Microbial protein synthesis

Preformed amino acids can stimulate efficiency of microbial growth (Hackmann and Firkins, 2015b), but the majority of microbial protein in ruminants fed standard protein diets still is from amino acids synthesized using assimilated ammonia, which is underestimated when plant N contaminates bacterial samples (Ahvenjarvi etal., 2018). Microbial protein synthesis has been reviewed in efforts to improve our ability to improve both accuracy and precision of predicting microbial protein flow to the duodenum (Firkins et al., 2007; Hartinger et al., 2018). However, assessing microbial protein synthesis in vivo still has considerable variation that awaits explanation (White et al„ 2016), so further improvements in accuracy and precision should improve mechanistic integration of RDP with ruminally available carbohydrate for beef (Galyean and Tedeschi, 2014) and dairy (Van Amburgh et al., 2015; White et al., 2017b) cattle. Ammonia concentration must be adequate to optimize microbial protein synthesis (Schwab and Broderick, 2017). However, ammonia concentration is hard to predict (Firkins etal., 2007) and clearly depends on adequacy for microbes (Ahvenjarvi and Huhtanen, 2018). Amino-N could be limiting the efficiency of microbial protein synthesis, particularly when there is high rumen availability of carbohydrate; increased availability of preformed AA could increase efficiency of growth enough to coincidentally increase assimilation of ammonia into AA (Firkins etal., 2007). Thus, there is a need to integrate N metabolism with ruminal carbohydrate availability. Assimilation of ammonia can increasingly rely on ATP as ammonia decreases (Section 2.2), and these reactions also might influence how unique ATP-yielding pathways can be used for advantage (Hackmann et al., 2017) compared with assimilation of central metabolites into cell constituents. Increasing the ratio of energy used for cell growth:maintenance functions needs further attention (Hackmann and Firkins, 2015b).

Microbial protein production typically is measured as a net value of synthesis and degradation. Protozoal outflow vs. retention and autolysis likely is increased with faster passage rate from the rumen (Firkins et al., 2007) but has received limited research attention and particularly with respect to sequestration by isotrichids (also called holotrichs) but not entodiniomorphs (Diaz et al., 2014) and biomass:cell (Firkins and Yu, 2015). Previous models assumed no outflow of protozoal protein, which can cause significant error in predicting intestinal supply of microbial amino acids, particularly lysine (Sok et al., 2017; Fessenden et al., 2017). Those authors also noted differences in amino acid profile of the fluid- and particulate-phase bacteria. Moreover, increased bacterial protein flow was associated with the latter bacteria passing with rumen-undegraded fiber (Sauvant and Noziere, 2016). Consequently, passage versus recycling is intricately linked with proteolysis and ammonia assimilation and will therefore be discussed subsequently.

Intra-ruminal recycling of microbial protein

Normal turnover of microbial protein includes synthesis through cell growth to replace those cells that pass to the duodenum plus those that recycle within the rumen. Cell lysis is much more energetically wasteful than recycling of ammonia (Firkins et al., 2007). Protozoa have long been considered the main culprits contributing to excess proteolysis and recycling of previously synthesized microbial protein (Hristov and Jouany, 2005). However, recycling is challenging to assess in vivo because of incomplete mixing of 1SNH4 doses. Labeling feed fractions with 15N fertilizer and an extensive sampling scheme allowed a model fit in dairy cattle (Ahvenjarvi et al., 2018). They noted that most of the protozoal protein was derived from bacterial protein. However, intra-ruminal N recycling was estimated at 22%, which is a value much lower than previous estimates primarily using sheep. Those authors suggested about 15% of the omasal microbial N was protozoal N, which agrees with other reports (Fessenden et al., 2019; Sok et al., 2017) and contrasts with models that assume no outflow (i.e. 100% of protozoal recycling via sequestration), yet they also noted about 40% selective retention of protozoa. As discussed previously, most of the selective retention would be expected to be from isotrichids, which might have been abundant in that study but were not enumerated.

Intra-ruminal recycling probably has been inflated by overestimating both 1) bacterial cells consumed through predation and 2) protozoal biomass, which also factors through an increased amount of bacteria ingested. Bacterial predation is overestimated when assessed with these cultures because protozoa typically were starved before being dosed only bacteria; omitting feed allows easier accounting for the counts of dosed bacteria, but it also would deprive normal substrate. As summarized by Williams and Coleman (1992), clearance of infused £. coli declined by 0, 0,18, and 68% when mixed entodinia were fed 10, 52,260, and 1670 rice grains per protozoan cell, respectively. Similar limitations persist when comparing faunated versus defaunated (with vs. without protozoa) animals; that is, the difference between these observations ignores interactions by protozoa with prokaryotes and fungi. Clearly, these types of observations are very helpful (Newbold et al., 2015), but care must be exercised in interpretation.

The original estimations of protozoa up to 50% of biomass might have stemmed from papers that overestimated protozoal volume using geometric formulas that did not fit their shape (Wenner et al., 2018), did not account for bacterial contamination (Sylvester et al., 2005; Firkins et al., 2007), had marker problems (Sok et al., 2017), and deterministically discussed sequestration of all protozoa in the rumen, whereas only the isotrichids are highly sequestered (Diaz et al., 2014). Based on cell counts, Karnati et al. (2007) determined that generation time of protozoa (primarily entodiniomorphids) in lactating dairy cattle was approximated by the ruminal retention time of particulate matter. An elaborate model based on greater cell counts in the rumen versus omasum suggests greater sequestration and intra-ruminal recycling (Hook etal., 2017) but also ignores that omasal counts can be diluted by passage of unmixed drinking water and autolysis of cells resulting from backflow of abomasal acid can occur between the time of animal euthanization and sample collection (Firkins and Yu, 2006). Protozoa might be up to 50% of the microbial biomass for sheep fed at low intakes, but this 50% value should not necessarily be extrapolated to production situations (Firkins et al., 2007). Hence, 25% of the biomass is likely a more appropriate value, at least for dairy cattle (Ahvenjarvi et al., 2018).

Protozoa do consume bacteria, fungal zoospores, and (in lower amounts) smaller protozoa (Hartinger et al., 2018) and digest them incompletely (Hristov and Jouany, 2005), which lessens efficiency of microbial protein synthesis. Because of an increasing expectation for fungal involvement in fiber degradation (Edwards et al„ 2017), the antagonism by protozoa against fungi is likely not beneficial to the host. Potential selective predation of certain bacteria, including fibrolytics(ParkandYu,2018b), contrasts with the general expectation for protozoa to improve fiber digestibility (Newbold et al., 2015). A meta-transcriptomics screening of ruminal contents supported the importance of protozoa and fungi in fiber digestibility (Comtet-Marre et al., 2017). This work suggests revisiting the standard assumption of the main benefit from protozoa being indirect (e.g. 02 consumption or rapid starch ingestion by protozoa). Perhaps, selective inhibition to limit the sequestering isotrichids (Newbold et al., 2015) would be beneficial, yet roadblocks in strategies such as vaccines still remain (Hartinger et al., 2018).

Approaches to selectively limit protozoa need to prevent oversimplification of results. For example, counts of protozoa were directly associated with the ratio of methane produced per unit of dry matter intake; however, meta-regression also documented that those counts also were inversely associated with fiber digestibility and dry matter intake (Guyader et al., 2014). Dry matter intake must be considered with methanogen-suppression approaches (Ungerfeld,

2018). Similarly, dietary protein (presumably through RDP) is positively associated with dry matter intake in dairy cattle (Zanton, 2016). Protozoa might outcompete bacteria (their protein sources) with decreasing dietary protein and more reliance on blood urea-N (Oelker et al., 2009). The advantage of urea-N diminishes with increasing concentration of dietary protein, particularly to dietary protein (and presumably RDP) concentrations needed to maintain milk production, as discussed in section 2.1.

Reliance on increasing blood urea-N to subsidize decreased dietary crude protein concentration is a questionable strategy for dairy cattle because of the potential depression of feed intake but also because typically the duodenal amino acid profile is more favorable for production compared with the original feed protein (Schwab and Broderick, 2017). Depressed microbial protein supply to the duodenum compared with its predicted supply either limits productivity or requires a safety factor for increased feeding of RUP; unfortunately, RUP is much more expensive than RDP, more variable in intestinal digestibility (even more so with respect to individual AA), and also contributes extensively to blood urea recycling to the gut and urinary N excretion (Batista et al., 2017). Those authors documented a strong potential for beef cattle to transfer blood urea-N transfer into bacterial protein.

Case study to advance understanding of protozoa-mediated proteolysis and intra-ruminal recycling of microbial protein

Williams and Coleman (1992) reported that single cultures of protozoa cleared bacteria 1.5- to 17.6-fold faster than did protozoa collected ex vivo. Thus, Belanche et al. (2012) set up an intricate ex vivo sampling protocol to quantitatively and mechanistically assess intra-ruminal recycling of microbial protein. Among the many conclusions outside of the current scope, the greater bacterial lysis in the protozoa recovered on the 20-pm sieve was from a combination of higher specific activity and higher abundance of biomass (Fig. 5). The authors noted that this fraction represents the more bacterivorous small entodinia such as Entodinium caudatum. These populations now can be assessed with an increasing repertoire of species-specific primers (Ishaq et al., 2017), which provide an alternative to traditional counting methods (Kittelmann et al., 2015). However, counts can be combined with microscopic measurements to estimate cell volume, with the assumption that volume is proportional to metabolic activity. Authors should consider real-time videography and the ellipsoid shape tool available in imaging software, which better capture width:depth ratios that vary among taxa but probably also by inhibitors against protozoa (Wenner et al., 2018).

Extending this approach, fractions could be split into a control that is starved during filtering and washing (as done in this study) to be compared to

Lysis of bacterial N

Figure 5 Lysis of bacterial N (determined by pre-enriching ruminal fluid-associated bacteria with ,4C) administered to protozoal fractions that were fractionated by successively smaller pore sizes of nylon mesh (Belanche et al., 2012). The activities are as reported (Actual; blue dashed line is the weighted mean of 12.0 ng bacterial N lysed per 100 g protozoal N/h) or were re-derived by the current authors by weighting the contribution of each fraction's recovered protozoal N as a percentage of the sum of all fractions (Normalized; red dashed line is mean of 16.7% = 100%/6 fractions).

the same protozoa also fed purified (N-free) cellulose, starch, and glucose (Ye et al., 2018). Would lysis of isotopically labeled bacteria be different if those labeled bacteria were dosed by themselves compared with being dosed at the same time as purified carbohydrates? Feeding substrate likely lessens predation of fluid-associated bacteria because of competition for intracellular space available for digestive vacuole formation (Diaz et al., 2014).Those authors used beads that mimic the size and cell wall charge of bacteria, which might be an alternative for researchers who cannot use radioisotopes to assess bacterial predation. In that report, increasing bead dosing rate increased bead uptake, but bead uptake was decreased when co-introduced with normal substrate. These findings also beg the question: does measured bacterial lysis increase with increasing dosage of bacteria? If so, what bacterial dosing rate represents in vivo conditions in which fluid-associated bacteria are in lower abundance and probably lesser functional importance than particulate-phase bacteria that are adherent to plant fragments that also are protozoal substrate?

Ex vivo approaches can advance our knowledge of bacterial lysis in the rumen of cattle differing in protozoal population abundance, community type, and activity per cell. Ciliate protozoa generally can be categorized into four community types based on generic distribution (Kittelmann et al., 2016) and are relatively static in community composition among cows and are more resilient after diet change compared with prokaryotes (Mizrahi and Jami, 2018). Even so, some diet changes can have remarkable shifts. For example, coconut oil profoundly inhibited most protozoa except genus Epidinium (Reveneau et al., 2012), whereas this greater tolerance has an unknown mechanism.

The predominant genus Entodinium varies considerably by cell size, niche, and bacterivorous activity (Williams and Coleman, 1992). Does the generic protozoal population structure (often estimated by generic counts) adequately reflect diet change?

Case study using sequencing approaches with protozoal in vitro or ex vivo cultures

Although single cultures of protozoa have to be qualified with respect to relevance to in vivo populations, they can provide very useful mechanistic information. Single cultures of Entodinium caudatum and Epidinium caudatum generally had recovered sequences of bacteria and archaea that were typical of the rumen but not necessarily in the same relative abundance compared with ruminal sequencing studies (Park and Yu, 2018a,b). Even so, relative sequence abundance of Proteobacteria was unusually high from bacteria recovered in close proximityto protozoa in single cultures; subsequent similar results fromex vivo cultures provide additional support for an important protozoal-associated role of this taxon. Although they could be preferred prey, Park and Yu (2018b) suggested that Proteobacteria might be able to resist proteolytic degradation by the protozoan and are likely endosymbionts. Do these endosymbionts contribute to 'protozoal' proteolysis or more general ecology in a significant and variable way?

Many questions could be refined with both ex vivo and in vitro approaches. Do ciliate protozoa help provide peptides and AA as substrates for HAB (increasing wasteful deamination) or, conversely, do protozoa selectively predate HAB (decreasing wasteful deamination by HAB) because those bacteria are more likely to be fluid associated than particulate associated (Firkins et al., 2007)? Current approaches inconsistently associate protozoal counts or 18S rRNA gene copies with feed efficiency (Delgado et al., 2019). However, those authors noted that ciliate protozoal interactions with bacteria and archaea are complex and probably only can be sorted out by combining transcriptomics with phylogenetic analyses. They also noted that gene databases lack protozoal representation, which should be rectified with emerging results from genomic sequencing (Park et al., 2018).

< Prev   CONTENTS   Source   Next >