Biology Reference
In-Depth Information
the receptor, confirming a requirement for bivalent binding. Wrighton and col-
leagues screened peptide phage libraries and one peptide was identified that
could bind and agonize EPOR [54]. This peptide, AF11154, had no homology
to EPO. It self associates into dimers to form a bivalent molecule that could
homodimerize EPOR in a 2:2 mimetic:EPOR complex [58]. Additional
sequence modification of this peptide resulted in EMP1, a 20-amino acid pep-
tide with an approximate 50-fold increase in affinity over the starting peptide
(Tab. 1). The affinity was increased more by covalent linkage of the peptide
[59]. The activity of this peptide was still significantly lower (500-fold) than
that of rHuEPO when tested in an in vitro bioassay. In addition, the in vivo
activity was very low (25,000-fold less than rHuEPO). This work demonstrat-
ed, however, that a molecule smaller than rHuEPO could successfully dimerize
and activate the receptor. An attempt to discover other EPO agonist peptides by
another group was also successful [61]. The in vitro activity was not increased,
however, and the size of this peptide was not decreased compared with EMP1.
One explanation for the low in vivo activity of mimetic peptides is their
rapid clearance. One group addressed this problem by creating a fusion protein
between EMP1 and a larger protein, plasminogen activator inhibitor
(PAI1)[60], resulting in an increase in molecular weight from 4.8 Kd to 66 Kd.
The in vivo activity was significantly increased (2500-fold); its in vivo activi-
ty, however, was still significantly less than that of rHuEPO (100-fold) and the
ability to be delivered orally was compromised by the size increase.
The peptide mimetics described above are significantly larger (4.2 Kd) than
the preferred size of an orally bioavailable compound (<0.6 Kd). These pep-
tides may be used to design lead compounds of smaller size. Some small mol-
ecule agonists have been isolated based on the EMP1 structure [56]; however,
their in vitro activities were low (Tab. 1). An independent approach was to
directly screen for small-molecule EPO mimetics that could dimerize EPOR.
Small molecule libraries containing compounds with two-fold symmetry were
screened to find dimerizing compounds that agonize the receptor [55, 63]. This
strategy did not result in discovery of agonist compounds, however. A small
molecule (compound 1, approximately 5 Kd) that bound but did not agonize
EPOR was discovered. Compound 1 was made active in vitro by oligomeriz-
ing it with a multivalent crosslinker resulting in a molecule (compound 5) con-
taining eight compound-1 molecules joined together. Compound 5 binding to
EPOR was increased somewhat (10-fold) over that of compound 1; however,
the size (6.4 Kd) was greater than that required to be orally bioavailable. In
addition, its in vitro activity was low relative to rHuEPO and the compound
had toxicity. The feasibility of the small-molecule approach to discovery of
small molecules that could agonize EPOR was demonstrated, however.
Further progress in development of small molecule EPO mimetics has been
slow. Progress has been made with the development of small molecules that
can agonize other cytokine receptors including granulocyte colony-stimulating
factor receptor [64] and insulin receptor [65]. The insulin mimetic is notewor-
thy in that it is orally active in rodents [66]. This work demonstrates that small
Search WWH ::




Custom Search