Increased expression of aggrecan and biglycan mRNA in Achilles tendinopathy.

OBJECTIVES: To determine the expression of mRNA encoding the proteoglycans aggrecan, versican, biglycan and decorin in mid-tendon samples of chronic painful Achilles tendinopathy and ruptured Achilles tendons, compared with normal tendons. METHODS: Total RNA isolated from frozen tendon samples (14 normal, 13 painful, 14 ruptured) was assayed by relative quantitative reverse transcription polymerase chain reaction for aggrecan, versican, biglycan and decorin mRNA, normalized using 18S rRNA. Differences between sample groups were tested by univariate analysis of variance with age as co-variate. RESULTS: In normal tendon samples expression of each of the proteoglycan mRNA decreased with increasing age. Decorin mRNA was the most highly-expressed of the proteoglycan mRNA, while versican mRNA expression was higher (3.8-fold) than that of aggrecan. In painful tendinopathy both aggrecan and biglycan mRNA expression increased (more than 10-fold and 5-fold, respectively) compared with normal tendon samples, but levels of versican and decorin mRNA were not significantly changed. In ruptured tendons the levels of aggrecan, biglycan and versican mRNA were not changed compared with normal tendon samples, but decorin mRNA decreased markedly. CONCLUSIONS: Increased aggrecan and biglycan mRNA expression in painful tendinopathy resembles the pattern in fibrocartilaginous regions of tendon, and may reflect an altered mechanical environment at the site of the lesion. Increased aggrecan mRNA expression may underlie the increase in glycosaminoglycan observed in painful tendinopathy.

Contrasting effects of fluoroquinolone antibiotics on the expression of the collagenases, matrix metalloproteinases (MMP)-1 and -13, in human tendon-derived cells.

OBJECTIVES: Fluoroquinolone antibiotics may cause tendon pain and rupture. We reported previously that the fluoroquinolone ciprofloxacin potentiated interleukin (IL)-1beta-stimulated expression of matrix metalloproteinases (MMP)-3 and MMP-1 in human tendon-derived cells. We have now tested additional fluoroquinolones and investigated whether they have a similar effect on expression of MMP-13. METHODS: Tendon cells were incubated for two periods of 48 h with or without fluoroquinolones and IL-1beta. Total ribonucleic acid (RNA) was assayed for MMP messenger RNA by relative quantitative reverse transcriptase polymerase chain reaction, with normalization for glyceraldehyde-3-phosphate dehydrogenase mRNA. Samples of supernatant medium were assayed for MMP output by activity assays. RESULTS: MMP-13 was expressed by tendon cells at lower levels than MMP-1, and was stimulated typically 10- to 100-fold by IL-1beta. Ciprofloxacin, norfloxacin and ofloxacin each reduced both basal and stimulated expression of MMP-13 mRNA. In contrast, ciprofloxacin and norfloxacin increased basal and IL-1beta-stimulated MMP-1 mRNA expression. Both the inhibition of MMP-13 and the potentiation of MMP-1 expression by fluoroquinolones were accompanied by corresponding changes in IL-1beta-stimulated MMP output. The non-fluorinated quinolone nalidixic acid had lesser or no effects. CONCLUSIONS: Fluoroquinolones show contrasting effects on the expression of the two collagenases MMP-1 and MMP-13, indicating specific effects on MMP gene regulation.

Versican splice variant messenger RNA expression in normal human Achilles tendon and tendinopathies.

OBJECTIVES: Versican is the principal large proteoglycan expressed in mid-tendon, but its role in tendon pathology is unknown. Our objective was to define the expression of versican isoform splice variant messenger ribonucleic acid (mRNA) in normal Achilles tendons, in chronic painful tendinopathy and in ruptured tendons. METHODS: Total RNA isolated from frozen tendon samples (normal n = 14; chronic painful tendinopathy n = 10; ruptured n = 8) was assayed by relative quantitative reverse transcriptase polymerase chain reaction (RT-PCR) for total versican, versican variants V0, V1, V2, V3 and type I collagen alpha1 mRNA, normalized to glyceraldehyde-3-phosphate dehydrogenase (GAPDH). Differences between sample groups were tested by Wilcoxon statistics. RESULTS: Painful and ruptured tendons showed a significant decrease (median 2-fold) in the expression of versican mRNA, in contrast to an increased expression (median 8-fold) of type I collagen alpha1 mRNA in painful tendons. Versican splice variants V0 and V1 mRNA were readily detected in normal samples, V3 levels were substantially lower, and V2 levels were more variable. Each of V1, V2 and V3 mRNA showed significant decreases in expression in painful and ruptured tendons, but V0 was not significantly changed. CONCLUSIONS: Changes in versican expression relative to that of collagen, and alterations in the balance of versican splice variants, may contribute to changes in matrix structure and function in tendinopathies.

Ciprofloxacin reduces the stimulation of prostaglandin E(2) output by interleukin-1beta in human tendon-derived cells.

OBJECTIVE: Fluoroquinolone antibiotics such as ciprofloxacin can induce tendon pathology and have various effects on tendon-derived cells in culture. We are investigating whether ciprofloxacin modifies signalling responses in tendon cells. METHODS: Human Achilles tendon-derived cells were preincubated with or without ciprofloxacin (50 mug/ml) and were then challenged with interleukin-1beta (IL-1beta, 1 ng/ml) for up to 48 h. Prostaglandin E2 (PGE2) output was assayed by ELISA. The expression of cyclooxygenase-2 (COX-2) was examined by Western blotting. RESULTS: IL-1beta stimulated a substantial and prolonged increase in the output of PGE2. Preincubation with ciprofloxacin reduced IL-1beta-induced PGE2 output at all times tested; the reduction at 48 h was 69% (99% confidence interval 59-79%; 15 experiments). Norfloxacin and ofloxacin also reduced PGE2 output. However, ciprofloxacin did not affect the induction of COX-2 by IL-1beta, measured at 4 or 48 h. CONCLUSIONS: Ciprofloxacin reduces IL-1beta-induced PGE2 output in tendon-derived cells. The reduction in PGE2 output could modulate various cellular activities of IL-1beta, and may be implicated in fluoroquinolone-induced tendinopathy.

Changes in tumorigenesis- and angiogenesis-related gene transcript abundance profiles in ovarian cancer detected by tailored high density cDNA arrays.

BACKGROUND: Complementary DNA array analysis of gene expression has a potential application for clinical diagnosis of disease processes. However, accessibility, affordability, reproducibility of results, and management of the data generated remain issues of concern. Use of cDNA arrays tailored for studies of specific pathways, tissues, or disease states may render a cost- and time-effective method to define potential hallmark genotype alterations. MATERIALS AND METHODS: We produced a 332-membered human cDNA array on nylon membranes tailored for studies of angiogenesis and tumorigenesis in reproductive disease. We tested the system for reproducibility using a novel statistical approach for analysis of array data and employed the arrays to investigate gene expression alterations in ovarian cancer. RESULTS: Intra-assay analysis and removal of agreement outliers was shown to be a critical step prior to interpretation of cDNA array data. The system revealed highly reproducible results, with intermembrane coefficient of reproducibility of +/- 0.98. Comparison of placental and ovarian sample data confirmed expected differences in angiogenic profiles and tissue-specific markers, such as human placental lactogen (hPL). Analysis of expression profiles of five normal ovary and four poorly differentiated serous papillary ovarian adenocarcinoma samples revealed an overall increase in angiogenesis-related markers, including vascular endothelial growth factor (VEGF) and angiopoietin-1 in the diseased tissue. These were accompanied by increases in immune response mediators (e.g. HLA-DR, Ron), apoptotic and neoplastic markers (e.g. BAD protein, b-myb), and novel potential markers of ovarian cancer, such as cofilin, moesin, and neuron-restrictive silencer factor (REST) protein. CONCLUSIONS: In-house production of tailored cDNA arrays, coupled to comprehensive analysis of resulting hybridization profiles, provides an accessible, reliable, and highly effective method of applying array technology to study disease processes. In the ovary, abundance of specific tumor markers, increased macrophage recruitment mediators, a late-stage angiogenesis profile, and the presence of chemoresistance-related markers distinguished normal and advanced ovarian cancer tissue samples. Detection of such parallel changes in pathway- and tissue-specific markers may prove a hallmark ready for application in reproductive disease diagnostic and therapeutic developments.

Hepatocyte growth factor (HGF) in ovarian epithelial tumour fluids stimulates the migration of ovarian carcinoma cells.

Hepatocyte growth factor (HGF) is a pleiotropic growth factor implicated in the growth and spread of some epithelial tumours. The epithelial cells of a proportion of ovarian tumours, and some ovarian carcinoma cell lines, express high levels of the HGF receptor, c-Met. In this study, we show that ovarian ascitic fluid as well as benign and malignant ovarian cyst fluids contain significant levels of HGF. Ovarian cyst and ascitic fluid stimulated the migration of the ovarian carcinoma cell line, SK-OV-3, and this was greatly reduced by the addition of an HGF-neutralising antibody. Non-malignant peritoneal fluid contained low levels of HGF, and did not stimulate migration of the SK-OV-3 cells. Our results show that HGF is present in benign and malignant ovarian cyst and ascitic fluid, and that HGF in ovarian tumour fluid may be a major inducer of ovarian carcinoma cell migration.

A vascular endothelial growth factor antagonist is produced by the human placenta and released into the maternal circulation.

Vascular endothelial growth factor (VEGF) is a potent secreted factor that promotes angiogenesis and maintains the integrity of the endothelium. Levels of VEGF are increased in many tumors and are elevated in women with pre-eclampsia, a serious disease of pregnancy. Here we show by in situ hybridization that the trophoblast contains the mRNA encoding a soluble version of the VEGF receptor known as Flt-1 (sFlt-1: initially described by Kendall and Thomas, PNAS 90:10705-10709). Binding assays and Western blotting of villus-conditioned media confirmed the production of sFlt-1. Serum from pregnant women was found to contain a VEGF-binding protein that was not present in serum from men or nonpregnant women. As determined by heparin affinity, column fractionation, and cross-linking, this protein was identical to sFlt-1. Taken together, these results show that the placenta secretes sFlt-1, which would be expected to be a VEGF antagonist. This is the first report of production of the sFlt-1 receptor in vivo, and it reveals a new mechanism for naturally regulating this potent angiogenic agent. The presence of such an antagonist suggests that regulation of VEGF action is essential to successful pregnancy. This has important implications for the activity of VEGF locally and systemically in other conditions.

Hepatocyte growth factor stimulates motility, chemotaxis and mitogenesis in ovarian carcinoma cells expressing high levels of c-met.

A proportion of ovarian carcinomas markedly overexpress the proto-oncogene c-met, which encodes the receptor for hepatocyte growth factor (HGF). HGF may either stimulate or inhibit the multiplication of its target cells, and may also promote motogenesis and morphogenesis. In this study, we established that the ovarian carcinoma-derived cell-line SK-OV-3 expressed about 20-fold higher levels of c-met protein than are expressed by a second line, CH1. This enabled us to test functional consequences of high-level expression of c-met in ovarian carcinoma cells. The addition of HGF to attached cultures of SK-OV-3 cells caused a change to a motile phenotype, that was evident after 4-6 hr and affected essentially all of the cells by 24 hr. When HGF was placed in the lower compartment of a migration chamber, it induced a 17-fold increase in the migration of SK-OV-3 cells to the lower surface of the filter. Finally, HGF stimulated the incorporation of [3H]-thymidine by cultures of SK-OV-3 cells incubated in medium containing either low (0.2%) or full (10%) FCS. None of these responses were obtained when HGF was added to CH1 cells. We conclude that high levels of c-met expression in ovarian cancer cells may lead to a range of responses to HGF that would promote tumour growth and dissemination.

Expression and localization of the vascular endothelial growth factor family in ovarian epithelial tumors.

Vascular endothelial growth factor (VEGF) is an angiogenic factor secreted by various tumors, including epithelial tumors of the ovary, and is involved in tumor progression and maintenance. The significance and function of other members of the VEGF family in the ovary has not yet been elucidated. In the present study, we have defined the expression of mRNA encoding VEGF-B, VEGF-C, and placenta growth factor (PIGF), compared with that of VEGF mRNA, in normal ovary and a range of ovarian epithelial tumors. Analysis by reverse transcription-PCR indicated that mRNA encoding VEGF (isoforms 121 and 165), VEGF-B (isoforms 167 and 186), and VEGF-C, but not PIGF, were present in all ovarian tissues examined. By in situ hybridization, neither VEGF-C nor PIGF transcripts were detected in any of the samples. The expression pattern of VEGF-B mRNA was generally similar to that of VEGF mRNA, in that transcripts were readily detected in the epithelial cells of all histologic types of ovarian carcinoma, but could not be detected in normal or benign tumor epithelium. Specific differences in the expression of the two genes were noted in areas of tumor necrosis, in which the expression of VEGF mRNA, but not VEGF-B mRNA, was further enhanced, and in a sample in which VEGF-B mRNA was strongly expressed in tumor-associated macrophages that did not hybridize with the riboprobe to VEGF mRNA. These results imply that a second member of the VEGF family, VEGF-B, may play a significant role in the angiogenesis, progression, and maintenance of ovarian carcinomas.

Identification of a functional promoter element in the 5'-flanking region of the rat cMG1/TIS11b gene.

The cMG1 gene was originally identified as a mitogen-stimulated primary response gene. However, in contrast to genes such as c-fos and TIS11, cMG1 is also expressed at significant levels before and after the transient elevation induced by agonists. We have sequenced a 1.3 kb rat genomic cMG1 clone, which includes 931 bp upstream of the transcription start site identified by primer-extension analysis. A 1033 bp fragment, including this 5'-flanking sequence, directed the expression of the reporter gene chloramphenicol acetyltransferase (CAT) in transfected NIH-3T3 cells. Progressive 5'-to-3' deletion indicated that an element located between -138 and -114 was responsible for most of this basal CAT expression. DNA mobility-shift assays showed that the sequence between -143 and -105 contained binding sites for cellular proteins, the principal complexes involving nucleotides between -119 and -105. We conclude that these complexes may represent the transcription factor-DNA element interactions that determine basal cMG1 expression.

Insulin and insulin-like growth factor I stimulate expression of the primary response gene cMG1/TIS11b by a wortmannin-sensitive pathway in RIE-1 cells.

The addition of insulin or insulin-like growth factor I (IGF-I) to RIE-1 cells increased the expression of the primary response gene cMG1; dose-response analysis suggested that this effect was mediated largely through type 1 IGF receptors. Insulin/IGF-I did not affect the expression of the cMG1-related genes TIS11 and TIS11d, whereas epidermal growth factor, angiotensin II or 12-O-tetradecanoyl phorbol-13-acetate stimulated the expression of all three genes. Incubation with wortmannin (WM) prevented the insulin/IGF-I-induced elevation of cMG1 mRNA, but not that induced by the other mitogens or the stimulation of mitogen-activated protein kinase by insulin. We conclude that WM-sensitive phosphatidylinositol 3-kinase may be involved in the specific stimulation of cMG1 expression by insulin/IGF-I.

Activation of AT1 angiotensin receptors induces DNA synthesis in a rat intestinal epithelial (RIE-1) cell line.

Proliferation of the rat intestinal epithelial cell-line, RIE-1, has previously been shown to be stimulated by certain polypeptide growth factors acting via receptors that possess intrinsic tyrosine kinase activity. In this study, we show that the octapeptide hormone angiotensin II (AII), apparently acting through the AT1 G-protein-coupled receptor, is also a mitogen for RIE-1 cells. Maximal stimulation of DNA synthesis and cellular proliferation occurred at an AII concentration of 10-100 nM, with half-maximal stimulation at 1 nM. The mitogenic response to AII was completely inhibited by the AT1 angiotensin-receptor antagonist, DuP753, but not by the AT2-receptor antagonist, PD123319. The early signalling responses activated by AII in RIE-1 cells include increased production of inositol phosphates, a transient increase in the intracellular concentration of free calcium, an activation of protein kinase C, and a rapid change in the pattern of cellular protein-tyrosine phosphorylation. These results implicate an activation of the inositol lipid signalling pathway via the AT1 receptor subtype in the AII-stimulated mitogenic response of this normal epithelial cell line.

Mitogen-induced expression of the primary response gene cMG1 in a rat intestinal epithelial cell-line (RIE-1).

cMG1 is a primary response gene first identified in a rat intestinal epithelial (RIE-1) cell-line [(1990) Oncogene 5, 1081-1083]. A number of mitogens, including epidermal growth factor (EGF), angiotensin II (AII), serum and insulin rapidly induced 2- to 6-fold increases of cMG1 mRNA in RIE-1 cells, while transforming growth factor-beta caused a small reduction. Cyclic AMP-elevating agents blocked the increase of cMG1 mRNA induced by EGF. The AII-stimulated increase of cMG1 mRNA was blocked by the depletion of protein kinase C, whereas the EGF-stimulated increase was not affected, indicating that protein kinase C-dependent and -independent signalling pathways stimulate cMG1 expression.

Mitogen-stimulated events in nuclei of Swiss 3T3 cells. Evidence for a direct link between changes of inositol lipids, protein kinase C requirement and the onset of DNA synthesis.

Two different clones of Swiss 3T3 cells belonging to the same original cell line have been obtained, one of which was unresponsive to mitogenic stimulation (e.g. insulin-like growth factor-I, bombesin, insulin-like growth factor-I + bombesin), while the other clone showed a very high rate of DNA synthesis under identical conditions as demonstrated by 5-bromodeoxyuridine incorporation. Both types of cells expressed the IGF-I receptor and showed high contact inhibition. When highly purified nuclei from responsive cells, treated for a short time with bombesin and insulin-like growth factor-I or insulin-like growth factor-I alone, were incubated with [gamma-32P]adenosine triphosphate, the labelling of phosphatidylinositol-mono- and diphosphate decreased when compared to controls, while this transient effect did not appear in the nuclei from unresponsive cells. Similarly nuclear protein kinase C is activated only in responsive cells. Therefore, it seems that a direct link exists between polyphosphoinositide metabolism, protein kinase C activation and the early events leading to cell division, since the rapid changes in the labelling of both phosphatidylinositol mono- and di-phosphate occur only in nuclei from Swiss 3T3 cells, which respond to the mitogenic stimulus determined by insulin-like growth factor-I on its own, or in combination with bombesin.

Mitogens regulate the production of insulin-like growth factor-binding protein by Swiss 3T3 cells.

Quiescent Swiss 3T3 cells can be stimulated to reenter the cell cycle by various mitogens used in synergistic combinations with insulin-like growth factors (IGFs). The cells constitutively secrete an IGF-binding protein (IGFBP), which can modulate the interaction of IGFs with their receptors and could, therefore, alter cellular responsiveness to IGFs. We have now characterized the IGFBP secreted by Swiss 3T3 cells and tested whether its secretion is regulated by heterologous mitogens. Ligand blotting using [125I]IGF-I revealed a major IGFBP of 40,000 mol wt, and treatment of the cells with tunicamycin reduced the mol wt of this protein to about 32,000. mRNA from Swiss 3T3 cells hybridized to a 32P-labeled oligonucleotide (50-mer) complementary to rat IGFBP-3. Taken together, these results indicate that the principal IGFBP secreted by Swiss 3T3 cells is probably the N-glycosylated IGFBP-3. Production of this IGFBP by Swiss 3T3 cells was stimulated by 50-150% by the mitogens bombesin, vasopressin, platelet-derived growth factor, epidermal growth factor, and 12-O-tetradecanoylphorbol 13-acetate and also by IGF-I. The increased production of IGFBP was first detected after 4-6h of incubation and was then maintained for 48-72 h. Agents that elevate intracellular cAMP and the glucocorticoid dexamethasone reduced IGFBP output. In cells in which protein kinase-C had been down-modulated, the stimulation of IGFBP output by 12-O-tetradecanoylphorbol 13-acetate was abolished, but the stimulation induced by the other mitogens was not prevented. Thus, the production of IGFBP by Swiss 3T3 cells can be regulated by a number of different signalling pathways.

Increase in milk secretion and mammary blood flow by intra-arterial infusion of insulin-like growth factor-I into the mammary gland of the goat.

The close-arterial infusion of free insulin-like growth factor-I (IGF-I; 1.1 nmol/min) for 6 h into the pudic artery supplying one mammary gland of lactating goats caused a 25 +/- 6% (mean +/- S.E.M., n = 6) increase in the rate of milk secretion of that gland. The increase in the rate of milk secretion in the adjacent noninfused gland (14 +/- 4%) was not significantly different from that observed during saline infusion (4 +/- 5%). Blood flow to the infused gland was increased from 378 +/- 26 ml/min 1 h before to 487 +/- 56 ml/min approximately 5 h after the start of the infusion of IGF-I, declining to 420 +/- 44 ml/min approximately 2 h after the end of the infusion. The total concentration of IGF-I (free and bound) in milk of the infused gland was significantly higher than that of the non-infused gland. The concentrations of IGF-I in carotid arterial plasma samples increased during IGF-I infusion from a mean value of 32 +/- 2 nmol/l before to a maximum of 49 +/- 3 nmol/l 5 h after the infusion commenced. Circulating concentrations of total IGF-I declined slowly after the infusion with an estimated half-life of 5 h. Infusion of saline alone did not alter mammary blood flow or the concentration of total IGF-I in milk or plasma. The results indicate that the infusion of free IGF-I into the mammary arterial supply enhances milk secretion and mammary blood flow in intact, conscious goats.(ABSTRACT TRUNCATED AT 250 WORDS)

Receptors for epidermal growth factor and insulin-like growth factor-I on preimplantation trophoderm of the pig.

125I-labelled epidermal growth factor (125I-EGF) and 125I-labelled insulin-like growth factor-I (125I-IGF-I) bound to trophoderm cells from pig blastocysts obtained on days 15-19 of pregnancy. Specific binding was detected on freshly isolated cell suspensions and on cells cultured for several days. The binding of 125I-EGF was inhibited by increasing concentrations of EGF, but not by various other growth factors and hormones. Chemical cross-linking of 125I-EGF to its receptors using disuccinimidyl suberate (DSS) revealed a radiolabelled band of relative molecular mass 160,000, similar to that identified as the EGF receptor in other cell types. The binding of 125I-IGF-I was inhibited by both IGF-I and insulin, indicating that the receptors were either type I IGF receptors or insulin receptors. Cross-linking of 125I-IGF-I to serum-free supernatants from trophoderm cultures showed that the cells secreted an IGF-binding protein, giving a complex of relative molecular mass about 45,000. The presence of receptors for EGF and IGF/insulin suggests that these factors could be involved in regulating the growth and development of the early blastocyst.

Single-cell analysis of the mitogen-induced calcium responses of normal and protein kinase C-depleted Swiss 3T3 cells.

Single-cell fluorescence image analysis has been used to characterize the mitogen-induced increases in intracellular free [Ca2+] ([Ca2+]i) in control and protein kinase C-depleted Swiss 3T3 cells. More than 80% of the control cells exhibited fast, transient responses to bombesin, vasopressin, or prostaglandin F2 alpha (PGF2 alpha). In contrast, the [Ca2+]i responses induced by platelet-derived growth factor (PDGF) were markedly more heterogeneous, slower, and often biphasic, with fewer cells (60-70%) responding. The peak [Ca2+]i values obtained in response to each mitogen showed substantial variation between cells. Brief pretreatment of the cells with 12-O-tetradecanoyl phorbol 13-acetate (TPA) reduced the [Ca2+]i responses to bombesin, but did not affect the responses to PDGF. Long-term pretreatment of the cells with TPA to down-modulate protein kinase C resulted in substantially prolonged [Ca2+]i responses to bombesin, vasopressin, and PGF2 alpha, but had no such effect on the responses to PDGF. We conclude that differences between the [Ca2+]i responses to bombesin and PDGF, previously reported using cell populations, reflect differences occurring in individual cells, and that the [Ca2+]i responses to bombesin, vasopressin, and PGF2 alpha (but not PDGF) are subject to feedback inhibition via protein kinase C.

Increased secretion of insulin-like growth factor I into milk of cows treated with recombinantly derived bovine growth hormone.

Six lactating, non-pregnant Jersey cows were given subcutaneous injections of recombinantly derived bovine growth hormone for 7 d. Milk yield was increased by 4.5 kg/d on d 7, compared with the average yield of 10.7 +/- 0.4 kg/d (mean +/- s.e.m.) for the 7 d preceding treatment. Concentrations of insulin-like growth factor I (IGF-I) in the milk increased from 0.44 +/- 0.04 nmol/l (mean +/- s.e.m.) during the 7 d preceding treatment to 1.6 +/- 0.2 nmol/l on d 7 of treatment. Taking the increase in milk yield into account the total increase in the secretion of IGF-I into milk of one udder half was 6-fold. Plasma concentrations of total IGF-I rose from 15.5 +/- 1.3 nmol/l (mean +/- s.e.m.) on the day preceding treatment to 56.9 +/- 3.6 nmol/l (mean +/- s.e.m.) on d 7 of treatment. Mammary plasma flow increased from 1.6 +/- 0.09 to 2.2 +/- 0.06 l/min.udder half over the same time. Estimates of the amount of IGF-I that reached the mammary gland gave values of 24 and 116 nmol/min.udder half before and during treatment respectively. IGF-I in milk of treated cows was associated predominantly with proteins ranging from 40,000 to 150,000 mol.wt, but a significant proportion (19%) of the total IGF-I was present in the free unbound form. IGF-I crosslinking studies revealed the presence in milk of one specifically labelled band at 31,000 mol.wt.