Encapsulated nitrate replacing soybean meal in diets with and without monensin on in vitro ruminal fermentation

: This study assessed the association between encapsulated nitrate product (ENP) and monensin (MON) to mitigate enteric methane (CH 4 ) in vitro and possible effects on ruminal degradability, enteric fermentation characteristics, and microbial populations. Six treatments were used in randomized complete design in a 2×3 factorial arrangement with two levels of MON (0 and 2.08 mg/mL of buffered rumen fl uid) and three levels of ENP (0, 1.5 and 3.0%). The substrate consisted of 50% Tifton-85 hay and 50% concentrate mixture (ground corn and soybean meal). ENP replaced soybean meal to achieve isonitrogenous diets (15% CP). No ENP×MON intera ction was observed for any measured variable (P > 0.05) except for the relative abundance of F. succinogenes (P = 0.02) that linearly increased in diets with MON when ENP was added. The ENP addition decreased CH 4 production (P < 0.01) without affecting (P > 0.05) truly degraded organic matter nor the relative abundance of methanogens. Hydrogen production was reduced with MON (P = 0.04) and linearly decreased with ENP inclusion (P = 0.02). We concluded that use of nitrate is a viable strategy for CH 4 reduction, however, no additive effect of ENP and MON was observed for mitigating CH 4 production.


INTRODUCTION
Meth ane (CH 4 ) production in the rumen is an inherent part of the digestive process of ruminants (Beauchemin et al. 2008). Reduction of the CH 4 production can be achieved by use of feed additives that affect methanogenic microorganisms (Beacon 1988) or allow alternative hydrogen (H) sink, competing with CH 4 production (Ungerfeld & Kohn 2006). Ionophores, such as monensin (MON), decrease the concentration of Gram-positive bacteria and protozoa populations (Guan et al. 2006) and can reduce CH 4 production between 27 and 31% (Guan et al. 2006). MON promotes selection of succinate-producing bacteria, reduces the number of H 2 -producing bacteria and stimulates the production of propionate (Chen & Wolin 1979). On the other hand, nitrate (NO 3 -) has a higher affi nity for H 2 than CO 2 (Leng 2008). Thus, when NO 3 is present in the rumen, its reduction to nitrite (NO 2 -) and ammonia (NH 4 ) is favored over the production of CH 4 (Ungerfeld & Kohn 2006 , are toxic to microbes, altering the microbial population and lowering feed digestion (Zhou et al. 2011). Therefore, encapsulated slowrelease forms of NO 3 for ruminants seems to decrease the risk of toxicity (Lee et al. 2017). This occurs because slow release forms provide the possibility of gradual adaptation of microbes to NO 3 − and NO 2 − , improving the feed degradation, since NO 3 metabolism in the rumen can be improved when microbes are acclimatized to NO 3 − (Leng 2008).
Our hypothesis is that NO 3 can interact with MON manipulating rumen fermentation and reducing CH 4 production because of changes to ruminal microbiota. Besides that, the use of an encapsulated form of NO 3 may reduce the risk of toxicity by NO 3 and MON interaction. Thus, the aim of this study was to evaluate the in vitro interaction between MON and encapsulated NO 3 on CH 4 mitigation potential and ruminal microbiota.

Experimental design and treatments
A completely randomized design in a 2 × 3 factorial arrangement with two levels of monensin (MON: 0 and 2.08 mg/mL of buffered rumen fluid) and three levels of encapsulated nitrate product (0, 1.5 and 3.0% in dietary DM) was used. Encapsulated nitrate product (ENP) replaced soybean meal to achieve three isonitrogenous diets (15% Crude Protein, CP) formulated with 50% Tifton-85 hay (Cynodon spp) and 50% concentrate (corn and soybean meal) ( Table I). The experimental diets were selected and formulated according to crude protein (CP) requirements for growing and weight gain in lambs (NRC 2007). The forage concentrate ratio aimed at providing adequate substrate for microbial growth and ENP levels was selected according to previous studies from our research group, in which we found that even using an encapsulated form of nitrate, levels higher than 4.5% can cause toxicity and impair the microbial microorganisms (Natel et al. 2019). The diets were ground in a Willey mill (Marconi, Piracicaba, SP, Brazil) to pass through a 1 mm screen while the ENP was incubated in the encapsulated original formula, at doses corresponding to 0; 1.0% and 2.0% NO 3 in dietary DM (Table I).
For the treatments with MON inclusion, a stock solution of pure MON (M5273; Sigma-Aldrich Co., St. Louis, MO, USA; Molecular Weight 692.850) was prepared by diluting 15.6 mg in 1.0 mL absolute ethanol, stored at -10°C. Then, 10 µL of stock solution was added to each incubation glass flask 15 minutes before incubation, as described by Araujo et al. (2011). The final concentration of MON was 0. 156 mg/75mL of buffered rumen fluid (2.08mg/L). This dosage was chosen because it had previously been found to decrease gas and CH 4 production, increase propionate, and decrease acetate concentration with minimal effects on OM degradation (Araujo et al. 2009(Araujo et al. , 2011.

Incubation conditions and gas production
An in vitro gas production technique (Theodorou et al. 1994) adapted to a semi-automatic system (Maurício et al. 1999) with further modifications (Bueno et al. 2005, Longo et al. 2006) and using  (Table I) was weighted in #F57 ANKOM filter bags (ANKOM, Technology Corporation, Fairport, USA) (Soltan et al. 2017) and put into serum glass flasks (160 mL of total volume and 85 mL of head space) with 50 mL of incubation medium (Menke´s buffered medium) and 25 mL of inoculum. Two incubation flasks per inoculum per treatment served as analytical units and were sealed with 20 mm butyl septum stoppers (Bellco Glass Inc, Vineland, NJ, USA), manually mixed and incubated in a forced air oven at 39°C (Marconi MA35, Piracicaba, SP, Brazil) for 24 hours. In addition, for each inoculum, blank flasks (containing #F57 ANKOM filter bag without substrate, inoculum and medium) were included to correct the values of gas production and degradability, and a laboratory internal standard substrate (Tifton hay) was included to monitor incubation conditions (Soltan et al. 2017).
Head space gas pressure was measured at 2, 4, 8, 16, and 24-hour intervals after the start of incubation. Total volume of gas produced in each flask was determined following the equation V = (7.365 × P; n = 500; R 2 = 0.99) where: V = gas volume (mL) and P = measured pressure (psi) (Araujo et al. 2011). Total accumulated gas production (TGP) after 24-hour incubation was considered the sum of partial gas production at each time interval and deducting the values of gas production by blanks.
For CH 4 determination, 2 mL of gas were sampled and stored in 10 mL vacuum tubes after each gas measurement, resulting in a pool sample of each flask. A 5 mL-surgical syringe (Becton Dickson Indústria Cirúrgica LTDA, Curitiba, Paraná, Brazil) was used for gas sampling. After each gas sampling, flasks were vented, mixed, and returned to air oven. After 24 hours, flasks were placed in cold water (4°C) to cease fermentation and the #F57 ANKOM filter bags were removed. The CH 4 concentration in the collected gas was determined in the pool sample of each flask as described in Araujo et al. (2011)  Restek, Bellefonte, PA, USA). The temperatures of column, injector, and flame ionization detector were 60, 200, and 240 o C, respectively. Helium at 10 mL/min was the carrier gas. CH 4 concentration was determined by external calibration using an analytical curve (0, 30, 90, and 120 mL/L) prepared with pure CH 4 (White Martins PRAXAIR Gases Industriais Inc., Osasco, SP, Brasil; 99.5 mL/L purity). The production of CH 4 (CH 4 P) was calculated according to Longo et al. (2006) according to the following equation CH 4 P, mL = (Total gas, mL + Head space, 85 mL) x CH 4 concentration, mL/mL.

Ruminal degradability, fermentation characteristics, and microbial populations
At the end of the incubation period, the #F57 ANKOM filter bags removed from the flasks were treated with neutral detergent solution (NDS) for 1 hour at 90°, washed with hot water, acetone, and DM and ash were determined. The truly degraded organic matter (TDOM) was calculated as the difference between incubated organic matter (OM) and the remaining not degraded OM (Blümmel et al. 1997), and the same was performed with incubated and not degraded DM to determine the truly degraded dry matter (TDDM). Values of TGP and CH 4 P were expressed in basis of TDOM (mL/g TDOM) and TDDM (mL/g TDDM).
For quantifying the relative abundance of microbial microorganisms, the incubation liquid was collected and stored in frozen condition (-80°C) prior to DNA extraction. The DNA extraction from the buffered rumen fluid samples was performed using a commercial kit PowerLyzerTM PowerSoil (Mo Bio Laboratories, Inc., Carlsbad, CA, USA) and according to the manufacturer recommendations. The quantification of the relative abundance of methanogenic microorganisms as: Archaea, Selenomonas ruminatium and Wolinella succinogenes (nitrate-and nitrite-reducing bacteria), Ruminococcus flavefaciens and Fibrobacter succinogenes was performed using specific primers in real-time PCR (Table  II). The relative expression of each microbe was calculated as described by Denman & McSweeney (2006). The quantity of each microbe was expressed as a percentage relative to the total quantity of bacterial 16S rDNA in combined rumen fluid, according to Zhou et al. (2012).

Statistical analyses
Data statistical analyses were performed on SAS software version 9.4 (SAS Institute Inc.). The analytical units (two incubation flasks per inoculum per treatment) were averaged prior to the statistical analysis, and four inocula per treatment (n=4) were used as true statistical repetition. A mixed linear model using restricted maximum Likelihood (Restricted Maximum Likelihood: REML) in the MIXED procedure was used to analyze the response variables. The model included the fixed effect of MON, level of ENP, and the interaction between both effects (MON×ENP). Regression analyses (REG procedure) were performed considering ENP levels within diets. The least square means for MON, ENP level, and the interaction between them (MON×ENP) were obtained by LSMEANS procedure and when a significant fixed effect (P < 0.05) was identified, the respective means were compared by F and Tukey-Kramer tests.
The inclusion of MON in the diet did not affect the number of protozoa and the relative abundance of archaea Methanogens and ruminal bacteria (P > 0.05). The nitrate and nitritereducing bacteria showed a linear increase following the ENP addition: W. succionogenes (0.02, 0.34 and 2.03; y = 0.0194 + 0.642*ENP; P L = 0.01) and S. ruminantium (0.01, 0.04 and 0.06; y= 0.004+0.0196*ENP; P L = 0.01). No effect of ENP addition (P > 0.05) was observed on the number of protozoa or the relative expression of archaea Methanogens (Table V).

DISCUSSION
The NO 3 addition was effective to reduce methane production linearly, which is in agreement with previous studies in vitro (Anderson et al. 2008, 2010, Capelari et al. 2018, Natel et al. 2019, Zhang & Yang 2011 and in vivo (Brown et al. 2011, Klop et al. 2016, Lee et al. 2017, Newbold et al. 2014). There are two major mechanisms in which NO 3 reduces CH 4 production: 1) CH 4 is decreased by the competition for H 2 between NO 3 and methanogenesis, in a thermodynamically favorable process to methanogenesis (Lee et al. 2017) the toxicity of NO 3 and NO 2 on methanogenic microorganisms (Božic et al. 2009, Iwamoto et al. 2002, Sar et al. 2005, Zhou et al. 2011. In this study, the potential reduction of CH 4 was between 21.6% and 47.1% with the addition of 1.5 and 3% ENP (% MS) when compared to control (0% ENP), which agrees with Leng (2010) that showed a decrease of 16-50% with the use of NO 3 in ruminant diets. This result indicates that NO 3 reduction (consumption of H + ) was the major mechanism for lowering CH 4 production because of a reduced availability of H 2 to archaea methanogens. This hypothesis is confirmed by the linear reduction in the use of H 2 when ENP was added to the diet. In a stoichiometric approach: the complete reduction of NO 3 to NH 3 consumes 4 mol of H 2 , which is the same number of H 2 molecules necessary for methanogens reduce CO 2 to CH 4 (Capelari et al. 2018). Thus, when NO3 is present in the rumen, H2 is effectively used to reduce NO3 to NO2 and have this reduced to NH3, contributing to reduce CH4 production (because of the lack of H2) (Ungerfeld & Kohn 2006). The theoretical mitigation potential of NO 3 assumes that all NO 3 added is reduced to NH 3 (Li et al. 2013) in a way that 1 mol of NO 3 -(62 g) added in ruminant diets reduces 1 mol of CH 4 formation (22.4 L). In this study, doses of 5 and 10 mg of NO 3 in 500 mg of substrate were used, which theoretically should reduce CH 4 production around 3.36 and 6.44 mg/g TDOM, respectively. However, the linear CH 4 reduction Another explanation for the reduction of CH 4 production could be the direct NO 3 and NO 2 toxicity on the methanogens population (Božic et al. 2009, Sar et al. 2005. However, in this study, no reduction was found in the relative expression of methanogenic microorganisms nor in the number of protozoa with ENP addition, indicating that there was no direct effect of NO 3 on these populations. On the other hand, studies have shown that at least part of the CH4 that was reduced in the in vitro assays of Capelari et al.   of propionic acid with a reduction in acetic, butyric, and lactic acid, in CH4 and CO2 gases, and in ammonium (Bertipaglia 2008). In this experiment there was no influence of MON on the number of bacteria, except F. succinogenesis, nor in CH 4 production. It is possible that the amount of MON used was not able to act on the metabolism of gram-positive bacteria to reduce their number, which would have implied in an increase of gram-negative bacteria (such as F. succinogenesis). However, reductions in metabolic H 2 production were observed when MON was added as a consequence of a decrease in butyrate, with tendency to an increase in propionate acid production. Stoichiometrically pyruvate conversion to propionate requires a net input of H 2 per mol of fermented glucose, thereby reducing hydrogen supply (Janssen 2010) and the formation of acetate and butyrate release two moles of CO 2 and four moles of H 2 per mole of fermented glucose. (Kohn & Boston 2000). Thus, volatile fatty acid production rates determine ruminal hydrogen supply, which is used for methane production (Elliot & Loosli 1959).
The MON addition also affected isobutyrate and isovalerate acids production, without changing the total production of SCFA, acetateto-propionate ratio nor the CH 4 production. Since the inhibition of isoacids indicates attenuation of deamination, the reduction of isovaleric production following the addition of MON may be related to the reduction of ruminal deamination and the inhibition of NH 3producing bacteria (Russel & Strobel 1988).
In our study MON inclusion did not affect the nitrate-and nitrite-reducers (S. ruminantium and W. succinogenes). Chen & Wolin (1979) also observed no effect of MON on S. ruminantium population. However, higher dose of MON may affect gram-positive bacteria like D. detoxificans, major bacterial groups in the acquisition of tolerance by ruminants that are gradually adapted to nitrotoxins (Anderson & Rasmussen 1998, Majak 1992. According to Capelari et al. (2018) the combination of encapsulated NO 3 plus MON numerically increased the levels of NO 2 in rumen fluid after 24-hour incubation, suggesting a possible undesirable influence of MON on nitrate reduction. Thus, results should be interpreted with care. An interaction between ENP and MON was observed on the relative abundance of F. succinogenes, a gram-negative bacterium, suggesting that through different mechanisms, additives might change the rumen microbiota. Gram-negative bacteria have an outer membrane that prevents MON from reaching the cell membrane and is therefore more resistant to MON than gram-positive bacteria (Strobel & Russell 1989). Although MON did not significantly reduce gram-positive bacteria, it is possible that MON and ENP (more specifically NO 2 -) may have inhibited the general activity of gram-positive bacteria in the medium, thereby increasing gram-negative bacteria numbers, such as F. succinogenes.
The ENP inclusion did not reduce the number of protozoa nor inhibit archaea methanogens but increased the relative expression of nitrate and nitrite reducing bacteria such as S. ruminantium and W. succinogenes. Lin et al. (2011)

CONCLUSIONS
The CH 4 reduction by ENP addition reflected the effect of NO 3 acting as a H 2 sink. However, the reduction on CH 4 production was lower than expected. The additive effect of ENP and MON was not confirmed on reducing CH 4 nor affecting nitrate-and nitrite-reducing bacteria, but an increase on the relative abundance of gram-negative bacteria (F. succinogenes) was observed.