BackgroundThe fatty acid mixture of human milk is ideal for the newborn but little is known about its composition in the first few weeks of lactation. Of special interest are the levels of long-chain PUFAs (LCPUFAs), since these are essential for the newborn’s development. Additionally, the LCPUFAs arachidonic acid (AA), eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) are precursors for lipid mediators which regulate inflammation.MethodsWe determined the composition of 94 human milk samples from 30 mothers over the first month of lactation for fatty acids using GC-MS and quantified lipid mediators using HPLC-MS/MS.ResultsOver the four weeks period, DHA levels decreased, while levels of γC18:3 and αC18:3 steadily increased. Intriguingly, we found high concentrations of lipid mediators and their hydroxy fatty acid precursors in human milk, including pro-inflammatory leukotriene B4 (LTB4) and anti-inflammatory and pro-resolving lipoxin A4 (LXA4), resolvin D1 (RvD1) and resolvin E1 (RvE1). Lipid mediator levels were stable with the exception of two direct precursors.ConclusionsElevated levels of DHA right after birth might represent higher requirements of the newborn and the high content of anti-inflammatory and pro-resolving lipid mediators and their precursors may indicate their role in neonatal immunity and may be one of the reasons for the advantage of human milk over infant formula.

The conserved oligomeric Golgi (COG) complex is a tethering factor composed of eight subunits that is involved in the retrograde transport of intra-Golgi components. Deficient biosynthesis of COG subunits leads to alterations of protein trafficking along the secretory pathway and thereby to severe diseases in humans. Since the COG complex affects the localization of several Golgi glycosyltransferase enzymes, COG deficiency also leads to defective protein glycosylation, thereby explaining the classification of COG deficiencies as forms of congenital disorders of glycosylation (CDG). To date, mutations in COG1, COG4, COG7 and COG8 genes have been associated with diseases, which range from severe multi-organ disorders to moderate forms of neurological impairment. In the present study, we describe a new type of COG deficiency related to a splicing mutation in the COG5 gene. Sequence analysis in the patient identified a homozygous intronic substitution (c.1669-15T>C) leading to exon skipping and severely reduced expression of the COG5 protein. This defect was associated with a mild psychomotor retardation with delayed motor and language development. Analysis of different serum glycoproteins revealed a CDG phenotype with typical undersialylation of N- and O-glycans. Retrograde Golgi-to-endoplasmic reticulum trafficking was markedly delayed in the patient's fibroblast upon brefeldin-A treatment, which is a hallmark of COG deficiency. This trafficking delay could be restored to normal values by expressing a wild-type COG5 cDNA in the patient cells. This case demonstrates that COG deficiency and thereby CDG must be taken into consideration even in children presenting mild neurological impairments.

Inhibition of constitutive active signaling pathways, which are a characteristic phenomenon for many tumors, can be an effective therapeutic strategy. In contrast, oncogenic transcription factors, often activated by mutational events, are in general less amenable to small-molecule inhibition despite their obvious importance as therapeutic targets. One example of this is alveolar rhabdomyosarcoma (aRMS), in which specific translocations lead to the formation of the chimeric transcription factor PAX3/FKHR. Here, we found unexpectedly that the transcriptional activity of PAX3/FKHR can be inhibited by the kinase inhibitor PKC412. This occurs via specific phosphorylation sites in the PAX3 domain, phosphorylation of which is required for efficient DNA-binding and subsequent transcriptional activity. Consequently, we show that PKC412 exerts a potent antitumorigenic potential for aRMS treatment both in vitro and in vivo. Our study suggests that posttranscriptional modifications of oncogenic transcription factors can be explored as a promising avenue for targeted cancer therapy. [Cancer Res 2008;68(10):3767-76]

S100A5 is a novel member of the EF-hand superfamily of calcium-binding proteins that is poorly characterized at the protein level. Immunohistochemical analysis demonstrates that it is expressed in very restricted regions of the adult brain. Here we characterized the human recombinant S100A5, especially its interaction with Ca 2؉ , Zn 2؉ , and Cu 2؉. Flow dialysis revealed that the homodimeric S100A5 binds four Ca 2؉ ions with strong positive cooperativity and an affinity 20 -100-fold higher than the other S100 proteins studied under identical conditions. S100A5 also binds two Zn 2؉ ions and four Cu 2؉ ions per dimer. Cu 2؉ binding strongly impairs the binding of Ca 2؉ ; however, none of these ions change the ␣-helical-rich secondary structure. After covalent labeling of an exposed thiol with 2-(4-(iodoacetamide)anilino)-naphthalene-6-sulfonic acid, binding of Cu 2؉ , but not of Ca 2؉ or Zn 2؉ , strongly decreased its fluorescence. In light of the three-dimensional structure of S100 proteins, our data suggest that in each subunit the single Zn 2؉ site is located at the opposite side of the EF-hands. The two Cu 2؉ -binding sites likely share ligands of the EF-hands. The potential role of S100A5 in copper homeostasis is discussed.Calcium, a versatile second messenger of extracellular signals inside the cell, binds to a multitude of cytosolic Ca 2ϩ -binding proteins each of which can, in turn, regulate several effector proteins. The S100 protein family (18 different members) constitutes the largest group of Ca 2ϩ -binding proteins of the EF-hand type (1, 2). At least 13 S100 genes are clustered on human chromosome 1q21, leading to the designation S100A1 to S100A13 for the protein products of these genes (3). Their expression is cell-and tissue-specific, and different human diseases have been associated with deregulated expression (4, 5). S100 proteins are non-covalent homodimers with the notable exception of the heterodimeric S100A8-S100A9. Each monomer possesses 2 EF-hands as follows: a classical C-terminal EFhand with a canonical Ca 2ϩ -binding loop of 12 amino acids and a N-terminal EF-hand with a loop of 14 amino acids which is specific for S100 proteins. This structural difference likely is the reason for the large difference in the Ca 2ϩ affinities of the N-and C-terminal EF-hands. The affinities of S100 proteins for Ca 2ϩ are in general rather low with [Ca 2ϩ ] 0.5 1 values of 100 -300 mM (for review see Refs. 2 and 6), and in 6 out of 8 studied cases binding of Ca 2ϩ displays positive cooperativity. Zn 2ϩ binds to several S100 proteins; the Zn 2ϩ and Ca 2ϩ sites are distinct and can modify the affinity for Ca 2ϩ . S100B binds four Zn 2ϩ ions with concomitant 10-fold increases of the Ca 2ϩ affinity (7). High affinity binding of two Zn 2ϩ to the S100A12 dimer leads to induction of two high affinity Ca 2ϩ -binding sites (8). S100A3 binds eight Zn 2ϩ , but without effect on the affinity for Ca 2ϩ (9, 10). S100A2 binds four Zn 2ϩ with high affinity, in a manner antagonistic to Ca 2ϩ (11). Recently it was reported...