Reagent proteins are usually required in much lower amounts han target proteins. Some can even be purchased commercially in sufficient quantities to meet the required need. Others, because of price or the required quantit expression. But, since only small quantities are usually required (<10mg), it is possible to choose an expression system with fea- tures that will favor efficient and rapid expression. Furthermore the expression scale can be minimized. The bottom line is that reagent proteins should be the least resource intensive to produce One should avoid trying to overproduce reagent proteins or scaling them to quantities that will never be used In contrast tic pre he most demanding in terms of resource. Therapeutic proteins have intrinsic biological properties like medical drugs. The ulti mate objective for expression of a therapeutic protein is the pro- duction of clinical-grade protein approaching or exceeding gram per liter quantities. For most expression systems this is not readily achievable. Other than bacterial and yeast expression, the most robust system for producing these levels is the Chinese hamster ovary(CHO)system. Due to the lack of proper post-translational modifications(e.g, glycosylation)in bacteria and yeast, CHO cell expression is often the only choice to achieve sufficient expres sion. Examples of therapeutic proteins, produced in CHO cells, include humanized monoclonal antibodies(Trill, Shatzman, and Ganguly, 1995), tPA (tissue plasminogen activator; Spellman et al 1989), and cytokines(Sarmiento et al., 1994). In many cases months are spent selecting and amplifying lines with appropriate growth properties and expression levels to meet production criteria What Do You know about the gene and the gene product Infor homologues or orthologues, enables one to make an educated guess as to what is the best eukaryotic expression system to use Is there anything published in the literature about the gene, or is it completely uncharacterized? Do we know in what tissue he gene is expressed, based on either Northern blot analysis or by quantitative or semiquantitative RT-PCR measures? Other factors to determine are whether the protein to be expressed is secreted, cytosolic, or membrane-bound. If it is a receptor, is it a homodimer, heterodimer, multimeric, single, or multispanning 496 Trill et alReagent proteins are usually required in much lower amounts than target proteins. Some can even be purchased commercially in sufficient quantities to meet the required need. Others, because of price or the required quantity, may necessitate recombinant expression. But, since only small quantities are usually required (<10mg), it is possible to choose an expression system with fea￾tures that will favor efficient and rapid expression. Furthermore the expression scale can be minimized. The bottom line is that reagent proteins should be the least resource intensive to produce. One should avoid trying to overproduce reagent proteins or scaling them to quantities that will never be used. Therapeutics In contrast to reagent proteins, therapeutic protein agents are the most demanding in terms of resource. Therapeutic proteins have intrinsic biological properties like medical drugs. The ulti￾mate objective for expression of a therapeutic protein is the pro￾duction of clinical-grade protein approaching or exceeding gram per liter quantities. For most expression systems this is not readily achievable. Other than bacterial and yeast expression, the most robust system for producing these levels is the Chinese hamster ovary (CHO) system. Due to the lack of proper post-translational modifications (e.g., glycosylation) in bacteria and yeast, CHO cell expression is often the only choice to achieve sufficient expres￾sion. Examples of therapeutic proteins, produced in CHO cells, include humanized monoclonal antibodies (Trill, Shatzman, and Ganguly, 1995), tPA (tissue plasminogen activator; Spellman et al., 1989), and cytokines (Sarmiento et al., 1994). In many cases months are spent selecting and amplifying lines with appropriate growth properties and expression levels to meet production criteria. What Do You Know about the Gene and the Gene Product? Information about the gene product or for that matter, its homologues or orthologues, enables one to make an educated guess as to what is the best eukaryotic expression system to use. Is there anything published in the literature about the gene, or is it completely uncharacterized? Do we know in what tissue the gene is expressed, based on either Northern blot analysis or by quantitative or semiquantitative RT-PCR measures? Other factors to determine are whether the protein to be expressed is secreted, cytosolic, or membrane-bound. If it is a receptor, is it a homodimer, heterodimer, multimeric, single, or multispanning 496 Trill et al