Previously we reported that stable transfection of human UDP-glucose pyrophosphorylase (hUGP2) rescued galactose-1-phosphate uridyltransferase (GALT)-deficient yeast from "galactose toxicity." Here we test in human cell lines the hypothesis that galactose toxicity was caused by excess accumulation of galactose-1-phosphate (Gal-1-P), inhibition of hUGP2, and UDP-hexose deficiency. We found that SV40-transformed fibroblasts derived from a galactosemic patient accumulated Gal-1-P from 1.2+/-0.4 to 5.2+/-0.5 mM and stopped growing when transferred from 0.1% glucose to 0.1% galactose. Control fibroblasts accumulated little Gal-1-P and continued to grow. The GALT-deficient cells had 157+/-10 micromoles UDP-glucose/100 g protein and 25+/-5 micromoles UDP-galactose/100 g protein when grown in 0.1% glucose. The control cells had 236+/-25 micromoles UDP- glucose/100 g protein and 82+/-10 micromoles UDP-galactose/100 g protein when grown in identical medium. When we transfected the GALT-deficient cells with either the hUGP2 or GALT gene, their UDP-glucose content increased to 305+/-28 micromoles/100 g protein (hUGP2-transfected) and 210+/-13 micromoles/100 g protein (GALT-transfected), respectively. Similarly, UDP-galactose content increased to 75+/-12 micromoles/100 g protein (hUGP2-transfected) and 55+/-9 micromoles/100 g protein (GALT-transfected), respectively. Though the GALT-transfected cells grew in 0.1% galactose with little accumulation of Gal-1-P (0.2+/-0.02 mM), the hUGP2-transfected cells grew but accumulated some Gal-1-P (3.1+/-0.4 mM). We found that 2.5 mM Gal-1-P increased the apparent KM of purified hUGP2 for glucose-1-phosphate from 19.7 microM to 169 microM, without changes in apparent Vmax. The Ki of the reaction was 0.47 mM. Gal-1-P also inhibited UDP-N-acetylglucosamine pyrophosphorylase, which catalyzes the formation of UDP-N-acetylglucosamine. We conclude that intracellular concentrations of Gal-1-P found in classic galactosemia inhibit UDP-hexose pyrophosphorylases and reduce the intracellular concentrations of UDP-hexoses. Reduced Sambucus nigra agglutinin binding to glycoproteins isolated from cells with increased Gal-1-P is consistent with the resultant inhibition of glycoprotein glycosylation.

No abstract

Adherence of Streptococcus mutans to smooth surfaces has been attributed to the production of sucrose-derived d -glucans. However, several studies indicate that the bacterium will adhere in the absence of sucrose. The present data confirmed that S. mutans adherence to saliva-coated hydroxyapatite beads in the absence of sucrose is described by the Langmuir equation. The nature of the sucrose-independent adherence was studied with the Persea americana agglutinin as a selective adherence inhibitor. Pretreatment of the bacterium with P. americana agglutinin caused a 10-fold reduction in adherence, and the inhibition was not reversed with the addition of sucrose. Pretreatment of S. mutans with proteases also reduced adherence, regardless of the sucrose content, whereas periodate oxidation and glucanohydrolase treatment of the bacteria reduced sucrose-mediated adherence to the levels found for sucrose-independent adherence. The P. americana agglutinin, glucanohydrolase, and pepsin pretreatment of the cells did not eliminate sucrose-induced agglutination. Scanning electron microscopy showed that short streptococcal chains were bound to saliva-coated hydroxyapatite crystals in the sucrose-independent system, whereas the presence of sucrose caused larger bacterial clumps to be found. A two-reaction model of S. mutans adherence was developed from these data. It is proposed that one reaction is attachment to the tooth pellicle which is mediated by cell-surface proteins rather than glucans or teichoic acids. The other reaction is cellular accumulation mediated by sucrose-derived d -glucans and cell surface lectins. A series of sequential adherence experiments with P. americana agglutinin as a selective inhibitor provided presumptive evidence for the validity of our model of S. mutans adherence.

Classical galactosemia (G/G) is caused by the lack of galactose-1-phosphate uridyltransferase (GALT) activity. A more common clinical variant, Duarte/Classical (D/G) produces partial enzymatic impairment. Although neonatal death due to G/G galactosemia has been largely eliminated by population-based screening and intervention, long-term outcome in some is associated with impaired growth, ovarian failure, dyspraxic speech and neurologic deficits. At least 32 variants in the nucleotide sequence of the GALT gene have been identified and 9 have transferred impaired GALT activity to transformed cells in transfection experiments. We here define the prevalence and biochemical phenotype of two mutations. An A to G transition in exon 6 of the GALT gene converts a predicted glutamine at codon 188 to an arginine (Q188R), and introduces a new HpaII cut site into the gene which enables population screening by polymerase chain reaction. An A to G transition in exon 10 in the GALT gene produces a codon change converting an asparagine to aspartic acid at codon 314 (N314D) and adds an AVA II cut site. We screened a large population for the Q188R and N314D sequence changes to investigate the prevalence of Q188R in G/G galactosemia, the effect of homozygosity for Q188R on outcome, and the prevalence and biochemical phenotype of the N314D sequence change. We found that the Q188R mutation has a prevalence of 62% in a predominately Caucasian population of 107 patients with G/G galactosemia. Homozygosity for Q188R was associated with a poor clinical outcome in a subgroup of these patients. The N314D mutation is associated with the Duarte biochemical phenotype with extraordinary concordance.

No abstract