The stem cells of small intestinal crypts: where are they?

Recently, there has been resurgence of interest in the question of small intestinal stem cells, their precise location and numbers in the crypts. In this article, we attempt to re-assess the data, including historical information often omitted in recent studies on the subject. The conclusion we draw is that the evidence supports the concept that active murine small intestinal stem cells in steady state are few in number and are proliferative. There are two evolving, but divergent views on their location (which may be more related to scope of capability and reversibility than to location) several lineage labelling and stem cell self-renewing studies (based on Lgr5 expression) suggest a location intercalated between the Paneth cells (crypt base columnar cells (CBCCs)), or classical cell kinetic, label-retention and radiobiological evidence plus other recent studies, pointing to a location four cell positions luminally from the base of the crypt The latter is supported by recent lineage labelling of Bmi-1-expressing cells and by studies on expression of Wip-1 phosphatase. The situation in the human small intestine remains unclear, but recent mtDNA mutation studies suggest that the stem cells in humans are also located above the Paneth cell zone. There could be a distinct and as yet undiscovered relationship between these observed traits, with stem cell properties both in cells of the crypt base and those at cell position 4.

Epithelial stem cell-related alterations in Trichinella spiralis-infected small intestine.

OBJECTIVES: Infection of mice with the parasite Trichinella spiralis leads to small intestinal inflammation, characterized by changes in mucosal architecture and subpopulations of epithelial cells. This model has been used to explore changes in the epithelial proliferative cell population and expression of transforming growth factor-beta (TGF-beta). MATERIALS AND METHODS: Histochemical and immunohistochemical studies were undertaken in duodenal samples. Location and number of Ki-67-positive cells were assessed using Score and Wincrypts program. Changes in mRNA transcripts were studied by real-time RT-PCR. RESULTS: T. spiralis infection induced an increase in total number of proliferative (Ki-67-positive) cells per half crypt on day 2 post-infection. Transcription of Math1, a transcription factor required for secretory cell differentiation in the intestine, was up-regulated on days 6-18 post-infection. At these time points, numbers of Paneth cells at the crypt base were also increased and the epithelial proliferative zone was shifted up the crypt-villus axis. Transcription of TGF-beta isoforms within the small intestine was up-regulated on days 6 and 12 post-infection, but anti-TGF-beta antibody treatment had no effect on T. spiralis-induced changes in mucosal architecture or increase in Paneth/intermediate cells. CONCLUSIONS: T. spiralis infection promotes an initial increase in small intestinal epithelial proliferation and subsequent cell differentiation along the secretory cell lineage. The resulting increase in numbers of Paneth cells at the crypt base causes the proliferative zone to move up the crypt-villus axis. Further studies are required to determine the significance of an increase in the expression of TGF-beta transcripts.

Characterization of putative stem cells in isolated human colonic crypt epithelial cells and their interactions with myofibroblasts.

Colonic epithelial stem cells are believed to be located at the crypt base where they have previously been shown to express musashi-1. The colonic stem cell niche, which includes extracellular matrix and myofibroblasts (together with other cell types), is likely to be important in maintaining the function of the progenitor cells. The aims of our studies were to characterize stem cells in isolated and disaggregated human colonic crypt epithelial cells and investigate their interactions with monolayers of primary human colonic myofibroblasts. In unfractionated preparations of disaggregated colonic crypts, musashi-1 positive cells preferentially adhered to colonic myofibroblasts, despite the presence of excess blocking anti-beta(1)-integrin antibody. These adherent epithelial cells remained viable for a number of days and developed slender processes. Cells with side population characteristics (as demonstrated by ability to expel the dye Hoechst 33342) were consistently seen in the isolated colonic crypt epithelial cells. These side population cells expressed musashi-1, beta(1)-integrin, BerEP4, and CD133. Sorted side population crypt epithelial cells also rapidly adhered to primary colonic myofibroblasts. In conclusion, in preparation of isolated and disaggregated human colonic crypts, cells with stem cell characteristics preferentially adhere to primary human colonic myofibroblasts in a beta(1)-integrin-independent fashion.

Adult small intestinal stem cells: identification, location, characteristics, and clinical applications.

There are few systems which enable adult tissue stem cells to be studied. However, the gastrointestinal tract with its high degree of polarity, well-defined cell migratory pathways, and dynamic cell replacement is a model tissue providing unique opportunities for stem cell study. Lineage tracking indicates that all cell replacement originates at well-defined stem cell positions, with an associated slower cell cycle. Radiobiological studies suggest a hierarchical stem cell compartment (actual and potential stem cells). Actual stem cells have an intolerance of genotoxic damage and die via apoptosis. Stem cells also selectively sort the old and new DNA strands at division, retaining the replication error free strands in the stem cell daughter. High genotoxic sensitivity and selective sorting of old and new DNA strands, provides extremely effective protective mechanisms against both replication and random errors. This provides a new explanation for the low cancer risk in the small intestine.

Regional localisation of p53-independent apoptosis determines toxicity to 5-fluorouracil and pyrrolidinedithiocarbamate in the murine gut.

Pyrrolidinedithiocarbamate (PDTC) enhanced the activity of 5-fluorouracil (5-FU) in a colorectal cancer xenograft model. Pyrrolidinedithiocarbamate also reduced gastrointestinal toxicity associated with 5-FU therapy in large but not small bowel. We sought to clarify the basis of this differential enteric toxicity. Apoptosis and mitosis were assessed on a cell positional basis in small and large intestinal crypts of p53 wild-type (+/+) and p53 null (-/-) mice 6, 12, 24, 36, 48 and 72 h after the administration of high (200 mg kg(-1)) or low (40 mg kg(-1)) dose 5-FU+/-250 mg kg(-1) PDTC. Regimens were chosen to model a single human dose and a weekly schedule. The effects of another antioxidant N-acetylcysteine (NAC) were also investigated. Large intestinal crypts affect apoptosis purely by p53-dependent mechanisms, whereas small intestinal crypts are able to initiate both p53-dependent and -independent pathways following treatment with 5-FU. Pyrrolidinedithiocarbamate and NAC antagonised p53-dependent but potentiated p53-independent apoptotic activity. Consequently, the proportion of surviving clonogens increased in the large but not in the small intestine. Regional availability of p53-dependent and -independent apoptotic pathways in small and large intestine together with separate modulation of these pathways by antioxidants explains the different regional enterotoxicity following 5-FU therapy.

Teduglutide ([Gly2]GLP-2) protects small intestinal stem cells from radiation damage.

Glucagon-like peptide-2 and its dipeptidyl peptidase (DP-IV) resistant analogue teduglutide are trophic for the gastrointestinal epithelium. Exposure increases villus height and crypt size and results in increased overall intestinal weight. As these effects may be mediated through stimulation of the stem cell compartment, they may promote intestinal healing and act as potential anti-mucositis agents in patients undergoing cancer chemotherapy. A study was initiated to investigate the protective effects of teduglutide on the murine small intestinal epithelium following gamma-irradiation using the crypt microcolony assay as a measure of stem cell survival and functional competence. Teduglutide demonstrated intestinotrophic effects in both CD1 and BDF1 mouse strains. In BDF1 mice, subcutaneous injection of GLP-2 or teduglutide (0.2 mg/kg/day, b.i.d.) for 14 days increased intestinal weight by 28% and resulted in comparable increases in crypt size, villus height and area. Teduglutide given daily for 6 or 14 days prior to whole body, gamma-irradiation significantly increased crypt stem cell survival when compared with vehicle-treated controls. The mean levels of protection over a range of doses provided protection factors from 1.3 to 1.5. A protective effect was only observed when teduglutide was given before irradiation. These results suggest that teduglutide has the ability to modulate clonogenic stem cell survival in the small intestine and this may have a useful clinical application in the prevention of cancer therapy-induced mucositis.

Intestinal crypt clonogens: a new interpretation of radiation survival curve shape and clonogenic cell number.

Estimates of the clonogen content (number of microcolony-forming cells) of murine intestinal crypts using microcolony assays show an apparent dependence on the radiation dose used in the assay of clonogen content. Crypt radiation survival curves often show increased curvature beyond that expected on the basis of the conventional linear-quadratic model. A novel form of crypt survival curve shape is proposed based on two contributory mechanisms of crypt killing. Six previously published sets of microcolony data were re-analysed using a dual-kill model, where target cells are killed by two contributory mechanisms, each described by a linear-quadratic function of dose. The data were analysed as two series--high-dose rate and low-dose rate irradiation. The data were fitted to the models using direct maximization of a quasi-likelihood, explicitly allowing for overdispersion. The dual-kill model can reproduce both the apparent dose-dependence of the clonogen estimates and the high-dose curvature of the dose-response curves. For both series of data the model was a significantly better fit to the data than the standard linear-quadratic model, with no evidence of any systematic lack of fit. The parameters of the clonogenic cell component of the model are consistent with other studies that suggest a low clonogen number (somewhat less than five) per crypt. The model implies that there is a secondary mechanism decreasing clonogen survival, and hence increasing clonogen number estimates, at high doses. The mechanisms underlying the modification of the dose-response are unclear, and the implied mechanisms of, for example, slow growth, induced either directly in the surviving cells or indirectly through stromal injury or bystander effects are only speculative. Nevertheless, the model fits the data well, demonstrating that there is greater kill at high doses in these experimental series than would be expected from the conventional linear-quadratic model. This alternative model, or another model with similar behaviour, needs to be considered when analysing in detail and interpreting microcolony data as a function of dose. The implied low number of < or = 5 of these regenerative and relatively radioresistant clonogenic cells is distinct from a similar number of much more radiosensitive precursor stem cells which undergo early apoptosis after doses around 1 Gy.

Acromegaly and colorectal cancer: a comprehensive review of epidemiology, biological mechanisms, and clinical implications.

Acromegaly is an endocrine disorder characterised by sustained hypersecretion of growth hormone (GH) with concomitant elevation of insulin-like growth factor (IGF)-I, and is associated with malignancy and premature mortality from cardiovascular and respiratory diseases. In particular, there may be an increased risk of colorectal neoplasia, but the exact extent of this is contentious. Colonoscopy-based studies of adenoma prevalence rates in acromegalic patients are misleading, but population-based studies on colorectal cancer risk are more consistent - a meta-analysis estimated a pooled risk ratio of 2.04 (95 % CI: 1.32, 3.14). Possible mechanisms underlying this increased risk include direct actions as a consequence of elevated levels of circulating GH and IGF-I and/or other perturbations within the IGF system. Other possible mechanisms include altered bile acid secretion, altered cellular immunity, hyperinsulinaemia, shared genetic susceptibility and increased bowel length. However, most explanations only offer indirect evidence, and the expectation of acromegaly as a natural model of colorectal carcinogenesis has not materialised. From a clinical perspective, it seems reasonable to consider a once-only colonoscopic screening at approximately age 55 years, but potential risks and benefits should be balanced.

Cell kinetic studies in the murine ventral tongue epithelium: thymidine metabolism studies and circadian rhythm determination.

The oral mucosa is a rapidly replacing body tissue that has received relatively little attention in terms of defining its cell kinetics and cellular organization. The tissue is sensitive to the effects of cytotoxic agents, the consequence of which can be stem cell death with the subsequent development of ulcers and the symptoms of oral mucositis. There is considerable interest in designing strategies to protect oral stem cells and, hence, reduce the mucositis side-effects in cancer therapy patients. Here we present details of a new histometric approach designed to investigate the changing patterns in cellularity in the ventral tongue mucosa. This initial paper in a series of four papers presents observations on the changing patterns in the labelling index following tritiated thymidine administration, which suggest a delayed uptake of tritiated thymidine from a long-term intracellular thymidine pool, a phenomenon that will complicate cell kinetic interpretations in a variety of experimental situations. We also provide data on the changing pattern of mitotic activity through a 24-h period (circadian rhythms). Using vincristine-induced stathmokinesis, the data indicate that 54% of the basal cells divide each day and that there is a high degree of synchrony in mitotic activity with a mitotic peak occurring around 13.00 h. The mitotic circadian peak occurs 9-12 h after the circadian peak in DNA synthesis. The data presented here and in the subsequent papers could be interpreted to indicate that basal cells of BDF1 mice have an average turnover time of about 26-44 h with some cells cycling once a day and others with a 2- or 3-day cell cycle time.

Cell kinetic studies in murine ventral tongue epithelium: cell cycle progression studies using double labelling techniques.

The dorsal and ventral epithelia on the murine tongue exhibit very pronounced circadian rhythms in terms of the cell cycle. These rhythms are such that three injections of tritiated thymidine 3 h apart spanning the circadian peak in S phase cells labelled between 40 and 50% of the basal cells. Injection of bromodeoxyuridine generally gave slightly lower labelling indices. Approximately the same proportion (54% of the basal cells) could be accumulated in metaphase over a 24-h period using vincristine as a stathmokinetic agent. The experiments reported here using mouse ventral tongue epithelium use double-labelling approaches to address the question: what proportion of the approximately 50% of the basal cells that are proliferating have a 24-h cell cycle and can therefore be labelled by a similar labelling protocol the following day? The results suggest a heterogeneity amongst the proliferating basal cells, similar to the heterogeneity proposed for the dorsal tongue epithelium. Although not all the basal component has been accounted for, the data presented here suggest that about 20% of the basal cells may have a cell cycle time of 24 h, about 30% appear to have a longer cell cycle time (48 or 72 h), while about 20% of the basal cells appear to be postmitotic maturing G1 cells, awaiting the appropriate signals for migration into the suprabasal layer.

Cell kinetic studies in the murine ventral tongue epithelium: the effects of repeated exposure to keratinocyte growth factor.

Keratinocyte growth factor (KGF) stimulates proliferation and differentiation in various epithelial systems. Three daily subcutaneous injections of 125 microg of this protein into mice induce dramatic changes in the histology and histometric measurements of the ventral tongue epithelium. The thickness of the epithelium is increased two-fold and the number of cells beneath a 1-mm length of the surface is increased 1.6-fold. KGF also induces a four-fold increase in the number of S phase cells labelled with tritiated thymidine in the basal layer on the third day after KGF administration. The increase in thickness and cellularity persist for at least 4 days after the end of the KGF injections. However, there is a dramatic fall in the number of S phase cells detected by 3HTdR pulse labelling 2 days after the end of the KGF treatment. There are indications that by 7 days after the 3-day regimen of KGF treatment, both thickness and cellularity have fallen back to near control levels. Continued exposure to KGF over a period of 7 days does not result in any further increases in thickness, cellularity or proliferation. In fact, the proliferation decreases on the fifth, sixth and seventh days of KGF injection to control values on day 7. These changes in the epithelium following KGF treatment suggest that the thicker and more cellular epithelium may be more able to cope with an exposure to a cytotoxic agent and hence be protected in comparison with normal or vehicle-treated epithelium.

Cell kinetic studies in the murine ventral tongue epithelium: mucositis induced by radiation and its protection by pretreatment with keratinocyte growth factor (KGF).

Radiation kills or reduces reproductive capacity of proliferating cells, including stem cells. In the oral mucosae this can result in a decline in the number of cells in the tissue which, if severe enough, will result in the formation of an ulcer when the cellularity essentially reaches zero. We have used histometric measurements of cellularity following exposure to radiation in mouse ventral tongue epithelium as a model for oral mucositis (ulcer development). Here we provide further measurements of cellularity changes in the basal layer and in the epithelium as a whole at various times following 15, 20 or 25 Gy doses. The protective effects of prior treatment with keratinocyte growth factor (KGF) are also investigated. 20 Gy of 300 kV X-rays has become our standard reference dose and the changes in cellularity seen following this dose are highly reproducible, with minimum values being observed 6 days following irradiation. A higher dose results in a greater reduction of cellularity, although the minimum value also occurs at 6 days. A lower dose (15 Gy) results in a much shallower curve, with a minimum value being observed about 1 day earlier. These changes in cellularity can be related to the less sensitive index of mucositis, namely epithelial thickness. There is also a sharp peak in proliferation about 1 day after the minimum in cellularity, i.e. on day 7. The peak following a lower dose of radiation occurs a little earlier and, following the higher dose, the peak tends to be broader. Previous work and data presented in the preceding paper in this series has shown that KGF, given over a period of 3 days, results in a dramatic increase in epithelial thickness in oral mucosa, including the ventral tongue. As a result of the increased cellularity induced by KGF given before radiation, a delay in the fall in cellularity results, which is the consequence of the increased number of cells in the epithelium at the beginning of the study.

A continuum model to analyse spatial cell fluxes from positional labelling data applied to cell migration in oral mucosa.

A novel method is presented which allows the estimation of the velocity of migrating cells from positional 3HTdR labelling data in spatially organized epithelial tissues. In a continuum approach, wave-like profiles of labelling index (LI) data, which travel away from the basal layer, are followed and compared with experimental LI profiles. The method yields estimates of migration velocity, cell flow, and turnover time. Results for the ventral tongue mucosa in a group of 55 BDF1 mice that were labelled at the same time of the day and culled at different time points within a 24-h period have been analysed. The results show a strong circadian rhythm in the migration velocity and the related parameters.

Simulation of cell proliferation in mouse oral epithelium, and the action of epidermal growth factor: evidence for a high degree of synchronization of the stem cells.

Computer simulation has been carried out to help to determine the cell-proliferative mechanisms underlying data gathered from a double-labelling experiment on the dorsal tongue of the mouse. Good fits to the data have been obtained by assuming that there is a high degree of synchrony in the stem cells, which have a 24-h cell cycle time, and that daughters of these cells undergo two further divisions, with mean cell cycle times of 48 h, before differentiating. This results in one-seventh of proliferative cells being stem cells, which ties in well with the concept of epidermal proliferative units. There is no need to assume that S-phase duration changes diurnally. The administration of epidermal growth factor seems to increase the degree of synchrony. In such systems, the influx to S-phase and the efflux from it have very sudden short peaks, which it is impossible to observe unless observations are taken very frequently. There are therefore implications for the designs of experiments that attempt to study diurnal rhythms or the effect of factors that disturb the normal proliferative pattern of cells.

Early cellular events in colorectal carcinogenesis.

Colorectal cancer develops through a multistage process recognizable at a histopathological level by progression from normal mucosa to invasive carcinoma (the adenoma-carcinoma sequence). For many years, it has been hypothesized that increased cell proliferation in the colonic crypt represents the earliest recognizable stage in this sequence. This perspective is now changing. While several human studies have reported increased crypt cell proliferation in samples from at-risk patients, there are many inconsistencies and paradoxes in their conclusions. In addition, it is appreciated that the process of apoptosis (programmed cell death) is vital for normal crypt homeostasis and its impairment may be an early event in the neoplastic process. It is now believed that aberrant crypt foci (ACFs) represent the earliest step in colorectal carcinogenesis. Two ACF types are identifiable: hypercellular and dysplastic. Increased proliferative activity may be seen in both, but the dysplastic entity is most relevant to carcinogenesis. Animal and human studies support the notion that ACFs grow by crypt fission leading to the formation of microadenomas. Adenomas are monoclonal expansions of an altered cell, but very early lesions may be polyclonal. There are outward and inward theories of polypoid growth, and evidence to support both mechanisms. The ACF assay has become a useful tool to detect carcinogens in animal studies but has been less frequently used in human studies. For future cancer chemopreventive and risk assessment studies in humans, the identification and quantification of ACFs should be considered a more effective intermediate marker of risk than the determination of crypt cell proliferation alone.

Expression of Id helix-loop-helix proteins in colorectal adenocarcinoma correlates with p53 expression and mitotic index.

Id helix-loop-helix (HLH) proteins function as regulators of cell growth and differentiation and when overexpressed can induce malignant transformation. In a series of 34 cases of primary human colorectal adenocarcinoma, immunoreactivity for Id1, Id2, and Id3 was found to be significantly elevated in tumor compared with normal mucosa (P = 0.001 for Id1 and Id2; P = 0.002 for Id3). No elevation of Id expression was observed in 17 cases of adenoma. Expression of Id1 and to a lesser extent of Id2 was correlated with mitotic index (P = 0.005 for Id1; P = 0.042 for Id2) in human adenocarcinomas, and expression of all three Id proteins was correlated with p53 immunoreactivity (a marker of mutational 'inactivation' of p53 function; P = 0.002 for Id1; P = 0.006 for Id2; P = 0.016 for Id3). In normal intestinal mucosa of p53-null mice and in spontaneous tumors arising in Min+/- mice, expression of all three Id proteins was also found to be up-regulated. Antisense oligonucleotide blockade of Id protein expression inhibited the proliferation of human adenocarcinoma cells. Enforced, ectopic expression of the E47 basic HLH (bHLH) protein in human adenocarcinoma cell lines efficiently sequestered endogenous Id proteins as Id-bHLH heterodimers, as shown by coimmunoprecipitation and subcellular colocalization studies. This led to growth arrest of the cells. Enforced overexpression of a mutant E47 protein, deficient in transactivation and DNA binding function, also partially inhibited cell growth. Taken together, these data imply that deregulated expression of Id proteins in colorectal adenocarcinoma arises at least in part as a consequence of loss of p53 function and contributes to the uncontrolled proliferation of tumor cells in colorectal cancer.

Protection against mucosal injury by growth factors and cytokines.

This article provides an overview of published studies in which growth factors and cytokines were used to modify the sensitivity of intestinal stem cells to a dose of radiation. In these experiments, growth factors were used to manipulate the sensitivity of stem cells in the gastrointestinal tract to reduce the severity of gastrointestinal mucositis in cancer therapy patients. Transforming growth factor beta3, interleukin 11, and keratinocyte growth factor were used. All three agents, given according to appropriate protocols, can result in a threefold to fourfold increase in the number of intestinal stem cells that survive a dose of radiation therapy. This result was assessed by using the crypt microcolony assay of stem cell functional capacity. The changes in stem cell survival that were observed resulted in increased animal survival. This increased survival was taken as a surrogate for improvement in patient well-being. The severity of diarrhea, a marker of functional impairment, was concomitantly reduced.

In vitro labelling studies and the measurement of epithelial cell proliferative activity in the human oral cavity.

Samples (110) of human mandibular gingiva and buccal mucosa, harvested from patients undergoing third-molar surgery, were subjected to in vitro labelling with tritiated thymidine, bromodeoxyuridine, or a sequential double-labelling technique comprising tritiated thymidine followed by bromodeoxyuridine, in order to determine the efficacy of a new incubation labelling technique, and to characterize S-phase labelling indices in human oral mucosa. Whilst, there were no demonstrable differences in labelling indices obtained by single thymidine, single bromodeoxyuridine or double labelling, there was a significant difference between anatomical sites, with higher S phase labelling observed in buccal mucosa (mean LI 11.7) than mandibular gingiva (mean LI 8.5; P<0.01). There was, however, no significant correlation between individual labelling indices and patient age, sex or the time of day when tissue was harvested. The in vitro labelling technique provides a reliable and quantifiable method of characterizing proliferative labelling indices in the human oral cavity. Further investigation is being carried out to profile wider age and anatomical ranges and to utilize the double-labelling technique to calculate S-phase durations and cell-cycle times. These profiles may have a future role in the assessment of oral mucous membrane disease.

Caspase activation during spontaneous and radiation-induced apoptosis in the murine intestine.

The aim of this study was to characterize the activation of caspase-3 along the crypt/villus axis in the normal and irradiated intestine and to compare active caspase-3 expression with existing apoptosis detection techniques. Small and large intestine were removed from mice at various time points after exposure to 8 Gy gamma-radiation. Positive apoptotic cells stained with an antibody against active caspase-3, haematoxylin and eosin (H&E) or TUNEL were scored in histological sections of small and large intestinal crypts and villi. In the control intestine, active caspase-3 expression was rarely observed; however, expression was markedly increased following exposure to radiation and was predominantly confined to apoptotic bodies. Measurement of apoptosis in intestinal crypts using active caspase-3 expression gave similar results to apoptosis detected from H&E-stained sections. In the normal villus, active caspase-3 expression was observed infrequently and did not significantly increase following radiation, consistent with a lack of apoptotic body formation from H&E sections. This study indicates that caspase-3 is activated in intestinal crypts but not in villi following gamma-radiation. Active caspase-3 detection compared favourably with existing immunological techniques, suggesting that it is a suitable alternative method for apoptosis quantification.

Paradoxical elevations in serum IGF-II and IGF binding protein-2 in acromegaly: insights into the regulation of these peptides.

OBJECTIVE: Circulating insulin-like growth factor (IGF)-II and IGF binding protein-2 (IGFBP-2) are frequently altered, often in parallel, in numerous pathologies including neoplastic disease but little is known about their normal regulation. This study compared serum IGF-II and IGFBP-2 distributions between acromegalics and a large normal adult population to explore possible determinants. PATIENTS: Sixty acromegalic patients undergoing screening colonoscopy (age range 25-81 years); normative data from 306 healthy adults (age range 20-89 years). MEASUREMENTS: Serum IGF-I, IGF-II, IGFBP-2 and IGFBP-3 were measured in healthy adults and acromegalics. Mean growth hormone (GH) levels were obtained for acromegalic patients. Differences were compared using t-tests (unadjusted) and multiple regression models (adjusted for age and gender). Correlations were expressed as Pearson's coefficient (r). RESULTS: For acromegalic patients, GH was significantly correlated with IGF-I (r = 0.50; P < 0.001) and IGFBP-3 (r = 0.29; P = 0.03) but not IGF-II or IGFBP-2. Contrary to expectations, mean IGF-II and IGFBP-2 levels were significantly raised in the acromegalics compared with normals [adjusted mean difference (95% CI) = 226 (181, 271) microg/l and 305 (200, 410) microg/l, respectively]. Ten acromegalic patients had colorectal neoplasia but their presence did not contribute to the elevations in serum IGF-II and IGFBP-2. The (IGF-I + IGF-II)/IGFBP-3 molar ratios were remarkably constant in both healthy adults and acromegalics, but the relationships of the ligands individually with IGFBP-3 were not linear: as IGFBP-3 increased, IGF-I also increased whereas IGF-II initially increased but then decreased. IGFBP-2 did not correlate with IGF-II, but molar concentration significantly correlated with the IGF-II/IGFBP-3 molar ratio (r = 0.40; P = 0.001). CONCLUSIONS: Serum IGF-II and IGFBP-2 levels were paradoxically elevated in acromegalics, independent of the presence of colorectal neoplasia. The (IGF-I + IGF-II)/IGFBP-3 molar ratio appears to be pivotal in determining IGF-II values, which, in turn, expressed as a ratio of IGFBP-3, is related to IGFBP-2. These observations offer new insights into the regulation of these peptides.