Fetal serine fluxes across fetal liver, hindlimb, and placenta in late gestation IRENE CETIN, PAUL GIACOMO MESCHIA,

V. FENNESSEY, AND FREDERICK

JOHN W. SPARKS, C. BATTAGLIA

Departments of Pediatrics, Physiology, and Pharmacology, University of Colorado School of Medicine, Denver, Colorado 80262; and Department of Obstetrics and Gynecology, University of Milan, Milan, Italy Cetin, Irene, Paul V. Fennessey, John W. Sparks, Giacomo Meschia, and Frederick C. Battaglia. Fetal serine fluxes across fetal liver, hindlimb, and placenta in late gestation. Am. J. Physiol. 263 (Endocrinol. Metab. 26): E786E793, 1992.-Eleven studies of fetal serine fluxes were performed in chronically catheterized fetal lambs by continuous infusion of [l- l:C] - and [U- 14C]serine into a fetal brachial vein. At tracer serine steady state, samples were collected from the fetal abdominal aorta, umbilical vein, fetal hepatic vein, and fetal femoral vein and from the maternal femoral artery and uterine vein. Analyses were performed for plasma serine and glycine concentration, for serine and glycine lsC mole percent enrichment, and for whole blood 14C0, and 0, concentrations. Uterine and umbilical blood flows were also measured. The placenta had a significant net uptake of fetal serine (2.1 t 0.5 pmolmin-‘*kg-‘, P < 0.01). Fetal plasma serine disposal rate (DR) was 42.5 t 3.9 prnol mine1 kg-‘. CO, production from decarboxylation of fetal plasma serine represented 7.9 t 0.5% of DR, or 10.1 t 1.2 pmol CO,*min-‘*kg-‘. Fetal plasma glycine enrichment was 59.7 t 4.9% of fetal plasma serine enrichment. There was a significant loss of tracer serine from the fetal circulation into the placenta accounting for -45% of infused tracer. Fifteen percent of this was converted to glycine and released into the umbilical circulation. There was a significant uptake of tracer serine by both fetal liver and fetal hindlimb with a significant CO, production by both sites with serine oxidation predominantly in the carcass. These results indicate a high fetal serine disposal rate in the lamb, with rapid fetoplacental serine exchange, resulting in a net uptake of fetal serine by the placenta. Significant placental glycine production from fetal plasma serine was demonstrated, supporting the concept of interorgan cycling of serine and glycine between the fetal liver and placenta. l

placental glycine production; fetal serine disposal rate; placental serine uptake; serine oxidation ACIDS can be supplied to the fetus by delivery from the placenta into the fetal circulation and/or can be produced within the fetus. Despite their obvious importance for protein synthesis and accretion in a rapidly growing organism, there have been few studies of fetal amino acid fluxes for those amino acids capable of being synthesized within mammalian tissues. Our attention was drawn to fetal and placental serine metabolism because of a number of studies suggesting unique characteristics for fetal serine metabolism. Previously we and others (8, 9, 13) had reported that in the late-gestation fetal lamb there is no measurable umbilical uptake of serine. Furthermore, at midgestation there is a large net flux of serine from the fetal circulation into the placenta (1, 17). These observations indicate that all fetal serine requirements are met by fetal production. In a study of the extraction of amino acids across the fetal liver, Marconi et al. (13) reported that there was a large net efflux of serine from the fetal liver, an observation AMINO

E786

0193-1849/92

$2.00

Copyright

which has not been found across the liver of adult mammals (4, 6, 24, 25). Marconi et al. suggested the possibility of interorgan cycling of serine and glycine between the fetal liver and placenta. More recently, we have measured glycine fluxes in the late-gestation fetal lamb (5). Those studies demonstrated significant serine production from glycine within the fetal liver. The present study utilizes multiple-tracer methodology to quantify fetal and placental serine fluxes in chronically catheterized, late-gestation fetal lambs. The goal of the study was to quantify fetal rates of serine production and oxidation. In addition, the utilization of fetal plasma serine for glycine production in the fetus and placenta was determined. The study provides further confirmation for the concept that fetal requirements for some amino acids are met primarily by endogenous production rather than placental transport. MATERIALS

AND

METHODS

Biological preparation. Ten pregnant mixed-breed ewes (Columbia-Rambouillet) were studied between 118 and 137 days of gestation (full term = 147 days). All were singleton pregnancies. Surgery was performed after 24 h of fasting under pentobarbital sodium sedation and spinal anesthesia. Polyvinyl catheters were placed in both the maternal and fetal circulation: 1) in the uterine vein draining the pregnant horn (V); 2) in the fetal abdominal aorta via a pedal artery (a); 3) in the fetal femoral vein (f); 4) in the common umbilical vein (v); 5) in either the right or left fetal hepatic vein (h) via the thoracic inferior vena cava; 6) in the fetal brachial vein (b); and 7) in the maternal femoral artery (A). An amniotic catheter was placed for antibiotic administration. Catheters were placed as previously described (13). The catheters were exteriorized through a subcutaneous tunnel to a plastic pouch fixed on the ewe’s flank and irrigated daily with heparinized saline (30 U/ml). Ampicillin (500 mg) was administered intra-amniotically for 3 consecutive days. After surgery, all ewes resumed normal feeding and were maintained on an ad libitum diet of alfalfa pellets, water, and mineral supplements. In all cases the daily food intake exceeded 0.8 kg/day. In each animal the experiments were initiated at least 5 days after surgery. Experimental protocol. Eleven studies were performed in 10 animals. During the study period the animals were conscious and provided free access to food and water. An infusate consisting of 90 mg of [l-‘:C]serine (99% atom% excess lC, Cambridge Isotope Laboratories, Woburn, MA), 100 Ci of [U-14C]serine (ICN Radiochemicals, Irvine, CA), and 3 g of antipyrine was prepared in 30 ml of sterile saline solution. A priming dose consisting of 6 ml of this solution was administered followed immediately by a constant infusion of the same solution in the fetal brachial vein at a rate of 0.10 ml/min for 150-240 min. Infusion rates were selected to deliver - 1.0 mol min-l . kg fetus-’ of [l-‘“Cl serine and 250,000 dpmemin-’ kg fetus-l of [U-14C]serine.

0 1992 the American

l

l

Physiological

Society

Downloaded from www.physiology.org/journal/ajpendo at Washington Univ (128.252.067.066) on February 15, 2019.

FETAL

SERINE

FLUXES

IN LIVER,

Zero-time samples were collected from all vessels before the start of the infusion. Fetal arterial blood samples were withdrawn serially beginning at 15 min after the start of the infusion to demonstrate achievement of steady state for fetal plasma serine enrichments. Steady state was achieved by 1 h of infusion. Four steady-state draws were collected simultaneously from A, V, a, v, h, and f at intervals of at least 30 min. All flux calculations were based on the mean values from the four sets of samples for all variables measured. At the end of the experimental period the ewe and fetus were killed with rapid intravenous injection of T-61 euthanasia solution. The total infusion time was noted, and the infusate was saved for analysis of [l-W] serine concentration and [U-l*C] serine radioactivity. The l*C infusate radioactivity infusion rate and time were used to calculate the total amount of [U-l*C] serine administered to the fetus. At autopsy the pregnant uterus, placental cotyledons, and study fetus were separated and weighed. The position of the catheters was confirmed. In two studies the fetus and the placenta were homogenized immediately after weighing and stored at -70°C for further tissue analysis. As previously described (5, l3), hepatic venous samples were considered valid if both the following conditions were met: 1) autopsy confirmed the anatomic position of the catheter and 2) antipyrine and 0, content data confirmed the functional position of the catheter (i.e., that the sample was not contaminated by inferior vena caval blood). For antipyrine, the value of a fraction (f,) calculated as described below (Eq. 1) had to be >0.6. For 0, content, a similar fractional value was calculated as follows: (umbilical venous-hepatic 0, content difference)/ (umbilical venous-fetal arterial 0, content difference); 0.9 was considered an upper limit for this value. The data of three studies were excluded from the calculation of hepatic fluxes on the basis of these criteria. Chemicd analyses. Blood samples for determination of ‘*CO, were collected in oiled glass syringes lined with dried heparin, and whole blood ‘“CO, was measured as previously described (22). Blood samples for gas analyses were drawn into heparinized glass capillary tubes. Hemoglobin concentration and 0, saturation were measured immediately with an automatic direct-reading photometer (Radiometer OSM-2), and blood 0, content was calculated. Fetal blood samples for amino acid analyses, mass spectral isotope ratio analyses, and antipyrine concentrations were collected in plastic syringes lined with dried EDTA. Antipyrine was measured on whole blood utilizing the Technicon AutoAnalyzer (23). Blood samples for amino acid analyses and mass spectral isotope ratio analyses were immediately centrifuged at 4°C for 15 min, and the plasma was frozen at -70°C. Plasma amino acid concentrations were determined on supernatants after sulfosalicylic acid precipitation and pH adjustment to 2.2 with LiOH. Serine and glycine concentrations were determined with a JEOL-200A amino acid analyzer using a norleucine internal standard. Reproducibility of analyses was t2%. Fetal and placental tissue homogenates were acid hydrolyzed as previously described (14). After hydrolysis, the samples were dried, redissolved in buffer (pH 2.0), and counted in a Tri-Carb 460 C Hewlett-Packard scintillation counter. Radioactivity was corrected for quenching by an external standard method. Radioactive tracer serine and glycine in tissue were measured with a JEOL-6AH amino acid analyzer. Serine and glycine were separated by using a lithium buffer system, and the fractions containing the eluted amino acids were then analyzed by liquid scintillation counting. Muss spectral isotope ratio analysis. The mass spectral isotope ratio analysis was performed following the methodology

HINDLIMB,

AND

E787

PLACENTA

previously described for leucine (12) and glycine and serine (5). Briefly, amino acids were separated from plasma by ion-exchange chromatography and the amino acid-rich fraction was evaporated to dryness. t-Butyldimethylsiloxy derivatives were formed by adding 0.75 ml of a derivatizing reagent [15% N(t-butyldimethylsilyl)-N-methyltrifluoroacetamide, 1.5% t-butyldimethylchlorosilane imidazole, and 83.5% acetonitrile] to each evaporated fraction followed by sonication for 5 min to achieve complete mixing. The mixture was then incubated at 100°C for 2.5 h, transferred to standard Hewlett-Packard autosampler vials, sealed, and analyzed. Analyses were performed on a Hewlett-Packard autosampler 5890 gas chromatograph (GC) fitted with a 76731’ Autosampler and interfaced to a 5970 mass-selective detector and a Hewlett-Packard-9000 (series 300) work station. The column utilized was DB-1 (J & W Scientific, Folsom, CA), 12 m long, with head pressure of 0.1 lb/in.? The temperature conditions for the GC oven were an initial temperature of 100°C and a 2O”C/min ramp to a final temperature of 300°C. The injector was held at 25O”C, and the transfer line between the GC and the mass spectrometer was maintained at 280°C. A split-splitless mode interface was used to inject the samples with a split time of 0.2 min. Multiple ion detection was performed on the derivatized sample. Peaks for [ 12C]- and [ lsC]serine were recorded at m/z 390 and 391, respectively. Peaks for [ 12C] - and [ Ylglycine were recorded at 246 and 247, respectively. These peaks represent the loss of a t-butyl (m/z 57) moiety from the molecular ion of each amino acid. Triplicate analyses were performed on each sample, and the mean value was used in all further calculations. Calculations. Umbilical blood flow (Qf) was calculated according to the antipyrine steady-state transplacental diffusion method of Meschia et al. (16). 0, uptake and the net ‘*CO, loss via the umbilical circulation were calculated by application of the Fick principle. The umbilical uptake of serine was calculated as the product of blood flow x the umbilical arteriovenous difference in plasma serine concentration, since plasma and erythrocyte concentrations are tightly correlated (5). However, we have shown that glycine does not equilibrate freely in fetal blood, and the nonexchangeable pool within erythrocytes could lead to an overestimation of glycine umbilical uptake of -20% when plasma arteriovenous differences of glycine concentration are used instead of whole blood concentrations. Thus glycine umbilical uptake was calculated using the glycine distribution between erythrocytes and plasma previously reported (5) and the fetal hematocrit. The mole percent enrichments (MPE) of plasma serine and glycine (SerMr’” and GlyMpE) were calculated as the difference in peak area ratios between unenriched and enriched samples. Plasma tracer serine and glycine concentrations (Serl” and Glyl:j) were calculated as the product of SerMr’” and GlyMpE times serine and glycine plasma concentrations (Serc and Gly”), respectively. As previously described (5, 13), the fractions of umbilical venous (f,) and hepatic arterial and portal (f,) blood flow determining the hepatic input (hi) were calculated from antipyrine blood concentrations in the umbilical vein (AZ), the fetal artery (AZ’), and the hepatic vein (At) f, = (A; - A;)/(A;

- AZ)

(I)

and f, = 1 - f,

(2)

The mixed hepatic input concentrations of plasma serine and glycine (SerLi and Gly~i) of plasma tracer serine and glycine (Serk: and Gly$ and of blood ‘*CO, (14C02 hi) were calculated from the sum of the relative contributions from the f, and fa.

Downloaded from www.physiology.org/journal/ajpendo at Washington Univ (128.252.067.066) on February 15, 2019.

E788

FETAL

SERINE

FLUXES

IN LIVER,

The hepatic plasma concentration differences were calculated as the difference between the plasma concentrations in the hepatic input and in the hepatic vein (hi - hv). The hepatic input concentrations (Chi) were calculated as follows c,; = f, ’ c, + f,

l

c,

(3)

The fetal plasma serine disposal rate (DRlsser) was calculated using the rate of serine infusion into the fetal circulation (ryer), the enrichment of the infusate (Ser”P”), and the enrichment of serine in fetal arterial plasma (Ser,MP”) = [ ($er . Ser”PE) /Ser,MPE] _ rFer

DR’:‘Ser

(4)

The loss of tracer serine molecules (rfsFr) from the fetal circulation to the placenta was calculated as the product of Qf times the arteriovenous umbilical concentration difference of plasma tracer serine rcS;r 7 =

Qf

(Serilj - Seri”)

(5)

The rate of 14C0, production from tracer serine excreted via the umbilical circulation was measured as the product of &f times the concentration difference between umbilical arterial and venous 14C0, blood concentrations

HINDLIMB,

AND

PLACENTA

that acrossthe whole fetal carcassand that the carcassflow-toumbilical flow ratio was -0.9 (20) f carcass ox

=

[(14C0,,

-

‘“CO,,)

l

o.9]/(‘4co,

a

-

14C0,

“)

(10)

The fraction of fetal glycine derived from fetal plasmaserine (Gly/Ser) was calculated as the ratio between the fetal arterial plasma enrichment of glycine (GlyfpE) and the fetal arterial plasmaenrichment of serine (Ser,MpE). Statistics. Each samplewas analyzed in triplicate for MPE and blood 14C0,. The mean of the determinations was then calculated. For each study day, mean values from the four steady-state samplesobtained from all vesselswere usedto calculate all fluxes. Results from all experiments are reported as meanst SE. Paired Student’s t test was used to calculate the significance of concentration differences between vessels.Linear regression analysis was used to examine the correlation between fetal plasma serine arterial concentration and other measuredvariables. RESULTS

Mean gestational age was 127 t 2 days, fetal weight was 2,925 t 180 g, uterine blood flow was 1,393 t 164 ml/min, and umbilical blood flow was 208 t 14.8 ml/min. Fetal O2 uptake averaged 355.8 t 23.3 prnol. mine1 . kg-l. rSer = Qf . (‘“CO, a - 14C0, J 0 Serine and glycine plasma concentrations. Table 1 reports data for serine and glycine arterial plasma concenThe serine oxidation fraction (ftTer), representing the proheportion of infused tracer serine carbon excreted as 14C0, via the trations and their differences across the umbilical, patic, and hindlimb circulations and for serine and umbilical circulation, was defined as glycine umbilical uptakes. Samples from the fetal hepatic f 14Ser = r 14Ser lr f4Ser (7) vein were collected in six experiments simultaneously with samples from the fetal artery and umbilical vein. The production of CO, derived from the carbon of fetal The hepatic venous samples were from the right hepatic plasma serine (Rz;‘) was estimated as vein in two cases and from the left hepatic vein in the Rz3r = 3 (fzzer . DR’:‘Ser) remaining four. The relative contribution of umbilical (8) venous and fetal arterial and portal blood to hepatic blood The release of tracer CO, via the hepatic circulation in rela- flow was estimated from the antipyrine measurements tion to 0, uptake was calculated as the ratio between tracer CO, (Eqs. 1 and 2) and was used to calculate the hepatic input release and 0, uptake by the liver. The hepatic oxidation fracconcentrations. The concentrations of serine in umbilical tion (f ii,“‘), representing the fraction of tracer CO, produced by venous plasma were significantly lower than in umbilithe liver in relation to the release of tracer CO, by the whole cal arterial plasma with a significant uptake of serine by fetus, was calculated as the placenta (-2.08 t 0.5 pmol/kg). In contrast to serine, 0.20 glycine concentrations were significantly higher in umfliver = [V4C0, h” - 14C0, hi) / (0, hi (9) ox (14C0,, - 14C0,J / (0, ” - 0, J bilical venous than in arterial plasma, with a significant umbilical uptake (5.6 t 0.7 pmol/kg). In terms of hi-hv with 0.20 representing the estimated fraction of hepatic 0, differences there was a significant (P < 0.01) release of consumption in relation to fetal O2 consumption (3). serine by the fetal liver (-36.0 t 7.5 PM), concomitant to The fraction of tracer CO, produced by the fetal carcassin (P < 0.01) uptake of glycine (+24.1 t 5.7 relation to the production of tracer CO, by the whole fetus a significant PM). Both serine (17.4 t 7.7) and glycine (20.49 t 4.5) was estimated by assumingthat the arteriovenous dif(f iircass) ference of the tracer acrossthe hindlimb was representative of plasma arteriovenous concentration differences across OX

OX

OX

-

0,

hv)]

l

Table 1. Serine and glycine concentration differences across umbilical, hepatic, and hindlimb circulation and umbilical uptakes Serine, Arterial concn (4

hean *SE

P a, fetal artery;

766.8 10 t45.3

Umbilical diff (a - v)

10.26 10 t2.51 co.01 v, umbilical vein;

Glycine,

PM Hepatic diff (hi - hv)

Hindlimb diff (a - 4

-36.03 6 t7.46

17.41 6 27.72 CO.05

co.01 hi, hepatic

input;

hv, hepatic

Arterial concn (4

386.9 10 t22.4

Umbilical diff (a - 4

-32.87 10 t3.18