INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

INSULIN-LIKE GROWTH FACTORS (IGF-I and -II) are related in structure to proinsulin and have developmental and growth stimulatory effects as well as insulin-like metabolic actions (28). In the circulation, the IGFs are stabilized in growth hormone (GH)-dependent ternary complexes with IGF binding protein (IGFBP)-3 or IGFBP-5 (8, 45) and the acid-labile subunit (ALS). Compared with other IGF · IGFBP complexes, this ternary complex is thought to cross the capillary barrier relatively poorly. This suggests a significant role for ALS in regulating the passage of IGFs from the circulation into the extracellular tissue compartment, thereby modulating their metabolic and other biological activities. The liver is the principal source of circulating ALS, which is synthesized by hepatocytes in a GH-dependent manner (12, 41).

Hepatocytes in adult rats are normally quiescent, but they switch to a replicative state controlled by various paracrine/autocrine (and endocrine) growth factors and cytokines, culminating in an almost synchronous round of replication 24 h after partial hepatectomy (PHx) (19, 21, 31). cAMP and cGMP are acutely upregulated post-PHx and may also be important intermediate effectors of regeneration (20). Within 7-10 days, liver mass is fully restored, and hepatic cells stop growing, perhaps under the influence of factors such as interleukin (IL)-1 and transforming growth factor (TGF)- (31). TGF- function in liver regeneration is probably regulated by another member of the IGF axis, the IGF-II/mannose 6-phosphate receptor, with which it is upregulated and coexpressed after PHx (10, 27, 38). Certain factors that are induced in the first hour of liver regeneration, e.g., cAMP (20), also directly suppress hepatocyte ALS expression (17). Furthermore, GH binding to hepatocyte membranes (26) and serum insulin levels (43) decrease rapidly after PHx. We have also shown that the regulation of ALS in isolated hepatocytes by GH is dependent on the presence of insulin (15), therefore suggesting that ALS may be negatively regulated during the early stages of liver regeneration.

Although serum levels of IGF-I fall markedly after PHx, this effect has been attributed to the reduced food intake of PHx animals (35). Together with an observed lack of effect on hepatic IGF-I mRNA levels (33), this appears to rule out a role for IGF-I in liver regeneration. Although IGF-I receptor gene expression and IGF-I binding to hepatocytes are reported to increase after PHx, this occurs with a delayed time course, peaking at 48-72 h after PHx, some 24-48 h after the start of hepatocyte proliferation (11, 36). Although evidence for a direct involvement of IGF-I in hepatocyte proliferation is not good, the possibility remains that there is a role for the IGF-I axis in whole body metabolic compensation in rats with regenerating livers. Interestingly, Unterman and Phillips (46) found that the cartilage sulfation factor "somatomedin" activity of whole rat serum is greater in PHx than in sham-operated animals, suggesting increased IGF bioavailability in PHx. Our hypothesis is that this may result from suppression of ALS levels. The circulating IGFs have a significant insulin-like glucoregulatory potential that we propose is regulated primarily by ALS through the IGF · IGFBP ternary complexes (4, 7). The marked acute suppression of circulating insulin levels immediately after PHx has been proposed to induce the gluconeogenic capacity of the liver and maintain physiological normoglycemia (23). However, it is unclear how glucose uptake by peripheral tissues is regulated during this relatively hypoinsulinemic period.

Because ALS has a potential role in modulating the glucoregulatory activity of IGF-I and is negatively regulated in vitro by certain cytokines and factors that are acutely upregulated during the first few hours of liver regeneration, we studied the regulation of hepatic ALS gene expression and serum levels post-PHx in direct relation to IGFBP-3, IGF-I, IGFBP-1, insulin, and hepatic GH receptor (GHR).

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Partial hepatectomy. Female Wistar rats (10 wk old, 245 ± 20 g) were used for all experiments. PHx was carried out under isoflurane/NO₂ anesthesia by laparotomy and then excision of the median and left lateral lobes of the liver (25). Sham controls were laparotomized, and the liver was gently pulled through the incision, manipulated, and replaced. To avoid differences in nutritional intake between PHx and sham controls, the animals were fasted for the entire 24 h after surgery. In earlier studies, we found that the PHx animals eat little during the first 24 h after surgery and that the amount they eat is very variable; consequently, a regime of pair-feeding was not used in this short-term study. Blood was removed by cardiac puncture, and livers were excised and immediately frozen in liquid nitrogen from four PHx and sham-operated animals at 3, 6, and 12 h after surgery and six PHx and five sham-operated animals at 24 h after surgery. All protocols were approved by the Institutional Animal Care and Ethics Committee.

Preparation of rat hepatocytes. Hepatocytes were prepared from 10-wk-old (~250 g) female Wistar rats 24 h after sham and PHx operations by in situ perfusion of livers with collagenase and were plated at a density of 2 × 10⁶ cells/60-mm plate, as described previously (17). Additions of recombinant human GH (rhGH) diluted in Williams' E medium containing 0.2% BSA were made 24 h after the initial plating, and cells were then maintained for another 24 h. Media were collected for ALS radioimmunoassay, and RNA was prepared from the cells by the acid-phenol technique (14). DNA was determined from representative plates by fluorimetry as described before (29).

Northern analysis. Total RNA from each liver sample was prepared using the acid-phenol technique (14), and 20 µg were electrophoresed in 1% agarose gels containing 2.2 M formaldehyde. The integrity of the ethidium bromide-stained RNA was confirmed on an ultraviolet light box. The RNA was then transferred by capillary blotting to Zetaprobe GT membranes (BioRad, Hercules, CA) and cross-linked by baking at 80°C in a gel-drying apparatus.

Radioimmunoassays. Sera were prepared, stored at 20°C, and assayed using specific rat ALS, rat IGFBP-3, IGF-I, and insulin radioimmunoassays. ALS: Whole serum was assayed using a rat-specific assay developed in our laboratory (5). All sera for this study were quantitated in the same assay. Conditioned media were assayed in the same way. IGF-I: Whole sera were acid-ethanol extracted (70% EtOH, 0.2 M HCl) and precipitated by centrifugation, and the supernatant was neutralized (0.25 M Tris base) and assayed using an IGF-I-specific radioimmunoassays (40). IGFBP-3: Sera were quantitated in the same assay. Whole sera were acidified with 0.1 M HCl for 1 h at room temperature, neutralized with 0.1 M NaOH, and then assayed as described previously (22). Insulin: Serum insulin was measured using a rat-specific radioimmunoassay obtained from Linco Research (St. Charles, MO). Glucose: Serum glucose was measured using a hand-held glucometer (Roche Diagnostics).

Western blot analysis. Liver was homogenized in ice-cold 10 mM Tris · HCl, 1 mM Na₂EDTA, 250 mM sucrose, 50 µl/ml protease inhibitor cocktail (Sigma, St Louis, MO), 10 mM NaF, and 1 mM Na₃VO₄ (100 mg tissue/ml buffer). After centrifugation of the homogenate at 150,000 g for 20 min at 4°C, the supernatant was collected and stored at 80°C. The protein content of the supernatant was assayed using a BioRad protein determination kit. Forty micrograms of protein were run on an 8% polyacrylamide/Tris · HCl gel and transferred to nitrocellulose membranes. Membranes were screened for Stat5b protein by enhanced chemiluminescence (Pierce, Rockford, IL) using anti-Stat5b antibodies at 0.1 µg/ml (Santa Cruz Biotechnology, Santa Cruz, CA) in TBS containing 1% bovine serum albumin and 1% nonfat milk. Band intensities were semiquantitated using National Institutes of Health Image software.

Statistics. Results were analyzed by ANOVA, with P values calculated using Fisher's post hoc test, and repeated-measures ANOVA (Statview 5.0, Abacus Concepts, Berkeley, CA).

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Markers of liver regeneration. During the first 24 h, liver weight does not normally increase, but hepatocytes enter S-phase with DNA synthesis peaking at the end of this period (31). Accordingly, the weights of the regenerating livers in this study did not change significantly during the first 24 h post-PHx. However, a subgroup of PHx animals, not included in this study but followed out to 72 h, did show the expected increase in hepatic weight (Fig. 1A), suggesting that regeneration was proceeding normally during the 24 h of our study. In addition, we observed a marked transient induction of IGFBP-1 hepatic gene expression and serum levels at 3 h post-PHx (Figs. 1, B and C, and 2), which is consistent with data from previous studies (32).

View larger version (15K): [in this window] [in a new window]   Fig. 1.   Partial hepatectomy (PHx) leads to normal liver growth and rapid transient upregulation of insulin-like growth factor binding protein (IGFBP)-1 expression. A: wet liver weights after PHx. Weights of the regenerating livers in this study increased slightly during the first 24 h post-PHx. However, a subgroup of PHx animals, not included in this study but followed out to 72 h, did show the expected increase in weight. Hepatic IGFBP-1 mRNA levels (B; *P < 0.0001) and serum concentrations (C; *P = 0.043) show a marked transient induction at 3 h post-PHx but subsequently fall to baseline levels. Post-op, post-operation.

View larger version (54K): [in this window] [in a new window]   Fig. 2.   Hepatic gene expression of components of the growth hormone (GH)-IGF axis after sham and PHx surgery. Representative Northern blot of hepatic total RNAs from animals during a 24-h period after sham or PHx screened sequentially for acid-labile subunit (ALS), GH receptor (GHR), GH binding protein (GHBP), IGF-I, and IGFBP-1 mRNAs and 18S rRNA.

ALS expression. By 12-24 h post-PHx, there was a significant 50% suppression (P < 0.005) of ALS gene expression in PHx rats (Figs. 2 and 3A) and a 30% (P < 0.02) suppression of serum ALS levels (Fig. 3B) relative to sham controls. Relative to intact animals at 0 h, however, there was an ~60% suppression of both hepatic gene expression and serum levels of ALS at 24 h. Hepatic gene expression was suppressed to its lowest level (still ~50% of control levels) by 6 h after PHx; however, at 3 h, there was no difference between sham and PHx levels. In contrast, serum ALS levels were significantly suppressed (P < 0.02) relative to controls only 3 h after PHx. Overall, there was a significant suppression of both gene expression and serum levels of ALS in PHx animals during the study period (P = 0.002 and 0.009, respectively, repeated-measures ANOVA).

View larger version (20K): [in this window] [in a new window]   Fig. 3.   ALS hepatic gene expression and serum levels are suppressed after PHx. A: hepatic ALS mRNA levels from sham and PHx animals normalized relative to 18S RNA (*P < 0.005). B: ALS serum levels after sham and PHx surgery measured using a specific RIA (*P < 0.02).

IGF-I expression. IGF-I hepatic mRNA was suppressed by ~30% (P = 0.0001) in PHx relative to sham-operated control animals only at 24 h post-PHx (Figs. 2 and 4A). IGF-I mRNA consists of two major groups of transcripts at ~7 kb and at 0.7-1 kb. The regulation of these transcripts has not been examined in detail after PHx. It was found that the 7-kb transcript was significantly downregulated in PHx relative to sham animals by 12 h post-PHx (Fig. 4B). On the other hand, the low molecular weight transcripts took longer to become suppressed, with the difference between PHx and sham controls being significant at 24 h (Fig. 4B). Serum levels of IGF-I were not significantly different from sham controls during the entire 24-h period (Fig. 4C). However, relative to 0-h controls, there was a significant fall at 12 and 24 h in PHx (P < 0.003 and P < 0.01, respectively) but not in sham-operated animals.

View larger version (16K): [in this window] [in a new window]   Fig. 4.   IGF-I gene expression and serum levels after PHx. A: hepatic IGF-I mRNA levels from sham and PHx animals normalized relative to 18S RNA (*P = 0.0001). B: steady-state hepatic IGF-I transcript levels (, sham 7 kb; , PHx 7 kb; , sham 0.7-1 kb; , PHx 0.7-1 kb) become significantly suppressed only at 12-24 h post-PHx (*P < 0.05). C: serum IGF-I levels after sham and PHx measured using a specific RIA.

IGFBP-3 expression. Using a new quantitative assay (22), we observed an ~30% (P = 0.0024) drop in IGFBP-3 levels at 3 h post-PHx relative to sham controls, followed by a return back to sham control levels at 6 h (Fig. 5). Subsequently, there was no significant difference between sham controls and PHx animals. Hepatic IGFBP-3 mRNA could not be detected by Northern analysis.

View larger version (19K): [in this window] [in a new window]   Fig. 5.   Serum IGFBP-3 levels levels after PHx. Serum IGFBP-3 levels after sham and PHx measured using a quantitative solid-phase assay (*P < 0.003).

GHR gene and Stat5b expression. Because regulation of ALS is intimately linked with hepatic GH sensitivity, we examined the regulation of GHR gene expression in our animals to assess its role in regulating ALS and the IGF-I axis during liver regeneration. We observed significantly lower GHR mRNA levels in PHx relative to sham control animals, particularly at 3 h (suppressed by 56%, P = 0.012) but also at 12 h (52%, P < 0.007) and at 24 h (44%, P < 0.03) post-PHx (Figs. 2 and 6A). GH binding protein gene expression was found not to be significantly different between PHx and sham-operated animals throughout the course of the study (Fig. 6B).

View larger version (20K): [in this window] [in a new window]   Fig. 6.   Hepatic GHR but not GHBP gene expression is suppressed after PHx. Hepatic GHR (A) and GHBP (B) mRNA levels from sham and PHx animals normalized relative to 18S RNA (*P < 0.03).

View larger version (60K): [in this window] [in a new window]   Fig. 7.   Hepatic levels of total and phosphorylated Stat5b are suppressed by PHx. A: immunoblot analysis of Stat5b in 40 µg total protein from 3 livers at 12 h postsurgery from PHx and sham-operated animals. Levels of total (B) and biphospho (C) Stat5b are suppressed by ~50% in PHx liver compared with sham-operated animals (*P < 0.05, unpaired t-test). Band densities are expressed in arbitrary units.

Serum insulin and glucose. Efficient GH-induced expression of ALS by hepatocytes requires insulin supplementation (15) that probably relates to the close interaction between the signaling mechanisms of these two factors. As has been described by others (see Ref. 43) there was an acute fall in serum insulin after PHx, reaching a nadir at 3 h post-PHx that was significantly (P < 0.02) below that of the sham controls (Fig. 8A). Insulin levels in PHx then remained below but not significantly different from sham until returning to sham control levels at 24 h post-PHx. Similar to previous studies glucose levels were suppressed after both PHx and sham operations (Fig. 8B) although there was no overall significant difference between the groups (P > 0.15, repeated-measures ANOVA), and all rats remained relatively normoglycemic.

View larger version (22K): [in this window] [in a new window]   Fig. 8.   Serum insulin, but not glucose levels, are acutely suppressed after PHx. A: insulin levels in PHx animals are suppressed to 19% of sham control levels (P < 0.02) within 3 h of surgery and remain below those of controls until 24 h after surgery. B: glucose levels, although becoming suppressed relative to 0-h controls, are generally not different between PHx and sham-operated animals and do not reach hypoglycemic levels post-PHx despite 60% loss of liver mass.

Regulation of ALS by GH in isolated hepatocytes from sham and PHx animals. The differential regulation of ALS in PHx animals appeared to correlate with GH sensitivity of the liver. To further examine the mechanism for this differential regulation of ALS in PHx animals, we examined the relative response of hepatocytes isolated from sham-operated and PHx animals to a dose curve of rhGH (Fig. 9). We found that steady-state levels of ALS mRNA were induced in a similar dose-dependent manner in hepatocytes from both control and PHx animals (Fig. 9A). However, in relation to ALS secretion, hepatocytes from control animals were more responsive to GH than their counterparts from PHx animals. There was a significant dose-dependent effect (P = 0.018, repeated-measures ANOVA; Fig. 9B) with significance reached at 10 and 100 ng/ml GH (P = 0.056 and 0.027). Secreted ALS levels, normalized for DNA content, were ~24% lower in conditioned medium from PHx hepatocytes at both 10 and 100 ng/ml GH doses. Steady-state levels of GHR mRNA were not affected by the dose curve of rhGH (Fig. 9C), and there was no significant difference between levels of GHR mRNA in either group of isolated cells (P = 0.77, repeated-measures ANOVA).

View larger version (14K): [in this window] [in a new window]   Fig. 9.   ALS is regulated at the posttranslational level in hepatocytes isolated from regenerating liver. ALS gene expression (A), secreted ALS (B), and GHR gene expression (C) in hepatocytes isolated from sham- and PHx-operated rats treated with a dose curve of recombinant human GH (rhGH). ALS secretion is more responsive to GH in hepatocytes from sham-operated animals than in those from PHx animals (hepatocytes prepared from 6 rats in each group; P = 0.018, repeated-measures ANOVA).

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Our results demonstrate that, during liver regeneration, ALS expression is suppressed not only by lowered food intake but also by factors perhaps intrinsic to the mechanism of liver regeneration. The rapid suppression of ALS serum levels after PHx confirms that the liver is the primary source of serum ALS, despite the possibility that its gene expression may occur in other tissues (12). Our data also demonstrate the long half-life of ALS in the serum (measured in hours), although, because 30% of the liver remains in the animals, a precise determination of the clearance rate cannot be determined. In a sense, it is not surprising that serum levels of ALS are suppressed in PHx animals, because 60% of its major source, the liver, has been removed. What is novel is the finding that ALS gene expression is even more markedly suppressed than ALS protein levels in the serum (50 vs. 30%). The suppression of steady-state mRNA levels is even more significant, considering that these data have been expressed relative to a measure of liver mass (18S rRNA) and that the liver mass has been reduced by >60%. This differential between gene expression and protein levels may be explained partly by the long half-life of ALS in the serum. Our data also demonstrate that, in agreement with other studies (26, 33), the expression of IGF-I, another hepatic and GH-responsive gene, is only slightly, albeit significantly, suppressed after PHx. In addition, we found that there was differential regulation of steady-state hepatic levels of the two major groups of IGF-I transcripts with the ~7-kb transcripts becoming significantly suppressed 12 h before the low molecular weight transcripts. A similar effect has been described in fasted animals where IGF-I pre-mRNA splicing is delayed and the rate of degradation of cytoplasmic IGF-I mRNA is accelerated (47). Although our animals were effectively acutely fasted, PHx had an additional effect on IGF posttranscriptional processing. Interestingly, the response of serum IGF-I to these changes in GH responsiveness was somewhat different from hepatic gene expression, with no significant changes during the first 24 h.

A potential major contributor to the regulation of both ALS and IGF-I in PHx is the relative abundance of GHR. It was found that hepatic GHR steady-state mRNA levels were significantly suppressed relative to controls within 3 h of PHx. This is counter to the finding of Husman and Andersson (26), who found no change in GHR gene expression, but it relates well with their finding of suppressed binding of GH to hepatic membrane fractions. The strong correlation that we found between GHR gene expression and ALS gene expression and serum levels in sham animals dissociates in PHx animals (Table 1), perhaps demonstrating a GH signaling defect in vivo. This was confirmed by the finding that hepatic Stat5b levels are reduced by 50% in PHx animals. GH-induced IGF-I gene expression appears to require not only Stat5b but also hepatocyte nuclear factor (HNF)-1 (30). HNF-1 is not appreciably changed after PHx (see Ref. 43), perhaps explaining why IGF-I transcription is less sensitive than ALS to lowered GH sensitivity during liver regeneration.

Hepatocytes from PHx animals were significantly less sensitive to GH in terms of induction of ALS secretion than their counterparts from sham-operated animals, despite there being similar levels of GHR gene expression. This may be explained by the defect in posttranslational processing of the GHR, suggested by Husman and Andersson (26), or by defects in postreceptor signaling. Interestingly, in our in vitro study, ALS gene expression appears unaffected, indicating a defect in posttranscriptional processing in PHx hepatocytes. Hepatocytes are known to lose much of their capacity to bind GH after being isolated from the liver (3); however, they retain their ability to respond to GH for 2-3 days after isolation. It is possible that GHR levels on the cell surface reach a nadir that is maintained during this time and whose level is independent of the state of the cells at the time of isolation. This seems to be different from insulin and IGF-II receptors, which are found at equivalent levels on isolated cells compared with the liver in vivo (39, 40). Our findings suggest that there is a switch from transcriptional to posttranscriptional control of ALS in hepatocytes during the early stages of liver regeneration.

Factors other than GH may regulate ALS expression by the regenerating liver. It has been shown that cAMP and components of its transcriptional regulatory pathway (20, 42) and IL-1 (24) are upregulated within 1 h after PHx. We have shown that cAMP (17) and IL-1 (16) suppress both ALS gene expression and secretion by hepatocytes in vitro, suggesting that these factors may modulate ALS expression during liver regeneration. This effect of IL-1 is probably through suppression of GHR gene expression (44) and induction of suppressor of cytokine signaling-3, which has been demonstrated to suppress GH-stimulated ALS transcription (9).

Using a quantitative assay (22), we determined that IGFBP-3 was significantly suppressed up to 24 h after PHx relative to 0-h controls (as were 24-h sham controls). This result is similar to that of Phillips et al. (34), who used a Western ligand blot assay, although comparisons were made only relative to a 3-h sham control. However, IGFBP-3 was significantly suppressed only relative to sham-operated animals at 3 h post-PHx, after which IGFBP-3 levels appeared to rise transiently and became similar to those of sham-operated animals. These data suggest that the liver is probably not the main source of serum IGFBP-3. This idea is supported by other studies where IGFBP-3 has been found not to be produced in measurable amounts by isolated whole livers (37), probably because its gene expression is limited to the relatively small population of endothelial and Kupffer cells (48). Furthermore, a number of other tissues express higher levels of IGFBP-3 mRNA than the liver (1). Our data also suggest that the suppression of ALS levels during the first 24 h after both sham and PHx surgery probably regulates the suppression of serum levels of IGFBP-3 through decreased ternary complex formation and perhaps increased clearance from the circulation. Although a marked effect of PHx on IGFBP-3 was observed only in the first few hours after PHx, the significant correlation between ALS and this component of the ternary complex in sham-operated animals was lost in PHx animals despite a continued strong correlation between ALS and IGF-I in the serum (Table 1). This further suggests differential regulation of IGFBP-3 levels in the PHx rat, which is unusual because ALS and IGFBP-3 generally exhibit a strong interdependence (6). The strong correlation between ALS and IGF-I in PHx animals suggests a possible interdependence not involving IGFBP-3, as long as IGFBP-3 is in high enough concentration to allow ternary complex formation. However, only ALS serum levels become significantly suppressed in PHx relative to sham controls, suggesting that, in PHx rats, a greater proportion of IGF-I occurs in binary complexes that would increase its bioavailability (2). IGFBP-1 is rapidly upregulated after PHx to a maximum at ~3 h (Fig. 1C, Ref. 32), and IGFBP-4 is upregulated from ~12 h (18). Therefore, it is possible that the maintenance of IGF-I serum levels after PHx is a consequence of its becoming sequestered initially into IGFBP-1 complexes and then into IGFBP-4 complexes. However, within 3 h of PHx, IGFBP-3 levels drop from ~50 nM (2.3 µg/ml) to ~30 nM (1.3 µg/ml). At the same time, IGFBP-1 levels rise from ~0.5 nM (15 ng/ml) to 2 nM (50 ng/ml). This 1.5 nM increase in IGFBP-1 is unlikely to compensate for the ~20 nM suppression of IGFBP-3, suggesting that other factors are involved in maintaining IGF-I concentrations (which actually rise slightly) immediately after PHx.

A number of liver-specific immediate early genes upregulated by PHx encode gluconeogenic enzymes and are important for cellular metabolism (43). Conversely, serum insulin is acutely downregulated, as we and others have shown (43). Overall, this would have the effect of increasing hepatic glucose output as an adaptive response to the greatly reduced size of the liver and would maintain relative normoglycemia (43). Moreover, lowered insulin levels may compound the defect in GH signaling and further reduce hepatic ALS expression. Unterman and Phillips (46) made the unexpected observation that, although IGF-I serum levels are not altered appreciably during regeneration, sulfation factor activity (IGF-I bioactivity) was increased in PHx rats. Evidence from our study suggests that lowered serum ALS causes the observed increase in IGF-I bioactivity in the circulation of PHx rats. Although IGFBP-1 and IGFBP-4 may increase during this period, evidence suggests that IGF-I bound to these binding proteins is still bioavailable, whereas IGF-I bound in ternary complexes is not (2). This increase in bioavailability may be important during the hypoinsulinemic phase after PHx when IGF-I, whose sites of action are restricted to extrahepatic tissues through lack of hepatic receptors, may adopt the glucoregulatory role of insulin in peripheral tissues.

Overall, our results demonstrate that, during liver regeneration, ALS gene expression and serum levels are suppressed predominantly through dysfunction of GH signaling. However, serum levels of ALS correlate less strongly than gene expression with hepatic GHR mRNA levels, suggesting that factors additional to GH, such as suppressed insulin levels, may affect ALS regulation. In addition, the increased bioactivity of IGF-I in PHx relative to sham control serum may be explained by the suppression of ALS. Although further investigation is required, on the basis of the data presented in this paper, we hypothesize that this increased bioavailable IGF-I may adopt the glucoregulatory role of insulin in extrahepatic tissues during the first few hours after PHx.