Multiple phosphorylation of SCF degraded proteins: (Nash, 2001; Deshaies, 2001)

• Allow for a temporal threshold to be built into the destruction of the phosphorylated protein.
  • explanation: If 5 equivalent phosphorylations are needed to trigger the protein's destruction. Then destruction will initially be slow, while the first four sites are being phosphorylated. After a lag period the protein will be degraded more rapidly. Now suppose that the protein's destruction is driven by a single phosphorylation. There will no longer be a lag period between start of protein phosphorylation and the protein's destruction,, even if the phosphorylation is slowed down. Thus, multistep phosphorylation can build a time delay into the degradation of the target protein.
• Allow protein to ignore low levels of phosphorylation (background signaling), and then be able to respond decisively to a high level of phosphorylation (a true signal).
• How are multiple phosphorylations counted?
  • F-Box could have binding sites for all of the phosphorylation sites (unlikely for Cdc4-Sic1 intereaction)
    • Scatchard analysis and Hill plots of binding data reveal only one class of phosphopeptide binding sites on Cdc4 and no evidence of enthalpic cooperativity.
    • A single phosphopeptide derived from cyclin E competes with multiplyphosphorylated Sic1 for binding to Cdc4.
    • Three conserved arginine residues in the WD-40 domain that are required for the phosphopeptide binding activity of Cdc4 are predicted by structural modeling to form a single pocket.

     

  

• But maybe an SCFCdc4 complex can recognize more
  than one phosphate at a time after all. There are at least
  three possible ways to envision this. First, the phosphoultrasensitive
  peptide binding pocket may accommodate more than
  one phosphopeptide at a time. Note that the Scatchard
  analysis does not exclude the existence of multiple comdegradation,
  parably strong sites that bind phosphopeptides noncoresponse
  operatively, or weak secondary binding sites. Second
  (as suggested by the authors), Cdc4 may function as
  an oligomer (Kominami et al., 1998; Wolf et al., 1999),
  with each Cdc4 monomer contributing a binding site (or
  sites). Third, Cdc4 may possess one or more secondary
  phosphopeptide binding sites in addition to the arginineitive
  lined pocket. The basic idea is illustrated in Figure 3,
  where we show how bisphoshorylated Sic1 might intercascade,
  act with two binding sites on a monomeric Cdc4. The
  first phosphate binds the arginine pocket on Cdc4.
  Given that Sic1 is highly flexible (Nash et al. 2001), one
  can imagine the second phosphorylated residue flopand
  ping around within a sphere whose radius is
  5 nm.
  One phosphopeptide per 5 nm sphere translates to an
  effective concentration of about 3 mM—a 3000-fold in-
  crease over the normal concentration of Sic1 in budding
  yeast cells (
  1 M). Thus, the binding of the first Sic1
  phosphopeptide to the arginine pocket tethers the secsharpens
  ond phosphopeptide in such close proximity to Cdc4
  that even a very weak binding interaction becomes fasubsequent
  vorable (i.e., entropic cooperativity). In considering po-
  tential secondary binding sites, it is worth noting that
  other amino acids besides arginine can make energetiable
  cally favorable contacts with phosphate (Lu et al., 1999).
• But still,
  we are left with the question of why a sixth phosphorylasingle
  tion event appears to be so crucial. Here, the solution
  could lie in the relationship between binding energy and
  dissociation constants. The dissociation constant is
  proportional to the logarithm of the binding energy, not
  to the binding energy itself. Thus, each phosphate could
  decrease the Kd by a multiplicative factor of 2 or 10 or
  100, and the difference in how much complex can be
  formed by Sic15P versus Sic16P at physiological
  Cdc4 concentrations could be considerable.
• eye for Cln-CDK to aim at. For sake of argument, we
  hypothesize that the true degron comprises four phos-
  phorylations. If so, there would be only one way to form
  a stable complex between Cdc4 and a Sic1 molecule
  that contains four phosphorylation sites. If Sic1 contains
  five sites, there are 5 different ways to form quadruply
  phosphorylated Sic1 (and one way to form Sic15P),
  resulting in a potential increase in binding affinity of
  6-fold. If Sic1 contains six sites, there are 6!/4!2!  15
  different possible configurations of Sic14P, 6 configurations
  of Sic15P, and Sic16P (22 total). Thus, in-
  creasing the number of phosphorylation sites from 4 to
  6 can enhance the statistical likelihood of generating a
  complex between quadruply phosphorylated Sic1 and
  Cdc4 by 22-fold! This idea—a sort of combinatorial co-
• CDK sites in substrates serves to tune downstream responses
  to spatial or temporal gradients of CDK activity.
  A major challenge for the future will be to see how many
  “nanobioprocessors” akin to Cdc4 are embedded in the
  cell’s circuitry, and how they are wired together to calculate
  a cell’s biology.