Serum biomarker levels for musculoskeletal disease in two- and three-year-old racing Thoroughbred horses: A prospective study of 130 horses.

REASON FOR PERFORMING STUDY: Biomarkers have shown some in vivo promise for the detection of musculoskeletal injuries, but further study to assess biomarker levels in clinical orthopaedic disease is required. OBJECTIVE: To assess 7 serum biomarkers for the detection of musculoskeletal injuries. METHODS: Two- and 3-year-old racehorses were entered into the study (n = 238). Exit criteria were lack of training for >30 days, or completion of 10 study months. Data from horses with solitary musculoskeletal injuries and completion of >2 months were analysed. Musculoskeletal injury was considered intra-articular fragmentation (IAF), tendon or ligamentous injury (TL), stress fractures (SF) and dorsal metacarpal disease (DMD). Monthly lameness examination and serum collection were performed. Serum was analysed for glycosaminoglycan (GAG), type I and II collagen degradation (C1, 2C), type II collagen synthesis (CPII), type II collagen degradation (Col CEQ), aggrecan synthesis (CS846), osteocalcin (OC) as a marker of bone formation and (C-terminal telopeptide of type I collagen) CTX as a marker of bone degradation. RESULTS: Of the 238 horses 59 injured and 71 uninjured control horses met the analysis criteria. Based on injury no significant differences in the proportions were observed for age, gender or lesion type, although a higher proportion of injuries occurred at the beginning of the study. Of injured horses, 16 (27%) sustained an IAF, 17 (29%) a TL injury, 7 (12%) SF and 19 (32%) were diagnosed with DMD. There were significant changes seen in biomarkers based on the injury incurred when longitudinal samples were assessed. Furthermore, based on the serum biomarkers collected prior to injury, horses could be correctly classified as injured or uninjured 73.8% of the time. CONCLUSIONS: A unique biomarker pattern occurred before each injury and this was beneficial in classifying horses as injured or uninjured. POTENTIAL RELEVANCE: Biomarkers have the potential to be used as a screening aid prior to musculoskeletal injury.

Gene expression analysis of EpiDerm following exposure to SLS using cDNA microarrays.

There is a need to investigate the mechanistic basis of the human skin irritation response if relevant in vitro test systems for the predictive identification of skin irritation hazards are to be developed. Recent progress in genomics technologies mean that tools for the identification and investigation of important biochemical events in the processes of skin irritation are now available. The aim of this work was to identify genes (for further mechanistic investigation) which may be regulated in response to skin irritation, following exposure of the EpiDerm skin model to the known skin irritant sodium lauryl sulphate (SLS). EpiDerm cultures were treated in triplicate with a non-cytotoxic dose of SLS (0.1 mg/ml, as determined by the MTT assay and histological examination) for 15 min, 30 min, 1 h, 2 h, 3 h, 4 h and 24 h. Total RNA was extracted from the pooled EpiDerm cultures and used to probe Atlas human arrays (Clontech) covering approximately 3600 genes. Preliminary data indicated an up-regulation at early time points (15-30 min) of a number of genes involved in transportation (e.g. the sodium and chloride dependent taurine transporter) and receptors (e.g. ZAP70 and protocadherin 42 precursor). The gene encoding the UV excision repair protein and other DNA repair genes (e.g. DNA-directed RNA polymerase II) were up-regulated after 1-3 h, along with TGF beta 3 and other tumour suppressors, which play a role in cellular development and wound healing. At the later time points of 4-24 h, genes involved in protein translation (e.g. Cathepsin D receptor) and metabolism (e.g. CYP27A) were up-regulated. In addition, a number of genes were down-regulated in response to treatment with SLS, although these followed less of a time dependent pattern. These results indicate the differential regulation of a number of genes in response to treatment with SLS, some of which may provide additional clues to the molecular events underpinning the irritation response to this particular surfactant and possibly to other chemical irritants.

The use of genomics technology to investigate gene expression changes in cultured human liver cells.

The field of genomics has great potential in toxicology; however, the technology is still in its infancy and there are many questions that need to be addressed. In this study we focus on the use of toxicogenomics for the determination of gene expression changes associated with hepatotoxicity. The human hepatoma cell line HepG2 was used to assess the toxic effects of two well-studied hepatotoxins, carbon tetrachloride (CCl(4)) and ethanol (EtOH). Replicate dishes of HepG2 cells were exposed to two concentrations of CCl(4) and EtOH--doses which caused 20% and 50% cell death (as determined by the MTT assay) were chosen [0.18% and 0.4% (v/v) CCl(4); 2.5% and 5% (v/v) EtOH] and the cells exposed for periods of 2 and 24 h. mRNA was extracted and used to probe Atlas Human Toxicology II arrays (Clontech). Preliminary data revealed that following a 2-h exposure at the low doses of both compounds, few changes in gene expression were detected. However, after 24-h exposure of the cells to the same low concentration of both compounds, multiple changes in gene expression were observed, many of which were specific to the individual hepatotoxins, presumably reflecting their different mechanisms of action. CCl(4) treatment of HepG2 cells gave rise to treatment specific up-regulation of genes involved in extracellular transport and cell signalling, whereas EtOH treatment gave rise predominantly to down-regulation of genes involved in stress response and metabolism. In addition, changes in regulation of certain genes (involved in stress response and cell cycle) were common to both treatments. Exposure of HepG2 cells to higher doses of the hepatotoxins gave rise to more changes in gene expression at lower exposure times. These results strongly suggest that different mechanisms of hepatotoxicity may be associated with specific patterns of gene expression, while some genes associated with common cellular responses may be useful as early markers of toxicity.

Endocrine disrupters--testing strategies to assess human hazard.

During the last decade an hypothesis has been developed linking certain chemicals (natural and synthetic) to observed and suspected adverse effects on reproduction in both wildlife and humans. The issue of 'endocrine disruption' originally focused on chemicals that mimic the action of the natural hormone oestrogen. However, the concern is now encompassing effects on the whole endocrine system. In response to public awareness, regulatory agencies (including the US EPA) and the OECD are formulating potential testing strategies and have begun the process of validating defined tests to systematically assess chemicals for their endocrine-disrupting activities. In order to investigate chemicals that have the potential to cause endocrine disruption, a large number of in vitro and in vivo assays have been identified. In vitro test systems (particularly when used in combination) offer the possibility of providing an early screen for large numbers of chemicals and can be useful in characterising the mechanism of action and potency. In vitro assays in widespread use for the screening/characterisation of endocrine disrupting potential include hormone receptor ligand binding assays (determination of the ability of a chemical to bind to the hormone receptor), cell proliferation assays (analysis of the ability of a chemical to stimulate growth of oestrogen sensitive cells), reporter gene assays in yeast or mammalian cells (analysis of the ability of a chemical to stimulate the transcription of a reporter gene construct in cell culture), and the analysis of the regulation of endogenous oestrogen sensitive genes in cell lines. However, in vitro assays do not always reliably predict the outcome in vivo due to differences in metabolic capabilities of the test systems used and the diverse range of mechanisms by which endocrine disrupting chemicals may act. Therefore a complementary battery of short- and long-term in vitro and in vivo assays (that assess both receptor and non-receptor mediated mechanisms of action) seems the most appropriate way at present of assessing the potential endocrine disrupting activities of chemicals. At Unilever we have used a combination of in vitro assays (receptor binding, reporter gene and cell proliferation assays) together with short-term in vivo tests (uterotrophic assay in immature rodents) to examine the oestrogenic potential of a large number of chemicals. An evaluation of the advantages and limitations of these methods is provided. Finally, any potential test system needs to be validated and standardized before the information generated can be for the identification of hazard, and possibly for risk assessment purposes.

Safety evaluation of phytosterol esters. Part 1. Assessment of oestrogenicity using a combination of in vivo and in vitro assays.

Phytosterols are natural constituents of the human diet, and as part of an extensive programme of safety evaluation studies investigating their use as a novel food ingredient, the possible oestrogenic effects of phytosterols have been investigated using a combination of in vitro and in vivo assays. Competitive binding with the immature rat uterine oestrogen receptor (ER) has been used to measure the ability of phytosterols to bind to ERs while the transcriptional activation of oestrogen-responsive genes has been examined in an oestrogen-inducible yeast screen. Phytosterols did not display any activity in these in vitro assays. Uterotrophic assays have been conducted to investigate the potential for phytosterols to elicit an oestrogenic response when administered orally to immature female rats (n = 10) at doses of 0, 5, 50 or 500 mg/kg/day for 3 consecutive days. Phytosterols (a well characterized mixture of beta-sitosterol, campesterol and stigmasterol) and phytosterol esters (the previous phytosterol mixture esterified with fatty acids from sunflower oil) did not exhibit oestrogenic activity in the immature female rat using uterine wet weight as the endpoint. Beta-oestradiol (0.4 mg/kg/day) consistently produced a significant increase in uterus weights. Coumestrol (a known phytoestrogen) was also tested as a weak positive control and produced a dose response at doses of 20, 40 and 80 mg/kg/day in the uterotrophic assay. In conclusion, we have shown that phytosterols do not bind to the ER and do not stimulate transcriptional activity of the human ER in a recombinant yeast strain. In addition, there was no indication of oestrogenicity from the uterotrophic assay when the material was administered by oral gavage to immature female rats.

Interpretation of the in vitro proliferation response of mcf-7 cells to potential oestrogens and non-oestrogenic substances.

Chemicals may interact with the oestrogen receptor (ER) in mammalian cells and thereby exhibit oestrogenic activity which may result in effects on development and reproduction in both wildlife and humans. It has been suggested that in vitro assays provide the best way forward for investigating the oestrogenic potential of large numbers of chemicals. We compared the effects of a series of chemicals in an MCF-7 cell proliferation assay with results from similar assays reported in the literature in order to investigate the practical use of such assays. Alkylphenols and benzylbutylphthalate were found to cause proliferation of MCF-7 cells, which could be antagonized by tamoxifen. However, all the substances tested, including ethanol, progesterone and epidermal growth factor, were found to have some effect on cell proliferation, suggesting that cell proliferation may result from mechanisms other than via the ER. This is in contrast to some reports that MCF-7 cells only respond to oestrogens and highlights MCF-7 cell strain differences. Such differences and other reports in the literature of the effects of growth modulators on MCF-7 cells indicate that interpretation of results from MCF-7 cell proliferation assays needs to be considered carefully. Comparison between the results of in vitro assays and appropriate in vivo assays is also required in order to evaluate the significance of in vitro oestrogenicity assay results and an understanding of the mechanisms responsible for the biological effects is necessary.

Modulation of MCF-7 cell proliferative responses by manipulation of assay conditions.

The use of MCF-7 cell proliferation assays has been proposed as one of a number of available in vitro assays to determine the oestrogenic potential of chemicals. However, there is concern that differences between strains of MCF-7 cells and protocols used may lead to significant interlaboratory variability, particularly in their response to the natural oestrogen 17beta-oestradiol, and inability to detect weak oestrogens (Villalobos et al., 1995). This investigation adds to the debate by demonstrating that manipulation of assay conditions may modulate the proliferative response of a particular MCF-7 cell strain to 17beta-oestradiol. MCF-7 cells were put in 12-well plates, then exposed for 6 days to 17)3-oestradiol in EMEM without phenol red using serum stripped of endogenous oestrogens (EMEM/H-). A maximum proliferation of 200-250% control (assessed by fluorescein diacetate uptake) was observed, which is lower than some reported values (up to eightfold increase) and would probably reduce the sensitivity of detection of a weak oestrogen. Preincubation of cells in EMEM/H- for 48 hr, prior to the addition of 17beta-estradiol, resulted in an increased maximum proliferation of 400% control or more. Differences were also seen in the shape of the time course of response to 10(-9)m 17beta-oestradiol; preincubated cells exhibited exponential growth after 6 days whereas those exposed directly after plating plateaued after 4 days. The increased response to 17beta-oestradiol may be due to oestrogen receptor induction following transfer of the cells to oestrogen-free medium. This increase in responsiveness of the cells may increase the sensitivity of this strain of MCF-7 cells for the detection of oestrogenic chemicals.

The nature and significance of multiple isoforms of the oestrogen receptor in breast tumours.

Oestrogen receptors (ERs) in breast tumours are highly heterogeneous. In previous studies we have shown that at least four isoforms may exist. These migrate in isoelectric focusing (IEF) gels to isoelectric points (pI values) 6.1, 6.3, 6.6 and 6.8. Of these the first (pI 6.1) corresponds to the 8S isoform as detected by sucrose gradient fractionation, while the others all sediment at 4S. In a series of 66 breast tumours it was found that those at pI 6.3 and pI 6.8 were significantly correlated with the presence of progesterone receptors. To characterize the isoforms more fully, ER isoforms labelled by [3H]oestradiol binding were fractionated by IEF. The results were compared with those obtained after sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE) and immunoblotting using the H222 anti-ER monoclonal antibody. In other experiments, tumour ER isoforms were covalently labelled with [ring-3H] tamoxifen aziridine and separated by IEF. The individual isoforms were electroeluted from the IEF gel and further analysed by SDS-PAGE and non-denaturing PAGE. In summary, the evidence shows that the isoforms of pI values 6.3, 6.8 and 6.6 have molecular masses of 50, 65 and 70 kDa respectively. In addition, all three of these isoforms, i.e. the pI 6.3, 6.8 and 6.6 isoforms, could form dimers. We conclude that the three isoforms sedimenting at 4S have the capacity to form dimers and thus may have the potential for binding to oestrogen response elements in the genome.

Oestrogen receptor isoforms, their distribution and relation to progesterone receptor levels in breast cancer samples.

Oestrogen receptors (ER) in breast cancer tumours are highly heterogeneous. In this study, the variability in the profile of ER isoforms and its relation to progesterone receptor (PgR) levels in breast tumours has been studied. Using high resolution isoelectric focusing (IEF) 4 ER isoforms can be detected with pI values of 6.1 (corresponding to the 8S ER), and 6.3, 6.6 and 6.8 (all of which have a sedimentation pI values of 6.1 (corresponding to the 8S ER), and 6.3, 6.6 and 6.8 (all of which have a sedimentation coefficient of approximately 4S in sucrose density gradients). Data were obtained on the soluble receptors from supernatants of 66 ER-positive primary breast tumour homogenates using high resolution IEF. In 43 of these samples PgR levels were also measured. The isoform at pI 6.6 was present in 97.0% of tumours, the isoform at pI 6.1 in 83.3%, the pI 6.3 isoform 39.4% of tumours and the pI 6.8 isoform in only 33.3% of tumours. Only 12.1% of tumours studied contained the full complement of ER isoforms (pI 6.1, 6.3, 6.6 & 6.8). The ER isoforms at pI 6.1 & 6.8 were only found in PgR-positive (> 10 fmol PgR/mg protein) tumours. Some tumours contained only a single ER isoform at pI 6.6 or 6.1, but those at pI 6.3 and 6.8 were never found singly. Tumours containing 3 or 4 ER isoforms had significantly higher levels of PgR (> 90 fmol/mg protein) than those with only 1 or 2 (P < 0.001). The presence of ER isoforms at pI 6.3 and pI 6.8 also significantly correlated with high levels of PgR (P < 0.001). This variability in the ER isoform profile of breast tumours and their correlation with PgR levels may have a bearing on prognosis and tumour response to endocrine therapy.

4 S oestrogen receptor isoforms and their distribution in breast cancer samples.

The variability in the profile of oestrogen receptor (ER) isoforms in breast tumours has been studied. Using low-resolution isoelectric focussing (IEF), two major ER isoforms with isoelectric point (pI) values of 6.1 and 6.6 could be identified, with corresponding sedimentation coefficients in sucrose density gradients of 8 S and 4 S respectively. Using high-resolution IEF or immunoblotting, the pI 6.6 form (4 S) was shown to be composed of three different species, with pI values of 6.3, 6.6 and 6.8, while the oligomeric pI 6.1 protein (8 S) did not show charge heterogeneity. Data were obtained on the soluble receptors from supernatants of 42 ER-positive primary breast tumour homogenates using high-resolution IEF to obtain ER isoform profiles. It was found that 54.7% of tumours contained the isoforms at pI 6.6 and 6.1, while only 11.9% contained the full complement of isoforms (pI 6.1, 6.3, 6.6 and 6.8). Of the tumours studied, 11.9% contained isoforms of pI 6.1, 6.6 and 6.8, with 14.3% containing isoforms with pI 6.1, 6.6 and 6.3. Very few tumours contained only one isoform, with 4.8% of tumours containing a single isoform at pI 6.1 and 2.4% of tumours containing only the isoform at pI 6.6. All four ER isoforms were also shown to be present in some tumours by immunoblotting using antibody H222 and, in addition, high-resolution IEF indicated that all isoforms bind oestradiol, diethylstilboestrol and tamoxifen. The variability in the ER isoform profile may have a bearing on the known variability of tumour response to endocrine therapy and prognosis.