Gut 1995; 37: 630-638 Administration of insulin-like growth factor-I (IGF-I) peptides for three days stimulates proliferation of the small intestinal epithelium in rats C B Steeb, J F Trahair, L C Read Cooperative Research Center for Tissue Growth and Repair and Child Health Research Institute, North Adelaide, South Australia C B Steeb J F Trahair L C Read Correspondence to: Dr B Steeb, Child Health Research Institute, 72 King William Road, North Adelaide, South Australia, 5006, Australia. Accepted for publication 29 March 1995 Abstract It has previously been shown that longterm administration of insulin-like growth factor-I (IGF-I) or the analogue Long R3 IGF-I (LR3IGF-I) selectively stimulate growth of the gastrointestinal tract in gut resected, dexamethasone treated, and normal rats. In this study, the short-term effects of IGF-I administration on intestinal proliferation have been inves- tigated. Female rats (110 g, five-six/group) were infused for three days with 2.5 mg/kg/day of either IGF-I or LR3IGF-I and compared with vehicle treated or untreated control rats. LR3IGF-I but not IGF-I increased body weight and wet tissue weight of the small and large intes- tine (+20%), compared with controls. Tissue weight responses were independent of food intake and were reflected in the histology of the tissue. In LR3IGF-I treated animals, duodenal and ileal crypts length were increased by 13 and 22%, respec- tively, associated with an increase in crypt cell number. No such histological changes were seen in IGF-I treated rats. Tritiated thymidine labelling indices were signifi- cantly increased after administration of either IGF-I or LR3IGF-I (up to 14%) in both the duodenum and ileum. In IGF-I treated rats, increased nuclear labelling was not associated with an increase in the crypt compartment. In contrast, LR3IGF-I induced proportional increments in thymidine labelling and crypt size, sug- gesting that LR3IGF-I is not only more potent than the native peptide but also induced proliferative events more rapidly. In the colon, the thymidine labelling index was low, however, a non-significant increase in the number of cells labelled with thymidine was seen. These results suggest that within a three day treatment period intestinal mitogenesis is more advanced in animals treated with LR3IGF- I. The differences in proliferative response between the two peptides may be accounted for by variations in pharmaco- kinetics, clearance rates, and interactions with circulating and tissue specific binding proteins. (Gut 1995; 37: 630-638) Keywords: insulin-like growth factor, epithelium, proliferation, thymidine labelling index, crypt growth fraction. Insulin-like growth factors (IGF-I and IGF-II) are multifunctional polypeptides that possess the ability to exert insulin-like metabolic activity and regulate cell proliferation and differentiation in a variety of cell types and tissues. IGF-I mediates growth hormone dependent growth and stimulates somatic growth and that of visceral organs including the kidneys, thymus, adrenals, and the spleen. In the circulation, IGFs are bound to specific IGF binding proteins, which function as modulators of IGF action. They interact with their target tissues by membrane bound receptors to act in an autocrine, paracrine or endocrine manner, or all three. 1 IGFs are ubiquitously distributed and have been detected in tissue extracts as well as body fluids. IGF-I immunoreactivity has been shown in the human fetal stomach2 and in fetal rat intestine.3 In neonatal pigs, growth and maturation of the intestinal mucosa coincides with an increase in IGF-I immunoreactivity.4 Furthermore, IGF-I immunoreactivity has also been shown in saliva, gastric, pancreatic, and extrapancreatic excretions and in jejunal chyme.5 Several studies in animal models of gastro- intestinal adaptation have shown that adminis- tration of IGF-I peptides selectively stimulate growth of the gastrointestinal mucosa.68 More recently, we have shown that IGF-I peptides also play an important part in gastrointestinal growth and function in normal adult rats.9 In that study we gave increasing doses of IGF-I or LR3IGF-I, a potent analogue with an N-terminal extension showing reduced binding affinity to IGF-binding proteins as well as the type 1 receptor,10 for 14 days to normal, female rats. After the 14 days of treatment with either IGF-I or LR3IGF-I, body weight gain and gastrointestinal weights were significantly increased. The LR3IGF-I was severalfold more potent than the native IGF-I in all responses. Furthermore, wet tissue weights were increased in a dose dependent manner and the proximal small intestine was identified as the most responsive region. Detailed histo- logical analyses showed that IGF-I peptide administration stimulated the growth of the proliferative compartment (crypt) and the func- tional compartment (villi), so that at the end of the 14 day treatment period crypt depth and villus height were increased by up to 30% above control values. More importantly, however, IGF-I peptide administration induced a pro- portional increase in potentially proliferative 630 on 14 August 2019 by guest. Protected by copyright. http://gut.bmj.com/ Gut: first published as 10.1136/gut.37.5.630 on 1 November 1995. Downloaded from
Transcript
Gut 1995; 37: 630-638
Administration of insulin-like growth factor-I(IGF-I) peptides for three days stimulatesproliferation of the small intestinal epithelium inrats
C B Steeb, J F Trahair, L C Read
Cooperative ResearchCenter for TissueGrowth and Repairand Child HealthResearch Institute,North Adelaide, SouthAustraliaC B SteebJ F TrahairL C Read
Correspondence to:Dr B Steeb, Child HealthResearch Institute, 72 KingWilliam Road, NorthAdelaide, South Australia,5006, Australia.
Accepted for publication29 March 1995
AbstractIt has previously been shown thatlongterm administration of insulin-likegrowth factor-I (IGF-I) or the analogueLong R3 IGF-I (LR3IGF-I) selectivelystimulate growth of the gastrointestinaltract in gut resected, dexamethasonetreated, and normal rats. In this study, theshort-term effects ofIGF-I administrationon intestinal proliferation have been inves-tigated. Female rats (110 g, five-six/group)were infused for three days with 2.5mg/kg/day ofeither IGF-I or LR3IGF-I andcompared with vehicle treated oruntreated control rats. LR3IGF-I but notIGF-I increased body weight and wettissue weight of the small and large intes-tine (+20%), compared with controls.Tissue weight responses were independentof food intake and were reflected in thehistology ofthe tissue. In LR3IGF-I treatedanimals, duodenal and ileal crypts lengthwere increased by 13 and 22%, respec-tively, associated with an increase in cryptcell number. No such histological changeswere seen in IGF-I treated rats. Tritiatedthymidine labelling indices were signifi-cantly increased after administration ofeither IGF-I or LR3IGF-I (up to 14%) inboth the duodenum and ileum. In IGF-Itreated rats, increased nuclear labellingwas not associated with an increase in thecrypt compartment. In contrast, LR3IGF-Iinduced proportional increments inthymidine labelling and crypt size, sug-gesting that LR3IGF-I is not only morepotent than the native peptide but alsoinduced proliferative events more rapidly.In the colon, the thymidine labelling indexwas low, however, a non-significantincrease in the number of cells labelledwith thymidine was seen. These resultssuggest that within a three day treatmentperiod intestinal mitogenesis is moreadvanced in animals treated with LR3IGF-I. The differences in proliferative responsebetween the two peptides may beaccounted for by variations in pharmaco-kinetics, clearance rates, and interactionswith circulating and tissue specific bindingproteins.(Gut 1995; 37: 630-638)
Insulin-like growth factors (IGF-I and IGF-II)are multifunctional polypeptides that possessthe ability to exert insulin-like metabolicactivity and regulate cell proliferation anddifferentiation in a variety of cell types andtissues. IGF-I mediates growth hormonedependent growth and stimulates somaticgrowth and that of visceral organs includingthe kidneys, thymus, adrenals, and the spleen.In the circulation, IGFs are bound to specificIGF binding proteins, which function asmodulators of IGF action. They interact withtheir target tissues by membrane boundreceptors to act in an autocrine, paracrineor endocrine manner, or all three.1 IGFsare ubiquitously distributed and have beendetected in tissue extracts as well as bodyfluids. IGF-I immunoreactivity has beenshown in the human fetal stomach2 and in fetalrat intestine.3 In neonatal pigs, growth andmaturation of the intestinal mucosa coincideswith an increase in IGF-I immunoreactivity.4Furthermore, IGF-I immunoreactivity has alsobeen shown in saliva, gastric, pancreatic, andextrapancreatic excretions and in jejunalchyme.5
Several studies in animal models of gastro-intestinal adaptation have shown that adminis-tration of IGF-I peptides selectively stimulategrowth of the gastrointestinal mucosa.68 Morerecently, we have shown that IGF-I peptidesalso play an important part in gastrointestinalgrowth and function in normal adult rats.9 Inthat study we gave increasing doses of IGF-Ior LR3IGF-I, a potent analogue with anN-terminal extension showing reduced bindingaffinity to IGF-binding proteins as well as thetype 1 receptor,10 for 14 days to normal, femalerats. After the 14 days of treatment with eitherIGF-I or LR3IGF-I, body weight gain andgastrointestinal weights were significantlyincreased. The LR3IGF-I was severalfoldmore potent than the native IGF-I in allresponses. Furthermore, wet tissue weightswere increased in a dose dependent manner andthe proximal small intestine was identified asthe most responsive region. Detailed histo-logical analyses showed that IGF-I peptideadministration stimulated the growth of theproliferative compartment (crypt) and the func-tional compartment (villi), so that at the end ofthe 14 day treatment period crypt depth andvillus height were increased by up to 30% abovecontrol values. More importantly, however,IGF-I peptide administration induced a pro-portional increase in potentially proliferative
630
on 14 August 2019 by guest. P
rotected by copyright.http://gut.bm
j.com/
Gut: first published as 10.1136/gut.37.5.630 on 1 N
enterocytes as shown by immunohistochemicaldetection of cells positive for proliferative cellnuclear antigen (PCNA). The proportionalitybetween the proliferative and maturation com-partment of the crypt as well as between cryptsand the villi was maintained.9 These findingshave led to the conclusion that administrationof either IGF-I or LR3IGF-I peptide for a pro-longed period stimulate mucosal growth,resulting in a new steady state between cell lossand cell production.
This study was undertaken to identify theearly proliferative responses of the intestinalepithelium that led to this new steady stateafter IGF-I or LR3IGF-I administration. IGF-I and LR3IGF-I were compared to determineif the two peptides differed in their potency toevoke proliferative responses of the intestinalepithelium during this early period. Further-more, we investigated if the proliferativeresponses were similar in different regions ofthe intestine. In review of the results from theaforementioned study, a short-term infusionprotocol was used. The age and body weightsof the rats as well as the peptide dose andinfusion protocol have been replicated so thatdirect comparisons could be made betweenthe two studies.
Methods
RECOMBINANT IGF-I PEPTIDESRecombinant human IGF-I and the recombi-nant analogue LR3IGF-I was provided byGroPep, Adelaide, South Australia. The IGF-Ianalogue LR3IGF-I, has an arginine replacingglutamate at position 3 and an N-terminalextension comprising the amino acids Met-Phe-Pro-Ala-Met-Pro-Leu-Ser-Ser-Leu-Phe-Val-Asn.
EXPERIMENTAL DESIGN
AnimalsFemale, hooded Wistar rats of approximately100 g were obtained from the CSIRO,Division ofHuman Nutrition breeding colony.All animals were housed individually inTechniplast metabolism cages and maintainedat 25°C with a 12 hour light/dark cycle. Theywere fed a powdered diet containing 180 gcasein and 2.5 g methionine/kg body weight asthe nitrogen source. Water and food wereavailable ad libitum. The experimentalprotocol was approved by the Animal Careand Ethics Committee of the Women's andChildren's Hospital and followed theAustralian Code of Practice for the Care andUse of Animals for Scientific Purposes.
Experimental protocolIGF-I peptides were infused subcutaneouslyfor a three day period. All animals wereacclimatised to the metabolism cages for athree day period, followed by a four day pre-treatment period. During the pretreatmentperiod and the three days of peptide infusion,
daily measurements of body weight, food, andfluid intake as well as urinary and faecal outputwere taken at precisely 24 hour intervals.Between 0900 and 1200 hours on the morningafter the pretreatment period, the rats wereanaesthetised with 0.04 ml/kg Brietal forosmotic mini-pump (Alzet, Model 1 003D,Alza, Palo, CA, USA) implantation within thesubcutaneous scapular region. The pumpswere filled with either IGF-I or LR3IGF-I orcontained the vehicle alone (0.1 M aceticacid). Each rat received 2-5 mg/kg/day of eitherIGF-I or LR3IGF-I, thus at a mean pumpingrate of 0 99 ,lp/h, each rat received 278 ,ug/dayfor a three day period. There were six rats ineach of the vehicle, IGF-I, and LR3IGF-Itreated groups and a control group (n= 5),receiving no treatment or pump was alsoincluded. The study was divided into twoanimal trials. The first trial contained three ratsfrom each of the two peptide treatment groups,three vehicle treated rats, and three untreatedcontrol animals. The second trial containedthree rats from each of the peptide treated andvehicle treated groups and two untreatedcontrol animals. The trials were staggered byone day. The pumps were not primed beforeimplantation, so that in accordance with themanufacturer's instructions, the full pumpingrate would be reached approximately fourhours after insertion of the pumps between1300-1600 hours. No special postoperativecare of the animals was required.During the three days of peptide infusion,
daily measurements of body weights andmetabolic collections were continued. At theend of the three day treatment period between1300 and 1600 hours, each animal wasinjected with a single intraperitoneal injectionof 0 5 ,uCi/g body weight of tritiated thymidine(Amersham International, Buckinghamshire,England, specific activity 25 Ci/mol) in thesame order as pump implantation. Exactly onehour after the injection of the isotope, theanimal was stunned and decapitated, thus allrats received IGF-I peptides for the samelength of time. The abdomen was opened by amidline incision and the entire gastrointestinaltract was rapidly excised and placed onto anice cold glass slab. The stomach, small andlarge intestine and the caecum were isolated.The duodenum was separated from the smallintestine at the ligament of Treitz and itsempty weight and length was recorded. Theweight of the remaining small intestine, largeintestine, and the caecum were taken afterremoval of gut contents. All length measure-ments of intestinal segments were made byplacing the tissue horizontally onto a cold glassslab, avoiding the stretching of the tissue. Forhistological analyses, multiple tissue sectionswere collected from the proximal duodenum(starting 1 cm caudal to the pyloric sphincter)and the distal ileum, while large intestinalsamples were collected from the proximalcolon. The segments were rinsed in cold0-9% w/v NaCl and immediately fixed inBouin's fluid. Total gut weights and lengthwere calculated as the sum of all intestinalcomponents.
631
on 14 August 2019 by guest. P
rotected by copyright.http://gut.bm
j.com/
Gut: first published as 10.1136/gut.37.5.630 on 1 N
Histology and autoradiographyFor quantitative histological morphometry,tissue segments of the duodenum, ileum, andcolon were fixed in Bouin's fluid for four hoursand stored in 70% ethanol before processingfor routine paraffin wax embedding. Fromeach of the intestinal regions sampled, four tosix tissue segments were embedded in trans-verse orientation in the same mould and sixserially cut sections of 2 Km thickness wereprepared for each animal. The first of theserially cut sections from each animal was de-waxed, re-hydrated, stained with haematoxylinand eosin, and mounted with DePex (Gurr,BDH Chemicals, Kilsyth, Australia) for histo-logical analyses. The remaining five serially cutsections were used for autoradiography. Forthis purpose, they were de-waxed, re-hydrated,and briefly dipped in 10% lithium carbonate toreduce chemography during autoradiographicprocessing. The sections were then incubatedat room temperature for 30 seconds in auto-radiographic emulsion (LM-1, Amersham,Australia) at a dilution of 1:1 with distilledwater. After incubation, the slides were chilledon a pre-cooled tray for 10 minutes and airdried for two to three hours in the dark room.All slides were stored at 4°C in light tightphotographic slide containers and kept for2-20 days. After exposure, the slides weredeveloped in Ilford Phenisol Developer(Amersham, Australia) at room temperature(dilution 1:4) for six minutes and rinsed insodium thiosulphate (BDH Chemicals,Australia) for four minutes. The developedslides were rinsed for 15 minutes in de-ionisedwater, counterstained with haematoxylin, andmounted with DePex. Negative control slideswere included from animals not injectedwith the isotope. In addition, a single, 2 cmlong tissue segment from each region wasembedded facing serosal side down, so thattissue sections orientated longitudinal to thebowel lumen could be obtained. From seriallycut sections, tissue segments representing thecrypt:villus junction were identified andutilised to count the number of enterocyteslocated around the circumference of the crypt(crypt row count).
AnalysesHistological sections were examined with anOlympus BH-2 light microscope. Quantitativemorphometric analysis was conducted using adrawing tube attached to the microscope andmeasurements were taken using a digitisingtablet (Summa Sketch II, Summa graphics),coupled to an Apple Macintosh IIci computer.In the duodenum and ileum, crypt depth wasmeasured in 15 well oriented crypts. Thedepth of the colonic crypts was measured in 15crypts randomly selected, with care taken toavoid sections containing Peyer's patches.The dose of tritiated thymidine used in this
study produced clear labelling of S-phasenuclei, showing numerous black grainsdeposited over the nuclei and a negligiblebackground. A minimum of six grains pernuclei was used to define positively labelled
cells. Preliminary analyses determined that theslides had to be exposed for at least 10 daysto obtain adequate signal of the isotope.Accordingly, all analyses were carried out onsections exposed for 10 days. Proliferativeparameters were assessed in 30 crypts fromeach animal in the duodenum, ileum, andcolon. Analyses were confined to crypts wherethe entire length could be completely visu-alised and which contained a single layer ofepithelial cells only. In each crypt, a singlecolumn (right hand column) along the longi-tudinal axis of the crypt was assessed and thetotal number of cells and the number and posi-tion of tritiated thymidine labelled cells wasrecorded. For each ofthe intestinal regions, thelabelling index was calculated as the ratio oflabelled cells to total cell number for each cryptcolumn. In addition, the circumferential cellcount (crypt row count), measured as thenumber of epithelial cells around the circum-ference at the crypt:villus junction wasmeasured in the duodenum and ileum in serialcut sections from tangentially embeddedmaterial. The product of the crypt columncount and the crypt row count was used toestimate the total crypt cell population in allsmall intestinal regions.
For each animal, thymidine labelling indexdistribution profiles were established for theduodenum and the ileum. From these curves,the cell position within the crypt at whichmaximal thymidine labelling occurs and thecell position of half maximum thymidinelabelling was identified. The maturation com-partment of the crypt in which epithelial cellshave lost their proliferate capabilities andacquire their mature, functional properties wasidentified from these curves as the regionabove the last labelled cells within the crypt.To determine if IGF-I peptides increasethymidine incorporation into enterocytes inthe lower and mid-crypt region, the cumulativenumber of cells labelled with tritiatedthymidine up to cell position 19 was calcu-lated. The crypt growth fraction, whichidentifies the proportion of proliferating cellswithin the crypt, was calculated for eachanimal from the thymidine labelling distribu-tion profiles of 30 perfectly orientated cryptsby dividing the cell position at which halfmaximum labelling occurred by the totalnumber of cells per cell column.
Statistical analysesAll values in Tables and Figures are expressedas means (SEM). All groups were compared bya one way analysis of variance (ANOVA)and where significance was achieved (p<0.05)a post-hoc Dunnett's test (SuperANOVA,Abacus Concepts, Berkeley, CA) was appliedto identify variations between treatmentgroups and either vehicle treated or controlanimals. To examine the degree to which bodyweight gain and fluid balance vary after IGFpeptide treatment, data were analysed byproduct-moment correlations and significancewas tested with a t test with n-2 degrees offreedom.1'
632
on 14 August 2019 by guest. P
rotected by copyright.http://gut.bm
j.com/
Gut: first published as 10.1136/gut.37.5.630 on 1 N
TABLE I Body weight (g), and body weight gain (g/3 days) and average food con,(g/24 hours) in rats treatedfor three days with 2.5 mg/kg/day ofeither IGF-I or L.Rcompared with vehicle treated or untreated control rats
Treatment Body weight Gain Averagegroup starting (g) Final (g) (g/3 days) intake (~
All values are expressed as means (SEM). Statistical significance from the vehicle treatis indicated by *:p<0.01 and from the untreated control group by t:p<001, ANOVA.Dunnett's post-hoc test (two tailed).
ResultsThe body weight at the start of th(averaged 97.9 (1.0) g (n=23) for all ratfour days of acclimatisation to the met;cages the average body weight incre;109.9 (1.0) g. Rats were then randomistreatment groups such that no statsignificant differences in body weigiapparent between treatment groups (TInfusion of 2.5 mg/kg/day of LR3IGFthree day period resulted in a signihigher body weight (134.9 (1 1) g) co:with animals treated with vehicle (127g) or normal rats (control group, 122-3without an implanted mini-pump (TThe body weight gain in LR3IGF-Ianimals could not be attributed to an iin food intake, as food consumptiiapproximately 13 g/day for all groupsI). Accordingly, food conversion effcalculated as the ratio of average dail
0.7 r
0.6 h
a,(L)(D
4-W)
ccc
.)
0)
C.)
c
a)
._
a)
c0
C.)
h00
,o0
0.5
0.4
0.3 h T
0.2
0-1
T
0 C)VIL L
c
Figure 1: Food conversion efficiency (g body weight gainlgfeed) in normal adult rats treatedfor three days with 2.5mg/kg/day ofIGF-I or LR3IGF-I compared with eithervehicle or untreated control rats. Values are means (SEM)for 6 rats/group (n=5, control group). Statisticalsignificance from vehicle treated rats is indicated by*:p<O.OOO1 and by t.:p<0 0001 (untreated control group)as detected byANOVA.
weight gain to food consumption for the threeday treatment period was highly significant inthe LR3IGF-I treated group (p<0.0001) whencompared with either vehicle, control or IGF-Itreated animals (Fig 1). To determine if theaccelerated weight gain in the LR3IGF-Itreated animals could be attributed to fluidretention, the fluid balance (fluid intake/24hours minus fluid output/24 hours) wasestimated for all animals. Although fluid intakewas highly variable throughout the three daytreatment period in all groups, no statisticallysignificant difference was detected in eitherfluid intake or urinary output over the experi-mental period, arguing against fluid retentionas a possible mechanism for the extra weightgain seen in LR3IGF-I treated rats (results notshown). To examine the degree to which bodyweight gain and the fluid balance correlateafter IGF-I peptide treatment, product-moment correlation coefficients were calcu-lated for the combined data. No statisticalsignificant correlation was found between bodyweight gain and the fluid balance during thethree day treatment period in either treated oruntreated rats (robtained=0.27, rcritical=0A42 atp<005).
Gastrointestinal weight in the animals fittedwith the vehicle pump was similar to the gutweight of untreated control rats, showing thatimplantation of the pumps had no effect ongastrointestinal tissue growth. Comparisonbetween the vehicle group and the animalstreated with 2.5 mg/kg/day of LR3IGF-Ishowed, however, that the increased bodyweight gain in the LR3IGF-I group wasreflected in the wet tissue weights of theirgastrointestinal tissues. Total gut weight, smalland large intestinal weight were increased by19%, 22%, and 21% respectively comparedwith vehicle treated animals (Fig 2 (A), (C),and (D)), while stomach weight increased by12% after treatment with LR3IGF-I (Fig 2(E)). Selective action of LR3IGF-I on theintestinal tissues was evident when correctionsfor body weight gain were made, so that frac-tional gut weight (total gut weight/kg bodyweight) was significantly increased in LR3IGF-I treated animals (56-6 (1-4), p<O0O1) com-pared with either IGF-I (48.5 (1.3)) or vehicletreated control animals (503 (1. 1)), Fig 2 (B).The increase in gastrointestinal tissue weight inLR3IGF-I treated animals contrasts consider-ably with the results obtained for IGF-I treatedanimals, so that infusion of 2.5 mg/kg/day ofrecombinant IGF-I did not affect body weightgain or gastrointestinal tissue weights (Table Iand Fig 2 (A-E)). Increases in intestinal lengthwere not seen in any of the animals treated forthree days with the IGF-I peptides (results notshown).We have previously shown that longterm
administration of 2-5 mg/kg/day of IGF-Ipeptides to normal female rats selectivelystimulated the growth of intestinal mucosa, sothe main focus here was to assess the mito-genicity of IGF peptides during the initialperiod of peptide administration. In this study,administration ofLR3IGF-I for only three daysresulted in a statistically significant increase
633
on 14 August 2019 by guest. P
rotected by copyright.http://gut.bm
j.com/
Gut: first published as 10.1136/gut.37.5.630 on 1 N
Figure 2: Total gut weight (A), fractional gut weight (B), small intestinal weightlarge intestinal weight (D), and stomach weight (E) in normalfemale rats treateddays with either 2.5 mg/kg/day of insulin-like growth factor-I peptides compared uvehicle or untreated control rats. All values are means (SEM) with 6 rats/groups (control group). Significance from the vehicle group is indicated by *:p<001 andjcontrol groups by tf:p<0 01. A significant difference between LR3IGF-I and the s,ofIGF-I is indicated by ..p<0 01, as detected byANO VA, Dunnett's (two tailed
(p<O0Ol) in crypt depth, both in thdenum (+8%) and in the ileum (-compared with vehicle treated animalsII). The increased crypt depth in thiswas associated with a proportional incithe number of cells per crypt columnII). Furthermore, the circumferenti
count (crypt row count) was also significantlyincreased in both small intestinal segments,resulting in an overall increase in crypt cellpopulation by approximately 30% (Table II).Administration of LR3IGF-I also stimulatedproliferation of the colonic mucosa within thethree day infusion period. Colonic crypt depthin LR3IGF-I treated animals was marginallyincreased in comparison with vehicle treatedanimals but not control treated animals; as inthe small intestine, colonic crypt hyperplasiawas accompanied by a statistically significantincrease in the crypt cell column count (TableII).To further assess the mitogenic properties of
the IGF-I peptides on the intestinal epithe-lium, we constructed thymidine labelling dis-tribution profiles of the duodenum and ileumfor each animal. In the duodenum of control orvehicle treated animals a total of 29-30% ofthe crypt cells were labelled with the isotope(crypt labelling index), Table III. In bothgroups, the thymidine labelling indices were
low in the basal cell positions (cell position1-4). From cell position 5-16, however,labelling indices increased, reaching a maximallabelling of approximately 60% for control andvehicle treated animals (Table III and Fig 3(A) and (B)). After cell position 16, prolifera-tive indices declined to reach half maximumlabelling at cell position 20 and from cell posi-tion 29-35 in the crypt column, no labelledcells were apparent (Fig 3 (A) and (B)). Thusthe maturation compartment of the crypt was
identified from cell position 29 upwards. TheIL crypt growth fraction for the vehicle and
untreated control rats was calculated atapproximately 60%.
In the duodenum of the LR3IGF-I treatedanimals, the crypt cell labelling index wassignificantly increased (32.7%) compared withthe control groups (29-30%), Table III. As forthe control and vehicle groups, low prolifera-tive indices were evident in the first few cellpositions, however, increased crypt celllabelling was reflected in an increasedmaximum labelling, calculated at 66%, at cellpositions 5-16 (Table III and Fig 3 (D)).Furthermore, the number of thymidinelabelled cells up to cell position 19 was alsosignificantly greater in LR3IGF-I treated ratscompared with either vehicle or untreatedcontrol rats (Table III). After cell position 16,the per cent thymidine labelling declined moreslowly than in the vehicle or control group, sothat half maximum labelling was not reached
(C) until cell position 24. Thymidine labelling wasIfor three detectable up to cell position 36 in this group,vith showing that the maturation compartment(n=5, (cell position 36-42) was shifted upwards, iname dose proportion with the increase in the cryptsd). compartment. This shows that LR3IGF-I
administration for three days led to a signifi-ie duo- cant increase in the number of crypt cells,+ 13%), associated with an increase in the proportion of(Table cells labelled, in particular in the lower and
s group mid-crypt region (up to cell position 19) andrease in an increase in maximal labelling. Most import-(Table antly, however, LR3IGF-I also increased the
ial cell cell position of half maximum labelling in
634
on 14 August 2019 by guest. P
rotected by copyright.http://gut.bm
j.com/
Gut: first published as 10.1136/gut.37.5.630 on 1 N
TABLE II Histological parameters in the duodenum, ileum, and colon offemale rats treatedfor three days with or without 2.5 mg/kg/day ofIGF-I or LR3IGF-I
Treatment group
Histological parameter Control Vehicle IGF-I LR3IGF-I
DuodenumCrypt depth (pm) 260 (7) 255 (5) 257 (5) 293 (14)*tCrypt column count (no of cells) 34-1 (1-1) 35-1 (0.2) 33-7 (0.5) 41-7 (0.9)t§Crypt row count (no of cells) 19-3 (0.3) 20-0 (0.3) 19-8 (0.3) 21-8 (0.5)t§Crypt population (no of cells) 656 (14) 702 (13) 668 (9) 910 (22)t§
IleumCrypt depth (pm) 190 (8) 203 (7) 192 (5) 231 (4)t§Crypt column count (no of cells) 29-4 (0.4) 29-2 (0.6) 29-3 (0.6) 33-5 (0-8)t§Crypt row count (no of cells) 20-3 (0.4) 20-0 (0.1) 20-1 (0.3) 21-6 (0.4)t§Crypt population (no of cells) 596 (16) 582 (11) 588 (16) 723 (26)t§
Values are expressed as means (SEM). Statistically significant difference from the vehicle groupis indicated by *:p<0.05 and t:p<0 01. Difference from the control group is indicated byt:p<0 05 and §:p<0-01. n=6 Animals per group except for the control group where n=5.
proportion to the total number of crypt cells andhence, the crypt growth fraction remained atapproximately 60%, Table III and Fig 3 (D).The labelling distribution profiles in IGF-I
treated rats also showed an increase in theoverall crypt cell labelling index, which wasvirtually identical to that in LR3IGF-I treatedrats (Table III). In this group, the number ofcells labelled with tritiated thymidine in basaland mid-crypt enterocytes (up to cell position19) was also significantly increased (Table III),so that maximum thymidine labelling (66.7%)was reached within cell positions 5-16.Consequently, thymidine labelling in the lowerand mid-crypt region in IGF-I treated animalswas increased by 9% and 14% compared withvehicle treated or control rats, respectively.However, this failed to achieve statistical sig-nificance (p<0 067) (Table III and Fig 3 (C)).In contrast with the LR3IGF-I group, nuclearlabelling declined more rapidly after cell posi-tion 16 (similar to the vehicle and controlgroup) so that half maximum labelling wasreached at cell position 20 in IGF-I treatedrats, which was virtually identical to controlvalues (Table III and Fig 3 (C)). As the totalnumber of cells per crypt column was notchanged in this group, the overall crypt growthfraction was maintained at 58% and compar-able with the crypt growth fraction of the con-trol groups (Table III). Thus, the mechanismby which IGF-I induced epithelial proliferationin this short-term administration protocol
differed from that of the LR3IGF-I group.Despite the considerable increase in the pro-liferative pool, short-term IGF-I adminis-tration did not increase the cellularity of thecrypt. Increased crypt labelling was thereforeachieved by increasing the proportion ofproliferative cells in lower and mid-crypt posi-tions as shown by the increase in maximallabelling rather than a recruitment of prolifera-tive cells in higher cell positions, as seen in theLR3IGF-I group.
In the ileum, the overall crypt labelling indexwas lower than in the duodenum (Table III).This was reflected by lower thymidine labellingindices in the basal positions as well as a lowermaximal labelling. On the other hand, theposition of halfmaximum labelling was similarin the two intestinal regions (Table III). As forthe duodenum, LR3IGF-I treated animalsshowed an increased number of proliferativecells in higher cell positions (as shown by theupward shift in the 1/2 maximum labelling),while IGF-I treatment seemed to increase theproportion of proliferative enterocytes in themid-crypt cell positions (Table III), reaching amaximum thymidine labelling index of 60%compared with 54-57%/o in the control groups(Table III). This led to a non-significant(p<0 08) increase in the crypt growth fractionin this group (Table III).
Finally, the mitogenic response of thecolonic mucosa to IGF-I peptides was assessedin tissue segments from the proximal colon.Although, similar changes in wet tissue weightin LR3IGF-I treated rats were seen for thesmall and large intestine, thymidine labellingindices were not increased by IGF-I peptides(Table IV). Nevertheless, in LR3IGF-I treatedrats, an average of 3-7 cells per crypt werelabelled compared with 2.9 cells/crypt invehicle or untreated control rats (Table IV).Although, statistical significance was notreached, it is possible that IGF-I peptideseffects in the colon have been disguised by theinherent low proliferative activity of the colon.In this study, thymidine labelling indices were11-13% for all groups with no statistical dif-ference between treatment groups. The site atwhich histological samples for the colon weretaken show a great deal of heterogeneity incrypt morphology. For example, the proximal
TABLE III Proliferative parameters in the duodenum and ileum of rats treatedfor three days with 2.5 mg/kg/day ofIGF-Ior LR3IGF-I as compared with vehicle and untreated control rats
Treatment group
Proliferative parameter Control Vehicle IGF-I LR3IGF-I
DuodenumCrypt labelling index (%) 29-9 (1-4) 29-1 (0.6) 33-3 (0.7)t§ 32-7 (0o7)**Maximal labelling (%) 58-3 (2.9) 61-1 (2.2) 66-7 (1.9) 65-6 (2.0)Labelling up to cell position 19 (no of cells) 7-12 (0.38) 7-32 (0.09) 8-45 (0.07)*§ 8-34 (0-21)*§Cell position at l/2 maximum labelling 19-6 (0.6) 20-1 (0.5) 20-0 (0.9) 24-3 (0.8)*§Crypt growth fraction(%) 59.0 (0-7) 57-4 (1-3) 58-5 (2-1) 58-4 (1.9)
IleumCrypt labelling index (%) 262 (0.7) 26-0 (0.6) 29-2 (0.6)t- 29-4 (0.6)t§Maximal labelling (/o) 56-7 (3.5) 54-6 (1-5) 60-0 (2.0) 55-2 (2-1)Labelling up to cell position 19 (no of cells) 6-88 (0.08) 6-92 (0.02) 7-33 (0.06)t* 7-58 (0.13)ttCell position at 1/2 maximum labelling 19-3 (0.6) 18-7 (0.9) 20-0 (0.3) 23-3 (0.7)t§Crypt growth fraction (/o) 66-3 (2.4) 64-2 (1-6) 69-5 (0.5) 69-5 (1-6)
Values are expressed as means (SEM). Statistically significant difference from the vehicle group is indicated by *:p<0.05 andt:p<0 01. Differences from the untreated control rats are indicated by *:p<005 and §:p<0-01. n=6 Animals per group exceptfor control group where n=5. The crypt labelling indices represent the proportion of tritiated thymidine labelled cells per cryptcolumn and was calculated from 30 full length open crypt columns for each animal. Maximal labelling occurred in enterocytes inmid-crypt cell positions and was calculated from the top three thymidine labelled enterocytes within cell positions 5-16.
635
on 14 August 2019 by guest. P
rotected by copyright.http://gut.bm
j.com/
Gut: first published as 10.1136/gut.37.5.630 on 1 N
Figure 3: Tritiatedthymidine labellingdistribution profiles in tissuesections from the duodenumin untreated control rats(A), rats treated withvehicle (B) or rats treatedwith 2-5 mg/kg/day ofeither IGF-I (C) orLR3IGF-I (D). All valuesare represented as means(SEM) of 6 rats/group(n= 5, control and IGF-I).Sections from one animal inthe IGF-I group wereexcluded because of veryweak incorporation of theisotope. For all animals themean tritiated labellingindex (%) was calculatedfrom 30 full-length opencrypts. Maximal labelling(%) represents the averagepercentage of labelling inthe top three labelledenterocytes within cellposition 5-16. The cryptgrowth fraction wascalculatedfor each animalindividually as the cellposition of half maximumlabelling divided by thetotal number of cells percrypt column. The cryptgrowth fraction wascalculated as the cellposition of 1/2 maximumlabelling divided by thetotal number of cells percrypt column.
a 60CDcn= 50a0X' 400.
2 30E= 20
10
0
80
70
^ 60
CDc 50
s 400)
.5 30
._
= 20
10
Cell position Cell position
colon contains mucosal rugae and the crypts atthe apex of mucosal folds are longer than thecrypts at the base of the crypts. Consequently,the colonic crypt cell population could not beestablished because serial cut sections of longi-tudinal embedded material does not provide auniform display of crypts suitable for circum-ferential crypt row counts in this region.Because of the low proliferative activity in thecolon, labelling distribution curves were alsonot established for this region.
DiscussionWe have previously shown that longterminfusion of IGF-I and in particular LR3IGF-I
TABLE IV Proliferative parameters in the colon of rats treatedfor three days with2.5 mg/kg/day ofIGF-I or LR3IGF-I, compared with vehicle infused or untreated controlrats
Treatment group
Proliferative parameter Control Vehicle IGF-I LR3IGF-I
Values are expressed as means (SEM). Statistically significant difference from the vehicle groupis indicated by *:p<0.05. Differences from the untreated control rats are indicated by t:p<005.n= 6 Animals per group except for IGF-I and control group where n= 5. The crypt labellingindices represent the proportion of tritiated thymidine labelled cells per crypt column and wascalculated from 30 full length open colonic crypts of the proximal colon.
significantly enhanced mucosal growth innormal female rats.9 Treatment with eitherpeptide resulted in a crypt hyperplasia withproportional increments in the percentage ofcells labelled with PCNA, indicating that after14 days of treatment with IGF-I peptides anew balance between crypt cell production andcell loss had been established. In this study, wehave clearly shown that administration ofIGF-I or LR3IGF-I for three days to adultfemale rats elicits early proliferative events thatlead to the massive increase in mucosal massseen in the longterm infusion study. Moreover,comparing the total gut weight (wet tissueweight) of the rats treated for three days with2.5 mg/kg/day of LR3IGF-I with the total gutweight of rats treated for 14 days with the samepeptide dose shows that 44% of the weightgain has occurred during the three day treat-ment period. In agreement with our findings,Olanrewaju et al,12 have shown that infusion of10 nM of hrIGF-I for three days into the ileallumen of adult male Sprague-Dawley ratssignificantly increased mucosal mass (wettissue weight) and the mucosal cellularity (asmeasured by DNA, RNA, and protein contentof tissue homogenates). In addition, IGF-Iadministration effectively induced the growthrelated enzyme ornithine decarboxylase.12The mitogenic properties of IGF-I to
636
on 14 August 2019 by guest. P
rotected by copyright.http://gut.bm
j.com/
Gut: first published as 10.1136/gut.37.5.630 on 1 N
stimulate proliferation of intestinal epithelialcells has been well demonstrated in a numberof in vitro studies. IGF-I is a potent mitogenfor human fetal small intestinal cells.13Administration of IGF-I to IEC-6 cells, a cellline derived from rat jejunal crypts, stimulateDNA and protein synthesis. Likewise, IGF-Istimulates cellular proliferation in RIE-1 cells,an epithelial cell line derived from rat smallintestine.14 The tritiated thymidine uptake bycanine fundic epithelial cells is also stimulatedby IGF-I administration although EGF andinsulin are also able to elicit a mitogenicresponse in these cells, much higher concentra-tion of these growth factors are needed toachieve an equivalent effect.'5 In this study,administration of IGF-I and LR3IGF-I for athree day period stimulated the thymidineincorporation into crypt epithelial cells in vivo,supporting that in the adult rat, IGF-I peptidesare important intestinal mitogens.From the thymidine labelling distribution
profiles in the control groups, it was evidentthat maximal thymidine incorporation wasgreatest in the mid-crypt region (cell position5-16), which represents the zone of greatestcell production. As shown by Wright,'6 this isthe 'proliferative compartment proper' and thelabelling index obtained in this region in ourstudy compares well to the theoretical labellingindex of 60%. After treatment with IGF-I, anincrease in the number of proliferative cells inlower and mid-crypt cells positions showedthat more cells in this region of the crypt hadentered the cell cycle, leading to the observedincrease in the thymidine labelling index in thisgroup. In addition, the maximum labelling wasalso increased, which shows that more cells inthe proliferative compartment proper werecycling. The most probable mechanism bywhich such a rapid increase in proliferativeactivity may have been achieved is a reductionin the cell cycle in enterocytes in basal cellpositions. Cell cycle times in basal cellpositions, as measured by fraction labelledmitosis, stathmokinetic or continuous labellingmethods, are prolonged compared with cellcycle times of enterocytes in the upper cryptregion,'7 so that a fractional decrease in the cellcycle time in basal positions would greatlyincrease the total proliferative pool within thecrypt.The most interesting finding of this study
was the fact that although both IGF peptidessignificantly increased the thymidine incorpor-ation into the crypt enterocytes, increasedproliferative activity associated with anincrease in the size of the crypt compartmentwas only seen in the LR3IGF-I treated animals.Although we cannot exclude the possibilitythat the mechanism by which the two peptidesinduce proliferative events differs, it is highlylikely that the proliferative responses seen inanimals treated with LR3IGF-I represent amore 'advanced' stage of intestinal prolifera-tion compared with the IGF-I treated rats.This is supported by the fact that in rats treatedfor 14 days with either IGF-I or LR3IGF-I aconsiderable increase in size of the cryptcompartment was seen, for both groups. This
suggests that although both peptides initiateincreased thymidine incorporation after threedays, in LR3IGF-I treated animals this hasbeen already translated into an increase incrypt size, which in IGF-I treated animalsoccurs some time later. Thus, the proliferativeeffect seen in the rats infused with IGF-Ishould be detectable after administration ofLR3IGF-I for only one or two days.
Another point of interest was the findingthat administration of LR3IGF-I resulted in agreater proportion of cycling cells in uppercrypt cell positions, showing that migratingenterocytes had retained their proliferativecapacities and had not entered the maturationcompartment of the crypt. Moreover, it seemsthat in LR3IGF-I treated animals the prolifera-tive response was approaching the new steadystate equilibrium between cell production andcell loss that was seen in the rats treated for 14days with IGF-I peptides. This may have beenachieved by a change in enterocyte transittime.The accelerated proliferative effects of
LR3IGF-I may have been the result of severalinteracting factors. LR3IGF-I has a severalfoldlower affinity towards IGFBP-3, IGFBP-4,total rat plasma, and L6 myoblast bindingproteins.'8 Despite the reduced affinity tobinding proteins, however, LR3IGF-I hasshown a substantially greater bioactivity thanIGF-I in several functional assays associatedwith growth in L6 myoblasts and H35hepatoma cells.1i This increased potency isseen despite the fact that the analogue bindswith approximately fourfold lower affinity tothe type 1 receptor.'0 It is possible thatinfusion of 2.5 mg/lkg/day of LR3IGF-I mayhave increased the free IGF pool in plasma to agreater extent than infusion of the IGF-I. Thiswould lead to a greater amount of IGF mole-cules free to interact with their respectivereceptors stimulating intestinal mitogenesisthrough signal transduction pathways, thus inLR3IGF-I treated animals proliferative eventsmay have induced more rapidly than in IGF-Itreated rats.
Another contributing factor is the fact thatLR3IGF-I is also cleared more rapidly from thecirculation.19 For example, the metabolicclearance rate for LR3IGF-I is approximately1 1-fold higher than for IGF-I in adult femaleSprague-Dawley rats. 19 In addition, location ofradiolabelled IGF-I or LR3IGF-I in visceralorgans and tissues differs between the twopeptides.i9 The increased potency seen withLR3IGF-I may also.result from reduced inter-action of the analogue with locally producedbinding proteins. Most probably, a com-bination of complex interactions between theIGF-I ligand, endogenous IGF bindingproteins present in the serum and tissues,clearance rates, and tissue distribution and theavailability of the analogue to the IGF receptordetermined the rapid mitogenic response.
Direct action of IGF-I peptides on thegastrointestinal tract is supported by thelocalisation of IGF receptors along the entirelength of the rat intestine.20-24 Similarly, asingle class ofhigh affinity IGF-I receptors have
637
on 14 August 2019 by guest. P
rotected by copyright.http://gut.bm
j.com/
Gut: first published as 10.1136/gut.37.5.630 on 1 N
been localised to the muscular and mucosallayer of the gastrointestinal tract in rabbits25and both type 1 and type 2 receptors are pre-sent in the porcine small intestine.4 As shownby membrane receptor binding studies in therat, 1251-IGF-I binding to IGF-I receptors wasfourfold higher in proliferative crypt cells thanin villus cells and IGF-I receptor densities aregreater in the lamia propria than in the surfaceepithelium in adult rat intestine.23 This showsthat the mitogenic properties of IGF-I peptidesare more probably mediated by a direct interac-tion of the IGF-I ligand with the cell surfacereceptor. Indirect action of IGF peptides ongastrointestinal tissues can, however, not bedismissed and additional and synergistic inter-actions of IGF-I and other growth factors havebeen reported.'3 26
In agreement with previous studies,89 theproximal small intestine was identified as themost responsive region despite the fact that inthe rat, binding of 1251-IGF-I is lower in cryo-stat sections of the proximal small intestinecompared with the distal small intestine or thecolon.20 23 In this study, administration ofLR3IGF-I increased the wet tissue weight ofthecolon to a similar extent to that of the smallintestine, yet an increase in thymidine labellingwas not seen. The proliferative effect of IGF-Ipeptides may have been somewhat disguised bythe inherent low proliferative activity of thecolon. In the colon, variation in cryptmorphology, in particular in the proximalcolon, lead to a great deal of heterogeneity inmorphometric and cell kinetic parameters. Forexample, in the rat, colonic crypts becomelonger and more slender with distance awayfrom the ileocaecal junction. Furthermore,mucosal rugae are prominent in the proximalcolon but not in the distal colon and crypts atthe apex of the mucosal folds are longer thanthe crypts at the base of the folds (reviewed byWright and Alison, ref 16). Thus, to estimatemore accurately the proliferative response ofIGF-I peptides on the colonic mucosa, directmeasurements of cell cycle time or the crypt cellproduction rates, or both, need to be taken.
In summary, this study has shown thatadministration ofLR3IGF-I for only three daysstrongly stimulated intestinal proliferation innormal adult rats. LR3IGF-I but not IGF-Iincreased the wet tissue weight of intestinalcomponents independent of food intake. Inthis group, the crypt length and crypt cellnumber were increased in the duodenum andileum. Although thymidine labelling indiceswere increased for both peptides, no change inthe crypt length and crypt cell populationwas seen for IGF-I treated rats, suggestingthat LR3IGF-I is not only more potent instimulating intestinal proliferation but inducesmitogenesis more rapidly.
1 Lund PK. Insulin-like growth factors. In. Walsh JH,Dockray GH, eds. Gut peptides: biochemistry and physiol-ogy. New York: Raven Press, 1994: 587-613.
2 D'Ercole AJ, Hill DJ, Strain AJ, Underwood LE. Tissue andplasma somatomedin-C/insulin-like growth factor I con-centrations in the human foetus during the first half ofgestation. Pediatr Res 1986; 20: 253-5.
3 Romanus JA, Yang YWH, Adams SO, Sofair AN, TsengLYH, Nissley SP, et al. Synthesis of insulin-like growthfactor II (IGF-II) in foetal rat tissues: translation of IGF-II ribonucleic acid and processing of pre-pro-IGF-II.Endocrinology 1988; 122: 709-16.
4 Schober DA, Simmen FA, Hadsell DL, Baumrucker CR.Perinatal expression of type 1 IGF receptors in porcinesmall intestine. Endocrinology 1990; 126: 1125-32.
5 Chaurasia OP, Marcuard SP, Seidel ER. Insulin-like growthfactor I in human gastrointestinal exocrine secretions.Regul Pept 1994; 50: 113-9.
6 Lemmey AB, Martin AA, Read LC, Tomas FM, OwensPC, Ballard FJ. IGF-I and the truncated analogue des(1-3)IGF-I enhance growth in rats after gut resection. Am JfPhysiol 1991; 260: E213-9.
7 Vanderhoof JA, McCusker RH, Clark R, MohammadpourH, Blackwood DJ, Harty RF, et al. Truncated and nativeinsulin-like growth factor-I enhance mucosal adaptationafter jejunoileal resection. Gastroenterology 1992; 102:1949-56.
8 Read LC, Tomas FM, Howarth GS, Martin AA, Edson KJ,Gillespie CM, et al. Insulin-like growth factor-I and itsN-terminal modified analogues induced marked gutgrowth in dexamethasone treated rats. J Endocrinol 1992;133: 421-31.
9 Steeb C-B, Trahair JF, Tomas FM, Read LC. Prolongedadministration of IGF peptides enhances growth of gas-trointestinal tissues in rats. Am J Physiol 1994; 266:1090-8.
10 Francis GL, Ross M, Ballard FJ, Milner SJ, Senn C,McNeil KA, et al. Novel recombinant fusion-proteinanalogues of insulin-like growth factor I (IGF-I) indicatethe relative importance of IGF-binding protein andreceptor binding for enhanced biological potency. J MolEndocrinol 1992; 8: 213-23.
14 Corps AN, Brown KD. Stimulation of intestinal cell pro-liferation in culture by growth factors in human andruminant mammary secretions. Jf Endocrinol 1987; 113:258-90.
15 Chen MC, Lee AT, Soll AH. Mitogenic response of caninefundic epithelial cells in short-term culture to transform-ing growth factor and insulin-like growth factor I. Jf CliInvest 1989; 87: 1716-23.
16 Wright NA. The organization of epithelial cell populations.In: Appleton DR, Sunter JP, Watson AJ, eds. Cell pro-liferation in the gastrointestinal tract. Bath: Pitman Press,1980: 3-21.
17 Al-Dewachi HS, Wright NA, Appleton DR, Watson AJ.The cell cycle time in rat jejunal crypt. Cell Tissue Kinet1974; 7: 587-94.
18 Ballard FJ, Walton PE, Bastian S, Tomas FM, Wallace JC,Francis GL. Effects of interactions between IGFBPs andIGFs on the plasma clearance and in vivo biologicalactivities of IGFs and IGF analogs. Growth Regul 1993; 3:40-4.
19 Bastian SEP, Walton PE, Wallace JC, Ballard FJ. Plasmaclearance and tissue distribution of labelled insulin-likegrowth factor-I (IGF-I) and LR3IGF-I in pregnant rats. JfEndocrinol 1993; 138: 327-36.
20 Laburthe M, Royer-Fessard Ch, Gammeltoft S. Receptorsfor insulin-like growth factors I and II in rat gastrointesti-nal epithelium. Am Jf Physiol 1988; 254: G457-62.
21 Ryan J, Costigan DC. Determination of the histologicaldistribution of insulin-like growth factor I receptors in therat gut. Gut 1993; 34: 1693-7.
22 Young GP, Taranto TM, Jonas HA, Cox AJ, Hogg A,Werther GA. Insulin-like growth factor and the develop-ing and mature rat small intestine: receptors and biologi-cal actions. Digestion 1990; 46: 240-52.
23 Heinz-Erian P, Kessler U, Funk B, Gais G, Kiess W.Identification and in situ localisation of the insulin-likegrowth factor-II/mannose-6-phosphate (IGF-II/M6P)receptor in the rat gastrointestinal tract: comparison withthe IGF-I receptor. Jf Endocrinol 1991; 129: 1769-77.
24 MacDonald RS, Park JHY, Thornton WH. Insulin, IGF-Iand IGF-II receptors in rat small intestine followingmassive small bowel resection: analysis by binding, flow-cytometry and immunohistochemistry. Dig Dis Sci 1993;38: 1658-69.
25 Termanini BR, Nardi V, Finan TM, Parikh I, Korman LY.Insulin-like growth factor I receptors in the rabbit gas-trointestinal tract. Gastroenterology 1990; 99: 51-60.
26 Lynch SE, Colvin RB, Antoniades HN. Growth factors inwound healing: single and synergistic effects of partialthickness porcine skin wounds. Jf Clin Invest 1989; 84:640-6.
on 14 August 2019 by guest. P
rotected by copyright.http://gut.bm
j.com/
Gut: first published as 10.1136/gut.37.5.630 on 1 N
