Proliferating Cell Nuclear Antigen
(PDB ID: 3VKX) from Homo sapiens
Created By: Alexander Clark
Human proliferating cell
nuclear antigen (hPCNA, PDBID: 3VKX) is a member of a superclass of DNA
processivity clamps. This family of
proteins is able to associate with over 200 proteins that contain a conserved
amino acid sequence (1). PCNA is most often associated with
recruitment of proteins for DNA replication and the stabilization of particular
DNA polymerases, such as Polδ. PCNA is also responsible for binding proteins
involved in prevention of rereplication, bypass replication, prevention of
sister chromatid recombination, base mismatch repair and base excision repair
among other processes found in the nucleus (2). Most of the proteins that are known to
associate with PCNA bind to one of three identical, structurally conserved regions on the outside surface of PCNA (2).
The PCNA family contains
six identical domains that come together to form the three identical binding
regions on the outside of PCNA. In
eukaryotic cells PCNA takes on a homotrimer complex, with each monomer composed
of two separate domains. The three
monomers bind head-to-tail to form a ring around DNA. The ring confirmation contains a set of
positively charged amphipathic α helices on the inside surface, adjacent to DNA, while the outer surface is primarily
β sheets with a
binding pocket responsible for associating with a conserved sequence on
associating proteins. The amino acids responsible for the positive charge on the amphipathic α helices are Lys-13, Lys-14, Lys-77, Lys-80,
Arg-146, Arg-149, Arg-210 and Lys-217. The interactions of the positively charged
residues with negatively charged DNA is thought to provide stability for the
PCNA homotrimer (2). The binding site on the outer surface of PCNA
consists of a patch of hydrophobic residues located on the C-terminal domain of
the monomer. The two domains of each
monomer are connected by a long sequence of amino acids called the interdomain connecting loop (IDCL). The IDCL and the
hydrophobic region are responsible for making the contacts needed with binding
proteins. The last six residues on the
C-terminal end of each monomer are also thought to associate with these
proteins; however, because of the highly dynamic character of these residues, a
crystallographic image of the interaction has not been recorded (3).
A highly conserved
PCNA-interacting protein box (PIP-box) motif is found in all of the proteins
that bind to the hydrophobic pocket of PCNA. The sequence of these residues follows the
pattern: Q-x-x-[ILM]-x-x-[FY]-[FY](2). Protein
p21 has a PIP-box with the highest known binding affinity to PCNA (PDB ID: 2ZVV). A key finding of the interaction between p21 and
PCNA is a hydrophobic association of Phe-150 and Tyr-151 in p21 with the
hydrophobic pocket of PCNA (4).
Studies have found that the closer these phenylalanine and tyrosine
residues are to the hydrophobic pocket of PCNA, the stronger the affinity (5). In
addition to hydrophobic packing, Tyr-151 forms a hydrogen bond with Tyr-133 on the IDCL of PCNA (4). The
interactions of the PCNA IDCL with Tyr-151 and Phe-150 are the source of the
strong affinity for p21. Because of this
strong affinity, p21 has been proposed as a potential cancer therapeutic, which
competitively binds to the active site of PCNA and blocks the binding of Polδ, arresting cells in the S phase (2).
The interacting PIP-box protein sequence
forms a 3/10 helix, hydrogen bonding with both the IDCL and beta sheeted
surface of PCNA. Hydrogen bonding occurs
around the hydrophobic pockets that are found on the surface of the protein (4). On
the surface of the PCNA active site, Asp-29 and Gln-125 contribute as hydrogen
bond acceptors while Gly-69 interacts with the C-terminal end of the PIP-box as
a hydrogen bond acceptor. The IDCL
contributes both hydrogen bond acceptors and donors from the carboxylic oxygen
and the nitrogen-linked hydrogen on the protein backbone. Asp-120 and Glu-124 provide hydrogen bond
acceptors for N-linked hydrogens on the PIP-box backbone. These residues
constitute the most important residues in the interaction of PCNA with
proteins.
Crystallization of PCNA with 3,3',5-triiodothyronine
(T3) shows that this drug can competitively compete for the same binding pocket as PIP-box containing proteins (6). However T3 induces thyroid activity, so a
small T3 derivative void of iodine, T2 amino alcohol (T2AA), was developed to
bind the pocket while not affecting thyroid activity. From crystallization of T3 with PCNA and the
similar cellular response of T2AA treatment, it is supposed that T2AA makes
contact with hPCNA at Met-40, Gln-131 and Pro-234. The
sulfide of Met-40 forms a weak dipole interaction with the hydrogen on a
positively charged nitrogen species.
Hydrophobic association occurs between Pro-234 and an aromatic ring of T2AA. The strongest interaction is between a
hydrogen bond donor on Gln-131 in the IDCL and a hydroxyl of T3. The in vivo treatment of cells with T2AA has been
shown to arrest cells in the S-phase of the cell cycle by preventing Polδ from binding PCNA, preventing Polδ from localizing to the DNA template.
All PCNA molecules
contain Lys-164, which has been shown to be an important residue in translesion
DNA synthesis (TLS) (7). A lesion is a type of error in base-pairing
of the DNA structure. The presence of a lesion causes replication to stop. TLS is a mechanism by which cells use
damage-tolerant polymerases that lack proofreading activity to replicate DNA
over a lesion. When Polδ stops, monoubiquitylation
of Lys-164 is triggered on the outer surface of PCNA. Ubiquitin is able to bind non-classical DNA
polymerases that are able to proceed over lesions (8). The conformation of the non-classical DNA
polymerase on PCNA is often referred to as a tool-belt, because it hangs off
one of the IDCLs until activated. Recent
models propose that when a lesion is encountered, ubiquitin binds Lys-164, then
a non-classical DNA polymerase binds ubiquitin and takes the place Polδ on the DNA strand (7).
Although the structure of
PCNA clamps is conserved, the sequence of these proteins varies greatly. The differences that do appear have little effect
on the function of PCNA molecules. Also,
the molecular weight and isoelectric points are expected to be similar for all
PCNA molecules, because they bind similar proteins and are found in similar
environments in all cells. Human PCNA
(MW = 28,768.78 Da, pI = 4.57) (9) structure was compared to S. cerviciae PCNA (PDB ID: 3K4X, MW =
29,560.03 Da, pI = 4.44) (9) using the Dali server. The Dali Server calculates a Z score by
comparing intramolecular distances between residues to compare the structure of
two proteins. Structures with a Z-score
greater than two are said to have significant structural similarities. The resultant Z-score was 33.6 (10), indicating the molecules have a similar structure, which was expected.
Following this, the primary
structures of these two proteins were compared using PSI-BLAST. This process compares the primary structure
of two proteins and outputs a measure of sequence homology (E). If E is less than 0.05, the proteins are
considered to show significant sequence homology. The sequence homology result was 4E-62 (11). This low value was expected, as the proteins
provide the same function and are known to have similar protein interactions. While this measure indicates they are
significantly similar, there is only a 35% residue agreement between the two
proteins. This means that only about a third
of the residues in PCNA are conserved from S.
cerevisiae PCNA to hPCNA. Because of
the conserved function between these two proteins it can be elucidated that the
matching residues are those most important to protein function.
Despite the lack of
sequence conservation from hPCNA and S.
cerevisiae PCNA, it is important to look at how these proteins differ in
their respective binding pockets. Upon
observing the hydrophobic pocket found on the surface of hPCNA and comparing it
to S. cerevisiae PCNA, eight of the
eleven residues are conserved, roughly three-quarters of the residues in this
region. The three residues that differ
in this pocket are switches from hPCNA to S.
cerevisiae PCNA of Met-40 to Val-40, Phe-129 to Glu-129 and Tyr-250 to Phe-250,
respectively. All of these switches are
either from a hydrophobic residue to a hydrophilic residue, or vice-versa. This result indicates these residues do not
have to be hydrophobic in order for the hydrophobic associations observed
between the PIP-box and PCNA to occur.
The residues on the IDCL that are important for electrostatic
interactions with the PIP-box sequences are also functionally conserved. Two of the four residues, Asp-120 and Tyr-133,
are conserved from hPCNA to S. cerevisiae
PCNA. The two other important residues switch
from Glu-124 and Gln-131 in hPCNA, to Asp-124 and Leu-131 S. cerevisiae PCNA. The Glu-124
to Asp-124 switch provides the same functional groups for hydrogen binding with
the PIP-box motif. The Gln-131 to
Leu-131 will most likely decrease the binding affinity of some motifs to S. cerevisiae PCNA as compared to hPCNA.
Although there are many
similarities between hPCNA and S. cerevisiae
PCNA, S. cerevisiae PCNA contains
Lys-107 and Lys 127, important in S. cerevisiae
PCNA regulation (8). Lys-107, in addition to Lys-164, can be
ubiquinated and form additional interactions with Polδ. Also
Lys-164 and Lys-127 of S. cerevisiae
have been shown to become SUMOylated, which has not been observed in hPCNA (8). SUMOylation results in the suppression of
unwanted recombination events. It can be
speculated that the lack of SUMOylation in hPCNA is a result of the lack of
Lys-127, and that SUMOylation is not as critical for suppression of unwanted
recombination events.
Human PCNA has been shown
to associate with a large number of proteins, because of a conserved amino acid
sequence in these target proteins. The
variety of proteins PCNA can accommodate lends it to providing a scaffold for
binding multiple cofactors involved in DNA replication and processing. Additionally, the cyclic conformation of PCNA
allows for binding to multiple proteins simultaneously as a way of tethering replication and regulatory proteins close to the DNA molecule.
Recently a drug was developed that competitively inhibits the binding of
PIP-box containing proteins to hPCNA (6). Competitive inhibition by this molecule of Polδ binding has potential
as a cancer therapeutic. The drug was
found to arrest cells in the S phase of replication as DNA polymerase is unable
to bind PCNA. The structural homology of PCNA clamps is highly conserved while sequence homology differs significantly. However, the regions important for PIP-box
binding are highly conserved from hPCNA to S.
cerevisiae PCNA. The nature of this
similarity will most likely extend to other PCNA family members, as they are
known to interact with similar proteins and have similar structures.