The Molecule of the Month series explores the functions and
significance of selected biological macromolecules for a general
audience. These features, written and illustrated by Dr. David
S. Goodsell of the Scripps Research Institute, are available at
January, 2002 -- Oxygen and nutrients are delivered throughout our bodies through the watery
transport system of the blood. Using a liquid delivery method poses two challenges. First, it leaves
the entire body open to infection, since bacteria and viruses will be quickly distributed everywhere that
the blood goes. The immune system, with antibodies as the first line of defense, fights this danger.
Second, there is the constant danger of damage to the blood circulatory system. Blood is pumped
throughout the body under pressure, and any small leak could lead to a rapid emptying of the entire
system. Fortunately, the blood carries an emergency repair system: the blood clotting system.
When we are cut or wounded, our blood builds a temporary dam to block the damage, giving the
surrounding tissues time to build a more permanent repair.
Thrombin is at the center of this process of blood clotting. Blood clotting starts with molecules that sense
that something is wrong. For instance, the protein tissue factor is found on the surfaces of cells that
are not normally in contact with the blood. If tissue is cut, the blood flows out of the blood vessels and
encounter tissue factor. Then, a cascade of signaling starts, beginning with a few tissue factor molecules
and amplifying, like a pyramid scheme, into a response large enough to cure the entire problem.
Tissue factor activates a few molecules of Factor VII. These then activate a lot of Factor X. And finally,
these activate even more thrombin. Thrombin, when activated, then translates this signal into action. It
clips a little piece off of the large protein fibrinogen, causing it to assemble into large stringy networks.
These networks then trap lots of blood cells, forming the dark red scab that blocks the damage.
Thrombin is a serine protease: a protein-cutting enzyme that uses a serine amino acid to perform the
cleavage. Other examples of serine proteases are trypsin and chymotrypsin, enzymes involved in
digestion. Thrombin, however, is more specific than these gastrointestinal cleavage machines. It is
designed to perform the specific cleavage needed to activate fibrinogen, without digesting all the other
important proteins in the blood. The active site can be seen in the structure of activated thrombin from
PDB entry 1ppb, at the base of a deep groove. A key serine amino acid is actived by a histidine here.
An aspartate also helps in the activation.
Of course, blood clotting must be carefully regulated, otherwise the blood would be clotting in all of the
wrong places. Errors in blood clotting have disastrous effects: improper blood clots in the heart can
cause heart attacks and misplaced blood clots in the brain cause strokes. Thrombin is controlled in
two ways. First, it is built as an inactive precursor, shown in part in PDB files
2pf2. The inactive
form has several extra domains that are clipped off when the protein is activated. Calcium ions bind to
specially modified glutamate amino acids. The strong positive charge on these ions tether the protein to
the surfaces of blood vessels, so that thrombin stays put. Since thrombin is not free to spread, blood clots,
once they start, will not spread everywhere. Only the thrombin right next to damage will be activated.
Second, once thrombin is activated, as in PDB entry
1ppb, it lasts only seconds,
also limiting the clot to area of damage.
Blood clots are not always wanted. For instance, many people take small doses of aspirin, under the
direction of their doctors, to reduce the chance of the blood clots that cause heart attacks. Aspirin acts
on the protein cyclooxygenase, which is important in another aspect of clot formation that uses small
cell fragments called platelets. The rat poison warfarin, not commonly used these days, blocks the
formation of the modified glutamate amino acids that hold calcium ions. The unfortunate rats then die
because of uncontrolled blood clotting. Leeches, as you might expect, also detest blood clots, because
it means the end to their meal. They make special proteins that block thrombin (or other enzymes),
stopping the formation of the clot. One example, a protein called hirudin, is shown in PDB entry
This leech protein blocks the active site of thrombin perfectly.
PDB entry 1mkx is a
perfect structure for exploring thrombin. It contains two molecules of the protein,
one in inactive form (chain K) and one activated (chain H and L). In order to activate the protein, the
protein strand must be cleaved between two segments on one side. Then, the two new ends separate
and the whole protein relaxes into the active form. In the active form, the key catalytic serine amino acid
changes position and points straight out into the active site, ready to perform the cleavage.
A list of all thrombin structures in the PDB as of January, 2002, is available at
For additional information on thrombin, see
Nitrogenase: The Nitrogen-splitting Anvil
Nitrogen-fixing bacteria have the ability to convert nitrogen gas into
ammonia, which is easily combined with other raw materials to form the
building blocks of proteins and nucleic acids. This process requires extreme
measures, because nitrogen gas is so stable. The industrial process used to
create ammonia requires high temperatures and pressures of 300
atmospheres, along with catalysts. In nitrogen-fixing bacteria, the enzyme
nitrogenase drives the reaction with a large quantity of ATP, and uses a
collection of metal ions, including an unusual molybdenum ion, to perform
Nitrogenase is composed of two components, such as those found in PDB entry
The MoFe protein contains all of the machinery to perform the reaction,
but requires a steady source of electrons. The reaction requires the addition of six electrons for each
nitrogen molecule that is split into two ammonia molecules. The Fe protein uses the
breakage of ATP to pump these electrons into the MoFe protein. In the typical reaction, two molecules
of ATP are consumed for each electron transferred. Nitrogenase also converts hydrogen ions to hydrogen
gas at the same time (this might be an obligatory part of the nitrogen-splitting reaction, or it might be a
simple side effect), thus consuming even more ATP in the process.
This is a large investment in energy, but well worth the effort if nitrogen is
not available in the environment. Fortunately, nitrogen-fixing bacteria are
found throughout the world, and are often found in partnerships with plants.
For instance, legumes build special nodules in their roots that provide a
perfect home for the bacteria. The plants provide shelter and even a few
essential nutrients, jealously guarding their guests, and the bacteria provide
a steady supply of nitrogen.
At the heart of nitrogenase is an unusual complex of iron, sulfur and a molybdenum ion, which
is thought to perform the nitrogen-fixing reaction. A string of cofactors feed electrons to this
MoFe-cluster. Electrons start at a pair of ATP molecules (two at each end of the dimeric complex),
flow inwards into the iron-sulfur cluster, then to the P-cluster, and finally to the MoFe-cluster.
A homocitrate molecule helps to stabilize this unusual metal ion. The P-cluster is in the middle
and the iron-sulfur cluster of the Fe protein is at the top. In spite of the detailed knowledge provided
by the beautiful structures of nitrogenase, such as that found in PDB entry
1n2c, the actual binding
site for nitrogen gas is still a subject of controversy and intense study.
The metal clusters are the centerpiece of nitrogenase, and are the major attraction on any tour of
the structures. PDB entry 1n2c
is a good place to start--it contains both the MoFe protein and two
copies of the Fe protein dimer bound on either end. The metal ions can be easily displayed using a
spacefilling representation, which reveals the iron-sulfur cluster, the P-cluster, and the FeMo-cluster
arranged in a row. The ATP binding site is revealed in this structure by using an unusual analogue of
ATP: an ADP molecule with an aluminum fluoride ion. Two of these molecules bind at each end,
forming a stable but inactive complex with the Fe protein, essentially gluing the Fe protein to the
FeMo protein so its structure can be solved.
A list of all nitrogenase structures in the PDB as of February, 2002, is available at
For more information on nitrogenase, see
Bacteriorhodopsin: A Light-driven Pump
Bacteriorhodopsin is a compact molecular machine that pumps protons across a membrane powered by green
sunlight. It is built by halophilic (salt loving) bacteria, found in high-temperature brine pools. They use sunlight to
pump protons outwards across their cell membranes, making the inside 10,000-fold more alkaline than the outside.
These protons are then allowed to flow back inwards through another protein, ATP synthase, building much of the
ATP that powers the cell.
Bacteriorhodopsin, as shown in PDB entry
1fbb, is composed of three protein
chains. It is found embedded in
dense arrays in the membranes of the bacteria. At the heart of each protein chain is a molecule of retinal,
which is bound deep inside the protein and connected through a lysine amino acid. Retinal contains a
string of carbons that strongly absorb light. When a photon is absorbed, it causes a change in the
conformation of the molecule. In bacteriorhodopsin, this is a change from a straight form to a bent form.
This change in shape powers the pumping of protons.
The capturing of light is so useful that these salt-loving bacteria actually build four different types of rhodopsins.
Bacteriorhodopsin is used to generate energy. Halorhodopsin, which may be seen in PDB entry
1e12, is also
a pump that funnels chloride ions instead of protons. It is in charge of keeping the internal concentration of
chloride at high levels that match the salty conditions outside the cell. The other two rhodopsins are sensory
rhodopsins, such as the ones in PDB entries
1jgj. These rhodopsins
sense bluish light and send
signals to the cell to move, finding an area with more useful, greenish light. All four of these rhodopsins are
built along similar lines, with a retinal molecule securely held inside a compact container of protein.
We also build several forms of rhodopsin and use them in our eyes for seeing light. As in bacteriorhodopsin,
our rhodopsin also contains a molecule of retinal. Bacteriorhodopsin, as seen in PDB entry
1fbb, can be
compared with rhodopsin from cows, as seen in PDB entry
1f88. Retinal is made in our
bodies from retinol,
or vitamin A, which is essential in the diet, since we cannot synthesize vitamin A on our own. When it absorbs
a photon, the retinal in rhodopsin changes shape from bent to straight--just the opposite of retinal in
bacteriorhodopsin! This change of shape then pushes the surrounding protein into a slightly different shape,
which is sensed by proteins inside the cell. Then, the message is passed through a cascade of proteins,
each sending the message to the next, finally launching a nerve signal to the brain. The process is so
sensitive that the eye can sense as few as 5 photons.
Many structures of bacteriorhodopsin are available in the PDB, showing many of the steps in the process
of absorbing light and pumping protons. Two snapshots can be found in PDB entries
The structure found in PDB entry
1c3w is in the ground state,
before it has absorbed light. The retinal is in the straight trans form. The structure from PDB entry
1dze shows the molecule after
Notice that the retinal now has a bent cis shape. This new shape has changed the orientation of the nitrogen
at the end of the retinal. It has also shifted the position of several protein amino acids that are along the
pathway of proton transfer. In particular, notice the large shift of arginine 82 at the bottom. Researchers
are working to discover how these changes in shape power the transfer of a proton from the top to the bottom,
through the middle of bacteriorhodopsin and across the bacterial membrane.
A list of all bacteriorhodopsin structures in the PDB as of March, 2002 is available at
For suggestions for additional reading about bacteriorhodopsin, see
The RCSB PDB (citation) is managed by two members of the Research Collaboratory for Structural Bioinformatics:
RCSB PDB is a member of the
The RCSB PDB is funded by a grant (DBI-1338415) from the
National Science Foundation, the
National Institutes of Health, and the
US Department of Energy.