![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
|
|
|
|
|
||
,3
* Departments of Nutritional Sciences and
Animal Sciences, University of Wisconsin-Madison, Madison, WI 53706
3To whom correspondence should be addressed. E-mail: njbeneve{at}ansci.wisc.edu.
 |
ABSTRACT |
|---|
 TOP ABSTRACT MATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
KEY WORDS: • piglets • nitrogen balance • nitrogen excretion • 15N recovery
Nitrogen balance data are often used to estimate protein and amino acid requirements (1,2). The nitrogen balance of an individual can be represented by the following equation:
 |
where Nin represents dietary nitrogen and Nout represents excretory nitrogen. A nongrowing individual, theoretically, would not be accruing protein, and should therefore have a nitrogen balance equal to zero if dietary protein is sufficient. A net accretion of protein is required during growth and pregnancy. Under these circumstances, it is important that an individual be in positive nitrogen balance. Nitrogen accretion is especially important in a premature infant, for example, whose glycogen and lipid stores are negligible and whose only substantial endogenous fuel source after birth is body protein (3).
Previous studies involving nitrogen metabolism produced discrepancies that question the validity of the nitrogen balance method. Balance studies in adult men reported positive nitrogen balances of 1.4–1.6 g N/d over 220- and 45-d experiments without corresponding increases in body weight (4,5). Also, no difference in body weight accretion was noted in premature infants given 2 or 9 g of protein/(kg · d) even though the infants given 9 g had a N retention 3 times that of those given 2 g (6). Further evidence of unaccountable nitrogen was shown in animal studies. Only 73 and 54% of infused alanine nitrogen was recovered in the urinary nitrogen and expanded body urea pool of small and large newborn piglets, respectively (7). The comparative slaughter method has also been used to study nitrogen metabolism in animals. This method allows a comparison of body nitrogen loss to nitrogen excretion by using reference animals to estimate initial total body nitrogen. A comparative slaughter study in starved piglets revealed that only 30–50% of body nitrogen loss was recovered as urinary nitrogen (8). Using both the nitrogen balance method and the comparative slaughter method in >400 pig experiments, nitrogen deposition estimated using the balance method yielded a mean that was 16% greater than that estimated using the comparative slaughter method (9), suggesting again a positive N gain that could not be verified.
These discrepancies suggest the existence of a nitrogen excretory product not detected by standard methods. The standard Kjeldahl method used for total nitrogen determination in balance and comparative slaughter experiments does not quantitatively recover oxidized nitrogen such as nitrate (NO3-) and nitrite (NO2-). Endogenous formation and excretion of nitrogen as NO3-, shown to occur in humans (10–12) and in germ-free rats (13,14), would therefore not be accounted for in nitrogen determinations via the standard Kjeldahl method. There are reports of gaseous nitrogen excretion as N2 by both conventional and germ-free mice and by humans (15), but no such evidence was found when rats were fed a 15N-labeled yeast slurry (16).
Because preterm infants have limited energy stores, an undetected nitrogen excretory product would have an effect on the estimation of the quantitative contribution of protein to the energy needs of these neonates. The newborn piglet serves as an appropriate model for studying the nutrition and metabolism of the human preterm infant because of similarities in body composition and physiology. Not only are newborn piglets and preterm infants both born with limited fat and glycogen stores (3), they also both have limited thermal insulation and similar means of thermoregulation (17).
The objectives of the following study were to determine the recovery of 15N in the body and urine of newborn piglets given an i.v. infusion of 15N L-alanine and to ascertain whether newborn piglets are capable of gaseous nitrogen excretion.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Chemicals and supplies. 15N L-Alanine was obtained from Cambridge Isotope Laboratories and its isotopic abundance [97.96 ± 1.13 atom percent (AP)4 ] was verified. Unenriched L-Alanine was obtained from U.S. Chemical Corporation and its isotopic abundance (0.370 ± 0.004 AP) was verified. U-14C-Alanine was obtained from Moravek Biochemicals and had a radiochemical purity of 98.5% based on the product data sheet. Concentrated sulfuric acid, mercuric oxide, potassium sulfate, iron, and zinc were obtained from Fisher Scientific. Amino acid standards used for Kjeldahl analysis were obtained from Sigma Chemical. Bromine, sodium hydroxide, ammonium sulfate, and 30% hydrogen peroxide were obtained from Mallinckrodt. Amino acid standards for HPLC analysis were obtained from Pierce Chemical. Octanol was obtained from Acros Organics. Lithium hydroxide was obtained from Matheson, Coleman & Bell. Molecular sieves (5Å) were obtained from Alltech. Right-angle stopcocks used to hold the molecular sieves were obtained from Chemglass. All other glassware used during the nitrogen isotope analysis was obtained from Kimble-Kontes. Glassware used for gas sampling and in the preparation of samples for nitrogen isotope ratio analysis was modified to fit our needs by Erway Glass Blowing. High-vacuum stopcock grease (Apiezon H) was obtained from Apiezon Products.
Selection of 15N L-alanine as the tracer.
Alanine was selected as the substrate because it is nontoxic at the levels infused and because there are multiple routes by which its nitrogen may enter the metabolism. It can transaminate with 
-ketoglutarate to form glutamate, which can undergo further transamination to form other amino acids or be acted upon by glutamate dehydrogenase to yield free ammonium. Alanine can also react with oxaloacetate via aspartate amino transferase (EC 2.6.1.1) to form aspartate and pyruvate. Thus, the two nitrogen atoms comprising urea, one contributed by aspartate and the other by free ammonium, can both be derived from alanine nitrogen. The potential for alanine to provide nitrogen through multiple routes of transamination should lead to the distribution of 15N throughout the free amino acid pool, thereby increasing the chance that the labeled nitrogen might enter an alternate route of nitrogen excretion. The potential for alanine N to be excreted was enhanced because only alanine was infused; thus net accrual of protein should have been prevented.
Experimental protocol. Newborn piglets (n = 4; 1.25 ± 0.27 kg) were killed immediately after surgery so that the natural abundance of 15N in piglets could be determined. All piglets were killed by injection of sodium pentabarbital (75 mg/kg body; Sigma Chemical) via the indwelling catheter, and the intact carcasses were then frozen. All 16 piglets (1.47 ± 0.27 kg) received water (8 mL/h) i.g. for the first 6 h of the experiment (baseline). Then the i.v. infusion of 15N enriched L-alanine was started at 8 mL/h and continued for 60 h. Piglets were switched back to water i.g. for the remaining 12 h (60–72 h) of the experiment (washout) and then killed.
The concentration of alanine in the infusate was calculated so that the rate of alanine infusion would meet 25% of the piglets’ resting energy expenditure (REE) (4) based on ATP equivalents (18). The nitrogen isotopic abundance of the infusion was 2.3 (n = 4), 2.8 (n = 10), or 3.3 (n = 2) AP 15N.
 |
The 15N abundance of a given infusion was accomplished by combining appropriate amounts of 15N L-alanine and unenriched L-alanine. When analyzed using MS, the 15N L-alanine and unenriched L-alanine had an abundance of 97.96 ± 1.13 and 0.370 ± 0.004 AP, respectively.
Alanine infusates were filtered through a sterile 0.2-µm filter (Micron Separations) before the experiment. The infusion rate of alanine was based on the expected production of ATP from its catabolism (16 mol ATP/mol alanine) and the amount of ATP required to meet 25% of the piglets’ REE. The REE of the piglet is equivalent to the ATP turnover; on the basis of the steady-state expired CO2, it has been calculated to be 138 mmol ATP/(h · kg0.75) (18). Although infusion pumps were calibrated before the experiment to a rate that would supply 25% of the piglets’ resting energy needs, intravenous back-pressure and partial collapse of infusion tubing with time resulted in a variation in the actual amount of alanine each piglet received. Overall, 23.1 ± 2.5% of the calculated resting energy needs were met. This variation in the amount of alanine infused was taken into account during statistical analysis of the data.
The infusion of 16 piglets was conducted in 5 separate trials because equipment and personnel limitations allowed for only a maximum of 4 piglets to be infused at once and because some piglets expired prematurely. The number of piglets and the level of 15N infused for each trial was as follows: Trial 1: 4 piglets, Infusion = 2.3 AP 15N; Trial 2: 3 piglets, Infusion = 2.8 AP 15N; Trial 3: 4 piglets, Infusion = 2.8 AP 15N; Trial 4: 2 piglets, Infusion = 3.3 AP 15N; Trial 5: 3 piglets, Infusion = 2.8 AP 15N.
Unused infusate was immediately frozen at the end of the infusion and stored at -10°C until analysis. One piglet expired early into the infusion in trial 2 and 2 piglets expired early into the infusion in trial 4. Because of the uncertainty of the 15N distribution in the body and urine, three different 15N enrichment levels were used in the infusate to enhance the probability of being able to measure the AP of samples.
To determine whether 15N recovery was dependent of the metabolism of alanine over 60 h, 2 piglets (1.34 ± 0.13 kg) were given a 9-mL bolus i.v. infusion of 3.50 mmol 15N L-alanine/kg body (97.96 ± 1.13 AP). After allowing the bolus to circulate and mix for 6 min, the piglets were killed and the intact carcasses frozen.
Development of a closed circuit respiration system to investigate the production of N-containing gases. The possibility of gaseous nitrogen excretion was studied by maintaining 4 of the 16 treatment piglets in a closed circuit respiration system (CCRS; Fig. 1) throughout the 60-h infusion period and 12-h washout period. Piglets were housed in an acrylic cylinder as before, but the cylinder was sealed by bolting grooved acrylic end-plates over compressible butyl-rubber O-rings (Lutz). Infusion and collection lines entered the chamber through 14-gauge needles that were welded into a fitting that then screwed into a polyethylene fitting (University of Wisconsin Physical Plant) on the posterior end-plate. Respiratory water (H2O) was sequestered in a 1-L glass cold-finger submerged in a 1:10, NaCl:ice mixture to achieve a temperature of -13°C. Respiratory CO2 was removed in an acrylic cylinder (i.d., 7.0 cm; o.d., 8.25 cm; 40 cm high) filled with soda-lime (Soda-Sorb, W.R. Grace & Company) and sealed with grooved stainless steel end-plates over butyl-rubber O-rings. The inferior 3 cm of this scrubber tower contained Drierite (W.A. Hammond) to remove any residual respiratory H2O not trapped by the cold-finger. All components were connected by stainless steel tubing (i.d., 0.45 cm; o.d., 0.65 cm) with stainless steel compression fittings (Swagelok) at the attachment points. As respiratory H2O and CO2 were removed from the gas phase, the decrease in pressure was equilibrated by the introduction of O2 from a 9-L spirometer (Warren E. Collins). To minimize the back-flow of gas from the system to the spirometer, a 25-cm length of stainless steel capillary tubing (i.d., 0.7 mm; o.d., 3.0 mm) was inserted at the point at which O2 exited the spirometer and entered the system. To reduce the chance of atmospheric nitrogen contamination from an inward leak, a 540-g lead donut was placed on top of the bell of the spirometer so that the system gas would be under positive pressure. The gas sample site was placed in the gas circulating line 20 cm before the entrance to the animal chamber and consisted of stainless steel tubing in a T-configuration with two 1-way ball valves to control the direction of gas flow during sampling. One arm of the sample-T led to an Edwards RV5 vacuum pump (Edward International) connected to a gauge; the other arm led to a glass sample bulb. For all evacuations, a pressure of <0.03 mbar was considered a sufficient vacuum. A flow-rate of 3 L/min was continually sustained in the system with a diaphragm pump (Neptune Dyna-Pump, Magnetek). The animal chamber was fitted with quick-disconnects (Swagelok) at its attachment points and could therefore be detached from the system. The total system volume (animal chamber + cold-finger + scrubber + tubing) was 9 L based on the dimensions of the components and assuming that Soda-Sorb occupied 50% of the scrubber tower volume.
|
Urine collection. Urine collection was possible via the urachal catheter. Urine was collected continuously over 3-h intervals throughout the baseline, infusion, and washout periods directly into 1 mol/L HCl in a graduated cylinder on ice. Samples were stored at -10°C until analysis.
Blood collection and analysis. At the beginning and end of the 6-h baseline period and every 12 h from that point on, 0.25 mL of blood was drawn via the indwelling arterial catheter. Blood samples were deproteinized by the method of Somogyi (19) and the supernatants stored at -10°C until analysis. The samples were spiked with uniformly labeled U-14C-alanine (1.6 x 106 dpm; 26.7 kBq) before deproteinization to estimate the recovery of blood alanine. Whole-blood protein-free supernatants were derivatized with phenyl isothiocyanate (PITC), and blood alanine concentration was determined by HPLC [Pico-Tag, Waters (20)]. A standard curve (peak area vs. pmol alanine) was generated with alanine standards (Pierce Chemical), and the alanine peak for each series of samples was verified by the addition of an alanine standard.
Collection of respiration chamber gases. Gas was sampled from the closed circuit respiration system in duplicate every 12 h with the first sample taken at the start of the infusion. Samples were collected and stored in 50-mL glass bulbs closed off from the atmosphere by an offset ground-glass stopcock with a 2-mm bore. The sample bulbs were attached to the sample site of the closed circuit respiration system via a 10/30 standard taper ground-glass joint. Both the stopcock and joint were lubricated with high-vacuum stopcock grease. Sample bulbs were evacuated and filled with Ar before the experiment. At the time of sampling, unmixed residual gas was first cleared from the sample-T by filling an evacuated 10-mL glass bulb with 1 volume of system gas. The sample bulb was then attached, evacuated, filled with 1 volume of Ar, and reevacuated before taking a gas sample from the system. The ground-glass stopcock along with the small sample volume required for MS analysis (<1 mL), allowed for repeated sampling from a sample bulb. Sample bulbs were stored on the benchtop for as long as 4 wk with no evidence of atmospheric contamination. Samples from the closed circuit respiration system were analyzed on a Finnigan Delta S mass spectrometer. A magnetic field scan produced a mass chromatogram ranging from masses 10 to 70 for each sample. The percentage of each gas in the sample was calculated from its corresponding mass peak height as a percentage of the sum of all peak heights in the sample.
Whole-animal homogenates. At the end of the 12-h washout period at 72 h, piglets were killed by an injection of sodium pentabarbital (75 mg/kg) and immediately frozen. The frozen piglets were cut with a meat saw into sections 2–5 cm in width. The pieces were then passed 3 times through a meat grinder (Hobart), twice through a 5-mm dye and once through a 3-mm dye. The mean recovery of the intact piglet following this grinding process was 96.6 ± 0.8%. Whole-piglet homogenates were mixed in an electric mixer (KitchenAid) and the homogenate subsampled. Frozen subsamples were then blended (Waring) in liquid nitrogen and stored at -78°C until analysis. A dry matter determination for each homogenate was done before liquid nitrogen blending by weighing out 6–7 g of the homogenate into aluminum weigh boats, placing them in a 100°C drying oven for 10 h, and then reweighing the dried samples. Previous experience showed that a constant weight is achieved by 5 h under these conditions.
Homogenate, urine, and infusate nitrogen analysis. Because the water content of body homogenates may have changed with grinding in liquid nitrogen and storage, dry matter was determined for each sample before analysis, and total body nitrogen was calculated on a dry matter basis. A constant percentage of each 3-h urine volume was used to make up the composite that was analyzed. There was one composite for the 6-h baseline period and one for each 12-h period from that point forward (1 baseline, 5 infusion, 1 washout).
Samples were analyzed for total nitrogen using a standard semimicro-Kjeldahl method adapted from the suggestions of Fleck and Munro (21). The amount of sample used for Kjeldahl analysis was one that contained 
3.5 mmol of nitrogen. Samples were digested for 6 h in duplicate in 10 mL concentrated sulfuric acid with 150 mg HgO added as a catalyst. A digestion temperature of 370°C was obtained with the addition of 3.4 g K2SO4. After digestion, samples were allowed to cool before the addition of 1 g Zn, 50 mL H2O, and 50 mL of 40% NaOH. Samples were then distilled into 25 mL of 0.1 mol/L H2SO4 until 75 mL of distillate was collected. The ammonia concentration of the resulting distillate was determined colorimetrically by the Berthelot reaction (22). Mean recovery of nitrogen from a histidine standard was 100.6 ± 2.1%.
Determination of oxidized N. The standard Kjeldahl was modified to include NO3- and NO2- by pretreating samples with H2O2 and Fe. The method was adapted from the potassium permanganate (KMnO4)-reduced iron Kjeldahl method outlined by Bremner and Mulvaney (23). H2O2 was used as an oxidizing agent because KMnO4 decreases urinary nitrogen values compared with the standard Kjeldahl method. H2O2 did not diminish urine nitrogen values and still gave complete recovery of nitrogen from NaNO2 and NaNO3 standards. To oxidize NO2- to NO3-, 5 mL of 1% H2O2 was added to samples and allowed to react for 1 min, followed by the addition of 10 mL of 18 mol/L H2SO4; 5 min later, 2.5 g of Fe powder was added to each digestion tube to reduce NO3- to NH4+, and 3 drops of octanol was added to prevent foaming. Samples were then gently heated (<100°C) for 1 h and allowed to cool before continuing with the standard Kjeldahl procedure as previously described. Mean recovery of nitrogen from NaNO3 and NaNO2 standards was 97.1 ± 0.7% and 101.0 ± 0.3%, respectively, using the modified Kjeldahl method. Total nitrogen and 15N recovery were determined by both methods in the urine and bodies of 4 piglets infused with 15N L-alanine.
Determination of nitrogen isotopes in homogenates, urine and infusates.
Isotope analysis of N (28N2, 29N2 and 30N2) was accomplished by conversion of the NH3 in the Kjeldahl distillate to N2 gas (24). This process (2NH3 + 3OBr- 
3Br- + 3H2O + N2) was done using a glass manifold system, which was designed such that 5 mol/L LiOBr and the sample could be placed in separate bulbs of a Y-tube and evacuated. After evacuation, the contents of the Y-tube were mixed and the N2 gas produced was collected in 0.5 nm molecular sieves that were submerged in liquid nitrogen. In the process of transferring N2 to the molecular sieves, H2O was removed via a cold-finger that was part of the manifold submerged in ethanol and dry ice. N2 was released from the molecular sieves by placing the apparatus in a 200°C sand bath for 15 min. The N2 released was immediately analyzed by MS. The isotopic abundance of 15N was derived from the proportions of 28N2, 29N2, and 30N2 and was expressed as AP 15N. The measurement was made relative to an air standard that was analyzed simultaneously.
Recovery of 15N from an enriched alanine standard after Kjeldahl digestion and conversion to N2 was 96.4 ± 0.9%. The possibility of cross-contamination between samples from 15N adhering to glassware was tested by running an unenriched alanine standard after each series of enriched samples. The 15N abundance of the alanine standard did not differ at any time point from the value measured for the alanine when it was analyzed apart from enriched samples (0.370 ± 0.004%).
Calculations.
The 15N enrichment of the bodies and urine of the piglets and of the infusion mixture was determined by calculating the atom percent excess (APE) 15N relative to the estimated natural abundance value. The isotopic abundance (AP) was first converted to an isotope ratio (R).
 |
A ratio difference (R') between the measured ratio of 15N in the enriched sample (Ren) and the estimated natural ratio of 15N in the sample (RN) could then be used to calculate the 15N enrichment of the sample (APE)
 |
 |
The natural abundance of 15N in piglet bodies (AP = 0.368 ± 0.001, RN = 0.003694) was estimated from the isotopic abundance of 15N in the bodies of 4 piglets killed immediately after surgery. The natural abundance of 15N in the urine of each treatment piglet was estimated from the enrichment of its baseline urine composite. The mean baseline abundance value across all treatment piglets was 0.372 ± 0.003 AP. The natural abundance of 15N in the infusions was estimated from the measured abundance of unenriched L-alanine (AP = 0.370 ± 0.004, RN = 0.003714). The Kjeldahl nitrogen value of a given sample (mmol NTotal) was used in conjunction with the APE value to calculate the mmol of excess 15N (E).
 |
Total recovery of 15N as a percentage of the total 15 N infused could then be calculated using the following equation:
 |
Statistical analysis. Differences were considered significant at P < 0.05. Analysis of blood alanine concentration and urinary 15N abundance data was done using the mixed procedure of SAS (version 7.0; SAS Institute). With autoregressive order one as the covariance structure, a repeated-measures model was fitted to assess the effects of trial, trial x time and pig (trial) (Random Effects) and the effects of infusion, time and infusion x time (Fixed Effects). Least-square means were generated for each data point and their differences used to evaluate significant differences from baseline. To determine whether a plateau was reached or was being approached, the fit of the data to a linear quadratic response was tested using the general linear models procedure of SAS with time as the linear and quadratic variable. Comparison of baseline and washout values for blood alanine concentration (mmol/L) was done by paired t test. Because of variation in the baseline values of mass 28 and mass 29 in the sampled gas, the statistical analysis for each was performed on the percentage change from baseline with time. Differences in slopes for the percentage change of mass 28 and mass 29 with time were evaluated using repeated-measures analysis in the mixed procedure of SAS. A model was fitted that assessed the fixed effects of treatment (with or without a pig), time (0–72 h) and their interaction, and also the random effect of chamber (Expts. 1–4). Comparison of 15N abundance in piglet bodies was assessed using contrast statements in the general linear models procedure of SAS. Total 15N recovery in the 16 infused piglets and 2 control piglets was analyzed with a t test (Ho: µ = 100%). Differences in 15N recovery with the standard and modified Kjeldahl method were analyzed using a paired t test. Values in the text are means ± SD.
|
RESULTS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
|
660 mL of chamber gas samples removed for analysis over the course of the experiment. Taking this into account, it was possible to calculate what the gas composition should have been at 72 h (72 hcalc). This calculation assumes that the system did not leak, that there were no changes in atmospheric temperature or pressure, and that there was no exchange of gas between the spirometer and atmosphere or through the capillary separating the spirometer and system when the two were at equal pressure. There was a difference of 1.01 ± 0.09 in the percentage of 28N2 without a pig in the system when comparing the observed composition at 72 h to the calculated composition. The difference in the observed O2 percentage and the calculated percentage was positive (3.09 ± 2.65) and the difference in the observed Ar percentage and the calculated Ar percentage was negative (-4.16 ± 3.01). The difference in the percentage of 28N2 with a pig in the system when comparing the observed composition at 72 h to the calculated composition was 1.61 ± 0.44. Again, the difference in the observed and calculated O2 percentages was positive (0.83 ± 0.54) and the difference in observed and calculated Ar percentages was negative (-2.48 ± 0.72). The positive difference in the amount of 28N2, in both cases, is indicative of an inward leak. The positive difference in the amount of O2 and the negative difference in the amount of Ar indicate an outward leak. These differences also suggest that gas was exchanged across the capillary separating the spirometer and system, which may explain why the magnitude of the differences for O2 and Ar was greater when the system was run without a pig. Without a pig in the system, the system and spirometer are presumably at the same pressure. The possibility of gas mixing across the capillary then becomes greater because the pressure in the chamber is not negative due to O2 consumption by the piglet. Due to variability in the amount of 28N2 and 29N2 initially present, all statistical analyses were performed on the absolute changes from the percentage at t = 0. The mean percentage change of 28N2 over 72 h for the 4 experiments was 1.3 ± 0.3 and 0.9 ± 0.1% with and without a pig in the system, respectively. The difference in the slope of the two regression lines over the 72-h experiment was marginally significant (P = 0.0969). The mean percentage change of 29N2 over 72 h for the four experiments with a pig in the system was 0.014 ± 0.005 and 0.013 ± 0.005% with and without a pig in the system, respectively. The slopes of the two regression lines did not differ (P = 0.8565). There was no evidence of a water-soluble gas in the cold-finger when its head-space gas was analyzed before and after thawing (data not shown). There was also no evidence of a Kjeldahl determinable form of nitrogen in the water of the cold-finger (data not shown).
Nitrogen isotope abundance of urinary N derived from infused 15N L-alanine. The mean 15N abundance (AP) in urine composites from the piglets with continuous urine output was adjusted for time and pig effects. The pooled SD for piglets infused with 2.3, 2.8, or 3.3 AP 15N were ± 0.086 (n = 4), ±0.099 (n = 9), and ± 0.599 (n = 2) AP, respectively. The mean 15N abundance of the baseline composite used to estimate the natural abundance of 15N in urine was 0.3724, 0.3719, and 0.3756 AP for the three groups of piglets, respectively. For all three infusion levels, the 15N abundance of the composite representing the first 12 h of the infusion was significantly different from baseline, and the 15N abundance of each composite over the remainder of the experiment was also significantly different from baseline. The maximum abundance of 15N in urine was a function of the amount of 15N alanine infused (2.3 vs. 2.8 vs. 3.3 AP infused = 1.3871 vs. 1.6319 vs. 1.8623 AP, respectively). Testing the fit of the urine composite data to a linear quadratic curve indicated that the 15N abundance of excretory nitrogen reached a plateau at 55.3, 55.6, and 52.5 h into the infusion for piglets infused with 2.3, 2.8, and 3.3 AP 15N, respectively. This suggests that a steady state between the rate of 15N infusion and that of 15N excretion was achieved.
The 15N abundance of the 12-h washout composite did not return to baseline levels for any of the 3 levels of 15N infusion. The washout composite included all urine samples from the final 12 h of the experiment, starting from the time the infusion was ended; therefore, the total nitrogen excreted in the last 3 h of the experiment and its isotopic abundance were not known. A calculation was made to determine whether the isotopic abundance of the final washout sample could have returned to baseline levels. The total nitrogen value used for each sample in the calculation was derived from its urea nitrogen value, assuming that the proportion of total urinary nitrogen comprised by urea nitrogen did not change among the 4 samples of the composite. For a return to baseline to have occurred, the mean isotopic abundance of each of the other 3 urine samples taken during the washout period would had to have been greater than the maximum abundance by 0.1 AP. It is therefore improbable that the isotopic abundance of excretory nitrogen had returned to baseline levels because an increase in the abundance after termination of the infusion would be unlikely.
Nitrogen isotope abundance of body N before and after infusion of 15N L-alanine. The natural abundance of 15N in the bodies of newborn piglets was 0.368 ± 0.001 AP. The 15N abundance of infused piglet bodies at 72 h was again a function of the amount of 15N alanine infused (2.3 vs. 2.8 vs. 3.3 AP infused = 0.451 ± 0.006 vs. 0.470 ± 0.012 vs. 0.497 ± 0.012 AP). Bodies of control piglets killed 6 min after an i.v. bolus infusion of 15N alanine had an isotopic abundance of 0.595 ± 0.011 AP. The bodies of all piglets infused with 15N had an abundance that was significantly different from one another and from the estimated natural abundance measured in uninfused piglets.
The percentage recovery of 15N infused as L-alanine in the body and urine. The mean recovery of 15N in the 15 infused piglets that produced urine throughout the experiment was 44.1 ± 5.5 and 49.1 ± 5.9% in the body and urine, respectively (Fig. 2). Urine data were limited for 1 of the 16 piglets because the piglet had urine output only during the first 12 h of the infusion and none thereafter, either via the urachal catheter or through natural voiding. This was presumably due to kidney failure. 15N recovery in this piglet was 94.4 and 1.5% in the body and urine, respectively. The mean total recovery of 15N in all 16 infused piglets was 93.3 ± 2.8% and was significantly different from 100% (P < 0.001). The mean total recovery of 15N in the control piglets killed 6 min after an i.v. bolus of 15N L-alanine was 101.5 ± 1.6% and did not differ from 100% (P > 0.1).
|
Comparison of the standard and modified Kjeldahl in estimating body and urinary N.
A comparison of the standard and modified Kjeldahl methods was done by measuring nitrogen recovery from amino acid and nicotinic acid standards and from salts containing nitrogen as either NO2- or NO3- (Table 2). Recovery of nitrogen from nicotinic acid and amino acids was 100% and did not differ between the two methods. Recovery of nitrite-N and nitrate-N was 20% using the standard method, but complete recovery was achieved using the modified method. Total nitrogen was also measured in the bodies and urine of starved piglets by both methods and expressed as a relative recovery setting the standard Kjeldahl value to 100%. In all cases, the relative recovery using the modified method was 100%, but recovery of NO3- nitrogen added to body homogenate was again only 20% using the standard method and 100% using the modified method. Because there was not a detectable difference between the 2 methods in either piglet bodies or urine, this was seen as evidence for a lack of production of nitrite or nitrate and a justification for using the standard Kjeldahl method in the present study.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
13.5 g N/d were studied over a 220-d period and were shown to accrue a mean of 1.4 g N/d without a gain in body weight (4). Another balance study using adult men found that when subjects were given a high-nitrogen diet (36 g N/d) over a 45-d period, they had an mean positive nitrogen balance of 1.6 g N/d without corresponding changes in body weight (5). Measurements of nitrogen intake and output in this balance study were very meticulous, with meals quantitatively consumed and nitrogen losses due to feces, urine, sweat, and integumentary loss all measured. In a study using premature infants (6) with a protein requirement of 3 g/(kg · d), no difference in body weight accretion was noted in infants given 2 or 9 g of protein/(kg · d). The 9 infants given 2 g protein/(kg · d) had a N retention of 221 ± 8 mg N/(kg · d), whereas the 9 given 9 g of protein/(kg · d) retained 644 ± 183 mg N/(kg · d). The range in the group given 9 g protein/(kg · d) was from 419 to 916 mg N/(kg · d) and did not overlap with the 215–232 mg N/(kg · d) of the group given 2 g protein/(kg · d). Discrepancies such as these have led some to suggest that a nitrogen excretory product not measured by standard methods is escaping detection. Recovery of infused L-alanine-N or body-N. The inability to totally recover infused 15N in the present study lends support to the notion that an alternate nitrogen excretory route is not being accounted for. Other studies of nitrogen metabolism in newborn piglets carried out in our laboratory yielded similar results. Davis et al. (7) administered a stair-stepped i.v. alanine infusion to newborn piglets at 0, 25, 50, and 75% of REE over 24 h followed by a 12-h washout period. Assuming that the piglet was 80% water and that urea was uniformly distributed in the water space, only 73 ± 5.5 and 54 ± 2.3% of infused alanine-N was recovered in the urine and expanded body urea pool of small (0.99 ± 0.16 kg) and large piglets (1.86 ± 0.16 kg), respectively. Applying the same calculation to the 16 piglets (1.47 ± 0.27 kg) from this experiment, but using the measured body water content for each piglet rather than an assumption of 80%, the mean recovery of infused alanine nitrogen in urine and the expanded body urea pool was 77.5 ± 16.0%. A comparative slaughter study in newborn piglets starved from 12 to 72 h of age also provided evidence of an alternate nitrogen excretory product (25). Only 30–50% of body nitrogen loss could be accounted for as urinary nitrogen loss. This represented a mean discrepancy of 1.50 g N/(60 h · kg) or 1.97 g N/(60 h · kg0.75). Nitrogen appears to be lost by unknown pathways and in unknown ways.
Evidence for nitrogen loss as N containing gas(es). That unaccountable nitrogen retentions could be explained by gaseous N2 excretion is challenging to prove or disprove because N2 makes up nearly 80% of the atmosphere. This makes it difficult to differentiate possible excretory N2 from atmospheric N2. Evidence of N2 excretion was reported in conventional mice maintained for 24 h in a closed-circuit respiration system flushed with an 80:20 He:O2 mixture and surrounded by a He-filled "buffer zone" (15). Differences between initial and final N2 concentrations were significant and ranged from 0.099 to 0.153%. Differences were also significant with germ-free mice and ranged from 0.05 to 0.086%. When humans prefed a high-protein, high-energy diet were studied in an analogous system but without a "buffer zone," differences ranged from 0.03 to 0.06%. Applying the same discrepancy seen in the human nitrogen balance study (1.4 g N/d) and using the reported system volume (1000 L), the expected change in N2 concentration in the human experiment would be 0.1%. This is assuming that the balance discrepancy was due entirely to nitrogen excretion as N2.
In the present study, the amount of 28N2 as a percentage of the CCRS volume would be expected to increase 29% over 72 h if the discrepancy observed in the comparative slaughter study (1.97 g N/60 h · kg0.75) was due entirely to N2 excretion. This was calculated using the system volume of 9 L and a piglet weight of 1.5 kg. The difference between the mean change in 28N2 with a piglet in the system and the mean change without a piglet in the system (control) was 0.4% and was marginally significant (P = 0.0969); however, it does not explain the discrepancy previously observed with comparative slaughter.
The observed increase may be due to gaseous excretion, but it could also be due in part to the 28N2 initially soluble in the tissues of the piglet not being completely removed by the end of the 6-h washout period. The observed increase in 28N2 above control levels was equivalent to 54 mL (0.6% of system volume). Based on the solubility of N2 in water (26) and fat (27) and assuming a 1.5-kg piglet to be 80% water and 1% fat, the amount of N2 soluble in the piglet before flushing was 18 mL (0.2% of system volume). The control experiment was not ideal because the seal on the animal chamber had to be broken and resealed to place the piglet in the system. Therefore, the system may not have leaked at the same rate for a given control experiment and the corresponding piglet experiment. Whatever the reason for the increase in 28N2 concentration, its magnitude was not large enough to explain previously observed discrepancies.
There was no evidence that 15N was metabolized to and excreted as 29N2. Similarly, Brown (16) was unable to detect increased levels of 29N2 in the expired air of adult rats fed a 15N-labeled yeast slurry. The inability to recover 6.7 ± 2.8% of the infused 15N prompted us to question whether a gaseous nitrogen excretory product could be in the water frozen out by the cold-finger or bound to the soda lime of the CO2 scrubber. To test for the possibility of a nitrogen-containing gas soluble in the water of the cold-finger, the head-space was flushed with Ar and sampled before thawing. After thawing at 37°C, a second head-space sample was taken and compared with the first. Kjeldahl analysis was also performed on the water frozen out by the cold-finger. There was no evidence of either a soluble gaseous nitrogen component or a Kjeldahl-determinable nitrogen component in the metabolic water. The possibility that a nitrogen- containing excretory product was bound to the soda lime of the CO2 scrubber was never investigated because the chemical composition of the soda lime [Ca(OH)2 73%, KOH 5%, NaOH 3%, H2O <19%, and ethyl violet indicator] made it incompatible for Kjeldahl digestion. The same brand of soda lime used in this experiment (Soda-Sorb) was shown to completely absorb nitrogen dioxide (NO2; mass 46) and partially absorb nitric oxide (NO; mass 30) (28). These conclusions are based on observed decreases in the concentration of both NO2 and NO after a mixture of the 2 gases was exposed to soda lime. These 2 gases as well as H2O and CO2 were not seen in the 10- to 70-mass unit scans routinely carried out in the current study. Two possibilities exist, i.e., the N-containing gases either are not produced or the scrubbers are removing them along with the H2O and CO2.
The results of this study do not rule out the possibility that a nitrogen-containing excretory product or metabolite may be escaping detection. Although the magnitude of the discrepancy is smaller compared with those observed in previous studies of nitrogen balance, the 15N loss observed in perinatal animals may not represent what occurs in older pigs or in other nonruminant animals adapted to high levels of dietary protein over a period of weeks or months. The existence of measurable levels of an undetected form of nitrogen would have serious implications in human and animal nutrition. Further investigation, using stable isotopes or other discriminate methodology, is warranted to determine whether additional nitrogen excretory products or metabolites do exist.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
2 Supported in part by Hatch funds (project # 3862) supplied by the University of Wisconsin College of Agricultural and Life Sciences.
4 Abbreviations used: AP, atom percent; APE, atom percent excess; CCRS, closed circuit respiration system; E, mmol 15N excess; PITC, phenyl isothiocyanate; R, isotope ratio; R', isotope ratio difference; REE, resting energy expenditure.
Manuscript received 6 October 2003. Initial review completed 31 October 2003. Revision accepted 10 January 2004.
|
LITERATURE CITED |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
1. National Research Council (NRC) Food and Nutrition Board (1989) Protein and amino acids. Recommended Dietary Allowances 10th ed. 1989:52-77 National Academy Press Washington, DC.
2. FAO/WHO (1973) Report of an Ad Hoc Expert Committee on Energy and Protein Requirements, no. 52. FAO Nutrition Meetings Report Series no. 522 :40-53.
3. Ziegler, E. E., O’Donnell, A. M., Nelson, S. E. & Fomon, S. J. (1976) Body composition of the reference fetus. Growth 40:329-341.[Medline]
4. Mitchell, H. H. (1949) Adult growth in man and its nutrient requirements. Arch. Biochem. 21:335-341.
5. Oddoye, E. A. & Margen, S. (1979) Nitrogen balance studies in humans: long-term effect of high nitrogen intake on nitrogen accretion. J. Nutr. 109:363-377.
6. Snyderman, S. E., Boyer, A., Kogut, M. D. & Holt, E. L., Jr. (1969) The protein requirement of the premature infant. I. The effect of protein intake on the retention of nitrogen. J. Pediatr. 74:872-880.[Medline]
7. Davis, J. A., Greer, F. R. & Benevenga, N. J. (2000) Urea production is increased in neonatal piglets infused with alanine at 25, 50, and 75% of resting energy needs. J. Nutr. 130:1971-1977.
8. Mickelson, B. D., Thomsen, J. A., Greer, F. R. & Benevenga, N. J. (1996) Nonprotein fuel decreases body dry matter loss but does not preferentially spare body protein. FASEB J. 10:A736 (abs.).
9. Just, A., Fernandez, J. A. & Jorgensen, H. (1982) Nitrogen balance studies and nitrogen retention. Leplace, J. P. Corring, T. Perat, A. eds. Digestive Physiology in the Pig 12:111-122 Les Colloques de l’INRA .
10. Mitchell, H. H., Shonle, H. A. & Grindley, H. S. (1916) The origin of nitrates in urine. J. Biol. Chem. 24:461-490.
11. Green, L. C., Luzuriaga, K. R., Wagner, D. A., Rand, W., Istfan, N., Young, V. R. & Tannenbaum, S. R. (1981) Nitrate biosynthesis in man. Proc. Natl. Acad. Sci. U.S.A. 78:7764-7768.
12. Tannenbaum, S. R., Fett, D., Young, V. R., Land, P D. & Bruce, W. R. (1978) Nitrite and nitrate are formed by endogenous synthesis in the human intestine. Science (Washington, DC) 200:1487-1488.
13. Witter, J. P., Balish, E. & Gatley, S. J. (1982) Origin of excess urinary nitrate in the rat. Cancer Res. 42:3654-3658.
14. Ward, F. W. & Coates, M. E. (1989) Influence of dietary protein and gut microflora on endogenous synthesis of nitrate and N-nitrosamines in the rat. Food Chem. Toxicol. 27:445-449.[Medline]
15. Costa, G., Ullrich, L., Kantor, F. & Holland, J. F. (1968) Production of elemental nitrogen by certain mammals including man. Nature (Lond.) 218:546-551.[Medline]
16. Brown, R. D. (1967) Use of 15N for Evaluation of the Hypothetical Excretion of Nitrogen by Animals. M. S. thesis 1967 University of Wisconsin Madison, WI.
17. Book, S. A. & Bustad, L. K. (1974) The fetal and neonatal pig in biomedical research. J. Anim. Sci. 38:997-1002.
18. Tetrick, M. A., Adams, S. H., Odle, J. & Benevenga, N. J. (1995) Contribution of D-(-)-3-hydroxybutyrate to the energy expenditure of neonatal pigs. J. Nutr. 125:264-272.
19. Somogyi, M. (1930) A method for the preparation of blood filtrates for the determination of sugar. J. Biol. Chem. 86:655-663.
20. Bidlingmeyer, B. A., Cohen, S. A., Tarrin, T. L. & Frost, B. (1987) A new rapid, high-sensitivity analysis of amino acids in food type samples. J. Assoc. Off. Anal. Chem. 70:241-247.[Medline]
21. Fleck, A. & Munro, H. N. (1965) The determination of organic nitrogen in biological materials. Clin. Chim. Acta 11:2-12.
22. Chaney, A. L. & Marbach, E. P. (1962) Modified reagents for determination of urea and ammonia. Clin. Chem. 8:130-132.[Abstract]
23. Bremner, J. M. & Mulvaney, C. S. (1982) Nitrogen-total. Page, A. L. eds. Methods of Soil Analysis 1982:595-624 Soil Science Society of America Madison, WI. .
24. Rittenberg, D., Keston, A. S., Rosebury, F. & Schoenheimer, R. (1939) Studies in protein metabolism. II. The determination of nitrogen isotopes in organic compounds. J. Biol. Chem. 127:291-299.
25. Mickelson, B. D., Thomsen, J. A., Benevenga, N. J. & Greer, F. R. (1995) Urinary nitrogen in piglets fasted from 12–72 hours of age accounts for 30–50% of body nitrogen loss. FASEB J. 9:A745 (abs.).
26. Battino, R. & Clever, H. L. (1966) The solubility of gases in liquids. Chem. Rev. 66:395-463.
27. Battino, R., Evans, F. D. & Danforth, W. F. (1968) The solubilities of seven gases in olive oil with reference to theories of transport through the cell membrane. J. Am. Oil Chem. Soc. 45:830-833.[Medline]
28. Ishibe, T., Sato, T., Hayashi, T., Kato, N. & Hata, T. (1996) Absorption of nitrogen dioxide and nitric oxide by soda lime. Br. J. Anaesth. 75:330-333.
This article has been cited by other articles:
|
|
 |
|
  J. C. Marini, D. G. Fox, and M. R. Murphy Nitrogen transactions along the gastrointestinal tract of cattle: A meta-analytical approach J Anim Sci, March 1, 2008; 86(3): 660 - 679. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. D. Mickelson, F. R. Greer, and N. J. Benevenga The Contribution of Body Protein to the Supply of Energy in Starved Newborn Piglets Is Not Preferentially Suppressed by Intravenous Provision of Glucose and Fat J. Nutr., November 1, 2005; 135(11): 2609 - 2615. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. A. Flis and M. A. Wattiaux Effects of Parity and Supply of Rumen-Degraded and Undegraded Protein on Production and Nitrogen Balance in Holsteins J Dairy Sci, June 1, 2005; 88(6): 2096 - 2106. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. C. Metges, K. J. Petzke, G. Backes, A. Elsner, P. Junghans, M. Derno, G. Nurnberg, and U. Hennig Response to lysine in a wheat gluten diet in adult minipigs after short-and long-term dietary adaptation as assessed with an indicator amino acid oxidation and balance technique J Anim Sci, April 1, 2005; 83(4): 833 - 841. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Bos, B. Stoll, H. Fouillet, C. Gaudichon, X. Guan, M. A. Grusak, P. J. Reeds, D. G. Burrin, and D. Tome Postprandial intestinal and whole body nitrogen kinetics and distribution in piglets fed a single meal Am J Physiol Endocrinol Metab, February 1, 2005; 288(2): E436 - E446. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
) and urine (
) respectively. The recovery in the piglet with limited urine production was 94.4 and 1.5% in the body and urine, respectively. The mean total 15N recovery in all 16 piglets (93.3 ± 2.8%) was differed from 100% (P < 0.001). The mean total 15N recovery in the bodies of 2 control piglets killed 6 min after a bolus i.v. infusion was 101 ± 1.6% and did not differ from 100% (P < 0.1).