Nucleotides (NT) are low molecular weight intracellular compounds, which play key roles in nearly all biochemical processes (Stryer, Reference Stryer1995). Nucleosides (NS) and NT are naturally present in all foods of animal and vegetable origin as free NS and NT, and as nucleic acids; soluble NS and NT, and RNA and DNA are present in milk from various mammals, contributing up to 20 % of the non-protein nitrogen (Gil & Uauy, Reference Gil and Uauy1989, Reference Gil, Uauy and Jensen1995; Leach et al. Reference Leach, Baxter, Molitor, Ramstack and Masor1995; Gil, Reference Gil and Liftschitz2001). Although NT deficiencies have not been related to any particular disease, dietary NT have been reported as beneficial for infants because they positively influence lipoprotein and PUFA metabolism, cellular and humoral immunity, and intestinal growth, development and repair (Carver & Walker, Reference Carver and Walker1995; Gil & Uauy, Reference Gil and Uauy1996; Cosgrove, Reference Cosgrove1998; Sánchez-Pozo et al. Reference Sánchez-Pozo, Rueda, Fontana and Gil1998; Gil, Reference Gil2002; Aggett et al. Reference Aggett, Leach, Rueda and MacLean2003; Buck et al. Reference Buck, Thomas, Winship, Cordle, Kuchan, Baggs, Schaller and Wheeler2004; Schaller et al. Reference Schaller, Kuchan, Thomas, Cordle, Winship, Buck, Baggs and Wheeler2004; Hawkes et al. Reference Hawkes, Gibson, Roberton and Makrides2006). However, there is a lack of information about the bioavailability of dietary NT and derivatives in infants.
Digestion of nucleic acids and NT clearly occur in man and animals. Most dietary NT are ingested in the form of nucleoproteins from which the nucleic acids are liberated in the intestinal tract by the action of proteolytic enzymes. Pancreatic ribonuclease and deoxyribonuclease degrade RNA and DNA into a mixture of mono-, di-, tri- and polynucleotides. In response to gastrointestinal hormones, intestinal Brunner's and Lieberkuhn's glands secrete phosphodiesterases, which supplement the action of pancreatic nucleases in producing mononucleotides (Gil & Uauy, Reference Gil and Uauy1996; Gil, Reference Gil and Liftschitz2001). Hydrolysis of NT to NS is carried out by intestinal alkaline phosphatase and ecto-5′-nucleotidase, which have a broad substrate selectivity (Anderson & Parkinson, Reference Anderson and Parkinson1997). The capacity for intestinal nucleic acid digestion and absorption varies between species, being higher in pigs compared with rabbits, dogs and cats (Barnard, Reference Barnard1969; Roth & Kirchgessner, Reference Roth and Kirchgessner1978; Greife, Reference Greife1984, Reference Greife1986). Using an intestinal homogenate from a 22-week-old human fetus, researchers qualitatively demonstrated that luminal and mucosal enzymes can digest RNA from human milk to cytidine, uridine and uric acid (Thorell et al. Reference Thorell, Sjoberg and Hernell1996). The efficiency of these enzymes to catabolize RNA and NT to absorbable products on a quantitative basis needs to be further investigated. A previous study from our group showed that cells from the intestinal epithelium are capable of hydrolysing RNA (Rueda et al. Reference Rueda, Gómez-León and Gil2002).
Because de novo synthesis of NT is an energy-intensive process, cells fulfil their NT requirements by NS salvage from both inside and outside the cell (Carver & Walker, Reference Carver and Walker1995; Griffith & Jarvis, Reference Griffith and Jarvis1996). Enterocytes have a limited capacity for de novo NT synthesis. Therefore, they rely heavily on NS transporters to salvage NS from their environment to meet their metabolic demands. To date, five Na+-dependent concentrative (N1, N2, N3, N4 and csg) transporters and two Na+-independent equilibrative (es and ei) transporters have been identified (Griffith & Jarvis, Reference Griffith and Jarvis1996). Only the purine-specific (N1; hCNT2) and the pyrimidine-specific (N2; hCNT1) Na+-dependent NS transporters have recently been found along the entire length of the fetal and adult small intestines (Ngo et al. Reference Ngo, Patil and Unadkat2001).
Evidence demonstrating the presence of concentrative and equilibrative transporters has come from a wide variety of experimental preparations from several species, including perfused tissues, isolated intact cells and membrane vesicles (Gil & Uauy, Reference Gil and Uauy1996; Griffith & Jarvis, Reference Griffith and Jarvis1996; Scharrer et al. Reference Scharrer, Rech and Grenacher2002). However, there is no information about the potential capacity of the neonatal intestine to absorb NS. Thus, the aim of the present study was to evaluate to what extent the suckling piglet intestine can hydrolyse RNA and NT to NS and absorb extracellular NS using cultivated intestinal explants.
Preparation of intestinal explants
The protocol for the present study was approved by the University of Granada (Spain) and piglets received humane treatment according to European Union regulations. A 20-d-old piglet supplied by the laboratory animal facility of the University of Granada was separated from the mother and fasted for 6 h. The animal was anaesthetized and killed by terminal bleeding of the jugular vein. The entire small intestine was removed quickly. A 60 cm-long segment of the small intestine from the ligament of Treitz was selected and considered as the proximal jejunum. The intestine segment was rinsed thoroughly with ice-cold saline solution, opened lengthwise and blotted dry. Explants (1 cm long) were obtained from the jejunum segment and put on Netwell inserts (Costar, Cambridge, MA, USA), one explant per insert.
Study of nucleoside utilization by intestinal explants
Intestinal explants were deposited on six-well cell culture plates (Costar), one insert per well. The explants were cultured in Williams' medium E (Sigma, St Louis, MO, USA) supplemented with 43·6 pmol/l (6 mU/ml) insulin (Sigma), 2 mmol/l glutamine (Sigma) and 100 IU penicillin, 0·1 mg streptomicin and 0·25 μg amphotericin B per ml (Sigma). The medium was not supplemented with fetal calf serum since in previous experiments it showed nuclease activity. The experimental design is shown in Fig. 1. RNA (60 mg/l) from calf liver (Sigma), a mixture of five NT (AMP, CMP, GMP, IMP, UMP; 7 mg/l each; Yamasa Shoyu Co., Choshi, Japan) and a mixture of NS (adenosine, cytidine, guanosine, inosine, uridine; 10 mg/l each; Sigma) were used as substrates, respectively. Four replicates of explants for each one of the substrates (RNA, NT and NS) and four replicates in the absence of any substrate were cultivated for 3 h. In the case of explants cultured in the absence of substrate, a negative control, which contained the medium without explant, was also cultured. When the medium was supplemented with RNA a positive control containing nuclease P1 (Sigma), and a negative control, without explant and nuclease, were cultured. When the medium was supplemented with the NT mixture we also cultured a positive control containing bacterial alkaline phosphatase (Sigma), and a negative control, without explant and enzyme. When the medium was supplemented with NS, a negative control, which contained the medium with NS but without any explant, was cultured.
Aliquots of the medium were taken at different times (0·5, 2, 5, 15, 30, 60, 180 min) and the nucleoside content was determined for explants cultured in the absence of substrate and for those cultured in the presence of the NS mixture. NS and NT contents were determined in the medium at those same periods when explants were cultured in the presence of RNA and the NT mixture. Nucleotide and nucleoside contents in the different culture supernatants were analysed by HPLC according to the method of Leach et al. (Reference Leach, Baxter, Molitor, Ramstack and Masor1995).
The four replicates of transplants for each one of the substrates (RNA, NT and NS) and the four replicates in the absence of any substrate were used for statistical analysis. A general linear model of variance for repeated measures was used to assess the effects of time on free NS and NT concentrations. When significant effects (P < 0·05) were found, post hoc comparisons of means were done using the adjusted Bonferroni t test. All statistical analyses were done with the Statistical Package for Social Sciences software (SPSS Inc., Chicago, IL, USA).
The total concentration of NS in the medium increased during the period of culture when explants were incubated in the absence of NS, indicating that the explants are liberating NS (Fig. 2). However, when explants were cultured in the presence of NS, the concentration of NS in the medium rapidly decreased, suggesting that NS are efficiently absorbed by the explants as well (Fig. 2).
The total concentration of NT decreased while the total concentration of NS increased, when explants were incubated in the presence of NT, suggesting the hydrolysis of NT to NS (Fig. 3). The total concentration of both NT and NS increased during the period of culture when explants were incubated in the presence of RNA, indicating that RNA was hydrolysed to NT and then to NS (Fig. 3).
Table 1 shows the concentration of individual NS when explants were cultured in the absence or presence of NS and Table 2 shows the concentration of individual NT and NS when explants were cultured in the presence of NT or RNA. Adenosine rapidly disappeared from the medium with a parallel increase in inosine but after 30 min of explant incubation inosine levels started to decrease. These changes were especially evident when explants were cultured in the presence of NS (Table 1). Guanosine and uridine also disappeared from the medium at similar rates. On the contrary, cytidine increased over time both in the absence or presence of any of the assayed substrates (Tables 1 and 2). Uric acid was not detected or it was at a very low concentration in the culture media (results not shown).
A, adenosine; C, cytidine; G, guanosine; I, inosine; U, uridine.
a,b,c,d,e Mean values with unlike superscript letters were significantly different (P < 0·05; a < b < c < d < e).
* Explants were cultured in Williams' medium E supplemented with 43·6 pmol/l insulin, 2 mmol/l glutamine and 100 IU penicillin, 0·1 mg streptomycin and 0·25 μg amphotericin B per ml. A mixture of five NS (A, C, G, I, U; 10 mg/l each) were used as substrates with four replicates of explants for each one of the substrates.
A, adenosine; C, cytidine; G, guanosine; I, inosine; U, uridine.
a,b,c,d,e Mean values with unlike superscript letters were significantly different (P < 0·05; a < b < c < d < e).
* Explants were cultured in Williams' medium E supplemented with 43·6 pmol/l insulin, 2 mmol/l glutamine and 100 IU penicillin, 0·1 mg streptomycin and 0·25 μg amphotericin B per ml. RNA (60 mg/l) or a mixture of five NT (AMP, CMP, GMP, IMP and UMP; 7 mg/l each) were used as substrates with four replicates of explants for each one of the substrates.
† Free NT measured.
‡ Free NS measured.
When RNA was used as substrate (Table 2) most NT concentrations remained constant during the period of study; only CMP increased during this period. However, when explants were cultured with NT (Table 2), the concentration of individual NT rapidly decreased while that of NS increased. Figure 4 shows the percentage of total NT disappearing in the course of the experiment, which appear as NS. After 180 min of explant incubation, about 12·1 % remained as NT, 64·4 % appeared as NS and a net percentage of 23·5 % of NT was taken up. Likewise, the percentages of UMP, IMP, GMP and AMP disappearing, and appearing as uridine, inosine, guanosine and adenosine, were 71·4, 95·1, 36·4 and 2·1, respectively, while cytidine increased 164·1 % from CMP.
The results found in the present study demonstrate that the intestine of suckling piglets is able to hydrolyse both RNA and free NT to NS in the absence of luminal content, and that uridine, inosine, guanosine and adenosine are efficiently taken up. Only cytidine is not absorbed at similar rates.
Since the piglet intestinal explants were washed before culture, it appears the hydrolysis of RNA can be achieved in the absence of pancreatic ribonuclease and that enterocyte phosphodiesterase should be the main enzyme responsible for the hydrolysis. The present result confirms previous findings reported by our group (Rueda et al. Reference Rueda, Gómez-León and Gil2002). In addition, Kuchan et al. (Reference Kuchan, Ostrom, Smith and Hu2000) studied the influence of purine intake on uric acid excretion in infants fed on soya infant formulas suggesting that healthy infants can digest RNA and subsequently absorb the ribonucleotides liberated.
Phosphodiesterase I has been found in the intestinal mucosa of animals and man (Hawley et al. Reference Hawley, Crip and Hodes1983; Morley et al. Reference Morley, Hawley, Ulbright, Byutler, Culp and Hodes1987). Thorell et al. (Reference Thorell, Sjoberg and Hernell1996) showed that RNA from human milk was hydrolysed by intestinal homogenates from a 22-week gestation human fetus. However, they could not ascertain to what extent the hydrolysis was due to luminal ribonucleases (pancreatic ribonuclease and biliary phosphodiesterase) or to endogenously secreted enterocyte phosphodieserases, because the homogenates were contaminated with the luminal content. Moreover, they could not exclude that part of the hydrolysis of RNA might be due to the presence of ribonucleases, 5′-nucleotidase and alkaline phosphatase from the mother's milk. These and other enzymes involved in NT metabolism have been identified in human milk. RNase activity has been associated with lactoferrin and RNase II and 5′- nucleotidase have been found in human milk (Hamosh, Reference Hamosh and Jensen1995). Since both alkaline phosphatase and ecto-5′-nucleotidase are present in the enterocyte membrane, we hypothesize that the hydrolysis of free NT, liberated from RNA or added to the culture medium, is mediated by both enzymes in the intestine of suckling piglets. However, in contrast to other studies related to NS transport, which used luminal vesicles, in the experimental set-up employed in the present study (explants put on Newel inserts with the mucosa facing the solution containing the substrate) we cannot exclude that both surfaces of the jejunum (mucosal and serosal) were exposed to substrates and this may complicate the interpretation of results.
The concentrations of RNA and NT used in the present study are within the range of those found in human and sow's milk. RNA in human milk has been reported to range from 100 to 5600 mg RNA/l, reflecting differences attributed to both stages of lactation and socioeconomic groups (Gil & Uauy, Reference Gil, Uauy and Jensen1995). More recently, total potentially available NS, including free NS, monophosphate and diphosphate NT, NT adducts of purine and pyrimidine, and RNA have also been reported in human milk with values being highest at earlier stages of lactation. The mean of the total potentially available NS ranged from 49 mg/l in colostrum to 87 mg/l in mature milk, with a grand mean of 67 mg/l (Leach et al. Reference Leach, Baxter, Molitor, Ramstack and Masor1995). Cytidine and uridine NT comprise the first and second largest free-nucleotide fraction. CMP, CDP-choline, UMP, and UDP adduct of glucose, galactose, N-acetyl-glucosamine and N-acetyl-galactosamine remain high throughout lactation in human milk (Gil & Uauy, Reference Gil, Uauy and Jensen1995). On the other hand, sow's milk contains about 450 mg/l of free NT, uridine derivatives being the predominant NT forms (Gil & Uauy, Reference Gil, Uauy and Jensen1995).
The hydrolysis of RNA renders a mixture of free NT and NS and that of NT a mixture of NS. The hydrolysis of both substrates was almost completed after 180 min of incubation, taking into account that liberated NS are efficiently absorbed. Although there is a small liberation of NS to the medium by intestinal explants, the rapid disappearance of exogenous NS contributes to a net absorption of these compounds by piglet gut.
Adenosine rapidly disappeared from the medium with a parallel increase in inosine. Thus adenosine is likely converted to inosine by the action of adenosine deaminase, which is known to be present in the intestinal mucosa (Witte et al. Reference Witte, Wiginton, Hutton and Aronow1991). Inosine levels decreased after 30 min of explant incubation. Guanosine and uridine also disappeared from the medium at similar rates. On the contrary, cytidine levels increased over time.
The true RNA digestibility in pigs has been reported to be about 99 % when an oral dose of 3 % RNA is used (Roth & Kirchgessner, Reference Roth and Kirchgessner1978) and large oral loads of purine and pyrimidine NT are readily absorbed by the small intestine in pigs (Greife & Molnar, Reference Greife and Molnar1984a, Reference Greife and Molnarb). The rapid digestion of nucleic acids and NT by the pig intestine has been associated with higher concentrations of pancreatic ribonuclease compared with other species (Barnard, Reference Barnard1969), and the rapid absorption appears to be due to the presence of transmembrane Na+-dependent NS contransporters with overlapping substrate specificity, one for purine NS (N1) and one for pyrimidine NS (N2) (Scharrer et al. Reference Scharrer, Rech and Grenacher2002). In contrast to pigs, in rats a considerable amount of purine derivatives seems to reach the hind gut (Greife, Reference Greife1986) and NS, particularly purine NS and bases, are intensively metabolized in the enterocytes (Bronk & Hastewell Reference Bronk and Hastewell1989a, Reference Bronk and Hastewellb; Stow & Bronk, Reference Stow and Bronk1993). In contrast to previous reports in rats, in the present study we were not able to detect uric acid in the culture media, thus suggesting that NS derived from purine NT hydrolysis are efficiently taken up by pig enterocytes and incorporated into the NT cell pools. Nevertheless, we cannot exclude that at least in part purine NT are catabolized to uric acid and later converted to allantoin by uric acid oxidase which is active in pigs (Wu et al. Reference Wu, Lee, Muzny and Caskey1989).
The human adult intestine expresses two Na+-dependent NS transporters, N1 and N2, and two Na+-independent transporters, es and ei, which exhibit differential sensitivity to S-(p-nitrobenzyl)-6-thioinosine (nitrobenzylmercapto-purine riboside) (Chandrasena et al. Reference Chandrasena, Giltay, Patil, Bakken and Unadkat1997). These N1 (hCNT2) and N2 (hCNT1) selectively transport purine and pyrimidine NS, respectively, and are found on the apical surface of adult jejunal enterocytes. The Na+-independent transporters, es and ei, are not present in fetus or adult enterocytes (Ngo et al. Reference Ngo, Patil and Unadkat2001). The apparent K m values for hCNT1-mediated transport of uridine and cytidine are 59 and 140 μm, respectively (Graham et al. Reference Graham, Leithoff, Coe, Mowles, Mackey, Young and Cass2000). This suggests that uridine is preferentially taken up by intestinal cells. The bioavailability of uridine is particularly crucial to the formation of pyrimidine nucleotide–lipid conjugates used to synthesize RNA and biomembranes (Connolly & Duley, Reference Connolly and Duley1999). NS transporters in pigs appear to be similar to those of man, the K m for the N1 (guanosine) and N2 (thymidine) transport being 58 and 91 μm, respectively (Scharrer et al. Reference Scharrer, Rech and Grenacher2002). However, pig NS transporters have higher K m and V max than man and other monogastic species (Griffith & Jarvis, Reference Griffith and Jarvis1996), which may be related to the high capacity of nucleic acid digestion observed in pigs (Barnard, Reference Barnard1969; Roth & Kirchgessner, Reference Roth and Kirchgessner1978; Greife, Reference Greife1984). In man, the K m of the N1 (hCNT2) transporter for guanosine ranges from 10 to 12 μm, which explains the high affinity for this NS (Graham et al. Reference Graham, Leithoff, Coe, Mowles, Mackey, Young and Cass2000), whereas that of pigs is about 5-fold (Scharrer et al. Reference Scharrer, Rech and Grenacher2002). We suggest that the suckling piglet intestine will express the N1 transporters as in adult pigs and man, and will take up exogenous guanosine, explaining the rapid disappearance of this NS from the medium. This is conceivable since uric acid was at a very low concentration 3 h after the intestinal explant incubation.
Under the conditions of the present study cytidine seems to be poorly absorbed by the piglet intestine in the presence of uridine. There is an efflux of cytidine from the intestinal explant to the culture medium, which might be due to intracellular catabolization of cytidine derivatives formed from absorbed uridine. It is well known that uridine is salvaged by uridine kinase to form UMP and later UDP and UTP which is used in part to form CTP and CDP derivatives (Sutle et al. Reference Sutle, Becroft, Webster, Scriver, Beaudet, Sly and Valle1995).
We cannot exclude the presence of equilibrative nucleoside transporters 1 and 2 (es, ENT1 and ei, ENT2) in piglet intestine. Although both transporters are broadly selective in man, ENT2 is a generally low-affinity NS transporter with 2·6-, 2·8-, 7·7- and 19·3-fold lower affinity than ENT1 for thymidine, adenosine, cytidine and guanosine, respectively. In contrast, the affinity of ENT2 for inosine is 4-fold higher than ENT1, which suggests that ENT2 might be important in transporting inosine (Ward et al. Reference Ward, Sherall, Mo and Tse2000).
In conclusion, the intestine of suckling piglets is able to hydrolyse RNA and free NT to NS, which are efficiently absorbed by the enterocytes, except cytidine.
The authors thank María Luisa Jiménez for her technical assistance and Cristina Cazorla for her editorial assistance. This study was funded by the European Union FEDER grant IFD97-0410 and by Abbott Laboratories. A. G. and R. R. contributed to the study design and wrote the different drafts. C. G.-L. was involved in obtaining the experimental results under the supervision of R. R. and all authors were involved in their discussion. No other authors have conflicts of interest. Carmen Gómez-León was the recipient of a fellowship from the Spanish government.