Vol. 281, Issue 5, E1063-E1072, November 2001
Parenteral nutrition with lipid or glucose suppresses liver
growth and response to GH in adolescent male rats
Andre
Sevette1,
Anthony J.
Kee1,
Anthony R.
Carlsson1,
Robert C.
Baxter2, and
Ross C.
Smith1
1 Department of Surgery, University of Sydney, and
2 The Kolling Institute of Medical Research, Royal North
Shore Hospital, Sydney, New South Wales 2065, Australia
 |
ABSTRACT |
Our aim was to
investigate the effects of modifying the carbohydrate-to-lipid ratio of
parenteral nutrition (PN) on body composition and the anabolic actions
of insulin-like growth factor I (IGF-I) and growth hormone (GH).
Adolescent male Sprague-Dawley rats were randomized to receive 7 days
of GH, IGF-I (3.5 mg · kg
1 · day1 for both)
or placebo while receiving high-carbohydrate PN (CHO-PN), high-lipid PN
(L-PN), or an oral diet (chow) (the PN protocols were isonitrogenous
and isocaloric). PN impaired muscle growth, which was reversed by GH in
the CHO-PN group only (P < 0.03). PN increased carcass
lipid (P < 0.02), the effect being greater in the L-PN
than in the CHO-PN group (P < 0.001). Visceral lean tissue growth was significantly impaired by PN (P < 0.001). IGF-I reversed this impairment, but GH had no effect. PN
impaired the normal increase in hepatic protein and DNA
(P < 0.001) and produced liver steatosis
(P < 0.001). However, this steatosis was less in L-PN
than in CHO-PN (P < 0.001). Serum IGF-I and the
acid-labile subunit (ALS) were decreased by PN (P < 0.001) and were not affected by GH during PN treatment. However, GH
significantly increased serum ALS concentrations in the chow-fed rats
(P = 0.032). In conclusion, modifying the CHO-to-L
ratio of PN had no significant effect on IGF-I action, but CHO-PN
increased the peripheral effect of GH. L-PN increased carcass lipid
significantly and decreased hepatic steatosis. Nevertheless, PN caused
significant liver steatosis and profound impairment of hepatic cell
growth, which was associated with relative hepatic GH resistance.
growth hormone; parenteral nutrition; organ composition; body
composition; insulin-like growth factor I treatment; steatosis; liver
impairment
| |
INTRODUCTION |
MALNUTRITION AND THE
LOSS of lean body mass are significant factors leading to poor
outcome in surgical and critically ill patients (1). Thus
parenteral nutrition (PN) remains an important therapy for many of
these patients who are unable to adequately nourish themselves by oral
nutrition. However, PN is plagued by a number of problems that limit
its wider use. These problems include atrophy of the intestinal mucosa
(31), impaired immune function (11), hepatic
dysfunction (38), and poor efficacy compared with oral
nutrition to promote growth and tissue restitution (30).
The poor efficacy of PN is compounded by the fact that many of the
patients requiring PN are stressed and therefore resistant to the
anabolic effect of nutrition on protein metabolism.
There has been much interest in the use of insulin-like growth factor I
(IGF-I) and growth hormone (GH) to improve the efficacy of PN and to
reduce the net loss of nitrogen in stressed patients. However, the
early promise of GH treatment in moderately stressed patients
(32, 36) has not been generally observed in more-stressed individuals (44). In the early stages of critical illness,
IGF-I levels are low despite the elevated GH concentrations. It is this suppression of IGF-I secretion that is thought to attenuate the protein
anabolic effects of GH in these conditions. Consequently, it has been
suggested that IGF-I may be more effective than GH in attenuating the
loss of muscle protein. Although studies in stress-induced states
(i.e., dexamethasone treatment and starvation) in rats (29,
43) and humans (8) indicate that IGF-I has protein-anabolic actions, this has not been consistently observed in
critically ill patients (44). Although the mechanism for this is still unknown, it is thought to be due to changes at the receptor or postreceptor level (17). There are a number of
studies suggesting that GH plus IGF-I may be more anabolic and have
fewer side effects than either hormone alone (23, 28).
In a recent study, we observed that young adolescent rats maintained on
PN were resistant to the anabolic effects of IGF-I on skeletal muscle
protein (21). This was despite a complete suppression of
the PN-induced visceral atrophy. The cause of this peripheral
resistance was not identified in this study, but the relatively high
lipid intake of these rats may have contributed (27).
Alternatively, the relatively high insulin concentrations that occur
during PN may have antagonized the peripheral effects of IGF-I.
High-carbohydrate PN is also associated with significant hepatic
dysfunction (e.g., cholestasis, steatosis), and this may contribute to
the decreased concentrations of liver-derived IGF-I and the acid-labile
subunit (ALS) [part of the 150-kDa IGF/IGF-binding protein-3 (IGFBP-3)
ternary complex] that occurs during PN treatment (21).
Increasing the lipid-to-carbohydrate ratio of PN is thought to reduce
its hepatotoxic effects and may attenuate development of hepatic GH
resistance and increase the actions of exogenous GH. Therefore, in the
present study, we have examined the influence of the
lipid-to-carbohydrate ratio on the peripheral and visceral actions of
exogenous IGF-I and GH. This has significant clinical implications, as
there is some debate about the relative benefits of "lipid"- vs.
"carbohydrate"-based PN.
| |
METHODS |
Recombinant human GH.
Recombinant human GH (rhGH) was kindly supplied by Pharmacia & Upjohn
(Stockholm, Sweden). It was administered subcutaneously, twice daily
(in the morning and evening) at a dose of 3.5 mg · kg body
wt1 · day1, based on the weight of
the rats at the start of PN infusion. Saline (placebo) injections were
given to rats not treated with GH.
Recombinant human IGF-I.
Recombinant human IGF-I (rhIGF-I; GroPep, Adelaide, Australia) was
dissolved in 0.1 M HCl, diluted with bovine serum albumin (1 mg/ml),
and added to fresh PN solution each day for continuous intravenous
infusion, as previously described (21). The dose of
rhIGF-I was 3.5 mg/kg body wt daily, based on the weight of the animals
at the start of treatment (day 0).
These doses of rhIGF-I and rhGH were found to be anabolic by other
investigators (28).
Experimental protocol.
The study design was approved by the Animal Care and Ethics Committee
of Royal North Shore Hospital and University of Technology, Sydney.
Male Sprague-Dawley rats were received (Gore Hill Animal Research
Laboratory, University of Technology, Sydney, Australia) at 4 wk of age
(75-90 g) and placed individually in metabolic cages in a light-
(12:12-h light-dark cycle) and temperature- (23-25°C) controlled
environment. The animals were then acclimatized for ~7 days in their
metabolic cages, during which time they were given free access to water
and rat chow (Gordon's Specialty Stock Food, Australia, providing 20%
protein, 6% fat, and 5% crude fiber and 12.0 MJ/kg of metabolizable
energy). When the animals reached 140-150 g, a catheter
was implanted aseptically into the superior vena cava through the right
external jugular vein, as described previously (22). This
method has been shown to provide PN to rats, free of sepsis, for
7-10 days.
After surgery, the animals were allowed 2 days of postoperative
recovery, when they received (0.8 ml/h) intravenous isotonic saline
infusions and were given continued access to chow and water.
On the morning of the 2nd postoperative day, the animals were
randomized to receive 75% lipid-25% carbohydrate parenteral nutrition
(L-PN), 5% lipid-95% carbohydrate PN (CHO-PN), or ad libitum chow
(Chow). Within each nutritional group, there were three subgroups
receiving GH, IGF-I, or placebo treatment. A baseline group of rats
were also included, which were killed on the 3rd postoperative day to
assess the changes in hormone levels and body and organ composition
with treatment. Both the chow-fed and baseline groups were infused with
isotonic saline at the same rate as the PN infusion to control for the
stress associated with intravenous infusion. PN was introduced in the
morning of the 2nd postoperative day, at one-half of the target rate,
for 1 day and thereafter for 7 consecutive days at the target rate of
PN infusion. Chow was withdrawn at the start of target rate PN
infusion. Animals were monitored daily for their body weight, urinary
and fecal output, and water and food (control groups) consumption.
PN solutions.
The PN solutions were prepared aseptically daily in a laminar flow
hood. The L-PN consisted of a 2:1:3 mixture of Synthamin 17 (Baxter
HealthCare, Sydney, Australia), 50% glucose (Baxter HealthCare), and
20% Intralipid (Kabi Pharmacia, Stockholm, Sweden). The CHO-PN was
made of a 20:38:2 mixture of Synthamin 17, 50% glucose, and 20%
Intralipid. Minerals and trace elements were added to all PN solutions,
and essential vitamins were supplemented on day 5 of PN.
The PN infusion started on postoperative day 2, at one-half
the target rate for the 1st day and thereafter for 7 consecutive days
at the target rate of infusion. The rate was adjusted according to the
body weight of the animal every morning. The PN solutions provided
daily requirements of energy (1.3 MJ · kg body
wt1 · day1), amino acid nitrogen
(1.27 g N · kg body
wt1 · day1), essential fatty acids,
minerals, and trace elements. The infusions of all the treatment groups
(saline/chow and all PN groups) were isovolemic (230 ml · kg
body wt1 · day1).
Animals were killed by intravenous pentobarbitone overdose (Nembutal,
Boehringer Ingelheim, Australia) on postoperative day 10.
After the animals were killed, 5-8 ml of blood were taken by
cardiac puncture and placed immediately on ice. The time from euthanasia until blood collection never exceeded 1 min. The catheters, silicone anchor plates, and jackets were removed from the animals, and
their weight was recorded. The liver, kidneys, lungs, heart, thymus,
stomach, small intestine, cecum, colon, spleen, and testes and right
gastrocnemius and soleus muscles were removed. The contents of the
intestine and stomach were washed out with isotonic saline. A length of
~25 cm of the terminal ileum was incised longitudinally, and the
mucosa was stripped from the serosa with the edge of a glass slide. All
of the organs were weighed and snap frozen in liquid nitrogen
immediately after their harvest. The contents of the bladder and the
intestines were recorded, and the eviscerated carcass weight was
calculated. The same person performed all scrapings and dissections to
reduce interoperator variation. Total visceral weight was estimated
from the weight of the heart, lungs, thymus, stomach, the whole of
large and small intestines, liver, kidneys, spleen, bladder, and testes.
Organ and carcass analysis.
The organs were minced and homogenized in deionized water by means of a
Polytron Tissue Homogenizer (model PT10St "OD" S, Kinematica,
Lucerne, Switzerland). The eviscerated carcass (consisting of all
remaining tissues after the aforementioned viscera were removed and
comprising the head, brain, tail, skin, feet, muscles, and bones) was
minced using a domestic kitchen mincer and was homogenized in ~400 ml
of isotonic saline with a blender (model BLE-37, Breville, Sydney,
Australia). Aliquots of organ and eviscerated carcass homogenates were
taken for total water, lipid, and nitrogen determination, as previously
described (22). Total protein content was calculated by
multiplying total nitrogen content by 6.25. The DNA content of the
liver was determined using the ethidium bromide method described by
Prasad et al. (37).
The body composition of the rats at the start of treatment was
estimated using their body weight and the average composition of the
reference group. By use of this value and the composition at end point,
the change in body composition was calculated.
Serum glucose was determined using a commercially available kit based
on the glucose oxidase method (Peridochrom, Boehringer Mannheim,
Sydney, Australia). Serum nonesterified fatty acids (NEFA) were
measured by an in vitro enzymatic (acyl-coenzyme A synthetase-acyl-coenzyme A oxidase) colorimetric method (Wako NEFA C
kit, Osaka, Japan). Serum insulin was determined by a rat-specific radioimmunoassay (RIA; Linco Research, St. Charles, MO). Rat serum IGFBP-1, ALS, and total IGF-I (endogenous rat IGF-I + rhIGF-I) concentrations were determined by an in-house RIA assay, as previously described (21, 22).
Statistical analysis.
Results are presented as means ± SE. The Levene's test was
performed to assure homogeneity of variance together with the
Kolmogorov-Smirnov (Lilliefors) test to check for normality of the data
(41). If the above tests were not satisfied, then the
results were normalized by logarithmic or square root transformation.
With the use of a fractional 2 × 2 × 3 factorial design
(5), with the factors being IGF-I, GH, and nutrition
(L-PN, CHO-PN, and Chow), ANOVA of means and interactions of factors
were examined (SPSS 8.0 software, Chicago, IL). If statistically
significant differences were detected by ANOVA, individual comparisons
were made between groups with the post hoc Fisher's least significant difference (LSD) test (
-level = 0.05).
Daily weights were analyzed by repeated-measures ANOVA, and again the
post hoc test was used to evaluate differences among specific means
(-level = 0.05).
| |
RESULTS |
Eighty-eight rats were required to complete the study, with a
total dropout rate of 20% (n = 18). Of these, 5 dropouts were perioperative and related to the anesthetic procedure, 2 were due to equipment failure, and only 11 were true dropouts related to the treatment protocol. The dropout rate was 9% in the saline and
L-PN groups and 25% in the CHO-PN group. The threefold greater dropout
rate in the last group was probably due to the higher osmolality of the
CHO-PN solution, resulting in more phlebitis and catheter blockage.
Body weight.
Daily body weights are shown in Fig. 1.
There was no difference among groups in the starting weight of the
animals at the time of catheterization (day 0).
View larger version (25K):
[in this window]
[in a new window]
|
Fig. 1.
Body weight. Chow, chow fed; PN, parenteral nutrition;
CHO, high carbohydrate; L, high lipid; GH, growth hormone; IGF,
insulin-like growth factor I. Values are means ± SE;
n = 7 for all groups, except for the
postcatheterization recovery period (days 1-3), where
the groups were pooled together. The least significant difference (LSD)
post hoc test was used to compare the mean weights of each group on
day 10:  P = 0.024, Chow vs. Chow + GH;  P = 0.053, CHO-PN vs. CHO-PN + IGF. Means
were also compared by repeated-measures ANOVA, and when pooled,
only the between-subject effects of IGF-I and oral nutrition were
significant (P = 0.046 and P < 0.001, respectively).
|
|
From the start of treatment (day 3), orally fed animals
gained weight at a constant rate, whereas for the parenterally fed animals there was an initial stunting of growth (1-2 days)
followed by a period of weight gain. This weight gain, however, was
significantly slower than that of the orally fed animals
(P < 0.001, repeated-measures ANOVA). There was no
significant difference in body weight on each day for the PN groups. GH
significantly increased the weight of the animals only in the orally
fed group (LSD test, P = 0.024). This was also true
when the results were analyzed by repeated-measures ANOVA for the
orally fed groups (P = 0.043). On the other hand, IGF-I
treatment had no significant effect on the end point body weight, but
there was a trend toward an increase in the high CHO-PN group with
IGF-I treatment (P = 0.053).
Peripheral tissue composition.
The weight and composition of the carcass and peripheral muscles (right
soleus and gastrocnemius) are shown in Table
1 and Fig.
2. Wet eviscerated carcass weight was
increased in all groups compared with the baseline group (reference
group, Table 1). However, this weight gain was greater in the orally
fed compared with the parenterally fed rats (P < 0.001; Fig. 2). Carcass protein was also significantly higher in the
orally vs. parenterally fed animals (P < 0.001; Fig.
2). When the GH groups were pooled, GH significantly increased the
estimated "change in carcass weight" (P = 0.001;
Fig. 2), but on post hoc analysis, it only increased the change in
carcass weight of the orally fed animals (LSD test, P = 0.005; Fig. 2). There was no effect of IGF-I or GH on the protein or
lipid content of the carcass.
View larger version (38K):
[in this window]
[in a new window]
|
Fig. 2.
Change in carcass weight and composition with treatment
(days 3-10). Values are means ± SE;
n = 7 for each group. *Significant difference between
L-PN and CHO-PN (P < 0.001). Significant effect of
GH at each nutritional group (P = 0.005). §Significant
difference between oral nutrition and PN (P = 0.034).
When GH groups were pooled, there was also an overall effect of GH on
the change in carcass weight with treatment (P = 0.001).
|
|
Pooled together, GH overall increased the lean tissue (weight, water,
and protein content) of gastrocnemius muscle, but on post hoc anlysis,
GH increased lean tissue weight only in the CHO-PN group (P
0.026) and the protein content of the orally fed group
(P < 0.001). IGF-I had no effect on gastrocnemius and soleus weight or composition (Table 1).
Oral feeding also significantly increased the protein content of
gastrocnemius compared with PN (P 0.006; Table 1).
There was an interaction between IGF-I and nutrition for gastrocnemius
weight and water content. This was due to a greater increase in weight
and water content in the PN groups than in the oral groups. There was
also a positive interaction between GH and nutrition for soleus weight
and for gastrocnemius weight and water content. This was due to a
greater increase in the CHO-PN with GH treatment compared with the L-PN
and oral groups.
Nonhepatic visceral composition.
The weight and composition of the nonhepatic viscera are shown in Table
2 (the results for the whole small
intestine were similar to those for small intestine mucosa; hence, for
simplicity, only the latter is included in the table). PN prevented the
normal gain in the visceral tissues (P < 0.001). IGF-I
attenuated this relative loss of visceral weight; however, the effect
of IGF-I was more pronounced in the CHO-PN group, where visceral weight gain was similar to that in the chow-fed controls, whereas in the L-PN
it was still significantly lower (P = 0.006). IGF-I
also significantly increased total visceral weight (P < 0.001) in the orally fed group. GH had no significant effect on the
total visceral weight in any of the groups.
Growth of lean tissue (water and protein) in the individual visceral
organs was also significantly impaired by PN (P < 0.001). In the thymus and small intestine (whole small intestine as
well as the mucosa), there was loss of lean tissue with PN, and in the
small intestine this loss was statistically significant (P 0.010). In the kidney and spleen, although lean tissue was not lost compared with the reference group, there was impaired growth compared with the chow-fed animals (P < 0.001).
When the GH groups were combined (factorial analysis), there was no
overall effect of GH on individual visceral composition except in the
spleen, where GH significantly increased lean tissue (P 0.006), but this was probably due to the strong effect of GH in the
orally fed group (Table 2).
There was an interaction of IGF-I with nutrition on the weight and
protein content of the spleen and the thymus, there being a greater
increase with IGF-I treatment in the Chow vs. the PN groups. In the
same organs, there was also an interaction of GH with nutrition due to
a significant increase in weight, water, and protein content with GH
treatment in the orally fed group (P 0.020), but not in
the PN groups. On post hoc analysis, IGF-I significantly increased the
lean tissue of the small intestine, the kidneys, and the spleen in all
three nutritional groups, but for the thymus this only occurred in the
CHO-PN and orally fed groups (Table 2).
GH had a much more modest effect on lean visceral tissue than IGF-I and
increased the lean tissue weight of the thymus and spleen only in the
chow-fed group (P 0.020). It also produced a
significant decrease in the protein content of the spleen and the
weight and water content of the thymus for the L-PN group (P
0.043).
Liver composition.
The liver composition at end point and the change in composition with
treatment are presented in Table 3 and
Fig. 3, respectively. Although there
was an increase in the liver weight in all groups, this gain was
significantly less in the PN groups compared with the orally fed group
(P < 0.001). Furthermore, in the orally fed animals,
this growth was due to lean tissue (protein and water) gain, whereas in
the parenterally fed animals, it was due to lipid and water gain. The
increase in DNA content of the liver, a marker of cell number, was also
significantly greater in the chow-fed than in the PN-fed animals
(P < 0.001), and there was no overall effect of
hormonal intervention (Table 3). In contrast to the orally fed group,
there was a significant increase in liver lipid for both PN protocols
(P < 0.001) and a significant decrease in protein gain
(P < 0.001). The CHO-PN group had also significantly higher hepatic lipid content than the L-PN group (P < 0.001).
View larger version (42K):
[in this window]
[in a new window]
|
Fig. 3.
Change in liver weight and composition with treatment
(days 3-10). Values are means ± SE;
n = 7 for each group, except for lipid content of
CHO-PN and CHO-PN + IGF groups, where n = 6, and
the CHO-PN + GH group, where n = 5. *Significant
difference between L-PN and CHO-PN (P < 0.001).
§Significant difference between oral nutrition and PN
(P < 0.001).
|
|
When the nutritional groups were combined, the only significant effect
of IGF-I was to decrease the lipid content of the liver (P = 0.021). There was no overall effect of GH on the
liver weight or composition, but on post hoc analysis, GH decreased
liver lipid in the chow-fed group (P = 0.025).
Serum parameters.
Serum insulin, IGF-I, IGFBP-1, and ALS are shown in Fig.
4 and NEFA in Table
4.
View larger version (41K):
[in this window]
[in a new window]
|
Fig. 4.
Serum IGF-I, its binding proteins, and insulin levels.
IGFBP-1, IGF binding protein-1; ALS, acid-labile subunit. Values
are means ± SE; n = 6-7 for each group,
except for the CHO-PN group, where n = 4 for IGFBP-1
results, and n = 5 for the Insulin and IGF-I results.
Results were normalized either by logarithmic (ln) or by square root
(SqR) transformation, before statistical analyses. *Significant
difference between the L-PN and CHO-PN (P 0.004).
Significant effect of GH (P 0.032). Significant
effect of IGF-I (P 0.002) on each nutritional group.
§Significant difference between oral nutrition and PN (P 0.034).
|
|
CHO-PN led to higher serum insulin levels than L-PN
(P < 0.001), and for both PN regimens these levels
were elevated compared with the chow-fed controls (P 0.034). Overall, there was no effect of GH on insulin levels, but there
was a significant decrease in insulin levels with IGF-I treatment
(P = 0.020). There was an interaction between
nutrition and IGF-I on insulin levels, this being due to a smaller
decrease in insulin levels with IGF-I treatment in the CHO-PN group.
Similarly, there was an interaction between GH and nutrition, but in
this case, the interaction reflected a greater increase in insulin
levels in the CHO-PN compared with the chow-fed group and decreased
levels in the L-PN group with GH treatment.
IGF-I levels were higher in the orally fed animals than in the
parenterally fed animals (P < 0.001), but there was no
difference between the two PN groups. Overall, total IGF-I levels were
increased with exogenous rhIGF-I treatment (P < 0.001), but there was no effect of GH on total IGF-I. The total IGF-I
levels were increased in the orally fed and L-PN groups with rhIGF-I
treatment (post hoc analysis, P < 0.020), but not in
the CHO-PN.
Levels of ALS, a GH-dependent, liver-derived protein, were also higher
in the orally fed groups (P < 0.001) compared with the
PN-fed animals. There was no significant overall effect of GH on ALS
levels, but overall IGF-I decreased ALS levels (P < 0.001). On post hoc analysis, IGF-I decreased ALS levels only in the PN
groups (P < 0.001 for L-PN and P = 0.004 for CHO-PN), and GH increased ALS levels only in the orally fed
group (P = 0.032).
Serum IGFBP-1 concentration (a marker of hepatic insulin sensitivity)
was significantly lower in the orally fed than in the PN group
(P < 0.001). It was also lower in the CHO-PN compared with the L-PN groups (P = 0.004). Overall, rhIGF-I
treatment increased IGFBP-1 levels (P < 0.001), but
there was no GH effect. There was an interaction between the type of
nutrition and hormonal treatment on IGFBP-1 levels due to an
rhIGF-I-induced increase in IGFBP-1 levels in the PN groups
(P = 0.001 for L-PN and P < 0.001 for
CHO-PN), but not in the orally fed group. Similarly, GH increased
IGFBP-1 levels only in the parenterally fed animals (P = 0.014 for L-PN and P = 0.071 for CHO-PN).
Serum NEFA was significantly increased with GH treatment
(P = 0.008), and this reached statistical significance
in the orally fed subgroup (LSD, P = 0.037). In
contrast, there was no overall effect of IGF-I treatment on NEFA
levels. Serum NEFA was also significantly higher in the L-PN group than
in the CHO-PN and the orally fed groups (P < 0.001),
and it was significantly lower in the CHO-PN compared with the orally
fed animals (P < 0.001).
| |
DISCUSSION |
In this study, PN administration to young peripubescent male rats
resulted in a marked suppression of the normal increase in hepatic
cellular DNA and protein that occurs in these actively growing animals.
These findings are unlikely to be due simply to PN-induced hepatic
steatosis, because the high-lipid PN regimen resulted in less hepatic
fat deposition than the high-carbohydrate PN but had no effect on DNA
or protein content. Furthermore, the hepatic steatosis was modest, but
there was marked suppression of cell growth. The consequences of this
are unclear, but in the current study, there was a suggestion of
hepatic GH resistance in the PN-fed rats, as GH treatment increased
serum ALS in the chow-fed rats but had no effect in the PN-fed groups.
It is also of note that the PN-induced visceral atrophy was reversed by
exogenous IGF-I, but IGF-I had no effect on the attenuated growth of
peripheral tissues. Furthermore, GH administration resulted in a gain
in gastrocnemius muscle protein, indicating that GH did maintain its
peripheral effects during PN.
Peripheral effects of GH and IGF-I treatment.
In agreement with several previous studies (2, 28), in
this study, GH treatment significantly increased skeletal muscle (gastrocnemius) protein mass in the CHO-PN and chow-fed groups. That
IGF-I concentrations were also not significantly altered by GH
treatment in the chow-fed rats and rats on CHO-PN suggests that the
increase in muscle protein was due either to a direct effect of GH or
to an increase in intramuscular IGF-I concentrations. Germane to this
point is the recent study by Liu and LeRoith (26), who
treated genetically engineered IGF-I-null mice with exogenous GH and
found no effect on postnatal body weight gain. This suggests that the
growth-promoting activity of GH is not due to its direct effect on
muscle growth but is more likely due to the paracrine/autocrine actions
of IGF-I in the muscle or to the interactions between IGF-I and its
binding proteins.
The reason for the lack of effect of GH on skeletal muscle protein gain
in the L-PN group is unclear but may be related to the lower insulin
concentrations with this treatment compared with CHO-PN. However, the
insulin concentrations were not decreased below control levels and not
below levels thought to suppress protein anabolism (33).
Alternatively, the high-lipid infusion, leading to increased adipose
fat deposition, may have attenuated the GH-stimulated IGF-I expression
in skeletal muscle, as has been observed in the obese Zucker rat
(34).
In the present study, the protein anabolic response to GH was much less
in the soleus compared with the gastrocnemius muscle. These two muscles
have very different myofibril type distribution, the soleus containing
predominantly slow, type I, oxidative fibers and the gastrocnemius
fast, type II, glycolytic fibers. In previous studies, the soleus
muscle has been shown to be generally less responsive to GH than the
gastrocnemius muscle (2, 16).
The lack of effect of IGF-I infusion on carcass and skeletal muscle
protein in chow- and parenterally fed rats agrees with results that we
obtained previously in rats maintained on 50% glucose-50% lipid PN
(21). This occurred despite an increase in total IGF-I
concentrations. Others have shown a significant effect of IGF-I
administration on carcass protein deposition in rats; however, the
greatest effect of IGF-I is usually seen in weight-losing catabolic
stress models (28, 42). This may partly explain the lack
of effect of IGF-I in the present study, as the PN-fed rats were
gaining weight, albeit at a slower rate than the chow-fed rats.
Alternatively, the age of the animals may be an important factor
explaining the lack of response of the carcass to IGF-I. Most previous
studies examining the effect of IGF-I or GH on lean tissue growth have
used young adult (8- to 10-wk-old) male rats (28, 42). The
rats in the present study were peripubertal (4- to 5-wk-old) males. It
is possible that exogenous IGF-I or GH is less effective in stimulating
carcass lean tissue in rapidly growing adolescent rats who already have
high GH and IGF-I concentrations. Support for this hypothesis comes
from a recent study by Rol De Lama et al. (39), where
exogenous GH was shown to stimulate skeletal growth in peripubertal
female rats but not male rats. They hypothesized that GH was unable to
enhance skeletal growth in peripubertal male rats because they were
already growing at a biologically maximal rate due to the preexisting
high GH levels. In the female, however, the increase in GH secretion at
the onset of puberty is less than the male's; therefore, females may
be more responsive to GH supplementation.
Infusion of L-PN led to a large increase in carcass lipid deposition.
This was due to the increase in substrate supply (reflected by
increased serum NEFA) and insulin concentrations, leading to uptake and
storage of lipid in adipose tissue (45). Although GH
administration had no effect on carcass lipid deposition, it resulted
in increased serum NEFA concentration. This is consistent with the
direct influence of GH on adipose tissue hormone-sensitive lipase,
where it promotes the release of NEFA. In contrast, high-dose exogenous
IGF-I has been shown to have short-term, insulin-like, antilipolytic
effects (24) and long-term lipolytic effects, the latter
presumably via suppression of insulin secretion (15). However, in the present study, 7 days of IGF-I treatment had no effect
on carcass lipid, even though insulin levels were suppressed more than
twofold in the L-PN group. The high lipid content in the L-PN overcame
any lipolytic effect of decreased insulin and/or exogenous IGF-I or GH.
Visceral effects.
The PN-induced loss of visceral mass (small intestine, kidneys, spleen,
and thymus) was prevented by IGF-I treatment during both PN regimens,
as we (21) and others (46) have previously observed. This occurred even in the CHO-PN group, where there was no
significant increase in serum total IGF-I concentrations. This
emphasizes the powerful proliferative effect of IGF-I on many visceral
organs. In contrast, GH had no significant effect on visceral mass
during PN but did increase lean tissue weight of thymus and spleen in
the orally fed animals. The mechanism for this is unclear, but there
was a trend toward raised serum IGF-I with GH treatment in the orally
fed rats, which may have contributed to increased lean tissues in these organs.
Hepatic steatosis occurred in both the CHO-PN and L-PN groups. The
increase in fat deposition was probably due to the increase in serum
insulin concentration with PN, which was exacerbated in the CHO-PN
group by high-level glucose infusion. Increased insulin secretion
caused by high level of glucose and amino acid infusion stimulates
hepatic de novo lipogenesis and inhibits fatty acid oxidation
(12). Serum NEFA is unlikely to have contributed greatly
to the liver steatosis, because it was decreased in the CHO-PN, which
had the highest level of steatosis, and increased in the L-PN, which
had less steatosis compared with the CHO-PN group. However, the
combination of increased insulin and serum NEFA levels in the L-PN
group may have worked in concert to produce steatosis in this group.
PN also had a profound effect on hepatic protein and DNA content. In
contrast to the chow-fed rats, where there was increased total liver
protein and DNA with increasing weight, there was no change in these
two components during 7 days of PN treatment. Increase in tissue
protein content can be due to an increase in cell number
(proliferation) or cell size (growth). In contrast, an increase in DNA
content usually suggests increased cell number, because the DNA content
of nonmalignant cells is normally constant. That both of these
parameters remained constant during PN suggests that PN leads to a
marked suppression of hepatocyte proliferation. Whether this is a
consequence of fatty acid infiltration is unclear; however, the
reduction in hepatic liver deposition in the L-PN compared with the
CHO-PN group had no effect on liver protein or DNA content. That the
rats in this study were not fully mature may have had an impact on the
suppressive effect of PN on liver growth, as children (38)
and young immature animals are more susceptible to PN-associated liver
damage (e.g., cholestasis) (9).
Effect of exogenous IGF or GH on serum hormones.
GH is thought to increase serum insulin by directly affecting
pancreatic islet cells (40). In the present study, GH
failed to increase insulin levels, although there was a trend toward an
increase in the orally fed and CHO-PN-fed animals. This is contrary to
the results of other studies (4, 18, 28), which found an
increase in insulin levels with GH treatment. The reason for this
variation could be the relative degree of immaturity of the pancreas in
our PN model compared with the level in latter studies. A well-reported
side effect of high-carbohydrate PN is hypotrophy of the pancreas
(14). Therefore, the impact of PN on the pancreas might be
more severe in our younger animal model, leading to a lack of effect of
GH on insulin secretion.
ALS is a liver-derived protein under GH and nutritional regulation.
With PN there was a general suppression of ALS and IGF-I levels and a
lack of response of serum ALS to exogenous GH treatment. Addition of
IGF-I to PN further decreased ALS levels, possibly through negative
feedback inhibition of endogenous GH by IGF-I, as seen in human studies
(23). The development of hepatic GH resistance leading to
lower IGF-I and ALS may contribute to the poorer growth of lean tissues
in animals and humans. Using a similar rat model, Ney et al.
(35) found that infusion of high-CHO PN to rats led to
significant hepatic steatosis without altering serum IGF-I
concentrations, although hepatic IGF-I mRNA was significantly depressed. Perhaps the peripubescent rats used in the present study
were more susceptible to PN-derived liver damage than young adult rats
(6). In this regard, it is well known that the incidence of PN-associated cholestasis is much higher in children than in adults.
Perhaps the central cause of the GH resistance in the rats in the
present study was the accumulation of bile salts rather than steatosis
itself. In fact, there are a number of reports suggesting that hepatic
GH resistance develops in cholestatic liver disease in children and in
rat bile duct cannulation models (7, 20). It remains to be
determined whether there is a link between PN-derived cholestasis and
hepatic GH resistance.
The hepatic production of IGFBP-1 is under complex regulation by many
factors, including insulin, nutrients (amino acids), glucocorticoids,
and cytokines. Insulin is a dominant suppressor of IGFBP-1 production
(13), although its activation by nutrient deprivation is
independent of insulin (25). Circulating IGFBP-1 is
thought to have an important glucoregulatory function (3) and can complex with IGF-I, cross the endothelial layer, and presumably inhibit IGF-I's actions. As a liver-derived protein suppressible by
insulin, IGFBP-1 serves as a useful index of hepatic insulin sensitivity. In our previous study (21), we found that
amino acid deficiency elevated serum IGFBP-1 levels, presumably through enhanced hepatic production. This has been confirmed in hepatocyte cell
culture experiments (19). In the current study, we found that IGFBP-1 was increased in the PN-fed animals compared with the
chow-fed controls, possibly blocking some of the effect of GH by
reducing free IGF-I levels (10). We also found
(3) that, compared with CHO-PN, the L-PN regimen led to
increased IGFBP-1 levels, presumably due to low carbohydrate infusion.
The results of this study also suggest that, during PN, there was some
degree of hepatic insulin resistance in the PN groups, particularly in
the CHO-PN group, where, despite high levels of serum insulin, there
was no suppression of IGFBP-1.
Perspectives.
PN-associated liver dysfunction (steatosis, cholestasis) is a major
complication of PN treatment, particularly in the young. The
association of major changes in liver composition with hepatic GH
resistance in parenterally fed rats in this study is potentially an
important observation. It may provide an explanation for the harmful
effects of PN after liver resection. Low serum IGF-I and ALS, and by
inference GH, may limit the effectiveness of PN to support the normal
growth of peripheral lean tissue and visceral organs.
Understanding the causes of these observations may have important clinical implications.
| |
ACKNOWLEDGEMENTS |
Sources of support for this study were G. J. Tattersall's Pty
Ltd., Pharmacia & Upjohn, the National Health and Medical Research Council, and the Cancer Surgery Research Foundation.
| |
FOOTNOTES |
Address for reprint requests and other correspondence: R. C. Smith, Dept. of Surgery, Univ. of Sydney, Wallace Freeborn Bldg., Royal North Shore Hospital, St. Leonards, NSW 2065, Australia (E-mail:
rsmith{at}med.usyd.edu.au).
The costs of publication of this
article were defrayed in part by the
payment of page charges. The article
must therefore be hereby marked
"advertisement"
in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Received 27 March 2001; accepted in final form 13 July 2001.
| |
REFERENCES |
1.
Apelgren, KN,
and
Wilmore DW.
Nutritional care of the critically ill patient.
Surg Clin North Am
63:
497-507,
1983[Web of Science][Medline].
2.
Bates, PC,
Loughna PT,
Pell JM,
Schulster D,
and
Millward DJ.
Interactions between growth hormone and nutrition in hypophysectomized rats: body composition and production of insulin-like growth factor-I.
J Endocrinol
139:
117-126,
1993[Abstract/Free Full Text].
3.
Baxter, RC.
Insulin-like growth factor binding proteins as glucoregulators.
Metabolism
44:
12-17,
1995[Web of Science][Medline].
4.
Belcher, HJ,
Mercer D,
Judkins KC,
Shalaby S,
Wise S,
Marks V,
and
Tanner NS.
Biosynthetic human growth hormone in burned patients: a pilot study.
Burns
15:
99-107,
1989[Web of Science][Medline].
5.
Bolton, S.
Factorial design.
In: Pharmaceutical Statistics: Practical and Clinical Applications, edited by Swarbrick J. New York: Dekker, 1981, p. 326-354.
6.
Briones, ER,
and
Iber FL.
Liver and biliary tract changes and injury associated with total parenteral nutrition
pathogenesis and prevention.
J Am Coll Nutr
14:
219-228,
1995[Abstract].
7.
Bucuvalas, JC,
Cutfield W,
Horn J,
Sperling MA,
Heubi JE,
Campaigne B,
and
Chernausek SD.
Resistance to the growth-promoting and metabolic effects of growth hormone in children with chronic liver disease.
J Pediatr
117:
397-402,
1990[Web of Science][Medline].
8.
Clemmons, DR,
Smith-Banks A,
and
Underwood LE.
Reversal of diet-induced catabolism by infusion of recombinant insulin-like growth factor-I in humans.
J Clin Endocrinol Metab
75:
234-238,
1992[Abstract].
9.
Duerksen, DR,
Van Aerde JE,
Chan G,
Thomson ABR,
Jewell LJ,
and
Clandinin MT.
Total parenteral nutrition impairs bile flow and alters bile composition in newborn piglet.
Dig Dis Sci
41:
1864-1870,
1996[Web of Science][Medline].
10.
Frystyk, J,
Grofte T,
Skjaerbaek C,
and
Orskov H.
The effect of oral glucose on serum free insulin-like growth factor-I and -II in healthy adults.
J Clin Endocrinol Metab
82:
3124-3127,
1997[Abstract/Free Full Text].
11.
Gogos, CA,
and
Kalfarentzos F.
Total parenteral nutrition and immune system activity: a review.
Nutrition
11:
339-344,
1995[Web of Science][Medline].
12.
Hall, RI,
Grant JP,
and
Ross LH.
Pathogenesis of hepatic steatosis in the parenterally fed rat.
J Clin Invest
74:
1658-1668,
1984.
13.
Holly, JM,
Biddlecombe RA,
Dunger DB,
Edge JA,
Arniel SA,
Howell R,
Chad T,
Rees LH,
and
Wass JAH
Circadian variation of GH-independent IGF-binding protein in diabetes mellitus and its relationship to insulin. A new role for insulin?
Clin Endocrinol (Oxf)
29:
667-675,
1988[Medline].
14.
Hughes, CA,
Prince A,
and
Dowling RH.
Speed of change in pancreatic mass and in intestinal bacteriology of parenterally fed rats.
Clin Sci (Colch)
59:
329-336,
1980[Medline].
15.
Hussain, MA,
Schmitz O,
Mengel A,
Glatz Y,
Christiansen JS,
Zapf J,
and
Froesch ER.
Comparison of the effects of growth hormone and insulin-like growth factor I on substrate oxidation and on insulin sensitivity in growth hormone-deficient humans.
J Clin Invest
94:
1126-1133,
1994.
16.
Isgaard, J,
Nilsson A,
Vikman K,
and
Isaksson OG.
Growth hormone regulates the level of insulin-like growth factor-I mRNA in rat skeletal muscle.
J Endocrinol
120:
107-112,
1989[Abstract/Free Full Text].
17.
Jenkins, RC,
and
Ross RJ.
Acquired growth hormone resistance in catabolic states.
Bailliere's Clin Endocrinol Metab
10:
411-419,
1996[Web of Science][Medline].
18.
Jenkins, RC,
and
Ross RJ.
Growth hormone therapy for protein catabolism.
Q J Med
89:
813-819,
1996[Abstract].
19.
Jousse, C,
Bruhat A,
Ferrara M,
and
Fafournoux P.
Physiological concentration of amino acids regulates insulin-like-growth-factor-binding protein 1 expression.
Biochem J
334:
147-153,
1998.
20.
Katz, S,
Pescovitz OH,
and
Grosfeld JL.
Growth failure and decreased levels of insulin-like growth factor I in obstructive jaundice are reversed by bile diversion.
J Pediatr Surg
26:
900-902,
1991[Web of Science][Medline].
21.
Kee, AJ,
Baxter RC,
Carlsson AR,
and
Smith RC.
Parenteral amino acid intake alters the anabolic actions of insulin-like growth factor I in rats.
Am J Physiol Endocrinol Metab
277:
E63-E72,
1999[Abstract/Free Full Text].
22.
Kee, AJ,
and
Smith RC.
The effect of the rate and route of nutrient delivery on total body and organ composition in rats.
Nutrition
12:
180-188,
1996[Web of Science][Medline].
23.
Kupfer, SR,
Underwood LE,
Baxter RC,
and
Clemmons DR.
Enhancement of the anabolic effects of growth hormone and insulin-like growth factor I by use of both agents simultaneously.
J Clin Invest
91:
391-396,
1993.
24.
Laager, R,
Ninnis R,
and
Keller U.
Comparison of the effects of recombinant human insulin-like growth factor-I and insulin on glucose and leucine kinetics in humans.
J Clin Invest
92:
1903-1909,
1993.
25.
Lewitt, MS,
and
Baxter RC.
Inhibitors of glucose uptake stimulate the production of insulin-like growth factor-binding protein (IGFBP-1) by human fetal liver.
Endocrinology
126:
1527-1533,
1990[Abstract/Free Full Text].
26.
Liu, JL,
and
LeRoith D.
Insulin-like growth factor I is essential for postnatal growth in response to growth hormone.
Endocrinology
140:
5178-5184,
1999[Abstract/Free Full Text].
27.
Liu, S,
Baracos VE,
Quinney HA,
Le BT,
and
Clandinin MT.
Parallel insulin-like growth factor I and insulin resistance in muscles of rats fed a high fat diet.
Endocrinology
136:
3318-3324,
1995[Abstract].
28.
Lo, HC,
Hinton PS,
Peterson CA,
and
Ney DM.
Simultaneous treatment with IGF-I and GH additively increases anabolism in parenterally fed rats.
Am J Physiol Endocrinol Metab
269:
E368-E376,
1995[Abstract/Free Full Text].
29.
Lo, HC,
Hinton PS,
Yang H,
Unterman TG,
and
Ney DM.
Insulin-like growth factor-I but not growth hormone attenuates dexamethasone-induced catabolism in parenterally fed rats.
J Parenter Enteral Nutr
20:
171-177,
1996[Abstract/Free Full Text].
30.
Loder, PB,
Smith RC,
Kee AJ,
Kohlhardt SR,
Fisher MM,
Jones M,
and
Reeve TS.
What rate of infusion of intravenous nutrition solution is required to stimulate uptake of amino acids by peripheral tissues in depleted patients?
Ann Surg
211:
360-368,
1990[Web of Science][Medline].
31.
MacFie, J.
Enteral vs. parenteral nutrition: the significance of bacterial translocation and gut-barrier function.
Nutrition
16:
606-611,
2000[Web of Science][Medline].
32.
Manson, JM,
Smith RJ,
and
Wilmore DW.
Growth hormone stimulates protein synthesis during hypocaloric parenteral nutrition. Role of hormonal-substrate environment.
Ann Surg
208:
136-142,
1988[Web of Science][Medline].
33.
McHardy, KC,
McNurlan MA,
Milne E,
Calder AG,
Fearns LM,
Broom J,
and
Garlick PJ.
The effect of insulin suppression on postprandial nutrient metabolism: studies with infusion of somatostatin and insulin.
Eur J Clin Nutr
45:
515-526,
1991[Web of Science][Medline].
34.
Melian, E,
Gonzalez B,
Ajo R,
Gonzalez N,
and
Sanchez FF.
Tissue-specific response of IGF-I mRNA expression to obesity-associated GH decline in the male Zucker fatty rat.
J Endocrinol
160:
49-56,
1999[Abstract].
35.
Ney, DM,
Yang H,
Smith SM,
and
Unterman TG.
High-calorie total parenteral nutrition reduces hepatic insulin-like growth factor-I mRNA and alters serum levels of insulin-like growth factor-binding protein-1, -3, -5, and -6 in the rat.
Metabolism
44:
152-160,
1995[Web of Science][Medline].
36.
Ponting, GA,
Halliday D,
Teale JD,
and
Sim AJ.
Postoperative positive nitrogen balance with intravenous hyponutrition and growth hormone.
Lancet
1:
438-440,
1988[Web of Science][Medline].
37.
Prasad, AS,
DuMouchelle E,
Koniuch D,
Oberleas D,
and
Park A.
A simple fluorometric method for the determination of RNA and DNA in tissues.
J Lab Clin Med
80:
598-602,
1972[Web of Science][Medline].
38.
Quigley, EM,
Marsh MN,
Shaffer JL,
and
Markin RS.
Hepatobiliary complications of total parenteral nutrition.
Gastroenterology
104:
286-301,
1993[Web of Science][Medline].
39.
Rol De Lama, MA,
Perez-Romero A,
Tresguerres JA,
Hermanussen M,
and
Ariznavarreta C.
Recombinant human growth hormone enhances tibial growth in peripubertal female rats but not in males.
Eur J Endocrinol
142:
517-523,
2000[Abstract].
40.
Sekine, N,
Wollheim CB,
and
Fujita T.
GH signalling in pancreatic beta-cells.
Endocr J
45, Suppl:
S33-S40,
1998.
41.
SPSS.
SPSS Advanced Statistics 7.5. Chicago, IL: SPSS, 1997, p. 358-359.
42.
Tomas, FM.
The anti-catabolic efficacy of insulin-like growth factor-I is enhanced by its early administration to rats receiving dexamethasone.
J Endocrinol
157:
89-97,
1998[Abstract].
43.
Tomas, FM,
Knowles SE,
Owens PC,
Chandler CS,
Francis GL,
Read LC,
and
Ballard FJ.
Insulin-like growth factor-I (IGF-I) and especially IGF-I variants are anabolic in dexamethasone-treated rats.
Biochem J
282:
91-97,
1992.
44.
Voerman, BJ,
Strack van Schijndel RJ,
Groeneveld AB,
de Boer H,
Nauta JP,
and
Thijs LG.
Effects of human growth hormone in critically ill nonseptic patients: results from a prospective, randomized, placebo-controlled trial.
Crit Care Med
23:
665-673,
1995[Web of Science][Medline].
45.
Williamson, DH.
Role of insulin in the integration of lipid metabolism in mammalian tissues.
Biochem Soc Trans
17:
37-40,
1989[Web of Science][Medline].
46.
Yang, H,
Ney DM,
Peterson CA,
Lo HC,
Carey HV,
and
Adamo ML.
Stimulation of intestinal growth is associated with increased insulin- like growth factor-binding protein 5 mRNA in the jejunal mucosa of insulin-like growth factor-I-treated parenterally fed rats.
Proc Soc Exp Biol Med
216:
438-445,
1997[Medline].
Am J Physiol Endocrinol Metab 281(5):E1063-E1072
0193-1849/01 $5.00
Copyright © 2001 the American Physiological Society
TOP
ABSTRACT
INTRODUCTION
