PEGylation of somatropin (recombinant human growth hormone): Impact on its clearance in humans

1. PHA-794428 is a PEGylated version of somatropin (human growth hormone). The pharmacokinetics of PHA-794428 have been studied in humans following single subcutaneous administration (dose range 10–500 µg kg^?1). In the same study the pharmacokinetics of somatropin were also determined following a 3.6 mg (51 µg kg^?1) subcutaneous dose. Comparison of the pharmacokinetics of both molecules indicates that PEGylation of somatropin with a 40 kD PEG results in a ten- to 20-fold increase in area under the curve and a similar increase in half-life when compared with somatropin in human (at equivalent subcutaneous doses).

2. Literature data indicate that somotropin is cleared by two mechanisms. The first processes is clearance by glomerular filtration. This is a passive, non-capacity-limited process. A second, capacity-limited, process is mediated by interaction with growth hormone receptors present in a number of tissues including the liver. It is hypothesized that PHA-794428 shares the same clearance mechanisms. However, the addition of the PEG moiety has modulated the clearance by both of these processes. Pharmacokinetic modelling of human serum concentration data obtained for these molecules strongly supports this hypothesis. The renal clearance is reduced due to the increased size of the molecule (Cl/F reduced from 9.6 to 0.1 l h^?1 for somatropin and PHA-794428, respectively). In addition, the reduction in growth hormone receptor affinity has reduced the clearance mediated by interaction with this receptor (somatropin K_m = 3.6 µg l^?1 and V_max = 104 µg h^?1/PHA-794428 K_m = 53 µg l^?1 and V_max = 84 µg h^?1).     

Unique immunochemical features and intracellular stability of platelet fibrinogen.

The characteristics of fibrinogen present in fresh and stored platelets have been immunochemically characterized and compared with plasma fibrinogen. Protein components of fresh platelets and platelets stored for 2 or 4 days appeared comparable with respect to mobilities and quantitative absorbance in polyacrylamide gels, and similar quantities of fibrinogen were present in platelet lysates at each time. Immunochemical analysis indicated however that the platelet fibrinogen after 2 or 4 days was not identical to the fresh platelet fibrinogen in the lysates. Differences were noted with respect to the expression of native antigenic determinants of plasma fibrinogen as well as the antigenic expressions of the Aα, Bβ and γ chains. The differences in immunochemical expression were both qualitative and quantitative in nature, and the stored platelet samples were less comparable to plasma fibrinogen than the fresh platelet lysate. No changes in the immunochemical expression of plasma fibrinogen was observed in these samples stored for the same periods of time. Comparison of plasma fibrinogen and platelet fibrinogen with respect to the immunochemical expressions of the native molecule and of the constituent chains indicated that the molecules were not identical. Marked differences were noted between fresh as well as stored platelet fibrinogen samples and reference plasma fibrinogen. These results suggest that platelet and plasma fibrinogen are not absolutely identical and establish the necessity for the use of fresh platelets for the isolation and detailed immunochemical and fine structural characterization of platelet fibrinogen.     

A novel vascular endothelial growth factor heparin-binding domain substructure binds to glycosaminoglycans in vivo and localizes to tumor microvascular endothelium

Vascular endothelial growth factor has an exon 7-encoded heparin-binding domain. To explore the expression of complementary ligands on endothelial surfaces in vivo and to assess potential for localization within the vascular tree, we introduced a truncated version of this domain (HBDt) into a modified M13 phage. Despite the small size and trace-level expression, this HBDt endowed the phage with affinity for heparin to which it bound in vitro. It also preferentially and selectively localized the phage in vivo to vascular endothelial surfaces, especially of tumors. Competition assays demonstrated that accumulation and localization of this phage was attributable to expression of the HBDt on the phage surface and sequence comparison suggests its novelty. We propose to use this novel HBDt structure to explore the expression of the ligand glycosaminoglycans within the vascular tree. This structure may facilitate directed delivery of therapeutic molecules.     

Prostate-specific membrane antigen directed selective thrombotic infarction of tumors

Prostate-specific membrane antigen (PSMA), a glutamyl preferring carboxypeptidase, is found in prostate and other carcinomas present on both tumor cells and associated microvascular lining cells. We find that the channel structures delineated by PSMA-expressing cells in human and rat prostate tumors are in functional continuity with the vasculature and thus form part of tumor microvasculature. The PSMA-positive cell-outlined channels are CD31 negative and mutually exclusive of CD31-positive cell-lined channels elsewhere in the tumor consistent with tumor cells adapted to a pseudoendothelial phenotype in vasculogenic mimicry. To assess the functional potential of such PSMA-lined microvasculature to selectively direct infarctive tumor therapy, we coupled the soluble extracellular domain of tissue factor to a PSMA catalytic site inhibitor to create a PSMA-directed selective tumor vascular thrombogen (STVT). This protein induced selective local in vivo infarctive necrosis of the rat Mat Lu prostate tumor when administered i.v. The combined administration of this STVT with low-dose doxorubicin produced a profound tumoricidal effect, resulting in complete eradication of some tumors. This is consistent with the therapeutic potential for a PSMA-directed STVT and expands the potential for selective infarctive ablation of tumors.     

Comparative immunochemical characterization of products of plasmin and leukocyte protease cleavage of human fibrinogen.

Plasmin-dependent as well as a plasmin-independent alternative pathways for fibrinolysis have been recognized, but means of discriminating between the products of these two pathways have not been established. To this end, a major product of the cleavage of human fibrinogen by neutral leukocyte proteases, a 270,000 MW derivative designated Fragment I, has been immunochemically compared with the D:E complex produced by plasmic cleavage. Both classes of cleavage fragments exhibited absolute antigenic deficiency relative to native fibrinogen. Differences in the expression of residual native fibrinogen determinants were distinctive and localized to both the D and the E domains of the molecule; quantitative differences were predominantly localized to the D domain, whereas differences in competitive inhibition slopes were restricted to the E domain. The cleavage-associated neoantigens, fg-D_neo and fg-E_neo, were expressed differently by the two types of cleavage products. In contrast to the complete expression of fg-E_neo by the plasmic D:E complex, this neoantigen was minimally expressed by Fragment I but somewhat comparable to that of plasmic fragment X. Fragment I was clearly distinguishable from the plasmic cleavage products with respect to the expression of fg-D_neo. The competitive inhibition slopes differed, and Fragment I was antigenically incomplete for fg-D_neo. It is concluded that the cleavage products of fibrinogen generated by neutral leukocyte proteases and by plasmin are immunochemically distinguishable; and assays for fg-D_neo may provide a means for discrimination of the pathways of fibrinolysis mediated by these enzymes.     

Glomerular Complement Components in Human Glomerulonephritis

154 of 255 individual human renal biopsies studied by immunofluorescence contained varying combinations of immunoglobulins (Ig), complement (C) components C1q, C3, C4, C5, C6, C8, C3 proactivator (C3PA), and/or properdin. 10 patients had linear deposits of Ig in glomeruli characteristic of antiglomerular basement membrane (GBM) antibodies; nine patients had C3 deposits (minimal in three) with generally lesser amounts of C1q, C4, C5, C6, and/or C8. 118 of the patients had granular deposits of Ig, suggesting immune complex glomerulonephritis; 114 of these had deposits of C3, usually accompanied by C1q, C4, C5, and/or C6. These observations indicate that the entire C sequence is deposited in glomeruli in most Ig-mediated glomerulonephritides. However, certain cases of anti-GBM glomerulonephritis with few or no C deposits may utilize pathways of injury independent of C.21 patients had granular C3 deposits without detectable Ig. C5, C6, and C8 were present in the majority of these patients while C1q was absent and scant C4 was observed in only two patients. The presence of only late-acting C components in the absence of Ig, C1q, and C4 suggests selective, possible nonimmune activation of the alternate C pathway. Finally, five patients had granular deposits of C3, C5, C6, and/or C8 diffusely in all or most glomeruli with a lesser number of glomeruli having additional focal granular deposits of Ig, C1q, and C4. This observation suggests that at least two patterns of C activation can occur simulatenously, possibly triggered by antecedent immune complex deposition and then perpetuated by an as yet undetermined mechanism.     

A hybrid fibronectin motif protein as an integrin targeting selective tumor vascular thrombogen

Targeted thrombotic eradication of solid tumors is a novel therapeutic strategy. The feasibility, efficacy, selectivity, and safety are dependent on multiple variables of protein design, molecular assembly, vascular target, and exclusive restriction of function to the tumor vasculature. To advance this strategy, we describe a design of an integrin targeting selective tumor vascular thrombogen. We adopted the fibronectin structural motif of tandem repeating modules with four type III repeat modules of fibronectin followed by two structurally homologous modules of the extracellular domain of tissue factor. This hybrid protein of six tandem modules recognizes integrins and selectively docks and initiates the thrombogenic protease cascade locally on the target cell surfaces. The protein is inactive in blood but is functionally active once assembled on integrin-positive cells. When administered i.v. to tumor-bearing mice, it selectively induces extensive local microthrombosis of the tumor microvasculature. The principles are addressed from the perspective of protein structural design for a class of selective tumor vascular thrombogen proteins that, through interaction with tumor angiogenic endothelium, elicit thrombotic occlusion rather than apoptosis or arrest of angiogenesis. This response can produce local tumor infarction followed by intratumoral ischemia-reperfusion injury, inflammation, and a local host tumor eradicative response.