Search

The N-terminal 15 amino acids of the major protein associated with inorganic pyrophosphatase activity in Bacillus subtilis WB600 are identical to those of B. subtilis ORF yybQ. This ORF was amplified from B. subtilis WB600 DNA by PCR and cloned into an overexpression vector in Escherichia coli. Induction of overexpression produced a soluble protein of 34000 Da by SDS-PAGE and by matrix-assisted laser desorption and ionization mass spectrometry. The overexpressed protein had a high specific activity for the hydrolysis of magnesium pyrophosphate, and was specifically and reversibly activated by Mn2+ ions. These properties are identical to those of inorganic pyrophosphatase purified from B. subtilis WB600. No significant similarity was found between the derived sequence of the B. subtilis yybQ-encoded protein and published sequences of identified inorganic pyrophosphatases of Eukarya, Bacteria or Archaea domains. However, there is significant similarity to three putative proteins of unknown function from the archaea Methanococcus jannaschii and Archaeoglobus fulgidus, and from Streptococcus gordonii. The genomes of B. subtilis, M. jannaschii and A. fulgidus do not contain sequences similar to those of hitherto known soluble inorganic pyrophosphatases. The present findings, together with a survey of the properties of inorganic pyrophosphatases from 38 different sources, suggest that the B. subtilis yybQ-encoded protein is the first fully characterized member of a new class of inorganic pyrophosphatase.

1. Manganese(II) buffers were set up with inorganic triphosphate, trimethylenediaminetetraacetate and tetramethylenediaminetetraacetate to study the Mn dependence of beta 1,4-galactosyltransferase (lactose synthetase) in preparations of rat mammary gland. 2. In intact particulate preparations, treated with the calcium ionophore A23187, lactose synthesis was abolished by chelators and restored by bivalent transition metal ions in a manner characteristic of activation site I of the pure enzyme. Ni(II) also activated, as did Mg at high concentration. 3. Only Mn(II) could restore endogenous rates, giving an apparent Km of 0.1-0.2 microM, and eliciting about 70% full activity without addition of a site II activator. 4. In purified Golgi membrane vesicles, Mn gave an apparent Km of 0.4 microM. This increased sharply to about 10 microM on permeabilization with filipin, lysis with detergents, solubilization with Triton X-100, or in the pure enzyme. Preparations of chemically undamaged Golgi vesicles, known to include a proportion of the enzyme on exposed membranes, exhibited both low-Km and high-Km components. 5. The response of particulate galactosyltransferase to apparently physiological concentrations of free Mn(II) ion is interpreted as either due to a sensitizing factor within the Golgi lumen, or to the accumulation of Mn at elevated concentrations. Alternatively, the high Km of the soluble enzyme may reflect proteolytic damage.

This paper assembles published and new material to trace out the compartmental aspects of milk sugar synthesis in the mammary gland. The generation of lactose, from glucose and UDP-galactose, in thetrans cisternae of the Golgi stack is well established from biochemical and electron microscopy work. New experiments show that in the rat essentially all this lactose is available for sialyllation to α-2,3-sialyllactose, which accumulates in the same compartment. UDP arising from utilization of UDP-galactose inhibits lactose synthetase, but is removed by nucleoside diphosphatase. Therefore lactose synthetase, sialyltransferase and nucleoside diphosphatase form a coupled enzyme system located on the luminal face of the Golgi membrane. Transport of UMP to the cytosol, followed by its re-conversion via UDP and UDP-glucose to UDP-galactose, defines the operation of a uridine nucleotide cycle that links Golgi and cytosol compartments in the support of milk sugar synthesis. Evidence for carrier-mediated uniport or antiport of UMP and UDP-galactose across the Golgi membrane is discussed, together with the evidence for transport of glucose and inositol by small water-filled pores that forbid back diffusion of lactose, galactinol or sialyllactose. Possible intra-Golgi signals, achieving short-term regulation of milk sugar synthesis, include glucose, submicromolar free Mn(II) ions and cationic activator proteins. Experimental evidence for intracellular and intra-Golgi glucose concentrations in the range 0.1–0.4 mM imply lack of saturation of lactose synthetase. It is shown that lactose synthetase within undamaged Golgi membrane vesicles is about 100 times more sensitive to Mn ions than in damaged or solubilized membranes. Both lactose synthetase and sialyltransferase are greatly stimulated by basic proteins, and similar lactose synthease-activating material is extractable from Golgi preparations. Development of these features may designate the sorts of control exerted upon these Golgi enzymes, and perhaps others. This may uncover the mechanisms by which hormonal and nutritional changes, and manipulation of mammary tissue in vitro, cause profound alterations in milk sugar production.

Openreading frame mj0608 of the Methanococcus jannaschii genome, recognized by its sequence similarity to that of the gene coding for class C inorganic pyrophosphatase in Bacillus subtilis, was cloned and over-expressed in Escherichia coli. The protein was purified and characterized by SDS-PAGE, M(r), and N-terminal sequence. Under suitable conditions it catalyzed the specific hydrolysis of PPi at about 600 micromol x min(-1) x mg(-1) at 25 degrees C, and at 8000 micromol x min(-1) x mg(-1) at 85 degrees C. Therefore this protein is a specific inorganic pyrophosphatase. The activities of Mg(2+), Mn(2+), Co(2+), and Zn(2+) ions as cofactors for hydrolysis of PPi were compared at pH 7.5 and 9.0. Unlike the class C pyrophosphatase of B. subtilis, this enzyme required no prior activation by low concentrations of Mn(2+) or Co(2+) ions. However, prior exposure to these ions afforded striking protection against inhibition by sodium fluoride, to which the enzyme was otherwise very sensitive.

The hydrolysis of magnesium pyrophosphate by inorganic pyrophosphatase from Bacillus subtilis required its specific, time-dependent, and prior activation by Mn2+ ions. This was reversed when Mn2+ ions were removed with EDTA. Free Mn2+ ions were not required for catalysis. Pyrophosphatase purified to near homogeneity gave a single main band of apparent M(r) 36,000 by SDS-PAGE, but of M(r) 34,000 by matrix-assisted laser desorption ionization-mass spectrometry. The native enzyme equilibrated at pH 7 between three distinct molecular forms. Exposure to Mn2+ generated a catalytically active trimer of specific activity about 5000 mumol pyrophosphate hydrolyzed/min/mg protein. Exposure to EDTA generated two catalytically inactive forms, a dimer at low ionic strength and a separate form, of uncharacterized multimeric nature, at molar concentrations of Na2SO4 or Li2SO4. The latter form was an intermediate in the dimer-trimer transition caused by addition or removal of manganese ions. Mn2+ reacted with this "intermediate" form, apparently by reversible association with two noninteracting binding sites of Kd approximately 0.005 and 0.35 microM, respectively. The properties of this enzyme may account in part for the unusual manganese requirements of B. subtilis and related species.

Purification of human liver arginase by chromatography on DEAE-Sepharose, CM-Sepharose, hydroxylapatite, and MonoS yielded protein of greater than 95% purity by sodium dodecyl sulfate-gel electrophoresis. Detailed kinetic studies of the interconversion of active and inactive forms of arginase showed the effects of metal ion addition and withdrawal, metal ion type, time, temperature, and pH. At pH 7 and 37°C, removal of Mn 2+ caused a first-order deactivation with half-life of 1 h. Reactivation was completed within 0.5 min (1 mM Mn 2+ ) or 90 min (ca. 6 nM Mn 2+ ). Activation by Mn 2+ showed a hyperbolic response, with K d for Mn 2+ of about 36 nM. Mn 2+ apparently displaced about 2 H + , resulting in sigmoid dependence upon concentration of OH − . Both the maximal velocity of catalysis and the K m toward arginine were markedly pH-dependent in the physiological range. The findings lead to a model where Mn 2+ allosterically activates arginase by a sequential, and pH-sensitive, mechanism. The combined pH sensitivities of activation, V max , and K m are likely to give arginase a role in mediating the demonstrated pH control of the ornithine cycle and hence in the regulation of body pH.

The manganese dependence of arginase was reinvestigated with extracts of mouse liver to see whether more physiological properties were displayed than have been reported for the purified enzyme. In a preincubation with Mn(II) ions at 37 °C the enzyme underwent a slow and reversible activation. At least 90–95% of the activation achieved was dependent on Mn 2+ . However, no Mn 2+ was required for catalytic activity in the assay. The activation showed little dependence upon pH over the range 6.5–9.5, whereas the catalytic activity increased 12-fold in apparent accord with the titration curve of an ionizable group of p K α 7.9. The Mn 2+ dependence of arginase activation obeyed Michaelis-Menten kinetics, with K d varying from 0.3 μ m at pH 6.8 to 0.08 μ m at pH 7.7. Free Mn 2+ concentrations were established in these assays with a trimethylenediaminetetraacetate-Mn buffer. V max increased about three-fold over this range. The calculated arginase activity at 0.05 μ m Mn 2+ increases about nine-fold over this physiological pH range. An enzyme model is proposed to explain these findings. The activity of arginase at “physiological” [Mn 2+ ] and the pronounced pH dependence conferred upon it are consistent with a recently revised role for the urea cycle in the control of bicarbonate and pH in the body. It appears possible that arginase loses Mn 2+ sensitivity during the usual purification.

1. 1. Glucose uptake and lactose synthesis were measured in acini prepared from lactating-rat mammary gland and suspended in medium 199. 2. 2. Changes in the rate of glucose uptake, due to different concentrations of medium glucose or to the inclusion of phloretin or theophylline. were accompanied by larger changes in the rate of lactose synthesis. 3. 3. No clear evidence for cyclic nucleotide control of lactose synthesis was obtained. 4. 4. Direct measurement of intracellular glucose concentration (0.45 ± 0.08 mM) showed that it must be a rate-limiting factor for lactose synthesis and that glucose transport from the plasma into the mammary cell must be rate-limiting for glucose metabolism in general.

Galactosyltransferase was purified from Golgi membranes of lactating-rat mammary gland and studied with respect to its physical and enzymic (lactose synthetase) properties. The enzyme occurred in both monomeric (43–46 kDa) and apparently dimeric (90 kDa) forms. It was very unstable except in the presence of phospholipid, detergent, or cations binding to site 2. The amino acid composition and the N-terminal sequence closely resembled that of the human and bovine milk enzymes, particularly in respect to a Pro-Pro-Pro-Pro sequence. Kinetic studies demonstrated a high-affinity Mn2+-binding site (1) essential for activity, and a low-affinity Mn2+-binding site (2) that could also bind spermidine or clupeine. Mn2+ binding at site 2 raised Vmax fivefold. Spermidine binding at site 2 enhanced Mn2+ binding at site 1, and influenced binding of glucose. At physiological glucose concentration, clupeine or spermidine activated nearly as well as 15 mM MnCl2 and are regarded as models of a natural cation activator that remains to be isolated. Evidence is given for an essential histidine residue in the galactosyltransferase. It is proposed that site 1 Mn2+ participates directly in the reaction mechanism, whereas site 2 is a regulator site allosterically activated by a basic protein.

Lactose was synthesised within the lumen of purified Golgi membrane vesicles, prepared from lactating rat mammary gland, from externally added glucose and UDP-galactose. An apparent Km of 1.5 mM was shown towards glucose at anomeric equilibrium, but only beta-glucose was utilised. Two apparent Km values, about 17 micro M and 112 micro M, were shown towards UDP-galactose. 5-D-Thioglucose, 6-deoxy-D-glucose and 6-deoxy-6-chloro-D-glucose were alternative substrates, acquiring alpha-lactalbumin dependence when the vesicles were lysed with detergent. Substrates independent of alpha-lactalbumin, or inhibited by it, included a wide range of N-acylated glucosamines as well as phenyl-beta-glucoside. The galactosylation of these by vesicle preparations could be ascribed to a proportion of leaky vesicles. Suitably low concentrations of Triton X-100 activated lactose synthesis by intact vesicles, indicating the membrane as a rate-limiting feature of the system.

1. At short incubation times, and under suitable osmotic conditions, the lactose synthesized by Golgi-derived vesicles of rat mammary gland is 85-90% particulate. Evidence is presented for its occlusion within the lumen of the vesicles. 2. Ovalbumin is used as a bulky active-site inhibitor to show that the active site of lactose synthase lies on the inner face of the Golgi membrane. 3. Phlorrhizin and phloretin inhibit lactose synthesis by such vesicles, indicating the presence of a glucose-transport system. 4. The relationship of this topography to the synthesis of N-acetylneuraminyl-lactose and to the secretion of milk sugars is discussed.

SummaryThe free glucose concentration in the aqueous phase of samples of goat, sheep, cow, rat and rabbit milk was about 0·1–0·3 mn, while that in human milk was about 2 mM. During starvation the glucose concentration of goat milk fell considerably (by about 80 % in 2 d) in parallel with the decreased rate of lactose production. With rats fedad lib., glucose concentration in the milk was greater at 12.00 h than at 18.00 h, when lactose synthesis has been shown to decrease. 3-O-Methyl-D-glucose injected into the goat mammary gland via the teat canal specifically entered the blood. These findings support the idea that glucose equilibrates across the apical membrane of mammary secretory cells, so that milk glucose concentrations reflect intracellular glucose concentrations.

1. 1. Washed, particle preparations of fully lactating rat mammary gland hydrolysed glucose-6-phosphate at 191 ± 28 nmol/min/g fresh tissue, and a wide range of other phosphates at greater or lesser rates, at pH 6.4 and 37°C. 2. 2. Activity was maximal at pH 5 and pH 10 or above, and minimal at pH 6.4–7.0. 3. 3. Hydrolysis of glucose-6-phosphate at pH 6.4 did not follow Michaelis-Menten kinetics, and the rate at physiological concentrations of glucose-6-phosphate was very low compared with rates of tissue glucose uptake. 4. 4. It is concluded that the observed activity is not due to a specific glucose-6-phosphatase (EC 3.1.3.9), and cannot be utilized as a microsomal marker.

1. myo-Inositol was galactosylated by UDP-galactose in the presence of alpha-lactalbumin plus rat mammary Golgi membranes enriched in galactosyltransferase (EC 2.4.1.22). 2. The isolated product migrated on GLC as two peaks of material, apparently identical to galactinol (galactosylinositol) isolated from rat milk. 3. These findings make it likely that in vivo lactose synthetase is responsible for converting inositol into galactinol within the Golgi lumen. 4. This is consistent with the ability of inositol to penetrate the Golgi membrane in vitro, and renders less likely a previously proposed role of beta-galactosidase (EC 3.2.1.23) in mammary galactinol synthesis.

Elucidation of the details of lactose synthesis, in particular its dependence upon alpha-lactalbumin and its location within the lumen of the Golgi apparatus, now allows one to ask useful questions pertaining to its regulation. Attention is directed towards galactosyltransferase itself (EC 2.4.1.22), which appears to be rate-limiting in the uridine nucleotide cycle that supports lactose synthesis, and to those factors that may affect its activity. In laboratory animals alpha-lactalbumin appears to be the major agent of regulation during lactogenesis but is not necessarily limiting at other times, whereas the increase in amount of galactosyltransferase seems largely to account for the rising yield of lactose during lactation. Studies with pinched-off Golgi membrane vesicles, together with measurements of intracellular chemical concentrations, suggest that beta-glucose and uridine diphosphategalactose do not saturate lactose synthesis and are, therefore potentially regulatory features of this process. Further aspects of lactose synthesis that may offer points of regulation include calcium ions, generation of protons within the Golgi lumen, and the generally rate-limiting nature of the Golgi membrane.