Egr-l is an immediate-early response gene induced by diverse signals that initiate growth and differentiation. Its cDNA sequence predicts a protein with zinc fingers. We have generated an antiserum to the Egr-1 gene product and identified it as an 80-kilodalton short-lived protein in serum-stimulated mouse fibroblasts. The rat Egr-l product has also been identified in nerve growth factor-induced PC12 cells. In addition, we show by cell fractionation and immunocytochemistry that the Egr-l protein is located in the nucleus. We also demonstrate that it is phosphorylated. In vitro-generated Egr-l protein binds with high affinity to the sequence CGCCCCCGC in a zinc-dependent manner.

We have previously described a patient in whom the breakpoint occurred within the first intron of the BCR gene and have cloned the 9q+ and 22q- junctions. We have now determined the nucleotide sequence around the breakpoints on both translocation products from this patient as well as the corresponding regions from the normal chromosomes 9 and 22. We have compared the sequence with that of the breakpoint regions in the Ph1-positive leukemic patients in order to check for the presence of conserved motifs. A + T-rich sequences and ALU repeat elements are the only sequence characteristics which appear to be very common around translocation regions. The chromosome 9 ABL sequences at or adjacent to the breakpoints present in the 22q- product show homology to the consensus ALU sequence while the chromosome 22 sequences do not, suggesting a non-homologous recombination mechanism. While no sequences are deleted, there is a two-base-pair "homology" at the junction. Therefore, staggered breaks followed by ligation and repair could be part of the mechanism involved in the process of translocation in some cases of Ph1-positive ALL.

Colonization of the vaginal mucosa with uropathogens from fecal flora is an important step in ascending urinary tract infections (UTIs) in women. Colonization is influenced by interactions between uropathogens, vaginal fluid, and epithelial cells. In this study, vaginal fluid from 21 women was assessed for effects on adherence of type 1 piliated Escherichia coli to the A431 cell line. Adherence to cells was enhanced by all fluid specimens when tested at low fluid protein concentrations in an in vitro assay. At higher concentrations, certain specimens maintained enhanced binding whereas others resulted in diminished binding. Increases in adherence were associated with increased binding of E. coli to vaginal fluid in vitro and with higher vaginal fluid pH. These results demonstrate that vaginal fluid significantly alters the adherence of type 1 piliated E. coli to epithelial cells in vitro and, therefore, should be studied as a potential modifier in bacterial colonization and UTIs in vivo.

To better understand the interactions involved in bacterial adherence and the role of mucus in the pathogenesis of urinary tract infections, we developed a system to study the binding of a recombinant Escherichia coli strain, HB101/pWRS1-17, expressing type 1 pili, to vaginal mucus collected from 28 women. Bacteria bound to differing extents to all specimens examined, and preincubation of bacteria with mannose inhibited binding by 50 to 89%. Additionally, all mucus samples showed reactivity with anti-mannose antibody, and the levels of reactivity correlated with the levels of bacterial binding, suggesting that the mannose-terminal saccharides present on these glycoproteins are the receptors for the binding of type 1-piliated bacteria. Mucus specimens collected over periods of 5 days and 12 weeks exhibited significant variation in bacterial binding, indicating temporal differences in the ability of vaginal mucus to act as a receptor for type 1-piliated E. coli. The results show that vaginal mucus can bind bacteria and may thus influence the initial attachment and subsequent colonization of the vaginal and urinary tract epithelium by E. coli.

The importance of Lewis blood group antigens in recurrent urinary tract infections has led to a more detailed study of the expression of these antigens in an in vitro culture system. Expression of A, B, H, and Lewis blood group antigens by primary cultures of ureter epithelial cells was analyzed by flow cytometry. Cells maintained blood group determinants for up to 7 weeks in culture over three passages. In several specimens, the cells that exhibited expression of Lewis A (Le(a) and Lewis B (Le(b)) antigens increased dramatically over 2-3 passages in culture. Lewis Y (Le(y)) antigen expression was positive on all primary cultures tested. The effect of fetal bovine serum (FBS) on blood group antigen expression was evaluated. Increased concentrations of FBS in the growth media increased the expression of Le(a) and Le(b) antigens in certain specimens, but did not affect expression of the other blood group antigens. The effect of FBS on blood group antigen expression of cultured cells depended on the Lewis blood type of the individual donating the specimen. To evaluate the heterogeneity often seen in Le(a) and Le(b) antigen expression, dual staining of cells for Le(a) and Le(b) antigens was performed. The results of these studies show for the first time that Lewis blood group antigens are expressed by cultured normal ureteral epithelial cells and that the pattern of expression of Le(a) and Le(b) changes over time. Furthermore, an individual cell may simultaneously express more than one Lewis antigen on its surface.