sorting extra milk test samples is also greatly appreciated. The senior author also thanks Prof.
Lourens Erasmus and Dr. Peter Robinson, fellow students (Jill Soderstrom, Stacey Tsang, Kristan Reed), and friends (Rodrigo)
in collecting and processing samples. Very special thanks go to Grace Cun for going above and beyond in both assistance and
support.
References
Akers, R.M., 2002. Lactation and the Mammary Gland. Iowa State Press, Ames, IA, USA.
AOAC, 1997. Official method of analysis of AOAC International, vol. I., 16th ed. Association of official Analytical Chemists, Inc., MD, USA.
AOAC, 2005. Official Methods of Analysis of AOAC International, 18th ed. Association of official Analytical Chemists International, Gaithersburg, MD, USA.
AOAC, 2006. Official Methods of Analysis of AOAC International, 18th ed. AOAC International, Arlington, VA, USA.
Appuhamy, J.A.D.R.N., Knapp, J.R., Becvar, O., Escobar, J., Hanigan, M.D., 2011. Effects of jugular-infused lysine, Met, and branched-chain amino acids on
milk protein synthesis in high-producing dairy cows. J. Dairy Sci. 94, 1952–1960.
Bach, A., Huntington, G.B., Stern, M.D., 2000. Response of nitrogen metabolism in preparturient dairy cows to methionine supplementation. J. Anim. Sci. 78,
742–749.
Bequette, B.J., Backwell, F.R.C., 1997. Amino acid supply and metabolism by the ruminant mammary gland. Proc. Nutr. Soc. 56, 593–605.
Bequette, B.J., Backwell, F.R.C., MacRae, J.C., Lobley, G.E., Crompton, L.A., Metcalf, J.A., Sutton, J.D., 1996. Effect of intravenous amino acid infusion on Leu
oxidation across the mammary gland of the lactating goat. J. Dairy Sci. 79, 2217–2224.
Burgos, S.A., Robinson, P.H., Fadel, J.G., DePeters, E.J., 2005. Ammonia volatilization potential: prediction of urinary urea nitrogen output in lactating dairy
cows. Agric. Ecosyst. Environ. 111, 261–269.
Canola Council of Canada, 2013. Annual Report. Canola Council of Canada, Winnipeg, MB, Canada http://www.canolacouncil.org/media/549926/2013
annual report.pdf (accessed 1109.14).
Cant, J.P., Trout, D.R., Qiao, F., McBride, B.W., 2001. Milk composition responses to unilateral arterial infusion of complete and histidine-lacking amino acid
mixtures to the mammary glands of cows. J. Dairy Sci. 84, 1192–1200.
ˇ
Cermáková, J., Kudrna, V., Illek, J., Blazˇková, K., Haman, J., 2012. Effects of a rumen-protected form of Met and a Met analogue on the lactation performance
of dairy cows. Czech J. Anim. Sci 57, 410–419.

N. Swanepoel et al. / Animal Feed Science and Technology 205 (2015) 10–22
21
Chalupa, W., Sniffen, C., 2006. Balancing rations on the basis of amino acids: the CPM-Dairy approach. In: Proc. 21st Annual Southwest Nutrition and
Management Conference, Tempe, AZ, February 23–24, pp. 96–109.
Chamberlain, D.G., Choung, J., Robertson, S., 1992. Protein nutrition of dairy cows receiving grass silage diets: effects of feeding a protein supplement of
unbalanced amino acid composition. J. Sci. Food Agric. 60, 425–430.
Chen, X.B., Gomes, M.J., 1992. Estimation of Microbial Protein Supply to Sheep and Cattle Based on Urinary Excretion of Purine Derivatives – An
Overview of the Technical Details. Int. Feed Res. Unit, Occasional Publ. Rowett Research Institute, Aberdeen, UK (edited and reprinted 1995)
http://www.macaulay.ac.uk/IFRU/pdf/chema.pdf (accessed 11.09.14).
Chen, Z.H., Broderick, G.A., Luchini, N.D., Sloan, B.K., Devillard, E., 2011. Effect of feeding different sources of rumen-protected Met on milk production and
N-utilization in lactating dairy cows. J. Dairy Sci. 94, 1978–1988.
Chilliard, Y., Cissé, M., Lefaivre, R., Rémond, B., 1991. Body composition of dairy cows according to lactation stage, somatotropin treatment, and concentrate
supplementation. J. Dairy Sci. 74, 3103–3116.
Ferguson, J.D., Galligan, D.T., Thomsen, N., 1994. Principal descriptors of body condition score in Holstein cows. J. Dairy Sci. 77, 2695–2703.
Gonzalez-Ronquillo, M., Balcells, J., Guada, J.A., Vicente, F., 2003. Purine derivative excretion in dairy cows: endogenous excretion and the effect of exogenous
nucleic acid supply. J. Dairy Sci. 86, 1282–1291.
Growing Great, 2015. http://www.canolacouncil.org/media/502091/growing great pdf.pdf (accessed 15.09.14).
Guinard, J., Rulquin, H., 1994. Effect of graded levels of duodenal infusions of casein on mammary uptake in lactating cows. 2. Individual amino acids. J.
Dairy Sci. 77, 3304–3315.
Haque, M.N., Rulquin, H., Andrade, A., Faverdin, P., Peyraud, J.L., Lemosquet, S., 2012. Milk protein synthesis in response to the provision of an ideal amino
acid profile at 2 levels of metabolizable protein supplies in dairy cows. J. Dairy Sci. 95, 5876–5887.
Haque, M.N., Rulquin, H., Lemosquet, S., 2013. Milk protein responses in dairy cows to changes in postruminal supplies of arginine, isoleucine, and valine.
J. Dairy Sci. 96, 420–430.
Jaurena, G., Moorby, J.M., Fisher, W.J., Cantet, R., 2005. Association of body weight, loin longissimus dorsi and backfat with body condition score in dry and
lactating Holstein dairy cows. Anim. Sci. 80, 219–223.
Karunanandaa, K., Goodling, L.E., Varga, G., Muller, L.D., McNeil, W.W., Cassidy, T.W., Lykos, T., 1994. Supplemental dietary fat and ruminally protected
amino acids for lactating Jersey cows. J. Dairy Sci. 77, 3417–3425.
Kaufman, S., 1971. The Phe hydroxylating system from mammalian liver. Adv. Enzymol. 35, 245–319.
Khorasani, G.R., Robinson, P.H., Kennelly, J.J., 1993. Effects of canola meal treated with acetic acid on rumen degradation and intestinal digestibility in
lactating dairy cows. J. Dairy Sci. 76, 1607–1616.
Kim, C.-H., Choung, J.-J., Chamberlain, D.G., 2001. Estimates of the efficiency of transfer of l-histidine from blood to milk when it is the first-limiting amino
acid for secretion of milk protein in the dairy cow. J. Sci. Food Agric. 81, 1150–1155.
Lapierre, H., Berthiaume, R., Raggion, G., Thivierge, M.C., Doepel, L., Pacheco, D., Dubreuil, P., Lobley, G.E., 2005. The route of absorbed nitrogen into milk
protein. Anim. Sci. 80, 11–22.
Liu, C., Schingoethe, D.J., Stegeman, G.A., 2000. Corn distillers grains versus a blend of protein supplements with or without ruminally protected amino
acids for lactating cows. J. Dairy Sci. 83, 2075–2084.
Moorby, J.M., Dewhurst, R.J., Evans, R.T., Danelon, J.L., 2006. Effects of dairy cow diet forage proportion on duodenal nutrient supply and urinary purine
derivative excretion. J. Dairy Sci. 89, 3552–3562.
Mulrooney, C.N., Schingoethe, D.J., Kalscheur, K.F., Hippen, A.R., 2009. Canola meal replacing distillers grains with solubles for lactating dairy cows. J. Dairy
Sci. 92, 5669–5676.
NFTA, 2006. Moisture Task Force Report, 2.2.2.5 Laboratory Dry Matter by Oven Drying for 3 hours at 105 ◦C., pp. 1–3, http://www.foragetesting.org/
files/NFTAReferenceMethodDM-09-18-06.pdf (accessed 1109.14).
National Research Council, 2001. Nutrient Requirements of Dairy Cattle, 7th rev. ed. National Academy Press, Washington, DC, USA.
Osorio, J.S., Ji, P., Drackley, J.K., Luchini, D., Loor, J.J., 2013. Supplemental Smartamine M or Metasmart during the transition period benefits postpartal cow
performance and blood neutrophil function. J. Dairy Sci. 96, 6248–6263.
Patton, R.A., 2010. Effect of rumen-protected Met on feed intake, milk production, true milk protein concentration, and true milk protein yield, and the
factors that influence these effects: a meta-analysis. J. Dairy Sci. 93, 2105–2118.
Polan, C.E., Cumming, K.A., Sniffen, C.J., Muscato, T.V., Vicini, J.L., Crooker, B.A., Clark, J.H., Johnson, D.G., Otterby, D.E., Guillaume, B., Muller, L.D., Varga, G.A.,
Murray, R.A., Pierce-Sandner, S.B., 1991. Responses of dairy cows to supplemental rumen-protected forms of Met and lysine. J. Dairy Sci. 74, 2997–3013.
Reynal, S.M., Broderick, G.A., 2005. Effect of dietary level of rumen-degraded protein on production and nitrogen metabolism in lactating dairy cows. J.
Dairy Sci. 88, 4045–4064.
Robinson, P.H., 2010. Impacts of manipulating ration metabolizable lysine and Met levels on the performance of lactating dairy cows: a systematic review
of the literature. Livest. Sci. 127, 115–126.
Robinson, P.H., Meyer, D., 2010. Total Mixed Ration (TMR) Sampling Protocol. ANR Publication 8413. ANR Publications, University of California, Berkeley,
CA, USA.
Robinson, P.H., Khorasani, G.R., Kennelly, J.J., 1994. Forestomach and whole tract digestion in lactating dairy cows fed increasing levels of canola meal
treated with acetic acid. J. Dairy Sci. 77, 552–559.
Robinson, P.H., Fadel, J.G., Ivan, M., 1996. Critical evaluation of diaminopimelic acid and ribonucleic acid as markers to estimate rumen pools and duodenal
flows of bacterial and protozoal nitrogen. Can. J. Anim. Sci. 76, 587–597.
Robinson, P.H., Chalupa, W., Sniffen, C.J., Julien, W.E., Sato, H., Watanabe, K., Fujieda, T., Suzuki, H., 1998a. Ruminally protected lysine or lysine and Met for
lactating dairy cows fed a ration designed to meet requirements for microbial and postruminal protein. J. Dairy Sci. 81, 1364–1373.
Robinson, P.H., Veira, D.M., Ivan, M., 1998b. Influence of supplemental protein quality on rumen fermentation, rumen microbial yield, forestomach digestion,
and intestinal amino acid flow in late lactation Holstein cows. Can. J. Anim. Sci. 78, 95–105.
Robinson, P.H., Chalupa, W., Sniffen, C.J., Julien, W.E., Sato, H., Fujieda, T., Ueda, T., Suzuki, H., 2000. Influence of abomasal infusion of high levels of lysine
or Met, or both, on ruminal fermentation, eating behaviour, and performance of lactating dairy cows. J. Anim. Sci. 78, 1067–1077.
Rulquin, R., Delaby, L., 1997. Effects of the energy balance of dairy cows on lactational responses to rumen-protected methionine. J. Dairy Sci. 80, 2513–2522.
Rulquin, H., Kowalczyk, J., 2003. Development of a method for measuring lysine and methionine bioavailability in rumen-protected products for cattle. J.
Anim. Feed Sci. 12, 465–474.
Rulquin, H., Pisulewski, P.M., 2006. Effects of graded levels of duodenal infusions of Leu on mammary uptake and output in lactating dairy cows. J. Dairy
Res. 73, 328–339.
Sakkers, M., Erasmus, L.J., Robinson, P.H., Meeske, R., Garrett, J.E., 2013. Determining in vivo ruminal stability of three ruminally protected nutrients in
lactating Jersey cows. Anim. Feed Sci. Technol. 185, 133–139.
SAS Institute, Inc., 2000. SAS/STAR® Software: Changes and Enhancements, Release 9.2. SAS Institute, Inc., Cary, NC, USA.
Schwab, C., 1995. Rumen protected amino acids. In: Tri-State Dairy Nutrition Conference, May, pp. 85–110.
Schwab, C., 2007. Protected proteins and amino acids for ruminants. In: Biotechnology in Animal Feeds and Animal Feeding. Wiley-VCH Verlag GmbH,
Weinheim, Germany, pp. 115–141.
Smith, D., 1969. Removing and Analyzing Total Nonstructural Carbohydrates From Plant Tissue. Wisconsin Agric. Exp. Sta. Res. Report 41.
Stensig, T., Robinson, P.H., 1997. Digestion and passage kinetics of forage fiber in dairy cows as affected by fiber-free concentrate in the diet. J. Dairy Sci. 80,
1339–1352.
Swanepoel, N., Robinson, P.H., Erasmus, L.J., 2010. Amino acid needs of lactating dairy cows: impact of feeding lysine in a ruminally protected form on
productivity of lactating dairy cows. Anim. Feed Sci. Technol. 157, 79–94.

22
N. Swanepoel et al. / Animal Feed Science and Technology 205 (2015) 10–22
Swanepoel, N., Robinson, P.R., Erasmus, L.J., 2014. Determining the optimal ratio of canola meal and high protein dried distillers grain protein in diets of
high producing Holstein dairy cows. Anim. Feed Sci. Technol. 189, 41–53.
Timmermans Jr., S.J., Johnson, L.M., Harrison, J.H., Davidson, D., 2000. Estimation of the flow of microbial nitrogen to the duodenum using milk uric acid
and allantoin. J. Dairy Sci. 83, 1286–1299.
Tyrrell, H.F., Reid, J.T., 1965. Prediction of the energy value of cow’s milk. J. Dairy Sci. 48, 1215–1223.
Vagnoni, D.B., Broderick, G.A., 1997. Effect of replacing alfalfa silage with high moisture corn on ruminal protein synthesis estimated from excretion of total
purine derivatives. J. Dairy Sci. 80, 1703–1712.
Valadares, R.F.D., Broderick, G.A., Valadares Filho, S.C., Clayton, M.K., 1999. Effect of replacing alfalfa silage with high moisture corn on ruminal protein
synthesis estimated from excretion of total purine derivatives. J. Dairy Sci. 82, 2686–2696.
Varvikko, T., Vanhatalo, A., Jalava, T., Huhtanen, P., 1999. Lactation and metabolic responses to graded abomasal doses of Met and lysine in cows fed grass
silage diets. J. Dairy Sci. 82, 2659–2673.
Wang, C., Liu, H.Y., Wang, Y.M., Yang, Z.Q., Liu, J.X., Wu, Y.M., Yan, T., Ye, H.W., 2010. Effects of dietary supplementation of Met and lysine on milk production
and nitrogen utilization in dairy cows. J. Dairy Sci. 93, 3661–3670.
Weeks, T.L., Luimes, P.H., Cant, J.P., 2006. Responses to amino acid imbalances and deficiencies in lactating dairy cows. J. Dairy Sci. 89, 2177–2187.
Williams, E.J., 1949. Experimental designs balanced for the estimation of residual effects of treatments. Aust. J. Sci. Res. Ser. A: Phys. Sci. 2, 149.
Wrinkle, S.R., Robinson, P.H., Garrett, J.E., 2012. Niacin delivery to the intestinal absorptive site impacts heat stress and productivity responses of high
producing dairy cows during hot conditions. Anim. Feed Sci. Technol. 175, 33–47.