Nitrogen metabolism of the intestine during digestion in a teleost fish, the plainfin midshipman (Porichthys notatus)

Accounting for the role of the gastro-intestinal tract in the ammonia and urea nitrogen dynamics of freshwater rainbow trout on long-term satiation feeding

The contribution of the gut to the ingestion, production, absorption and excretion of the extra ammonia and urea nitrogen (urea-N) associated with feeding ('exogenous' fraction) has received limited attention. Analysis of commercial pellet food revealed appreciable concentrations of ammonia and urea-N. Long-term satiation feeding increased whole-trout ammonia and urea-N excretion rates by 2.5-fold above fasting levels. Blood was sampled from the dorsal aorta, posterior, mid- and anterior sub-intestinal veins, as well as the hepatic portal vein in situ. Ammonia, urea-N and fluid flux rates were measured in vitro using novel gut sac preparations filled with native chyme. The sacs maintained the extreme physico-chemical conditions of the lumen seen in vivo. Overall, these results confirmed our hypothesis that the stomach, and anterior intestine and pyloric caecae regions play important roles in ammonia and urea-N production and/or absorption. There was a very high rate of urea-N production in the anterior intestine and pyloric caecae, whereas the posterior intestine dominated for ammonia synthesis. The stomach was the major site of ammonia absorption, and the anterior intestine and pyloric caecae region dominated for urea-N absorption. Model calculations indicated that over 50% of the exogenous ammonia and urea-N excretion associated with satiation feeding was produced in the anaerobic gut. This challenges standard metabolic theory used in fuel-use calculations. The novel gut sac preparations gained fluid during incubation, especially in the anterior intestine and pyloric caecae, owing to marked hyperosmolality in the chyme. Thus, satiation feeding with commercial pellets is beneficial to the water balance of freshwater trout.

Response of epithelial cell lines from the rainbow trout gut and gill to ammonia

Rainbow trout epithelial cell lines from the gill, RTgill-W1, and gut, RTgutGC, were exposed to NH4Cl at 18-21 °C in L15 (basal medium) with fetal bovine serum and were found to undergo cytoplasmic vacuolization and cell death, depending on NH4Cl concentration and exposure time. Vacuolization arose within 24 h of cultures being exposed to 10-100 mM NH4Cl, and vacuoles disappeared over 24 h after NH4Cl-exposed cultures were returned to just L15/FBS. RTgill-W1 appeared more sensitive to vacuolization, with one indicator being the maximal proportion of vacuolated cells in a culture, which approached 100% in 50 mM NH4Cl for 72 h. RTgill-W1 also were more sensitive to NH4Cl-induced cell killing. For 7-d exposures, the inhibitory concentrations (IC50s) for the 50% loss of cell viability as evaluated with Alamar Blue were 30 mM NH4Cl for RTgill-W1 and 80 mM for RTgutGC. In a wound-healing assay, RTgutGC cells in 0.1 and 1 mM NH4Cl were able to migrate and cover a 500-μm gap in 5 d, like the control, but in 50 mM NH4Cl healing was blocked. In 10 mM NH4Cl, repair was slowed but by 14 d the gap was covered with cells and many of these were vacuolated. Overall, the results provide a foundation for using these two cell lines to study the physiology and toxicology of ammonia in fish.

The physiological consequences of a very large natural meal in a voracious marine fish, the staghorn sculpin (Leptocottus armatus)

Little information exists on physiological consequences when wild fish eat natural food. Staghorn sculpins at 10-13°C voluntarily consumed 15.8% of their body mass in anchovies. Gastric clearance was slow with >60% of the meal retained in the stomach at 48 h, and was not complete until 84 h. At 14-24 h post-feeding, pH was depressed by 3 units and Cl- concentration was elevated 2-fold in gastric chyme, reflecting HCl secretion, while in all sections of the intestine, pH declined by 1 pH unit but Cl- concentration remained unchanged. PCO2 and total ammonia concentration were greatly elevated throughout the tract, whereas PNH3 and HCO3- concentration were depressed. Intestinal HCO3- secretion rates, measured in gut sacs in vitro, were also lower in fed fish. Whole-animal O2 consumption rate was elevated approximately 2-fold for 72 h post-feeding, reflecting 'specific dynamic action', whereas ammonia and urea-N excretion rates were elevated about 5-fold. Arterial blood exhibited a modest 'alkaline tide' for about 48 h, but there was negligible excretion of metabolic base to the external seawater. PaCO2 and PaO2 remained unchanged. Plasma total amino acid concentration and total lipid concentration were elevated about 1.5-fold for at least 48 h, whereas small increases in plasma total ammonia concentration, PNH3 and urea-N concentration were quickly attenuated. Plasma glucose concentration remained unchanged. We conclude that despite the very large meal, slow processing with high efficiency minimizes internal physiological disturbances. This differs greatly from the picture provided by previous studies on aquacultured species using synthetic diets and/or force-feeding. Questions remain about the role of the gastro-intestinal microbiome in nitrogen and acid-base metabolism.

An in vitro study of urea and ammonia production and transport by the intestinal tract of fed and fasted rainbow trout: responses to luminal glutamine and ammonia loading

Digestion of dietary protein in teleosts results in high ammonia levels within the intestinal chyme that may reach concentrations that are many-fold greater than blood plasma levels. We used in vitro gut sac preparations of the ammoniotelic rainbow trout (Oncorhynchus mykiss) to investigate the role of the intestine in producing and transporting ammonia and urea, with specific focus on feeding versus fasting, and on responses to loading of the lumen with 2 mmol L^-1 glutamine or 2 mmol L^-1 ammonia. Feeding increased not only ammonia production and both mucosal and serosal fluxes, but also increased urea production and serosal fluxes. Elevated urea production was accompanied by an increase in arginase activity but minimal CPS III activity, suggesting that urea may be produced by direct arginolysis. The ammonia production and serosal fluxes increased in fasted preparations with glutamine loading, indicating an ability of the intestinal tissue to deaminate glutamine and perhaps use it as an energy source. However, there was little evidence of urea production or transport resulting from the presence of glutamine. Furthermore, the intestinal tissues did not appear to convert surplus ammonia to urea as a detoxification mechanism, as urea production and serosal flux rates decreased in fed preparations, with minimal changes in fasted preparations. Nevertheless, there was indirect evidence of detoxification by another pathway, as ammonia production rate decreased with ammonia loading in fed preparations. Overall, our study suggests that intestinal tissues of rainbow trout have the ability to produce urea and detoxify ammonia, likely via arginolysis.

Impacts of low salinity exposure and antibiotic application on gut transport activity in the Pacific spiny dogfish, Squalus acanthias suckleyi

The role of the marine elasmobranch gastrointestinal tract in nitrogen-recycling and osmotic homeostasis has become increasingly apparent, with the gut microbial community likely playing a significant role converting urea, an important osmolyte in elasmobranchs, into ammonia. The Pacific spiny dogfish can experience and tolerate reduced environmental salinities, yet how this environmental challenge may affect the microbiome, and consequently nitrogen transport across the gut, is as of yet unknown. In the present study, excised gut sac preparations were made from dogfish acclimated to the following: full-strength seawater (C), low salinity for 7 days (LS), and after acute transfer of LS-acclimated fish to full-strength SW for 6 h (AT). Significantly reduced microbial derived urease activity was observed in the mucosal saline of gut sac preparations from the LS (by 81%) and AT (by 89%) treatments relative to the C treatment. Microbial derived cellulase activity from mucosal saline samples tended to follow similar patterns. To further ensure an effective decrease in the spiral valve microbial population, an antibiotic cocktail was applied to the mucosal saline used for in vitro measurements of ion, water, and nitrogen flux in these gut sac preparations. This caused a further 57-61% decrease in the mucosal saline urease activity of the C and LS treatments. Overall, we observed relatively little flux across the stomach for all measured parameters aside from water movement, which switched from a net efflux in control fish to a net influx in acutely transferred fish, indicative of drinking. While no significant differences were observed in terms of nitrogen flux (urea or ammonia), we tended to see the accumulation of ammonia in the spiral valve lumen and a switch from efflux to influx of urea in control versus acutely transferred fish. The increased ammonia production likely occurs as a result of heightened metabolism in a challenging environment, while the retention and acquisition of urea is suggestive of nitrogen scavenging under nitrogen-limiting conditions.

The gaseous gastrointestinal tract of a seawater teleost, the English sole (Parophrys vetulus)

There has been considerable recent progress in understanding the respiratory physiology of the gastrointestinal tract (GIT) in teleosts, but the respiratory conditions inside the GIT remain largely unknown, particularly the luminal PCO₂ and PO₂ levels. The GIT of seawater teleosts is of special interest due to its additional function of water absorption linked to HCO₃^- secretion, a process that may raise luminal PCO₂ levels. Direct measurements of GIT PCO₂ and PO₂ using micro-optodes in the English sole (Parophrys vetulus; anaesthetized, artificially ventilated, 10-12 °C) revealed extreme luminal gas levels. Luminal PCO₂ was 14-17 mmHg in the stomach and intestinal segments of fasted sole, considerably higher than arterial blood levels of 5 mmHg. Moreover, feeding, which raised intestinal HCO₃^- concentration, also raised luminal PCO₂ to 34-50 mmHg. All these values were higher than comparable measurements in freshwater teleosts, and also greater than environmental CO₂ levels of concern in aquaculture or global change scenarios. The PCO₂ values in subintestinal vein blood draining the GIT of fed fish (28 mmHg) suggested some degree of equilibration with high luminal PCO₂, whereas subintestinal vein PO₂ levels were relatively low (9 mmHg). All luminal sections of the GIT were virtually anoxic (PO₂ ≤ 0.3 mmHg), in both fasted and fed animals, a novel finding in teleosts.

The role of intestinal bacteria in the ammonia detoxification ability of teleost fish

Protein catabolism during digestion generates appreciable levels of ammonia in the gastrointestinal tract (GIT) lumen. Amelioration by the enterocyte, via enzymes such as glutamine synthetase (GS), glutamate dehydrogenase (GDH), and alanine and aspartate aminotransferases (ALT; AST), is found in teleost fish. Conservation of these enzymes across bacterial phyla suggests that the GIT microbiome could also contribute to ammonia detoxification by providing supplemental activity. Hence, the GIT microbiome, enzyme activities and ammonia detoxification were investigated in two fish occupying dissimilar niches: the carnivorous rainbow darter and the algivorous central stoneroller. There was a strong effect of fish species on the activity levels of GS, GDH, AST and ALT, as well as GIT lumen ammonia concentration, and bacterial composition of the GIT microbiome. Furthermore, removal of the intestinal bacteria impacted intestinal activities of GS and ALT in the herbivorous fish but not in the carnivore. The repeatability and robustness of this relationship was tested across field locations and years. Within an individual waterbody, there was no impact of sampling location on any of these factors. However, different waterbodies affected enzyme activities and luminal ammonia concentrations in both fish, while only the central stoneroller intestinal bacteria populations varied. Overall, a relationship between GIT bacteria, enzyme activity and ammonia detoxification was observed in herbivorous fish while the carnivorous fish displayed a correlation between enzyme activity and ammonia detoxification alone that was independent of the GIT microbiome. This could suggest that carnivorous fish are less dependent on non-host mechanisms for ammonia regulation in the GIT.

An in vitro analysis of intestinal ammonia transport in fasted and fed freshwater rainbow trout: roles of NKCC, K+ channels, and Na+, K+ ATPase

We examined mechanisms of ammonia handling in the anterior, mid, and posterior intestine of unfed and fed freshwater rainbow trout (Oncorhynchus mykiss), with a focus on the Na⁺:K⁺:2Cl^- co-transporter (NKCC), Na⁺:K ⁺-ATPase (NKA), and K⁺ channels. NKCC was localized by immunohistochemistry to the mucosal (apical) surface of enterocytes, and NKCC mRNA was upregulated after feeding in the anterior and posterior segments. NH₄⁺ was equally potent to K⁺ in supporting NKA activity in all intestinal sections. In vitro gut sac preparations were employed to examine mucosal ammonia flux rates (Jm_amm, disappearance from the mucosal saline), serosal ammonia flux rates (Js_amm, appearance in the serosal saline), and total tissue ammonia production rates (Jt_amm = Js_amm - Jm_amm). Bumetanide (10^-4 mol L^-1), a blocker of NKCC, inhibited Js_amm in most preparations, but this was largely due to reduction of Jt_amm; Jm_amm was significantly inhibited only in the anterior intestine of fed animals. Ouabain (10^-4 mol L^-1), a blocker of NKA, generally reduced both Jm_amm and Js_amm without effects on Jt_amm in most preparations, though the anterior intestine was resistant after feeding. Barium (10^-2 mol L^-1), a blocker of K⁺ channels, inhibited Jm_amm in most preparations, and Js_amm in some, without effects on Jt_amm. These pharmacological results, together with responses to manipulations of serosal and mucosal Na⁺ and K⁺ concentrations, suggest that NKCC is not as important in ammonia absorption as previously believed. NH₄⁺ appears to be taken up through barium-sensitive K⁺ channels on the mucosal surface. Mucosal NH₄⁺ uptake via both NKCC and K⁺ channels is energized by basolateral NKA, which plays an additional role in scavenging NH₄⁺ on the serosal surface to possibly minimize blood toxicity or enhance ion uptake and amino acid synthesis following feeding. Together with recent findings from other studies, we have provided an updated model to describe the current understanding of intestinal ammonia transport in teleost fish.

Zonation of Ca2+ transport and enzyme activity in the caeca of rainbow trout - a simple structure with complex functions

Trout caeca are vermiform structures projecting from the anterior intestine of the gastrointestinal tract. Despite their simple gross morphology, these appendages are physically distinct along the anterior-posterior axis, and ultrastructural evidence suggests zonation of function within the structures. Individual caeca from three sections (anterior, middle and posterior) were removed from the intestine of freshwater rainbow trout and investigated for ion transport and enzyme activity. Ca²⁺ absorption appeared as a combination of active and passive movement, with Michaelis-Menten kinetics observable under symmetrical conditions, and was inhibited by several pharmacological agents (ouabain, La³⁺ and a calmodulin antagonist). There was a decrease in ion transport function from adjacent to the intestine (proximal) to the distal tip of each caecum, along with decreasing transport from anterior to posterior for the proximal portion alone. Feeding increased the J _Max and K _M for Ca²⁺ absorption within all sections, whereas ion-poor water (IPW) exposure further increased the J _Max and K _M for Ca²⁺ transport in the anterior and middle sections. Increased Na⁺/K⁺-ATPase (NKA) and citrate synthase (CS) activity rates paralleled trends seen in Ca²⁺ transport. Feeding in freshwater and IPW exposure increased the glycolytic capacity of the caeca via increased pyruvate kinase (PK) and decreased lactate dehydrogenase (LDH) activity, while amino acid metabolism increased with IPW exposure through increased glutamate dehydrogenase (GDH) activity. Overall, feeding and IPW exposure each altered ionoregulation within the caeca of freshwater rainbow trout in a zone-specific pattern, with the anterior and proximal portions of the caeca being most affected. Increased carbohydrate and protein metabolism fueled the increased ATP demand of NKA through CS.

Nitrogen handling in the elasmobranch gut: a role for microbial urease

Ureotelic elasmobranchs require nitrogen for both protein growth and urea-based osmoregulation, and therefore are probably nitrogen-limited in nature. Mechanisms exist for retaining and/or scavenging nitrogen in the gills, kidney, rectal gland and gut, but as yet, the latter are not well characterized. Intestinal sac preparations of the Pacific spiny dogfish shark (Squalus acanthias suckleyi) incubated in vitro strongly reabsorbed urea from the lumen after feeding, but mucosal fluid ammonia concentrations increased with incubation time. Phloretin (0.25 mmol l^-1, which blocked urea reabsorption) greatly increased the rate of ammonia accumulation in the lumen. A sensitive [¹⁴C]urea-based assay was developed to examine the potential role of microbial urease in this ammonia production. Urease activity was detected in chyme/intestinal fluid and intestinal epithelial tissue of both fed and fasted sharks. Urease was not present in gall-bladder bile. Urease activities were highly variable among animals, but generally greater in chyme than in epithelia, and greater in fed than in fasted sharks. Comparable urease activities were found in chyme and epithelia of the Pacific spotted ratfish (Hydrolagus colliei), a ureotelic holocephalan, but were much lower in ammonotelic teleosts. Urease activity in dogfish chyme was inhibited by acetohydroxamic acid (1 mmol l^-1) and by boiling. Treatment of dogfish gut sac preparations with acetohydroxamic acid blocked ammonia production, changing net ammonia accumulation into net ammonia absorption. We propose that microbial urease plays an important role in nitrogen handling in the elasmobranch intestine, allowing some urea-N to be converted to ammonia, which is then reabsorbed for amino acid synthesis or reconversion to urea.

The interactive effect of digesting a meal and thermal acclimation on maximal enzyme activities in the gill, kidney, and intestine of goldfish (Carassius auratus)

Surrounding environmental temperatures affect many aspects of ectotherm physiology. Generally, organisms can compensate at one or more biological levels, or allow temperature to dictate processes such as enzyme activities through kinetic effects on reaction rates. As digestion also alters physiological processes such as enzyme activities, this study determined the interacting effect of thermal acclimation (8 and 20 °C) and digesting a single meal on maximal enzyme activities in three tissues of the goldfish (Carrassius auratus). Acclimation to elevated temperatures decreased branchial Na⁺, K⁺, ATPase (NKA) activity. In contrast, acclimation to elevated temperatures had no effect on citrate synthase (CS) or pyruvate kinase (PK) activity in any tissue, nor were renal NKA or glutamine synthetase (GS) activities impacted. Warm water-acclimation exaggerated the positive impact of digestion on intestinal and branchial NKA activities and intestinal GS activity only, but digestion had no effect in the kidney. CS and PK did not display intestinal zonation; however, there was a distinct increase towards the distal intestine in NKA and GS activities. Zonation of NKA was more prominent in warm-acclimated animals, while acclimation temperature did not affect intestinal heterogeneity of GS. Finally, the impact of tissue protein content on enzyme activity was discussed. We conclude that the intestine and gill of warm-acclimated goldfish exhibited an augmented capacity for increasing several enzyme activities in response to digestion while the kidney was unaffected by thermal acclimation or digesting a single meal. However, this amplified capacity was ameliorated by alterations in tissue protein content. Amplified increases in NKA activity may ultimately have implications for ATP demand in these tissues, while increased GS activity may beneficially increase ammonia-detoxifying capacity in the intestine.

A broader look at ammonia production, excretion, and transport in fish: a review of impacts of feeding and the environment

For nearly a century, researchers have studied ammonia production and excretion in teleost fish. Stemming from past investigations a body of knowledge now exists on various aspects including biochemical mechanisms of ammonia formation and specific routes and tissues used for ammonia transport, culminating in a current detailed theoretical model of branchial transport, including the molecular identities of the moieties involved. However, typical studies examining ammonia balance use routine laboratory conditions and fasted fish. While avoiding additional variables that influence nitrogen balance, these studies are arguably idealistic and do not reflect the natural variety of conditions that fish encounter. Further studies have revealed the impacts of extrinsic factors (e.g. salinity, pH, temperature) on ammonia handling in fasted fish whereas others have explored intrinsic factors, such as life history and developmental impacts. One routine challenge for ammonia balance that fish encounter is feeding and digestion. Fortunately, many new studies have revealed the impact of feeding and digestion on several aspects of ammonia balance; from production to excretion and to transport, and several have done so incorporating supplemental extrinsic and/or intrinsic factors. Together, these complex studies reveal realistic responses to multifactorial challenges encountered by animals in the wild and begin to provide a holistic view of ammonia balance in freshwater teleost fish.

Evolution of urea transporters in vertebrates: adaptation to urea's multiple roles and metabolic sources

In the two decades since the first cloning of the mammalian kidney urea transporter (UT-A), UT genes have been identified in a plethora of organisms, ranging from single-celled bacteria to metazoans. In this review, focusing mainly on vertebrates, we first reiterate the multiple catabolic and anabolic pathways that produce urea, then we reconstruct the phylogenetic history of UTs, and finally we examine the tissue distribution of UTs in selected vertebrate species. Our analysis reveals that from an ancestral UT, three homologues evolved in piscine lineages (UT-A, UT-C and UT-D), followed by a subsequent reduction to a single UT-A in lobe-finned fish and amphibians. A later internal tandem duplication of UT-A occurred in the amniote lineage (UT-A1), followed by a second tandem duplication in mammals to give rise to UT-B. While the expected UT expression is evident in excretory and osmoregulatory tissues in ureotelic taxa, UTs are also expressed ubiquitously in non-ureotelic taxa, and in tissues without a complete ornithine-urea cycle (OUC). We posit that non-OUC production of urea from arginine by arginase, an important pathway to generate ornithine for synthesis of molecules such as polyamines for highly proliferative tissues (e.g. testis, embryos), and neurotransmitters such as glutamate for neural tissues, is an important evolutionary driving force for the expression of UTs in these taxa and tissues.

Intestinal ammonia transport in freshwater and seawater acclimated rainbow trout (Oncorhynchus mykiss): evidence for a Na+ coupled uptake mechanism

In vitro gut sac experiments were performed on freshwater and 60% seawater acclimated trout (Oncorhynchus mykiss) under treatments designed to discern possible mechanisms of intestinal ammonia transport. Seawater acclimation increased ammonia flux rate into the serosal saline (Jsamm) in the anterior intestine, however it did not alter Jsamm in the mid- or posterior intestine suggesting similar mechanisms of ammonia handling in freshwater and seawater fish. Both fluid transport rate (FTR) and Jsamm were inhibited in response to basolateral ouabain treatment, suggesting a linkage of ammonia uptake to active transport, possibly coupled to fluid transport processes via solvent drag. Furthermore, decreases in FTR and Jsamm caused by low Na(+) treatment indicated a Na(+) linked transport mechanism. Mucosal bumetanide (10(-4) M) had no impact on FTR, yet decreased Jsamm in the anterior and mid-intestine, suggesting NH4(+) substitution for K(+) on an apical NKCC, and at least a partial uncoupling of ammonia transport from fluid transport. Additional treatments (amiloride, 5-(N-ethyl-N-isopropyl)amiloride (EIPA), phenamil, bafilomycin, 4',6-diamidino-2-phenylindole (DAPI), high sodium) intended to disrupt alternative routes of Na(+) uptake yielded no change in FTR or Jsamm, suggesting the absence of direct competition between Na(+) and ammonia for transport. Finally, [(14)C]methylamine permeability (PMA) measurements indicated the likely presence of an intestinal Rh-mediated ammonia transport system, as increasing NH4Cl (0, 1, 5 mmol l(-1)) concentrations reduced PMA, suggesting competition for transport through Rh proteins. Overall, the data presented in this paper provide some of the first insights into mechanisms of teleost intestinal ammonia transport.

An in vitro analysis of intestinal ammonia handling in fasted and fed freshwater rainbow trout (Oncorhynchus mykiss)

Ammonia transport and metabolism were investigated in the intestinal tract of freshwater rainbow trout which had been either fasted for 7 days, or fasted then fed a satiating meal of commercial trout pellets. In vivo, total ammonia concentrations (T amm) in the chyme were approximately 1 mmol L(-1) across the entire intestine at 24 h after the meal. Highest chyme pH and P NH3 values occurred in the posterior intestine. In vitro gut sac experiments examined ammonia handling with mucosal (Jmamm) and serosal (Js amm) fluxes under conditions of fasting and feeding, with either background (control ≤ 0.013 mmol L(-1)) or high luminal ammonia concentrations (HLA = 1 mmol L(-1)), the latter mimicking those seen in chyme in vivo. Feeding status (fasted or fed) appeared to influence ammonia handling by each individual section. The anterior intestine exhibited the greatest Jm amm and Js amm values under fasted control conditions, but these differences tended to disappear under typical post-feeding conditions when total endogenous ammonia production (Jt amm = Js amm - Jm amm, signs considered) was greatly elevated in all intestinal sections. Under fasted conditions, glutamate dehydrogenase (GDH) and glutaminase (GLN) activities were equal across all sections, but the ammonia-trapping enzyme glutamine synthetase (GS) exhibited highest activity in the posterior intestine, in contradiction to previous literature. Feeding clearly stimulated the total rate of endogenous ammonia production (Jt amm), even in the absence of a high luminal ammonia load. This was accompanied by an increase in GDH activity of the anterior intestine, which was also the site of the largest Jt amm. In all sections, during HLA exposure, either alone or in combination with feeding, there were much larger increases in endogenous Jt amm, most of which was effluxed to the serosal solution. This is interpreted as a response to avoid potential cytotoxicity due to overburdened detoxification mechanisms in the face of elevated mucosal ammonia. Thus T amm of the intestinal tissue remained relatively constant regardless of feeding status and exposure to HLA. Ammonia production by the gut may explain up to 18 % of whole-body ammonia excretion in vivo under fasting conditions, and 47 % after feeding, of which more than half originates from endogenous production rather than from absorption from the lumen.