2+-containing protein S inhibits extrinsic factor X-activating complex independently of tissue factor pathway inhibitor. J Thromb Haemost 2010; 8: 1976-85. Summary. Background: Protein S (PS) has direct anticoagulant activity, independently of activated protein C (APC

Summary. Background: Protein S (PS) has activated protein C-independent, direct anticoagulant activity (PS-direct). We reported that both multimers and monomers of affinity-purified PS have PS-direct similar to that in plasma, in contrast to another report. Objective: We extended our studies to establish the molecular forms and activity of plasma PS. Methods: Novel ELISAs were developed that could detect only multimeric, not monomeric, PS because they employed the same monoclonal antibody for capture and detection. PS forms were also examined on native PAGE immunoblots. A new activity assay for PS-direct was applied to plasma and gel-filtered plasma fractions. Results: Plasma PS multimers were clearly demonstrated using the ELISAs; 30-60% of free plasma PS appeared to be multimeric, a proportion similar to that of affinity-purified PS. On immunoblots, plasma PS multimers were more easily detected after gel filtration; plasma PS monomers and several apparent multimers comigrated with respective forms of affinity-purified PS. Antigen elution profiles after gel filtration of plasma revealed at least one major peak of apparent PS multimers (40-55% of free PS appeared multimeric). Biotinfactor Xa could bind to both plasma PS monomers and multimers. Strong plasma PS-direct was demonstrated, and plasma PS monomers, multimers, and PS-C4b-binding protein complexes each reconstituted PS-depleted plasma to similar levels of PS-direct. Conclusion: Our data are in disagreement with a report that monomeric purified PS has little PS-direct and that only monomeric PS exists in plasma. We find that both affinity-purified and plasma PS exist as monomers and multimers with similar PS-direct.

Iron shuttling in and out of cellular proteins and compartments comprises a chelatable or labile pool. Attempts at characterizing this pool have been largely unsuccessful. Using 3h labeling with tracer 59Fe, we identified a 1600 Da component in ultrafiltrates of cultured cell lysates that degraded to 600 Da over time (if not separated from the bulk of the cellular protein) and could be converted from 1600 to 600 Da with proteinase K. Both the 1600 and 600 Da components were negatively charged at pH 7+ and separated from cellular 59Fe‐labeled ferritin (Ft) in native PAGE. 59Fe in the unknown versus Ft varied inversely with total cellular Fe. However, analysis of Centricon 10 ultrafiltrates from human HepG2 and Caco2 cells showed that the Fe content did not vary appreciably with cellular Fe status (pretreatment with 180 uM Fe or 30 uM DFO for 24h, or untreated). Concentrations in this pool were 90‐ 200 ng/mg cell protein, with total iron contents ranging from 200 to1600 ng/mg. Mass spectrometry after laser fragmentation released multiple units of 44 Da (consistent with COO‐). We conclude that the cellular labile Fe pool stays quite constant despite changes in Fe status and may primarily consist of a complex of Fe with a small peptide plus carboxylate units. Supported in part by PHS Grant No. RO1 HD 46949.

We have determined that when cultured mammalian cells are exposed to 59Fe‐labeled 1–2 uM Fe(II)‐ascorbate for 3h, and cell lysates are applied to native PAGE, two radiolabeled peaks are shown. One is ferritin, the other of low molecular weight and filterable through a 10 kDa membrane. In the studies reported here, we determined the proportion of 59Fe in the unknown relative to ferritin in several cultured cell lines with and without 24h pretreatment with ferric ammonium citrate (10 ugFe/ml), or desferrioxamine (DFO) to increase and decrease iron availability. The ratio of 59Fe in the unknown to that in ferritin varied inversely with iron status. The unknown peak was present not just in cytoplasm but also in the lysosomes. After elution from native PAGE gels, two components were separated by size exclusion HPLC on Biosep 2000 coupled with ICP‐MS. This was also the case on an open Biogel‐P2 column, where Mr values of 1.8 and 0.6 kDa were obtained. Treatment of filtrates with proteinase K caused a shift in elution of the 1.8 kDa component to 0.6 kDa (detectable in native PAGE), indicating presence of peptide. Lectin analysis demonstrated a lack carbohydrate. Iron in the unknown components was directly available to DFO upon cell lysis, and could be chased out with non‐radioactive Fe. We hypothesize that these low molecular weight iron‐peptide complexes are part of the 'labile iron pool' in mammalian cells.