Stem cells in mammary tissue have been well characterised by using the mammary stem cell marker, cytokeratin (CK) 5 and the mature epithelial markers CK14, CK18 and CK19. As these markers have never been reported in cells from breastmilk, the aim of this study has been to determine whether mammary stem cells are present in expressed human breastmilk. Cultured cells from human breastmilk were studied by using immunofluorescent labelling and reverse transcription/polymerase chain reaction (RT-PCR). We found a heterogeneous population of cells with differential expression of CK5, CK14, CK18 and CK19. Further, by using the multipotent stem cell marker, nestin, we identified cells in culture that were positive only for nestin or double-positive for CK5/nestin, whereas no co-staining was observed for CK14, CK18 and CK19 with nestin. When cells isolated from breastmilk were analysed by using RT-PCR prior to culture, only nestin and CK18 were detected, thereby indicating that breastmilk contained differentiated epithelial and putative stem cells. Furthermore, fluorescence-activated cell-sorting analysis demonstrated, in breastmilk, a small side-population of cells that excluded Hoechst 33342 (a key property of multipotent stem cells). When stained for nestin, the cells in the side-population were positive, whereas those not in the side-population were negative. The presence of nestin-positive putative mammary stem cells suggests that human breastmilk is a readily available and non-invasive source of putative mammary stem cells that may be useful for research into both mammary gland biology and more general stem cell biology.

The mouse model of 2,4,6-Trinitrobenzene Sulfonic Acid (TNBS)-induced intestinal fibrosis allows for detailed study of the extracellular matrix changes that complicate Crohn's disease. Indomethacin induces intestinal fibrosis, while retinoic acid (RA) reduces liver fibrosis. Secreted protein acidic and rich in cysteine (SPARC), an extracellular matrix-modifying agent, may potentially link these opposing effects. Our aim was to determine the effects of indomethacin and RA and to evaluate their correlation to SPARC expression in the TNBS mouse model. CD-1 mice were randomised to TNBS enemas weekly for 2 or 8 weeks with or without indomethacin (0.2 mg/kg per day) or RA (100 microg/kg per day). At 2 weeks, indomethacin/TNBS enhanced and RA reduced inflammation, tissue destruction and fibrosis. The expression of SPARC was inversely related to fibrosis, but not to inflammation, in the TNBS-alone groups at 2 weeks; these differences were lost by 8 weeks. The results demonstrate that indomethacin increases TNBS-induced fibrosis in mice, while RA reduces it, and that SPARC may link these opposing effects.