Main

IUGR is a serious problem in both the developing and developed world. Many of these cases are believed to be due to placental insufficiency, decreasing the transport of oxygen and nutrients to the fetus(1). If the placenta of an IUGR fetus has a reduced capacity to transport nutrients, then supplemental nutrition to the mother would be of marginal value. Direct supplementation of the fetus has not only practical limitations, but medical and biologic constraints as well(2). The alternative, therefore, is to develop therapies to promote growth of the IUGR infant during the neonatal period.

Proper postnatal nutrition is, of course, the primary method of promoting growth. However, long-term studies have shown that infants receiving appropriate nutrition show catch-up growth during the first 6 mo, but seldom reach the same size as appropriate-for-gestational age peers(3). This suggests that proper nutrition by itself is not sufficient to restore normal growth and indicates that the most advantageous time for therapy is during these first 6 mo.

IGF-I is believed to be a major regulator of neonatal growth. IUGR infants are born with lower circulating IGF-I concentrations(4, 5), and their growth rate during the first year of life is directly related to their circulating IGF-I concentration(6). Experimental induction of IUGR in rats(7, 8) is also associated with a reduced IGF-I concentration. Therefore, we hypothesized that providing IUGR neonates with supplemental IGF-I would stimulate growth. Increased rates of growth have been demonstrated in mice when endogenous IGF-I production is increased either by the production of transgenics(9) or through genetic selection(10). Provision of exogenous IGF-I to neonatal rats(11) also increases growth.

To test our hypothesis, we experimentally induced IUGR in pigs by protein restriction of sows during the last half of pregnancy. We have previously shown that these neonates are approximately 15% smaller at birth(12) and grow more slowly during the neonatal period than control pigs(13), but we had no information on their IGF-I status. Therefore, our first objective was to determine whether these IUGR pigs were born with reduced concentrations of IGF-I. Second, we wished to test the hypothesis that supplementation with IGF-I would stimulate growth in IUGR neonates.

METHODS

Eight cross-bred, multiparous sows were fed 2.3 kg/d of either a control(13% protein; 32 MJ digestible energy; n = 4) or protein-restricted(0.5% protein; 32 MJ digestible energy; n = 4) diet from d 63 of pregnancy to parturition(13). After parturition, all sows had unrestricted access to the control diet. All pigs in each litter remained on the sow until 3 d of age. At d 3, 6 individuals/litter were selected to receive osmotic minipumps and participate in the 7 d infusion experiment. All remaining pigs/litter were killed, with 1 randomly selected for in vivo protein synthesis measurements and all others for body composition and tissue measurements. Throughout this report, d 0 of the experiment will refer to pigs that were 3 d of age and d 7 of the experiment to pigs that were 10 d of age.

Day 0. Six pigs/litter were selected to receive osmotic minipumps, and one pig/litter was used to measure in vivo protein synthesis. The pigs that received osmotic minipumps were paired by body weight, with the three pairs representing individuals that had heavy, average, and light body weights for the litter. Each control litter remained on their dam for the 7 d of the study; each IUGR litter was fostered onto a different sow that had received the control diet throughout pregnancy and had farrowed within 24 h of the protein-restricted sow. This fostering, in combination with the reduction of the litter to six, was done to help ensure that none of the pigs experienced any nutritional limitation during the period of study.

IGF-I infusion. Three pigs/litter (one pig/body weight pair) received an osmotic minipump (model 2001; ALZA Corp., Palo Alto, CA) containing saline and three pigs/litter received the same minipump containing recombinant human IGF-I (Kabi Pharmacia, Stockholm, Sweden). The IGF-I was solubilized in 10 mmol/L acetic acid and diluted with saline. The IGF-I was delivered at a rate of 4 μg/h for 7 d. All pigs were anesthetized with 0.4 mg/kg body weight acepromazine maleate (Promace, Aveco Co., Inc., Fort Dodge, IA) and 20 mg/kg body weight ketamine hydrochloride (Ketaset, Aveco Co., Inc., Fort Dodge, IA) for implantation of the minipumps. A small incision was made above the left scapula, and the minipump was tunneled under the skin. The pigs were returned to their dam (or foster dam) after all had fully recovered from the anesthetic (3-4 h). The protocol was approved by the Animal Care and Use Committee of Baylor College of Medicine and was conducted in accordance with the National Research Council Guide for the Care and Use of Laboratory Animals.

Blood measurements. Body weights were measured, and blood samples (3 mL; jugular venipuncture) were collected immediately before surgery and on d 1, 4, and 7 postsurgery. The blood was centrifuged, and plasma was collected and frozen at -20 °C. IGF-I was measured in plasma as described by Lee et al.(14), separating the IGFBP using Sep-pak chromatography (Waters, Milford, MA). These values were verified in plasma of control and IUGR pigs by acid gel chromatography following the procedures of Frey et al.(15). IGF-I antiserum was obtained courtesy of Drs. L. Underwood and J. J. Van Wyk through the NIDDK and the National Hormone and Pituitary Program. Glucose was measured by the hexokinase reaction (Roche Diagnostics, Branchburg, NJ) and insulin by RIA (Linco, Inc; St. Louis, MO). IGF-II was measured by RIA after acid/ethanol extraction to remove the IGFBP (assay kindly performed by Dr. J. Veenhuizen, Monsanto Corp., St. Louis, MO). Triiodothyronine and thyroxine were measured by RIA (ICN, Costa Mesa, CA). IGFBP were measured by ligand blot using125 I-IGF-II as described by Lee et al.(14) and quantified by scanning laser densitometry.

Protein synthesis. A flooding dose of L-[2,3,4,5,6-3H]phenylalanine (Amersham, Arlington Heights, IL) was used to measure the in vivo protein synthetic rate in three pigs/litter: one d 0, one d 7 saline-infused, and one d 7 IGF-infused following the procedure of Garlick et al.(16). In all cases, a polyethylene catheter (PE-60; Fisher, Houston, TX) was implanted in the jugular vein with the pig under isoflurane anesthesia (Anaquest, Madison, WI). After recovery (approximately 1 h), the pig was fed mature sow milk (25 mL/kg of body weight) hourly for 3 h. The isotope (28 MBq·kg-1 of body weight in a 150 mmol/L phenylalanine solution at a dose of 10 mL/kg of body weight) was injected over 2 min through the jugular vein. Blood samples were collected at 5, 15, and 30 min postinjection for measurement of [3H]phenylalanine-specific activity, and the pig was killed with an i.v. dose of sodium pentobarbital at 30 min postinjection. Immediately thereafter, sections of the liver, jejunum, spleen, heart, and gastrocnemius and longissimus dorsi muscles were dissected, snap frozen in liquid nitrogen, and stored at -70 °C. The specific activity of [3H]phenylalanine was measured by anion exchange chromatography(17), and the tissue fractional synthesis rates were calculated as described by Burrin et al.(18).

The fractional rate of protein synthesis (FSR; percent of protein mass synthesized/d) was calculated as: FSR (%/d) = [(SB/SA) ×(1440/t)] × 100, where SB is the specific activity of the protein-bound phenylalanine, SA is the specific radioactivity of the tissue-free phenylalanine at the midpoint of the infusion based on the specific radioactivity of the tissue-free phenylalanine at the time of tissue collection and the linear regression of the blood specific activity of the animal at 5, 15, and 30 min against time, and t is the time of labeling in min.

Tissue measurements. All remaining pigs/litter were killed by an overdose of sodium pentobarbital (Nembutal, Abbott Laboratories, North Chicago, IL) and the liver, jejunum, ileum, stomach, spleen, heart, brain, pancreas, lungs, kidneys, gastrocnemius, and a section of the longissimus dorsi muscle were removed and weighed. Tissues from one half of the pigs were subsampled for measurement of protein(19), DNA(20), and dry matter. For each animal killed on d 7, we estimated tissue protein accretion as the increase in protein content between d 0 and d 7. The initial (d 0) protein content of pigs killed at d 7 was estimated by multiplying their individual initial body weight by the average tissue protein content (mg/kg of body weight) of littermates that were killed at d 0. All tissues were returned to the body cavity after subsampling, and carcasses were frozen at -20 °C for measurement of body composition. Frozen carcasses were ground to provide samples for proximate analysis (A& L Mid West Laboratories, Inc., Omaha, NE), with the percentages of protein, lipid, and ash determined. Accretion over the 7-d treatment period was estimated as the increase in tissue mass between d 0 and d 7. The d 0 value was calculated by multiplying individual pig body weight at d 0 by the average value from the proximate analysis at d 0.

Statistical analysis. Litter was the experimental unit for number of pigs/litter, and d 0 body weight, IGF-I concentration, tissue weights, tissue protein and DNA concentrations, protein:DNA ratios, and body composition (n = 4). Means were analyzed by a one-way ANOVA, with pregnancy diet the main effect. Growth rates and tissue accretion between d 0 and 7 and d 7 body weight, tissue weights, tissue protein and DNA concentrations, protein:DNA ratios, and body composition were compared by an ANOVA with litter, pregnancy diet, and infusion treatment the main effects. Litter was not significant in any comparison and was, therefore, removed from the model, and the data were reanalyzed. Sample size for body and tissue weights and growth rate was: control/saline, 10; control/IGF-I, 10; IUGR/saline, 10; and IUGR/IGF-I, 12. Sample size for tissue protein and DNA was: control/saline, 5; control/IGF-I; 5; IUGR/saline, 5; and IUGR/IGF-I, 6.

Hormone and metabolite concentrations were compared by an ANOVA corrected for repeated measures with litter, day of infusion, pregnancy diet, and infusion treatment the main effects. Litter was not significant in any comparison and was, therefore, removed from the model, and the data were reanalyzed. IGFBP were analyzed in d 0 and d 7 plasma only and were compared by an ANOVA with pregnancy diet and infusion treatment the main effects. Sample size for blood analyses was: control/saline, 10; control/IGF-I; 10; IUGR/saline, 10; and IUGR/IGF-I, 12.

The d 0 fractional protein synthetic rates for each tissue were analyzed by a one-way ANOVA, with pregnancy diet the main effect. Day 7 fractional protein synthetic rates were compared by an ANOVA with litter, pregnancy diet, and infusion treatment the main effects. Litter was not significant in any comparison and was, therefore, removed from the model, and the data were reanalyzed. Means were compared using Fisher's least significant difference test and a p value of <0.05 was considered statistically significant.

RESULTS

Litter size was not different between the control and protein-restricted sows (7.8 ± 1.2 versus 9.8 ± 1.6, p = 0.26; mean ± SEM). The control pigs were significantly heavier at d 0 of infusion than were the IUGR pigs (1.74 ± 0.06 versus 1.54± 0.07 kg, p = 0.04). Body composition (%) was not different between control and IUGR pigs at d 0 (protein: 14.7 ± 0.3versus 14.2 ± 0.2, p = 0.54; fat: 3.1 ± 0.3versus 4.0 ± 0.4, p = 0.25; ash: 3.9 ± 0.1versus 3.9 ± 0.2, p = 0.88), but total body protein mass was greater in the control than the IUGR pigs (256 ± 11versus 219 ± 17, p = 0.05). IGF-I concentrations at d 0 were also significantly greater in the control than the IUGR pigs (20.4± 2.5 versus 12.9 ± 1.3 ng/mL, p = 0.01).

During the 7-d infusion period, the IGF-I concentration of all pigs increased (Fig. 1). The control/IGF-infused pigs had a significantly greater circulating concentration of IGF-I than all other pigs at all postinfusion time points. The IUGR/IGF-infused pigs and the control/saline-infused pigs had similar IGF-I concentrations throughout the 7 d of infusion. The IUGR/saline-infused pigs had significantly lower IGF-I concentrations than all other pigs at d 1 and d 4; however, by d 7 their IGF-I concentration was similar to that of the IUGR/IGF-infused and control/saline-infused pigs.

Figure 1
figure 1

Plasma IGF-I concentration of control (C) and IUGR neonatal pigs after receiving a continuous infusion of saline(Sal) or IGF-I from d 0 to d 7. Data were analyzed by an ANOVA corrected for repeated measures. C-IGF pigs had significantly greater concentrations than all other pigs at all time points; IUGR-Sal pigs had significantly lower concentrations than all other pigs at d 1 and d 4.

Full size image

Body weight at d 7 and weight gain (g/d) during the 7 d of infusion were not statistically different due to pregnancy diet or infusion treatment(Table 1). However, when the data were calculated as the percentage weight change from d 0 to d 7, both pregnancy diet and infusion treatment altered the rate of gain. The relative rate of weight gain in the IUGR pigs was significantly lower than that of the control pigs during the 7 d of study, and the infusion of IGF-I significantly increased the rate of weight gain approximately 10% in both the control and IUGR pigs. Weight gain of the IUGR/IGF-infused pigs was not significantly different from that of the control/saline-infused pigs.

Table 1 Body weight at d 7 and growth rate (g/d and% increase from d 0 to d 7) after receiving a continuous infusion of saline or IGF-I from d 0 to d 7 in control and IUGR neonatal pigs
Full size table

The composition of this gain was affected by both pregnancy diet and IGF-I infusion. Body composition on a percentage basis was not different among groups (Table 2); however, total body mass of protein(p = 0.05) and fat (p = 0.04) were greater in the control than in the IUGR pigs. There was a significant interaction of pregnancy diet and infusion treatment for the estimated accretion of both protein and fat(Table 2). This occurred because the IGF-I infusion significantly increased protein and fat accretion in the IUGR pigs, but not in the control pigs. Ash accretion was significantly greater in the control than the IUGR pigs and was significantly increased by the IGF-I infusion(Table 2).

Table 2 Body composition (% wet weight), total mass (g) at d 7, and the estimated accretion (g) between d 0 and d 7 of a continuous infusion of saline or IGF-I for control and IUGR neonatal pigs
Full size table

The weight, protein concentration, DNA concentration, and protein/DNA ratio were not significantly different in control and IUGR pigs in any of the sampled tissues at d 0 or d 7 (data not shown). However, when tissue protein accretion was estimated, pregnancy diet and infusion treatment were significant for the gastrocnemius muscle, with protein accretion being greater in the control pigs than in the IUGR pigs and increased by the IGF-I infusion(Table 3). The interaction of pregnancy diet and infusion treatment was significant in the liver, jejunum, and spleen because IGF-I infusion significantly increased protein accretion in the IUGR pigs, but not in the control pigs (Table 3). The increased protein accretion due to the IGF-I infusion was not associated with a significant increase in fractional protein synthetic rate; however, the fractional protein synthetic rate of both the liver and gastrocnemius were significantly greater in the IUGR than in control pigs (Table 3).

Table 3 The estimated accretion of protein between d 0 and d 7 and protein synthetic rate at d 7 of a continuous infusion of saline of IGF-I for control and IUGR neonatal pigs
Full size table

The infusion of IGF-I did not affect the circulating concentration of glucose, insulin, IGF-II, triiodothyronine, or thyroxine (Fig 2). At d 7, there was a significant increase in circulating glucose and insulin in the IUGR pigs compared with the control pigs (Fig. 2, a and b). In addition, the plasma triiodothyronine concentration at d 0 in control pigs was greater than that in IUGR pigs (Fig. 2d). The circulating concentrations of IGF-II declined and thyroxine increased between d 0 and d 7 in all pigs (Fig. 2, c and e).

Figure 2
figure 2

Plasma glucose (a), insulin (b), IGF-II (c), triiodothyronine (d), and thyroxine(e) of control and IUGR neonatal pigs after receiving a continuous infusion of saline of IGF-I from d 0 to d 7. Data were analyzed by an ANOVA corrected for repeated measures. The infusion of IGF-I had no effect on the circulating concentration of glucose or any of these hormones.

Full size image

The abundance of IGFBP-2 and IGFBP-3 was similar in control and IUGR pigs at d 0 (Fig. 3). By d 7, the abundance of IGFBP-2 in the control pigs had risen slightly, but not significantly (d 0 versus d 7, p = 0.10), and remained unchanged in the IUGR pigs (d 0versus d 7, p = 0.83). However, at d 7, the IGFBP-2 abundance was significantly greater in the control than in the IUGR pigs, but the infusion of IGF-I had no effect in either group (Fig. 3a). In contrast, IGFBP-3 abundance was significantly greater in both pregnancy diet groups by d 7, with a greater increase measured in the control than the IUGR pigs (Fig. 3b). The infusion of IGF-I also significantly increased IGFBP-3 abundance, with no significant interaction between the infusion of IGF-I and the dietary treatment during pregnancy (Fig. 3b). The bands for IGFBP-1 and BP-4 on the gel were extremely light and not quantifiable.

Figure 3
figure 3

The abundance of IGFBP-2 (a) and BP-3(b) in control and IUGR neonatal pigs before (d 0) and after (d 7) receiving a continuous infusion of IGF-I from d 0 to d 7. Data were analyzed by ANOVA. The infusion of IGF-I had no effect on IGFBP-2, but significantly increased the abundance of IGFBP-3.

Full size image

DISCUSSION

This study was designed to test the hypothesis that supplemental IGF-I stimulates growth in IUGR neonates such that they grow similarly to control neonates. Throughout the report we have been using the term growth as synonymous with an increase in tissue mass, not a linear increase. By comparing the growth of the IUGR/IGF-infused pigs with that of the control/saline-infused pigs, we found that the IGF-I infusion normalized the circulating IGF-I concentrations and absolute growth rate. This increased growth rate was due to increases in both protein and fat accretion, with much of the protein accretion occurring in the visceral tissues.

Intrauterine growth retardation. We induced intrauterine growth retardation experimentally by feeding sows a protein-restricted diet from d 63 of pregnancy to parturition. These sows produced pigs that weighed approximately 10% less at 3 d of age than sows fed a control diet. This weight reduction was comparable to that seen in previous studies(12). Plasma IGF-I concentration at birth was also reduced approximately 30-35% in the IUGR pigs, similar to the reductions seen in IUGR infants(4).

The IUGR neonates grew more slowly and maintained lower IGF-I concentrations than the control neonates during the week of study. This slower rate of growth is also seen in IUGR infants and is associated with reduced circulating IGF-I concentrations(6). Plasma IGF-I concentrations increased during the 7 d of study in both the control and IUGR neonates, consistent with the normal developmental rise reported previously for swine(14). Even though the circulating IGF-I concentration was lower at d 0 in the IUGR than in control neonates, this difference was not observed by d 7 of the study. This suggests that, under conditions of sufficient nutrition, IUGR neonates are able to attain normal circulating IGF-I concentrations, but that there is a delay in reaching these concentrations. Feeding a protein-restricted diet to growing rats decreases IGF-I concentrations due to both a decrease in IGF-I production and an increase in IGF-I clearance(21), so these mechanisms may have been functioning in these pigs.

Measurement of tissue protein synthetic rates provided evidence that the IUGR pigs were receiving sufficient nutrition, possibly owing to increased feed intake, during the 7-d study period. Because changes in feed intake acutely affect translational efficiency and chronically affect protein synthetic capacity in some tissues(18, 22), the increased protein synthetic rate in the liver of IUGR pigs at d 0 may reflect a stimulation of protein synthesis by increased feed intake.

IGF-I infusion. The IGF-I infusion increased circulating IGF-I concentrations throughout the 7 d of study in control and IUGR groups, with no effect on the circulating concentration of insulin, IGF-II, or the thyroid hormones. On d 7, however, whereas the control pigs infused with IGF-I continued to have elevated IGF-I concentrations, all IUGR pigs had similar IGF-I concentrations. It is not clear why this occurred in only the IUGR group, but we suggest that the infusion of IGF-I decreased endogenous production and/or increased IGF-I clearance in the IUGR pigs, as was measured in protein-restricted rats(21). Although IGF-I infusion increased IGFBP-3 concentrations in both control and IUGR groups, consistent with previous reports(23, 24), the generally lower BP-3 concentrations in IUGR pigs may have contributed to a lower half-life and, hence, a lower circulating IGF-I concentration in these pigs.

The rate of IGF-I infusion was a concern owing to the possibility that high doses would cause hypoglycemia(25). We chose an infusion rate that we believed would produce a plasma IGF-I concentration below this threshold, and this is supported by the normal circulating concentrations of both glucose and insulin measured in the infused pigs. The sow generally allows the pigs to suckle every 60-90 min(26), and this relatively constant intake of nutrients may have also helped the pigs to maintain euglycemia.

The provision of supplemental IGF-I increased the fractional weight gain by approximately 10% in both the control and IUGR neonates. The positive response by the IUGR neonates to IGF-I was somewhat expected, because their endogenous circulating IGF-I concentration was lower, but the similar response of the control pigs to elevated circulating IGF-I concentrations was unexpected. These results agree, in general, with those of several other investigators who have increased endogenous IGF-I by a variety of methods. However, organ hypertrophy did not occur in our pigs, whereas other authors have reported specific increases in the liver and heart(11), kidney and spleen(9, 27), brain(9, 11, 28), and pancreas(9) owing to an increased circulating concentration of IGF-I.

A number of measurements (tissue protein, DNA, protein synthesis, body composition) were made to determine what component of weight gain was stimulated by IGF-I. Total body protein, fat, and ash mass were not significantly affected by the IGF-I infusion; however, protein and fat accretion was 15-20% greater and ash 30% greater in the IGF-infused pigs. The largest increase was seen in the IUGR pigs, with IGF-I increasing protein and fat accretion 53% and ash 45% above IUGR pigs receiving the saline infusion. Tissue fractional protein synthetic rate of IGF-I-infused pigs was also numerically increased across all tissues, but not statistically significant. Small changes in the rate of protein synthesis can, however, result in large changes in protein accretion over time(29). Increased skeletal muscle and hepatic protein synthesis in response to IGF-I infusion have been reported in lambs(30), and it has been suggested that the endocrine role of IGF-I may be the regulation of whole-body protein metabolism(31). The increased protein accretion in IUGR/IGF-infused pigs was likely due to either a decrease in protein degradation or an increased protein synthesis earlier in the 7-d period.

Because the pigs were returned to the sow during the 7 d of study, we did not measure feed intake and cannot determine if the increase in growth was due to an increase in intake or an increase in growth efficiency. However, we purposely reduced each litter to six to minimize any limitation on nutrient intake, allowing the pigs to increase intake if needed. It is possible that the changes we have documented in the IGF-infused pigs could be explained by the IGF causing an increased intake, providing the IUGR pigs with additional nutrients. This may be further supported by the increased circulating insulin and glucose concentrations in both IUGR groups.

We conclude from these data that IGF-I infusion resulted in a modest increase in weight gain in both control and IUGR neonates. Moreover, provision of IGF-I to IUGR neonatal pigs was able to restore body weight and composition to normal within 7 d. The increased rate of growth was due to an increased rate of protein and fat accretion. This was, however, a short-term study, and whether these stimulatory effects would be maintained or compound during periods of chronic treatment is not known. However, these data suggest that IGF-I may be a potential growth promotant in neonatal pigs, with possible uses as a therapeutic agent for IUGR infants.