Delivery of protein therapeutics often requires frequent injections because of low activity or rapid clearance, thereby placing a burden on patients and caregivers. Using glycoengineering, we have increased and prolonged the activity of proteins, thus allowing reduced frequency of administration. Glycosylation analogs with new N-linked glycosylation consensus sequences introduced into the protein were screened for the presence of additional N-linked carbohydrates and retention of in vitro activity. Suitable consensus sequences were combined in one molecule, resulting in glycosylation analogs of rHuEPO, leptin, and Mpl ligand. All three molecules had substantially increased in vivo activity and prolonged duration of action. Because these proteins were of three different classes (rHuEPO is an N-linked glycoprotein, Mpl ligand an O-linked glycoprotein, and leptin contains no carbohydrate), glycoengineering may be generally applicable as a strategy for increasing the in vivo activity and duration of action of proteins. This strategy has been validated clinically for glycoengineered rHuEPO (darbopoetin alfa).

Platelet formation, occurring from bone marrow or lung megakaryocytes, has been difficult to study mechanistically. Recombinant human megakaryocyte growth and development factor (rHuMGDF), a recently described cytokine, has now been used to establish an in vitro system in which this important and little understood process occurs. CD34+ cells cultured with rHuMGDF develop into megakaryocytes which form long cytoplasmic extensions (proplatelets) that fragment into platelet-sized particles (in vitro platelets). Morphologically, in vitro and human plasma-derived platelets (control platelets) are virtually identical with respect to size, dense granule distribution and ultrastructural features. Functionally, in vitro and control platelets have similar aggregation and activation responses, and similarly incorporate mepacrine into dense granules. These findings suggest that rHuMGDF is sufficient to generate platelet-synthesizing megakaryocytes from CD34+ cells and provide an experimental setting in which the study of human platelet formation can be adequately performed.

Plant mitochondria have the unique ability to directly oxidize exogenous NAD(P)H. We recently separated two NAD(P)H dehy-drogenase activities from maize (Zea mays 1.) mitochondria using anion-exchange (Mono Q) chromatography. The first peak of activity oxidized only NADH, whereas the second oxidized both NADH and NADPH. In this paper we describe the purification of the first peak of activity to a 32-kD protein. Polyclonal antibodies to the 32-kD protein were used to show that it was present in mitochondria from several plant species. Two-dimensional gel analysis of the 32-kD NADH dehydrogenase indicated that it consisted of two major and one minor isoelectric forms. lmmunoblot analysis of submitochondrial fractions indicated that the 32-kD protein was enriched in the soluble protein fraction after mito-chondrial disruption and fractionation; however, some association with the membrane fraction was observed. l h e membrane-impermeable protein cross-linking agent 3,3'-dithiobis-(sulfosuccinimidylpropionate) was used to further investigate the submitochondrial location of the 32-kD NADH dehydrogenase. l h e 32-kD protein was localized to the outer surface of the inner mitochondrial membrane or to the intermembrane space. l h e pH optimum for the enzyme was 7.0. l h e adivity was found to be severely inhibited by p-chloromercuribenzoic acid, mersalyl, and dicumarol, and stimulated somewhat by flavin mononucleotide.

An NADH dehydrogenase activity from red beet (Befa vulgaris 1.) root mitochondria was purified to a 58-kD protein doublet. An immunologically related dehydrogenase was partially purified from maize (Zea mays 1. 873) mitochondria to a 58-kD protein doublet, a 45-kD protein, and a few other less prevalent proteins. Polyclonal antibodies prepared against the 58-kD protein of red beet roots were found to immunoprecipitate the NAD(P)H dehydrogenase activity. The antibodies cross-reacted to similar proteins in mitochondria from a number of plant species but not to rat liver mitochondrial proteins. The polyclonal antibodies were used in conjunction with maize mitochondrial fractionation to show that the 58-kD protein was likely part of a protein complex loosely associated with the membrane fraction. A membrane-impermeable protein crosslinking agent was used to further show that the majority of the 58-kD protein was located on the outer surface of the inner mitochondrial membrane or in the intermembrane space. Analysis of the cross-linked 58-kD NAD(P)H dehydrogenase indicated that specific proteins of 64, 48, and 45 kD were cross-linked to the 58-kD protein doublet. The NAD(P)H dehydrogenase activity was not affected by ethyleneglycol-bis(P-aminoethyl ether)-N,N'-tetraacetic acid or CaCI,, was stimulated somewhat (21 %) by flavin mononucleotide, was inhibited by pchloromercuribenzoic acid (49%) and mersalyl (40%), and was inhibited by a bud scale extract of Platanus occidentalis L. containhg plataneth (61 %).In contrast to their mammalian counterparts, plant mitochondria are capable of oxidizing cytoplasmic NAD(P)H directly, coupling this oxidation to the electron transport chain (Moller and Lin, 1986). This occurs via exogenous NAD(P)H DHs located on the cytosolic face of the inner mitochondrial membrane (Palmer and Moller, 1982;Moller, 1986;Moller and Lin, 1986; Douce and Neuberger, 1989). The oxidation of endogenous mitochondrial matrix substrates (NADH and succinate) has been shown to take precedence over the oxidation of exogenous NAD(P)H (Dry et al., 1983;Day et al., 1985). Thus, it has been suggested that the exogenous NAD(P)H DHs may function in balancing the redox levels between NAD(P)H pools in the cytosol and the mitochondrion (Moller and Lin, 1986; Douce and Neuberger, 1989). As a result, the exogenous NAD(P)H DH could coordinate glycolytic flwc with flow through the Krebs cycle. If this is true, then regulation of exogenous NAD(P)H DH activity could greatly influence plant metabolism. Consistent with this hypothesis, Kromer and Heldt (1991) have shown that the oxidation by mitochondria of reducing equivalents generated during photosynthesis is vital for obtaining maximum photosynthetic rates. Exogenous NAD(P)H DH activity can be distinguished from the other mitochondrial NAD(P)H DH activities by its insensitivity to rotenone (Wilson and Hanson, 1969), stimulation by Ca2' and inhibition by EGTA (Coleman and Palmer, 1971), and sensitivity to platanetin (Ravanel et al., 1986). Although these characteristics ca...