Subscriber access provided by Kaohsiung Medical University
Article
Effects of condensed and hydrolysable tannins on rumen metabolism with emphasis on the biohydrogenation of unsaturated fatty acids. Mónica Costa, Susana Alves, Alice Cappucci, Shaun R Cook, Ana Duarte, Rui Caldeira, Tim A. McAllister, and Rui J.B. Bessa J. Agric. Food Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jafc.7b04770 • Publication Date (Web): 01 Mar 2018 Downloaded from http://pubs.acs.org on March 3, 2018
Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of Agricultural and Food Chemistry is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.
Page 1 of 35
Journal of Agricultural and Food Chemistry
1
Effects of condensed and hydrolysable tannins on rumen metabolism with emphasis on
2
the biohydrogenation of unsaturated fatty acids.
3
4
Mónica Costa†, Susana P. Alves†, Alice Cappucci‡, Shaun R. Cook§, Ana Duarte†, Rui
5
M. Caldeira†, Tim A. McAllister§, Rui J.B. Bessa†*
6 7 8
†
9
Medicina Veterinária, Universidade de Lisboa; Avenida da Universidade Técnica 1300-
CIISA, Centro de Investigação Interdisciplinar em Sanidade Animal, Faculdade de
10
477, Lisboa, Portugal
11
‡
12
del Borghetto, 80 – 56124 Pisa, Italy
13
§
14
4B1, Canada
Dipartimento di Scienze Agrarie, Alimentari e Agro-ambientali, University of Pisa, Via
Agriculture and Agri-Food Canada, Lethbridge Research Centre, Lethbridge, AB T1J
15 16
*Corresponding author:
17
Rui J. B. Bessa
18
Faculdade de Medicina Veterinária, Universidade de Lisboa,
19
Avenida da Universidade Técnica, 1300-477, Lisboa, Portugal
20
Tel.: +351 213652876
21
Fax: +351 213652889
22
e-mail:
[email protected] 23
1 ACS Paragon Plus Environment
Journal of Agricultural and Food Chemistry
24
Abstract
25
The hypothesis that condensed tannins have higher inhibitory effect on ruminal
26
biohydrogenation than hydrolysable tannins was tested. Condensed tannin extract from
27
mimosa (CT) and hydrolysable tannin extract from chestnut (HT) or their mixture
28
(MIX) were incorporated (10%) into oil supplemented diets and fed to rumen fistulated
29
sheep. Fatty acid and dimethyl acetal composition of rumen contents and bacterial
30
biomass were determined. Selected rumen bacteria were analysed by quantitative real
31
time PCR. Lower (P < 0.05) rumen volatile fatty acids concentrations were observed
32
with CT compared to HT. Moreover, lower concentration (P < 0.05) of Fibrobacter
33
succinogenes, Ruminococcus flavefaciens, Ruminococcus albus and Butyrivibrio
34
proteoclasticus were observed with CT compared to HT. The extension of
35
biohydrogenation of 18:2n-6 and 18:3n-3 did not differ among treatments, but was
36
much more variable with CT and MIX than with HT. The trans-/cis-18:1 ratio in
37
bacterial biomass was higher (P < 0.05) with HT than CT. Thus, mimosa condensed
38
tannins had a higher inhibitory effect on ruminal metabolism and biohydrogenation than
39
chestnut hydrolysable tannins.
40
41
Keywords:
42
biohydrogenation; fatty acids; rumen; hydrolysable tannins; condensed tannins.
43
44
2 ACS Paragon Plus Environment
Page 2 of 35
Page 3 of 35
Journal of Agricultural and Food Chemistry
45
Introduction
46
The inclusion of tannins in ruminant diets has been reported to modulate
47
biohydrogenation of unsaturated fatty acids (FA) in the rumen.1 In general, the effects
48
of tannins on rumen microbiota have been attributed to the ability of these phenolic
49
compounds to form complexes with polymers of protein and carbohydrates and to
50
directly interact with bacterial cell membranes.2 However, the effects of tannins are
51
dependent on several factors including their molecular structure. Tannins can be broadly
52
classified into hydrolysable and condensed tannins. Hydrolysable tannins (mainly
53
gallotannins and ellagitannins) have a polyol as a central core, which is esterified with a
54
phenolic group, while condensed tannins (proanthocyanidins) are composed of flavan-
55
3-ol (epi)catechin and (epi)gallocatechin units that are connected by interflavonoid
56
linkages.3
57
Hydrolysable tannins are readily susceptible to depolymerisation in the rumen, whereas
58
the depolymerisation of condensed tannins by rumen bacteria has not been clearly
59
demonstrated.4 Depolymerisation can convert the oligomeric hydrolysable tannins into
60
low molecular weight monomeric subunits that exhibit a lower affinity for extracellular
61
or intracellular proteins, such as enzymes, or for cellular lipoproteins within the
62
membrane lipid bilayers of bacteria.5
63
Conversely, condensed tannins inhibit the activity and growth of microorganisms by
64
decreasing membrane permeability as a result of binding to lipoproteins, or even
65
disruption of membranes when added at high concentrations.2
66
However, it is possible that the monomeric subunits derived from hydrolysable tannins
67
also exert toxicity towards bacteria causing disturbances in membrane fluidity6, 7 which
68
are lessened through bacteria developing adaptative responses. In the presence of toxins
69
that affect membrane integrity, some bacteria are able to reduce membrane fluidity, 3 ACS Paragon Plus Environment
Journal of Agricultural and Food Chemistry
70
enriching their membrane lipids with trans-FA.8, 9 The generation of trans-FA and their
71
incorporation into microbial cells has been suggested to be one the main purposes of
72
ruminal bacterial biohydrogenation.10 In fact, biohydrogenation simultaneously reduces
73
the concentration of polyunsaturated FA in the rumen, lowering their direct toxicity to
74
bacterial membranes.11 In fact, increasing the level of trans-18:1 in the rumen could be
75
a mechanism whereby bacteria improve their ability to cope with the toxicity of
76
polyunsaturated FA.10 As the toxic effects of condensed tannins and hydrolysable
77
tannins on bacteria are probably mediated by distinct mechanisms of action, it is
78
possible that their effects on biohydrogenation also differ. In the present study, we
79
hypothesized that condensed tannins will exert a stronger inhibitory action on rumen
80
microbiota than hydrolysable tannins, leading to a corresponding general inhibition of
81
BH with large accumulation of C18 polyunsaturated FA. Conversely, we anticipated
82
that hydrolysable tannins would modulate, instead of inhibit biohydrogenation, leading
83
to a larger accumulation of trans-18:1 compared to condensed tannins.
84
Thus, a condensed tannin extract from mimosa (Acacia mearnsii) and a hydrolysable
85
tannin extract from chestnut (Castanea sativa) or their mixture were incorporated in an
86
oil-supplemented basal diet and fed to rumen fistulated sheep, to assess their effects on
87
FA composition of rumen digesta and bacteria, as well on the abundance of selected
88
rumen bacteria
89 90
Material and methods
91
Animal handling and management
92
Animal handling followed EU Directive 2010/63/EU concerning animal care and rumen
93
fistulated animals were used after approval of the ethical committee of the Faculty of
94
Veterinary Medicine, University of Lisbon. Four rumen fistulated ≈ 2-year old sheep 4 ACS Paragon Plus Environment
Page 4 of 35
Page 5 of 35
Journal of Agricultural and Food Chemistry
95
(52±3.0 kg) were used in a change-over design with 3 treatments and 4 experimental
96
periods. The sheep were kept in individual crates and bedded with wood shavings. The
97
dietary treatments were defined by the inclusion of tannin extract in common basal feed
98
ingredients: 1) Condensed tanins extract (CT) (100 g/kg dry matter (DM) of mimosa
99
tannin extract); 2) Hydrolysable tannins extract (HT) (100 g/kg DM of chestnut tannin
100
extract); 3) Mixture of CT and HT (MIX) (50 g/kg DM of each mimosa and chestnut
101
tannin extracts). Tannin extracts were incorporated in the diets at a high amount to
102
exacerbate their disturbing effects on rumen microbial ecosystem and metabolism. The
103
tannin extracts were commercial extracts containing 683 to 723 g/kg of condensed
104
tannins from Acacia mearnsii (Mimosa extract ME powder®; Mimosa Extract
105
Company (Pty) Ltd, Pietermaritzburg, South Africa) and ≥ 650 g/kg of hydrolysable
106
ellagitannins from Castanea sativa (Gallo Tanin B®; Lamothe – Abiet, Canejan,
107
France). Compound feeds containing the tannin extracts were pelleted and fed ad
108
libitum, and were provided twice daily (09:00 and 17:00) along with grass hay at 100
109
g/kg of DM intake. Grass hay contained 35 g/kg DM of crude protein and 764 g/kg DM
110
of neutral detergent fiber. The ingredients and chemical and fatty acid composition of
111
the pelleted supplement are presented in Table 1. Each experimental period consisted of
112
2 weeks of adaptation to the diets and 1 week of sample collection.
113 114
Sample collection and processing
115
Feed chemical composition was determined using the methods described by Oliveira, et
116
al. 12 Rumen sample collection occurred in the last week over 2 days with an interval of
117
3 days between them. On the first sampling day, rumen contents were collected directly
118
through the rumen cannula, from various regions in the rumen, before the morning meal
119
and used for isolation of bacterial biomass and FA analysis. On the second sampling
5 ACS Paragon Plus Environment
Journal of Agricultural and Food Chemistry
120
day, rumen contents sampling was repeated using the same procedure. On both
121
sampling days, rumen contents were also preserved for DNA extraction.
122
Planktonic bacteria (PB) and solid associated bacteria (SAB) pellets were obtained by
123
fractionation and differential centrifugation of rumen contents using a Beckman
124
Coulter, Avanti J-26 XPITM (Indianapolis, IN, USA) centrifuge, according to the
125
procedure reported by Bessa, et al.
126
included the use of four layers of surgical gauze to filter the rumen contents, as well as
127
the re-suspension and homogenization of rumen washed particles in saline solution with
128
carboximetilcellulose at 0.1%, which were then incubated at 39ºC for 15 min. At the
129
end, PB and SAB pellets were washed at least three times with saline solution until it
130
became clear.
131
The aliquots of total rumen contents used for FA analysis and DNA extraction, as well
132
the bacterial pellets, were immediately stored at -20ºC, freeze-dried (ScanVac CoolSafe,
133
LaboGene ApS, Lynge, Denmark) and preserved at -20ºC until analyzed. The aliquots
134
of total rumen contents for evaluation of fermentation activity were immediately filtered
135
through four layers of cheesecloth and then used either for pH measurement (Digital pH
136
meter “pH-2005”; JP Selecta S.A, Barcelona, Spain) or stored at -20ºC for volatile fatty
137
acids (VFA) analysis.
13
, with slight modifications. These modifications
138 139
Fatty acid and dimethyl acetals analysis
140
Volatile FA were analysed by gas chromatography with flame ionization detection (GC-
141
FID) using a Shimadzu GC-2010 Plus chromatograph (Shimadzu, Kyoto, Japan)
142
equipped with a Nukol capillary silica column (30 m; 0.25 mm i.d.; 0.25 µm film
143
thickness, Supelco Inc., Bellefonte, PA, USA), as previously described.12 For long chain
144
FA analysis of rumen content and bacterial pellets, 250 mg of each freeze-dried sample
6 ACS Paragon Plus Environment
Page 6 of 35
Page 7 of 35
Journal of Agricultural and Food Chemistry
145
was weighed and transesterified into FA methyl esters using a combination of basic
146
followed by acidic catalysis and with 19:0 (1 mg/mL) as the internal standard.14
147
Fatty acid methyl esters were separated by GC-FID using a Shimadzu GC-2010 Plus
148
chromatograph equipped with a SP-2560 capillary column (100 m, 0.25 mm i.d., 0.20
149
µm film thickness; Supelco Inc., Bellefont, PA, USA). Identification of FA methyl
150
esters and dimethyl acetals (DMA) was achieved by comparison of retention times with
151
those of authentic standards (FAME mix 37 components from Supelco Inc., Bellefont,
152
PA, USA, and a Bacterial FAME mix from Matreya LLC, Pleasant Gap, PA, USA).
153
Separation and identification was confirmed using gas chromatography-mass
154
spectrometry (GC-MS) using a Shimadzu GC-MS QP 2010 Plus chromatograph
155
(Kyoto, Japan) equipped with a SP-2560 column.15
156 157
Quantification of microbial 16S rRNA marker genes by qPCR
158
Total DNA was extracted from freeze-dried rumen contents (30 mg) using a Qiagen
159
QIAamp DNA Stool Mini Kit (Qiagen Inc., Valencia, CA, USA), according to
160
manufacturer’s instructions and with small modifications to enhance lysis of Gram-
161
positive bacteria.16 Supernatant (400 µl) was added to 30 µl of proteinase K (Qiagen
162
Inc., Valencia, CA, USA), 400 µl of AL buffer (Lysis buffer; Qiagen Inc., Valencia,
163
CA, USA) and 400 µl of ethanol. The eluted DNA was quantified using Quant-iTTM
164
PicoGreen® dsDNA Assay Kit (Invitrogen Canada Inc., Burlington, ON, Canada) with a
165
NanoDrop 3300 fluorometer (ThermoScientific, Wilmington, DE, USA).
166
The extracted DNA was subjected to real-time quantitative polymerase chain reaction
167
(qPCR), using a StepOnePlus thermocycler PCR (Applied Biosystems, Foster City, CA,
168
USA), for quantification of copies of 16S rRNA sequences specific for Fibrobacter
169
succinogenes subsp. S85, Prevotella bryantii B14, Ruminococcus albus 7,
7 ACS Paragon Plus Environment
Journal of Agricultural and Food Chemistry
170
Ruminococcus flavefaciens C94, Selenomonas ruminantium subsp. Lactilytica
171
(ATCC19205), Ruminobacter amylophilus (ATCC 29744), Butyrivibrio fibrisolvens
172
JW-11 and Butyrivibrio proteoclasticus PI-18. The PCR standards for the Butyrivibrio
173
group were created with genomic DNA from reference bacterial strains originated from
174
the Rowett Institute of Nutrition and Health, University of Aberdeen (UK) and cultured
175
in Abel Salazar Biomedical Science Institute (ICBAS) (Portugal). The standards for the
176
other bacteria were generated with plasmid DNA containing DNA amplicons derived
177
from strains of the Lethbridge Research Centre culture collection (Agriculture and Agri-
178
Food Canada). The genomic DNA of B. fibrisolvens and B. proteoclasticus was
179
extracted from 50 mg of freeze-dried bacterial pellets of pure cultures (ICBAS, Culture
180
Collection) using a Qiagen DNeasy Blood & Tissue Kit (Qiagen Inc., Valencia, CA,
181
USA), according to manufacturer’s instructions. The genomic DNA of other reference
182
strains was also extracted from bacterial pellets of pure cultures (Lethbridge Research
183
Centre, Culture Collection) followed by the preparation of plasmid standards, as
184
described by Wang, et al. 17. Detailed information about primers and cycling conditions
185
for F. succinogenes, P. bryantii, R. flavefaciens, S. ruminantium, R. amylophilus18 and
186
R. albus have already been reported.17,
187
used at concentrations of 2.0×108 to 2.0×106 and 1.0×106 to 1.0×102 copies per reaction
188
obtained by serial 10-fold dilutions. For genomic DNA standards, the number of copies
189
considered was covered in a 10-fold dilution series from 3.72×107 to 3.72×101 for B.
190
fibrisolvens and from 1.90×107 to 1.90×101 for B. proteoclasticus. For Butyrivibrio spp.,
191
the following conditions were applied: 95 ºC for 10 min; 40 cycles of 95 ºC for 15 s,
192
annealing temperature of 58 ºC for 1 min (B. fibrisolvens) or 55 ºC for 30 s with a final
193
extension step at 72 ºC for 30 s (B. proteoclasticus). For these two bacteria, the real-
194
time PCR reaction mixtures (15 µl) contained 1 × Taqman universal PCR master mix
19
For plasmid standards, the amplicons were
8 ACS Paragon Plus Environment
Page 8 of 35
Page 9 of 35
Journal of Agricultural and Food Chemistry
195
(Applied Biosystems®, Foster City, CA, USA), 10 µM of forward and reverse primers
196
for B. fibrisolvens and B. proteoclasticus at a final concentration of 100 and 300 nM,
197
respectively, 250 nM of both Taqman probes and 100 × purified BSA (BioLabs®
198
Ipswich, MA, USA). For the other bacteria, each qPCR reaction (25 µl) contained 1 ×
199
iQ SYBR Green Supermix (Bio-Rad Laboratories®, Hercules, CA, USA), 10 µM of
200
each primer at a final concentration of 300 nM and 100 × purified BSA (BioLabs®,
201
Ipswich, MA, USA). For all bacteria, BSA was used to overcome the presence of
202
tannins as PCR inhibitors. The primers and probes for Butyrivibrio spp. were designed
203
using Primer Express 2.0 software (Applied Biosystems®, Foster City, CA, USA) and
204
the homology of their sequences was searched against GenBank to confirm their
205
specificity. The sequences of the primers and probes were as follows: 5´-CCGCGTCA
206
GATTAGCCAGTT-3´
207
(reverse)
208
(probe) for B. proteoclasticus; 5´-ACACACCGCCCGTCACAC-3´ (forward), 5´-TCCT
209
TACGGTTGGGTCACAGA-3´
210
GGAGTTGGGAATGCCCGA-TAMRA-3´ (probe) for B. fibrisolvens. Primers were
211
synthesised by NZYTech (Lisbon, Portugal) and probes were synthesised by Stabvida
212
(Almada, Portugal). For all qPCR reactions, 2 µl of extracted DNA template was added
213
at 20 ng or 10 ng per reaction and amplified in duplicate.
and
(forward)
and
5´-GTAGGAGTTTGGGCCGTGTCT-3´
5´-FAM-CCAAAGCAACGATCTGTAGCCGGACTG-TAMRA-3´
(reverse)
and
5´-JOE-CATG
214 215
Calculations and statistical analysis
216
The ruminal biohydrogenation (% of disappearance) of dietary c9-18:1, 18:2 n-6 and
217
18:3 n-3 and the biohydrogenation completeness (18:0 concentration relative to
218
potential 18:0 concentration assuming a complete hydrogenation of substrates) were
219
calculated as previously reported.15 Membranes of rumen bacteria contain phospholipids
9 ACS Paragon Plus Environment
Journal of Agricultural and Food Chemistry
220
with alk-1-enyl (vinyl) ether chains (plasmalogens). The vinyl ether chains, quantified
221
as DMA, can be used as an internal marker of rumen bacteria.14, 20 Thus, the sum of the
222
DMA that were present at similar concentration in SAB and PB pellets (i.e. iso-14:0,
223
14:0, anteiso-15:0, 15:0 and 16:0 DMA) was used to estimate the rumen bacteria
224
biomass as follow: =
(0.5 × + 0.5 × )
225
Where, the Mrumen, MSAB and MPB are the concentration (mg/g DM) of the DMA marker
226
in the rumen, SAB and PB pellets, and biomass is expressed as mg bacterial DM/g
227
rumen DM.
228
Data were analysed using the MIXED procedure of SAS (SAS Institute Inc., Cary, NC,
229
USA) considering a change-over design with 3 diets, 4 sheep and 4 periods, according
230
to the follow model: = + + +
+ () + !
231
Where, yijk are the individual observations, u the overall mean, ti is the fixed effect of
232
treatment (i = 1, 2 and 3), aj is the random effect of animal (j = 1, 2, 3 and 4), pk the
233
random effect of period (k = 1, 2, 3 and 4), subsampling within experimental units (i.e
234
animal×period) (l = 1 and 2) and eijk the residual error. The variance homogeneity was
235
tested and, when significant, the variance heterogeneity structure was accommodated in
236
the model. The effects of treatments were evaluated using 2 orthogonal contrasts: C1)
237
contrasting the least square means of CT against HT, and C2) contrasting the average of
238
least square means of CT and HT against MIX (CT + HT)/2 vs. MIX). Statistically
239
significant differences or trends were set for P < 0.05 and for 0.05 < P 0.10) between CT and HT,
253
but the branched chain volatile FA (iso-butyrate and iso-valerate) presented higher (P
0.05) between HT and CT, a
267
higher (P = 0.023) concentration of 16:0 was found with HT than with CT. The
11 ACS Paragon Plus Environment
Journal of Agricultural and Food Chemistry
268
differences between CT and HT were more extensive for DMA. In fact, concentrations
269
of eight DMA (anteiso-15:0¸15:0, 16:0, 13:0, iso-14:0, 14:0, c9-18:1, c11-18:1) and the
270
sum of DMA (P < 0.022) were higher with HT than with CT. In general, for DMA,
271
MIX did not differ from the average of CT and HT, but concentrations of 14:0- and c11-
272
18:1-DMA were higher (P < 0.02) with MIX than the average of CT and HT.
273
The means of all C18 FA, except the 10-oxo-18:0, did not differ significantly between
274
CT and HT (Table 4). However, CT and MIX were consistently associated with greater
275
variances for the majority of C18 FA than HT, as supported by the larger standard error
276
of the means (Table 4). This higher variance arose as a result of occasional suppression
277
of biohydrogenation (i.e. animal 2, period 2, sampling day 1, treatment CT, animal 3,
278
period 1, sampling day 1, treatment MIX, animal 4, period 2, sampling day 1, treatment
279
MIX). In these cases, dietary unsaturated FA (i.e. c9-18:1, 18:2n-6, and 18:3n-3)
280
remained high, comprising more than 40% of total C18 FA in the rumen. Consequently,
281
the estimates of disappearance of unsaturated FA (i.e. biohydrogenation, %), as well as
282
the biohydrogenation completeness were numerical higher and much less variable with
283
HT than with CT or MIX. Despite such variance, the 10-oxo 18:0 was higher (P =
284
0.027) with CT than with HT.
285 286
Fatty acid and dimethyl acetals of rumen bacterial fractions
287
The total FA and C18 FA did not differ between SAB and PB populations nor among
288
treatments (Table 5). The FA and C18 FA content of bacterial pellets averaged 106 and
289
85 mg/g DM, respectively. The majority of FA in these fractions corresponded to C18
290
FA followed by 16:0, which also did not differ among treatments or bacterial
291
populations. Considering the FA composition in more detail, the sum of branched-chain
292
FA was higher (P = 0.019) in PB than in SAB, and this was due to the highest
12 ACS Paragon Plus Environment
Page 12 of 35
Page 13 of 35
Journal of Agricultural and Food Chemistry
293
concentrations of iso-13:0 and anteiso-15:0 in PB. The treatments did not affect the sum
294
of branched-chain FA but iso-14:0 and iso-16:0 were lower (P < 0.01) and iso-13:0
295
higher (P = 0.009) with CT than with HT.
296
Odd- and linear-chain FA were higher (P = 0.022) in PB than in SAB, mostly due to
297
15:0 and 17:0. The only odd- and linear-chain FA that differed between CT and HT was
298
the 23:0, which was higher (P = 0.003) with CT compared to HT. However, the
299
bacterial 15:0 and 17:0 contents were higher (P < 0.05) with MIX than the average of
300
contents of CT and HT.
301
Rumen bacteria presented a considerable amount of even chain FA with carbon chain
302
length ranging from 20 to 28. The concentrations of 22:0 and 24:0 were higher (P