Symposium review: The importance of the ruminal epithelial barrier for a healthy and productive cow^*

The feeding of rapidly fermentable grain-rich diets to high-yielding dairy cows is necessary to minimize metabolic disturbances in early lactation () and to maximize milk production in a cost-effective manner over the whole lactation period (; ). However, such feeding practices may overwhelm the digestive physiology of cattle by impairing chewing and rumination activity, lowering the flow of alkaline saliva and, hence, rumen buffering (). Furthermore, they may cause major disturbances in the ruminal ecosystem that affect, for example, the survival and selection of gram-negative bacteria (GNB; , ). Greater rates of fermentation acid production drive the reduction in ruminal pH causing SARA (). The simultaneous decrease in chewing and rumination activity reduces the flow of saliva and the supply of salivary buffer entering the rumen, thus increasing the reliance on SCFA absorption and passage of acid out of the rumen as strategies to buffer ruminal digesta (; ). A direct implication of feeding highly fermentable diets is a high risk of ruminal () and systemic (; ) metabolic disorders. Of metabolic disorders, SARA is considered a major animal health and welfare issue in intensive ruminant production systems (). Indeed, SARA is associated with lowered feed efficiency and significant production losses (), likely explained by decreased fermentation efficiency in the rumen () but also by systemic inflammation and related systemic effects of SARA. Subacute ruminal acidosis is initiated by a chain of metabolic alterations in the rumen (; ; ) characterized by increased SCFA concentration, low pH, and ruminal hyperosmolarity (; ; ). These alterations further lead to the release of toxic metabolites (e.g., biogenic amines; ) and large amounts of LPS and other microbial associated molecular patterns (MAMP) in the ruminal fluid (; ). All of these metabolic alterations have the potential to damage the RE barrier to facilitate the translocation of proinflammatory signals (e.g., LPS and histamine; Figure 1), which is further analyzed in the following subsections.

Substantial evidence supports the idea that grain-rich diets and SARA are involved in disruption of the RE barrier. For example, the study by showed that desmosomal proteins were downregulated during a gradual adaptation to diets containing 65% grain. In another study, a proteome analysis conducted by revealed that 40 proteins of the RE associated with key functions such as cellular stress, metabolism, and differentiation were upregulated, whereas 20 other proteins were downregulated following a 2-d induction of SARA. After 6 wk of feeding the SARA diet, however, these authors observed that only 11 proteins of RE were upregulated and 3 were downregulated, suggesting an adaptational process of RE proteins to high-concentrate diets. It has also been demonstrated that a high plane of nutrition and the resulting acidotic stress result in depletion of key protective factors of the RE cells, rendering them more vulnerable to cellular damage and inflammation (). In another study, imposed a rapid dietary change from a hay-based diet to one that contained 50% concentrate. Although acidosis was not induced in this study, tissue conductance, as a measure of permeability, linearly increased as calves were fed the 50% concentrate diet for 3, 7, 14, and 21 d. They also observed linear increases in the serosal-to-mucosal flux of Na⁺, further supporting an increase in ion permeability in response to dietary change. The tendency for mannitol flux, a more direct measure for paracellular permeability, in the same calves () to increase further suggests that changes were attributable to an opening of the paracellular space. As such, some of the changes observed resemble changes commonly reported during acidosis, suggesting that increased provision of rumen-fermentable OM, especially starch, may modulate RE barrier function even in the absence of ruminal acidosis.

The mechanisms behind the effects of SARA on the RE barrier have been partially elucidated. From an acute standpoint, it is clear that acidification of the RE plays a crucial role and decreases barrier function as indicated by increased permeability to paracellular permeability markers such as mannitol (; ). Only recently has it become clear that the simultaneously elevated SCFA concentration in the rumen plays a significant role as well. Experiments by and demonstrated that increases in permeability of the RE are moderate when luminal pH is lowered to pH 5.1 for several hours. However, the same pH induced significant decreases in protein abundance and TJ localization of TJ proteins in the presence of SCFA, coupled with profound increases in RE permeability (Figure 2). This indicates that the effect of luminal acidification requires potentiation by the co-presence of SCFA to modulate barrier function effectively (; ). Evidence for the effect of ruminal acidosis to alter barrier function has been partially confirmed in vivo using lactulose as a permeability marker (). In another study, goats were fed a high-grain diet, with 45 and 20% of the diet coming from corn and wheat, respectively, for 7 wk (). In that study, the authors reported reduced ruminal pH, increased SCFA concentration, and detectable free plasma LPS in the ruminal acidosis group only and reductions in CLDN-4, occludin, and zonula occludens with grain feeding relative to the hay-fed control. They also noted increased proinflammatory cytokine abundance, which suggests that, in vivo, low pH and high SCFA concentration may act coordinately with proinflammatory signals. Furthermore, although the outcome arising from high-grain feeding was well described, it is not clear whether consequences were the result of high-grain feeding or the ruminal acidosis that was induced, presumably repeatedly, with the dietary management. Similar methodological concerns have been raised challenging the interpretation of many studies regarding ruminal acidosis using a variety of induction protocols (). As such, studies ex vivo or in vitro are essential for isolating the contribution of those multifaceted effects occurring in vivo (; ).

The contribution of hyperosmolarity to RE barrier integrity has also been studied and confirmed ex vivo by exposing isolated pieces of the RE to luminal hyperosmolarity. Luminal hyperosmolarity results from the accumulation of fermentation products in the rumen when their absorption cannot keep pace with their production and is a frequent occurrence during SARA (; ). Hyperosmolarity induces an immediate increase in paracellular marker fluxes across the RE ex vivo (; ; ). This change in RE barrier has been explained by cell shrinkage with widening of the paracellular spaces (), and, in contrast to acidosis-induced RE damage, it is usually reversible upon return to normotonic ruminal conditions (; ). However, one experiment in which hyperosmolarity was induced by ruminal hyperosmolar potassium chloride solution infusion demonstrated that hyperosmolarity, in context with other challenges, can also promote cell destruction as evidenced by phagocytotic inclusions of cell debris in RE dendritic cells ().

Of the MAMP generated in the rumen during SARA, cell-free endotoxin has attracted special interest (; ). Endotoxin or LPS is an abundant proinflammatory molecule of the outer leaflet of the cell wall of GNB. showed that SARA increased the LPS concentration in ruminal fluid and blood, which stimulated the nuclear factor (NF)-κB and mitogen-activated protein kinase inflammatory pathways and upregulated the expression and production of proinflammatory cytokines [i.e., tumor necrosis factor (TNF)-α, IL-6, and IL-1β] in RE. , using a metabolomic approach, demonstrated that feeding large amounts of concentrate (up to 60%) in the diet increased not only endotoxins in the ruminal fluid (up to 14-fold) but also the release of various other chemical compounds. This event was associated with greater severity of systemic inflammation in cows, thus indicating involvement of multiple diet–host–microbial factors in the disruption of the barrier functions of the host's epithelial lining. Although LPS has received the vast majority of the focus, other components such as flagellin, lipoteichoic acid, and microbial DNA also have the potential to induce a proinflammatory response, at least in other epithelia (). The current view is that with low ruminal pH, proliferation of potentially pathogenic strains of GNB with certain virulence genes (e.g., Escherichia coli) and presence of their products (most likely SCFA, cell-free LPS, biogenic amines, and ethanol) negatively affect RE cadherins, contributing to the increase in permeability in the reticulorumen ().

It should be noted, however, that not only the concentration of endotoxin in ruminal fluid but also its toxicity play a role in the induction of inflammation. For example, compared with other GNB found in the rumen of cattle such as Megasphaera elsdenii, Fibrobacter succinogenes, Prevotella spp., and Bacteroides spp., the endotoxin of E. coli is much more toxic (). Therefore, the pathophysiology of SARA and the disruption of the RE barrier, as well as the activation of systemic inflammation, are expected to be greater when LPS of E. coli predominates in the rumen due to their greater virulence potential compared with other GNB (). This hypothesis is supported by data of the study by . They suggested that the numbers of E. coli with virulence factor genes for adhesion and biofilm formation increase markedly in the rumen under the low ruminal pH conditions induced by a grain-rich diet, indicating a role for them to facilitate the disruption of the barrier function of the RE.

The release of biogenic amines, especially of histamine, in the ruminal fluid of acutely acidotic ruminants and their possible implications in local and systemic effects of ruminal acidosis is one of the very early findings of experimental acidosis research (; ; ). In addition, later studies confirmed a negative relationship between ruminal histamine concentration and ruminal pH for the moderate forms of acidosis (i.e., SARA; ). were the first to show that application of histamine in relevant dosages (10 and 100 µM) impairs differentiation of RE cells in culture, which may result in impaired RE function and barrier integrity. Recently, histamine has further been shown to activate the inflammatory pathway of cultured RE cells via NF-κB (), which perceivably has consequences for RE function and integrity.

It is well known that moderate decreases in ruminal pH, within physiological limits, stimulate the RE absorption of Na⁺ () and SCFA (; ). By contrast, , ) demonstrated that more severe decreases in luminal pH (to pH 4.8 in vivo and pH 5.5 ex vivo, respectively) lead to rapid decreases in the absorption of electrolytes (Na⁺, Cl⁻, or Mg²⁺) across the RE. Studies that are more recent evidenced the latter also for SCFA; low ruminal pH () and exposure to ruminal acidosis () reduced SCFA absorption across the RE, which parallels the decreases in barrier function (; ; ). It is important to note that the increase in RE permeability and reduced SCFA absorption occur simultaneously and represent paracellular and transcellular processes, respectively. Morphologically evident swelling of RE cells and their mitochondria () verifies a global interference with diverse cellular functions and cellular energy utilization. Teleologically, the reduction for SCFA absorption in response to ruminal acidosis can be seen as a way to limit further intracellular acidification and the arising negative consequences of excessive intracellular SCFA supply on barrier function (; ). Hence, an effective barrier is necessary for both the controlled exposure and response to MAMP and other proinflammatory signals but also for efficient nutrient transport.

We must emphasize that the relationship between transport function and SARA is essentially bilateral; that is, SARA decreases RE transport function, but low RE transport function predisposes animals to SARA. Regarding the latter, were the first to show that sheep with lower absorptive capacity for SCFA responded to an oral glucose drench (5 g/kg of BW) with rapid development of SARA, whereas companion sheep with higher absorptive capacity regulated their ruminal pH above the SARA threshold. Later, demonstrated that cows with genetically greater proneness to SARA (acidosis index of 61.7 vs. 13.5 pH × min/kg of DMI) exhibited lower gene expression of Na⁺/H⁺-exchanger 3 in RE, which was coupled with lower SCFA absorption capacity and higher concentrations of SCFA in the ruminal fluid. By contrast, no differences were observed in SCFA absorption rates and the expression of different genes related to SCFA absorption and intracellular pH regulation in the RE when the inherent acidosis index difference was narrow (). Interestingly, abundance of Na⁺/H⁺-exchanger 3 in RE increased during SARA induction protocol over 5 d, further suggesting that adaptation of transport function is a main mechanism for counteracting the development of SARA ().

The current knowledge about the effects of high-grain feeding on the RE barrier and transport functions as described in the previous subsections has been modeled for a moderately severe case of acidosis in Figure 3A. This model also includes predictions for the recovery of RE functions following exposure to ruminal acidosis. The recovery of RE functions bears a higher degree of uncertainty because far fewer studies have been dedicated to the functional recovery of the RE after an acidotic insult. In the pioneering study of , where the isolated and washed reticulorumen of sheep was challenged with a pH of 4.8 in the presence of 85 mM SCFA for 1 h, absorption of Na⁺ and Cl⁻ was reduced acutely by more than 80% and required 5 d to return to the preacidosis values. Of note, the initial reaction upon return from acidotic to normal pH is a further increase in RE permeability (; ), and absorption of Cl⁻ appeared to reach its nadir 1 h after ruminal pH was back to physiological values (). The enhancement of RE permeability immediately following the acidotic insult may be explained by cell necrosis and the return of the swollen RE cells to normal cell volume, which further opens the paracellular space jointly with the damaged TJ.

The milder forms of SARA are not commonly associated with persistent increases in permeability or decreases in absorptive functions (Figure 3B). When RE of sheep that responded to an oral glucose challenge with developing SARA (153 min at pH 5.8; nadir pH 5.36) were investigated immediately after the acidotic insult, no differences in acetate and butyrate uptake () and no difference in RE mannitol permeability () were noticed compared with water-drenched control sheep. In a recent study, imposed a ruminal acidosis challenge by restricting calves to 25% of their voluntary DMI followed by the provision of 30% of the diet as pelleted barley grain. This protocol reduced pH in the rumen but also reduced pH in the large intestine (). also reported that ruminal acidosis decreased the width of the ruminal papillae but that 1 d after the ruminal induction protocol, differences in permeability of the rumen, omasum, duodenum, jejunum, ileum, cecum, and proximal colon were not detectable. Interestingly, they reported that flux of mannitol and inulin tended to be reduced for calves exposed to acidosis relative to the control. The authors speculated that the lack of decreases in permeability shortly after the acidotic challenge may be attributed to increased expression for several of the TJ or TJ-associated proteins that were evaluated. Thus, it is conceivable that adaptive mechanisms target to restore barrier function as quickly as possible after an acidotic insult. It may further appear that restoration of barrier function has priority over the restoration of absorptive functions; however, the latter is yet to be proven by suitable experiments.

Some of the early studies in acutely acidotic models suggested a role of systemic LPS endotoxins () and inflammatory mediators (e.g., histamine; ; ) in the pathophysiology and clinical signs of acidosis. Although it was initially thought that endotoxins and histamine act directly on peripheral organs to induce, for example, laminitis, it was later suggested that the primary site of action of LPS is the liver, which plays a key role in the development of systemic inflammation (). Meanwhile, it appears that systemic inflammation is present not only in acute acidosis but also during SARA. As such, acute-phase proteins [LPS-binding protein (LBP), serum amyloid A, or haptoglobin] as surrogate markers of systemic inflammation have been suggested as possible diagnostic aids for SARA (; ). However, systemic inflammatory responses may be partly blunted and highly variable in SARA. With the expectedly lower levels of LPS absorption during SARA, it is likely that many cows develop rapid LPS tolerance (; ). On the other hand, it has been shown that some cattle have inherited hyperresponsiveness to repeated low-dose LPS infusion and may even respond with increasing TNF-α plasma concentrations after repeated LPS exposure (). Other circumstances that may make inflammatory responses during SARA variable are concurrent illnesses. In the latter context, it has been shown that SARA aggravates responses to an intramammary LPS infusion such as decreased DMI, chewing activity, milk yield, and alterations of milk constituents ().

Another recent addition to our knowledge is that other regions of the GI tract, especially the large intestine, can also be involved in the systemic inflammatory response and should not be overlooked (; ). It has meanwhile become clear that grain overfeeding in the rumen can lead to increased bypass starch provision and fermentation in the hindgut, where hindgut acidosis may develop with similar negative consequences for hindgut epithelial functions and LPS absorption, as discussed in the previous sections for the RE ().

The model depicted in Figure 1 suggests the cascade of events of inflammation during SARA. It proceeds from concurrent LPS release and the disruption of the RE barrier in response to feeding grain-rich diets. Once translocated into the portal circulation, LPS reaches the liver and is recognized by specific pattern recognition receptors of local immune cells, most importantly through the toll-like receptor 4 (TLR-4), which initiates powerful immune reactions (; ). Effective binding of LPS to TLR-4 is facilitated by the serum component LBP that complexes LPS and catalyzes its transfer to CD14 (membrane bound or soluble), which then associates with and activates TLR-4. Co-activation of myeloid differentiation-2 by the lipid A component of LPS potentiates the activation of macrophages (Kupffer cells in the liver; ; ). The host signaling molecules released by activated immune cells include proinflammatory mediators such as nitric oxide, prostaglandins, histamine, and cytokines as a hallmark of inflammation (). During acidosis, histamine may not only be released from immune cells but may also be absorbed from the rumen, considering that histamine may accumulate in the rumen at low pH (; ; ) and that low pH greatly promotes the ruminal absorption of intact histamine (; ).

The downstream events of the TLR-4/myeloid differentiation-2 receptor complex include the release of proinflammatory cytokines such as TNF-α, IL-1, and IL-6, which use several adaptor molecules and lead to the activation of the central inflammatory transcription factor, NF-κB (). The cytokines trigger the production of acute phase proteins (serum amyloid A, LBP, and haptoglobin), which occurs primarily in hepatocytes but also in other organs, including the gut (). The majority of LPS taken up by Kupffer cells is subsequently subjected to detoxification by the liver and biliary excretion into the gut (Figure 1).

The cytokines released during inflammation are able to alter many organ functions because almost all cell types express cytokine receptors (; ). This may explain a pathophysiological link and the coincidence of SARA with other major diseases of cattle, such as displaced abomasum, laminitis, secondary ketosis, and liver diseases (, ; ). Furthermore, endotoxemia and inflammation may promote metabolic disturbances such as insulin resistance and oxidative stress and may promote catabolism and self-perpetuating immune-refractory states (; ; ). These metabolic disturbances could increase the susceptibility to secondary infectious diseases of cattle (e.g., mastitis and metritis) and infertility (; ). Indeed, poor metabolic health status and its associated lower performance and longevity are among the greatest challenges to the dairy industry at present.

The facts and conclusions described in the previous subsections represent an extracted picture from myriad data. However, there is variability in study outcomes that deserves some reflection here. In contrast to the highly reproducible models of acute acidosis that were mainly studied about 50 yr ago, in which acidosis most reproducibly led to the death of animals if severe enough (; ; ), studies on SARA have not consistently been eligible for simple and reproducible conclusions. For several aspects of diverse study outcomes, we refer readers to the recent review of , who analyzed the current challenges and controversies of SARA diagnosis. Apart from insufficiencies in diagnostic tools, which primarily relate to difficulties in correct ruminal pH assessment, study controversies originate from the multifaceted nature of SARA. The latter is partly based on individual differences in the susceptibility to SARA due to differences in DMI, chewing behavior, and ruminal motility and on many environmental, feed, and management factors that affect changes in blood, urine, and milk of SARA cows (. Individual variation in absorptive capacity for SCFA () and the pH regulatory capacity of the RE () as well as different tolerance toward a low-level LPS challenge () may further contribute to diverse outcomes of SARA challenges in different experimental settings and feeding situations, as discussed in previous subsections. Other limitations relate to the experimental settings themselves as reviewed previously by . These limitations include inadequate replication, low intakes of the experimental diets, and individual feeding of cattle that limits the ability to extrapolate the data to cattle fed under industry standards. Also, different production and disease states of animals in production settings are often difficult to model in experimental studies ().

A core element in the definition of SARA is the gradual overfeeding with rumen-fermentable carbohydrates relative to the adaptation state of the animal and its ruminal microbiota (; ). For most practical feeding situations, it follows that the occurrence of SARA can be deduced from the ratio between rumen-fermentable starch and physically effective NDF (). However, such a generally valid concept has limitations in that excess intake of easily fermentable fiber such as alfalfa pellets may likewise induce SARA despite negligible provision of starch. The type of SARA induced by alfalfa pellet seems to be different from the type of SARA induced by grain because the former has been shown to exclude systemic inflammation (; ). Furthermore, the replacement effects of sugar for starch are also incompletely understood. With pulse-dosing of larger amounts of sugar (3 kg), especially sucrose, ruminal pH depression was more severe than with pulse-dosing of comparable amounts of starch (). By contrast, replacing starch with sucrose or lactose in TMR did not cause any depression in ruminal pH (; ) and was associated with positive production responses such as increased milk production () and increased milk fat yield (; ). attributed such positive effects of sugar-containing diets, at least in part, to adaptation events of the RE—namely, an increased Cl⁻-competitive absorption of acetate and propionate. This supports the general concept that the adaptation of the absorptive capacity for SCFA during adaptation to highly fermentable diets is a key event that decreases the susceptibility to SARA (; ). It has further been demonstrated that such adaptation includes the enhancement of lactate absorption (; ), which appears completely absent or negligible in roughage-fed ruminants (). Lactate is produced especially during sudden increases in easily fermentable carbohydrates with irregular feeding practices (; ; ). Such irregular feeding practices are characterized by repeated bouts of mild acute (and partly lactic) acidosis that is different from “true” SARA caused by gradual daily overload with easily fermentable carbohydrates. Although both conditions are often summarized under the term SARA, the important difference is that the severity of acidosis may increase with each bout of mild acute acidosis (; ), partly due to persisting negative effects on RE transport and barrier functions (Figure 3A). By contrast, milder daily episodes of “truly subacute” SARA mostly imply an immediate and complete recovery of RE function between SARA episodes (). Moreover, repeated true SARA often triggers adaptations of the RE and ruminal ecosystem that may lower the severity of acidosis with each bout. In a recent study by applying 4 repeated periods of 7-d high-grain feeding interspersed by 7-d periods of high-forage feeding, ruminal pH was close to the SARA threshold (pH <5.6 for >3 h/d; ) only during the first and second challenges but not during the third and fourth periods despite higher ruminal SCFA concentrations in the latter 2 periods (). This indicates that acid–base regulatory mechanisms of the RE (and possibly of saliva) adapted to effectively regulate ruminal pH despite a higher SCFA load. Nonetheless, over time, true SARA may also lead to persistent damage of the RE barrier due to focal inflammation induced by repeated bacterial or LPS entry into the RE because early inflammatory responses have been observed when ruminal pH is <5.6 for only >1 h (; Figure 3B).

A newly emerging explanation for the variability in SARA readouts in different grain feeding situations and grain challenge models is the acknowledgment of hindgut acidosis when feeding high levels of rumen-resistant starch. A review by recently analyzed the similarities, diversities, and communication between the forestomach and the lower gut in different challenge situations, including the transition to high-energy diets. Their suggestion was that disruption of barrier function in the lower gut may promote systemic inflammation more intensely than disruption of the RE barrier. However, a main conclusion could be that decades of rumen-focused SARA research have dragged our attention away from the lower gut that is an integral part of the animal's GI tract and may suffer from severe acidosis that is missed by ruminal pH sensors and not amenable to supplementation of dietary buffers.