J. Phys. Chem.

B 2006, 110, 6987-6990

6987

Energy Transfer from Adenosine Triphosphate

John Ross*
Department of Chemistry, Stanford UniVersity, Stanford, California 94305

Downloaded by UNIV AUTONOMA DE BARCELONA on October 30, 2009 | http://pubs.acs.org

Publication Date (Web): March 16, 2006 | doi: 10.1021/jp0556862

ReceiVed: October 5, 2005; In Final Form: February 14, 2006

We suggest a direct molecular mechanism of energy transfer from adenosine triphosphate (ATP) in hydrolysis
and phosphorylation reactions, from chemical energy into mechanical energy. Upon hydrolysis of ATP, say
bound to a protein, the electrostatic energy of Coulombic repulsion of the ions adenosine diphosphate and
phosphate is available to assert a force on a neighboring molecular group in the protein and can do work on
that group, or as the ions recede from each without asserting such a force, they gain relative kinetic energy,
which, in the absence of dissipative collisions that turn this kinetic energy into heat, can be converted into
any other form of energy and work by an impulse, a collision with a neighboring group, without restrictions.
Either possibility can be used as a source of activation energy for reactions, as a source of energy to surmount
energy barriers in conformational changes, and as a source of work to be done, as in muscle. In some systems
where the Gibbs free energy change is fully utilized, all of this energy is turned into mechanical energy, and
we suggest a similar mechanism. From the literature we cite some experimental evidence and several quotations
indicative of the possibility of our suggestion.

I. Introduction
Adenosine triphosphate, ATP, is an universal energy carrier
in biological systems; it hydrolyzes and carries out many
phosphorylation reactions that proceed spontaneously. For the
reaction

ATP-4 + H2O f ADP-3 + HPO4-2 + H+

(1)

to the ester groups. All of these conditions are met by phosphoric

acid, and no other alternative is obvious. Furthermore, phosphoric acid can form monoesters of organic compounds that
can decompose by mechanisms other than nucleophilic attack,
a mechanism that allows them sufficient reactivity in intermediary metabolism... This remarkable combination of thermodynamic instability and kinetic stability was noted... by Lippmann.3
In this article we suggest a direct molecular mechanism of
energy release and tranfer in hydration and phosphorylation
reactions of ATP, from chemical energy into mechanical energy.
The emphasis is first on the Coulombic repulsion between ADP
and phosphate, and second, on the release of all the available
energy in ATP. We know of no other prior suggestion on this
specific topic.

the standard Gibbs free energy change is G ) -31 kJ/mol

at pH 7 in the presence of magnesium at pMg ) 3, and under
physiological conditions the Gibbs free energy change, G, is
about -50 kJ/mol.1
The free energy liberated in the hydrolysis of ATP is
harnessed to drive reactions that require an input of free energy,
such as muscle contraction. In turn, ATP is formed from ADP
and HPO4-2 when fuel molecules are oxidized in chemotrophs
or when light is trapped in phototrophs. This ATP-ADP cycle
is the fundamental mode of energy exchange in biological
systems.1 Reactions of this type1 are ubiquitious: for metabolism, for pumping ions against chemical and electrical gradients,
for generating conformational changes in proteins, etc.
The universality of the use of ATP is a marvel: what is its
basis; how and why was ATP (and similarly GTP) selected for
this role; how is energy stored in ATP; how is energy transferred
when ATP reacts?
Westheimer wrote an incisive article in 1978,2 addressing
some of these questions. We quote from his summary:
...the existence of a genetic material such as DNA requires a
compound for a connecting link that is at least divalent. In order
that the resulting material remain within a membrane, it should
always be charged, and therefore the linking unit should have
a third ionizable group. The linkage is conveniently made by
ester bonds, but, in order that the ester be hydrolytically stable,
that charge should be negative and should be physically close

The energy source in ATP has been described as due to a

high energy bond and is so presented in standard texts. The
source of the high energy has been assigned to Coulombic
repulsion of ADP and phosphate and to resonance effects.1
We gain some insight into the order of magnitude of the
Coulombic repulsion in two very simple, separate ways. First,
compare G of reaction 2, rewritten in greater detail, but with
Mg omitted with that of the hydrolysis of phosphoglycerate. In

* Author to whom correspondence should be addressed. E-mail:

john.ross@stanford.edu.

reaction 3 there is only hydrolysis of a phosphorus-oxygen

II. Proposal for Molecular Energy Release from ATP in

Hydrolysis and Phosphorylation Reactions

10.1021/jp0556862 CCC: $33.50 2006 American Chemical Society

Published on Web 03/16/2006

6988 J. Phys. Chem. B, Vol. 110, No. 13, 2006

Ross

TABLE 1: Estimates of K, the Relative Kinetic Energya

1.
2.
3.
4.

Downloaded by UNIV AUTONOMA DE BARCELONA on October 30, 2009 | http://pubs.acs.org

Publication Date (Web): March 16, 2006 | doi: 10.1021/jp0556862

zA

zB

K
(kJ/mol)

-3
-3
-1
-1

-1
-2
-1
-2

18
36
6
12

The symbols are defined in the text.

bond and no Coulombic repulsion of the post-hydrolysis

products. In reaction 2 there is also hydrolysis of a phosphorusoxygen bond followed by formation of relative kinetic energy
in the reaction products due to reduction of Coulombic repulsion.
As the two negative ions recede from each other to infinite
separation we may estimate the (maximum) relative kinetic
energy to be the difference in the standard Gibbs free energy
of reaction 2, -31 kJ/mol, and that of reaction 3, -10 kJ/mol,
which is about - 21 kJ/mol.1 A possible difference in resonance
energy in these two reactions is ignored in this estimate.
A second estimate can be made with the most simple
electrostatic calculation. The change in relative kinetic energy,
K, of two ions A and B, of like charge with magnitudes zA
and zB, with initial separation ri and final separation rj is

K )

( )

zAzBe2 1 1
40 ri rj

(4)

where 0 is the permittivity of free space, 8.85 10-12 F/m, 

is the relative permittivity of the medium, and e is the electronic
charge, 1.6 10-19 C. To calculate ranges of possible values
of K, we need to choose ranges of the final separation and
the permittivity of the medium in which the hydrolysis of ATP
takes place. We take the initial separation to be 0.3 nm, which
is approximately the P-P distance in ATP, and the final
separation to be 0.4 nm. The relative permittivity of water is
78, and we estimate that the relative permittivity of a hydrophobic environment, as in a protein fold, is 20.4 It could be
considerably lower, perhaps 2-4. In Table 1 we list several
possibilities, all for  ) 20 and the distances indicated: The
first line combines the charge of the proton, in reaction 2, with
the charge of the phosphate; the second line neglects the charge
of the proton. The next two lines consider the presence of
divalent Mg on ADP. Line 3 combines the charge of the proton
with that of the phosphate; line 4 neglects the charge of the
proton. To obtain the values of K for the relative permittivity
of 10 (80), multiply the values of K in the table by a factor of
2 (divide by 4); to obtain the values for the case of the final
separation distance at infinity, multiply the values in the table
by 4. For these ranges of conditions the entries in the table are
within an order of magnitude of the first estimate of 21 kJ/mol.
Thus the kinetic energy obtained from a reduction of the
Coulombic repulsion is in the range of two-thirds of the standard
Gibbs free energy change of hydrolysis and in the range of
somewhat less than half of the Gibbs free energy change under
physiological conditions.
The discussion presented here holds equally well for reactions
of ATP with pyrophosphate as one of the products.
An increase in ionic strength of the solution in which reaction2
occurs increases the shielding of the two negative ions and hence
reduces the Coulombic repulsion. (This shielding produces the
primary salt effect in the kinetics of ions.)
As the HPO4-2 ion moves away from the ADP ion due to
Coulombic repulsion, the relative kinetic energy of the two ions,
in the absence of other interactions, increases. If ADP is held

in place, for example, by binding to a segment of an enzyme,

then the HPO4-2 ion acquires all of this energy. If both ions
are free to move, then HPO4-2 acquires about 70% of the
relative kinetic energy.
The essence of the present proposal is:
Upon hydrolysis of ATP, the force due to the Coulombic
repulsion of the product ions can act on a neighboring molecular
group in the protein and can do mechanical work on it by
displacing it. If the ions recede from each other without asserting
such a force, then they gain kinetic energy, which can, prior to
dissipation into heat, be transferred in an impulse, a collision
with a neighboring group, into mechanical work. In either case
the mechanical energy generated can be wholly converted into
any other form of energy or work without any restrictions. If
ATP is tightly bound to an enzyme, then upon hydrolysis the
-phosphate may not be able to move away from ADP and the
Coulomb repulsion is retained as potential energy until the
binding is reduced.
In some systems the Gibbs free energy change available from
the reaction of ATP is fully utilized and all of this energy is
turned into mechanical energy. This energy conversion occurs
likely in the same way, due to the remaining electronic repulsion,
as for that part due to the Coulombic repulsion.
For full conversion from chemical to mechanical energy no
dissipation may occur. Dissipation may take place through
collisions in which mechanical translational energy is changed
to a statistical distribution, heat, or through various types of
inelastic collisions. Such events are precluded for very short
time scales. For a relative kinetic energy of the phosphate of
20 kJ/mol, for example, the phosphate ion moves about 0.1 nm
in 0.3 ps. If no other collisions occur in that short time interval,
as is likely, then the phosphate can transfer its kinetic energy
to a neighboring molecular group and can thus achieve, say, a
conformational change.
The mechanical energy being discussed, 20-50 kJ/mol, is
on the order of 10-20 times thermal energy, 0.25 kJ/mol, and
therefore thermal fluctuations can be expected to have a minor
effect on the present argument.
More sophisticated calculations are of course possible, yet
the simplest calculations yield interesting concordant results.
III. Utilization of Mechanical Energy Obtained from ATP
The mechanical energy, generated from chemical energy in
ATP, in the products to be formed in the hydrolysis and
phosphorylation reactions of ATP may serve different purposes.
First, the mechanical energy can produce work at the
molecular level. Human inventions of the conversion of chemical energy into work in a cyclic process require either a battery
connected to a motor, or require the production of heat and its
usage in thermal engines, at reduced efficiency.
Second, the mechanical kinetic energy can be directly used
to surmount energy barriers on the reaction coordinate. Compare
this efficient conversion of electrostatic energy into mechanical
energy, say, for surmounting a barrier or any other need for
energy, with providing a reaction complex, as in a unimolecular
reaction, with thermal energy (heat). This thermal energy is
generally distributed over all degrees of freedom statistically.
A fluctuation has to occur to place sufficient energy into a given
degree of freedom, say, one bond, for reaction to occur. It takes
time for such a fluctuation to take place, and hence thermal
excitation is less efficient (requires more energy and takes
longer) than activation by mechanical energy.
Third, the mechanical energy produced can be transferred to
bring about a conformational change in the protein that requires

Downloaded by UNIV AUTONOMA DE BARCELONA on October 30, 2009 | http://pubs.acs.org

Publication Date (Web): March 16, 2006 | doi: 10.1021/jp0556862

Energy Transfer from ATP

energy input simply by the force exerted by the Coulombic
repulsion between ADP and phosphate on neighboring groups
in the enzyme.
Fourth, the repulsion between the negative ions, both in
hydrolysis and in phosphorylation, enhances the rate of separation of the products and thus leads to rapid completion of the
reactions.
The suggestion made here for the conversion of chemical
into mechanical energy can encompass several different scenarios: (1) Suppose transfer of chemical energy occurs at a
given time, say, to effect a configurational change in the enzyme.
There may be subsequent steps of configurational changes
energized by the first one. (2) The transfer of chemical energy
may occur at different times. Suppose ATP is tightly bound to
an enzyme as in actomycin; hydrolysis may take place, but the
-phosphate cannot move away from ADP. In this case there
can be no reduction in the Coulombic repulsion and hence no
conversion of that potential energy into mechanical work. After
some time interval the binding may be reduced, some reduction
in the Coulomb repulsion may occur, and some mechanical work
may be done on neighboring groups. (3) Step 2 may occur more
than once, and thus energy from ATP may be distributed over
several steps of the enzyme cycle.
All these advantages lead to the universal use of ATP as an
energy source in plants and animals.
IV. Possible Experimental Evidence for the Present
Suggestion
In this section we seek experimental indications for the
utilization of the mechanical energy generated from the chemical
energy available from ATP as suggested here. There is no proof
of the suggestion, but the subsections list reasonable indications
of its possibility.
A. Efficiency of Utilization of Energy from ATP. In many
cases the energy available from ATP is not fully used; some
energy is dissipated. In other cases, however, the energy
available from ATP is fully utilized.5,6 In the latter case it is
clear that the mechanical energy generated from the repulsion
of the product ions is also fully utilized, and the free energy
change of the hydrolysis step is small. In refs 5 and 6 it was
shown that one ATP is utilized for each mechanical step in
myosin-V. The step size is 36 nm, and the maximum load is
2.5 pN. The product of these numbers translates to 54 kJ/mol,
well within experimental error of the full utilization of the Gibbs
free energy change of the hydrolysis of ATP under physiological
conditions. Another example of full utilization, near 100%
efficiency in ATP synthase, is given in ref 7.
The research described in refs 5 and 6 was done in vitro, in
water, which is likely a place for dissipation to occur due to
the collision of the phosphate ion with water molecules. Hence
it is worthwhile noting that no measurable dissipation of the
mechanical energy takes place in these examples.
B. Crystal Structure of Monomeric Actin in the ATP
State.8 The authors compare the crystal structures of the actin
monomer (G-actin) in the ATP and the ADP states... A
comparison of the structures in the two states reveals how the
release of the nucleotide -phosphate triggers a sequence of
events that propagate into a loop to helix transition in the DNase
I-binding loop in subdomain 2.
C. The Role of MeH73 in Actin Polymerization and ATP
Hydrolysis.9 ...the imidazole of (Me)H73 does not make direct
contact with the terminal phosphate, but it is in a position to
mediate conformational changes in the hairpin loops associated

J. Phys. Chem. B, Vol. 110, No. 13, 2006 6989

with nucleotide binding thereby governing hydrolysis and
polymerization.
D. Swing of the Lever Arm of a Myosin Motor at the
Isomerization and Phosphate Release Steps.10 The authors
use FRET to observe the working stroke, the tilt of a lever arm,
in the myosin motor. They show
that the hydroxyl terminal fluorophore swings at the isomerization step of the ATP hydrolysis cycle, and then swings back
at the subsequent step in which inorganic phosphate is released,
thereby mimicking the swing of the lever arm. The swing at
the phosphate release step may correspond to the working stroke,
and the swing at the isomerization to the recovery stroke... the
lever arm domain can generate sliding of myosin along actin
filaments in the appropriate direction for muscle action...
E. Analysis of the Conformational Change of Myosin
during ATP Hydrolysis Using Fluorescence Resonance
Energy Transfer.11 The phosphate release step shifts the
myosin from a weak to a strong binding conformation in the
interaction with actin and is the step believed to be associated
with the force generation during muscle contraction.
F. The Structural Basis of the Myosin ATPase Activity.12
Contrary to initial expectations, ATP hydrolysis in myosin is
not coincident with the force generating step... The energy
transduction step occurs during product release... Kinetic studies
indicate that the phophate ion is released prior to ADP during
the contractile cycle.
G. Photon Excitation of Myoglobin. Miller and co-workers
have published a series of articles13 on photon excitation of
myoglobin that leads to bond breaking. Thereupon they have
observed, by means of transient phase grating spectroscopy, that
the effect of this event is passed through a series of molecular
fragments. Collective modes of motion are excited within about
0.3 ps, which brings about a conformational change. The barrier
for this change is about 12 kJ/mol. In analogy, the imposition
of a force, or the impact (a collision) of a phosphate ion, with
high kinetic energy, unto a molecular fragment of a protein may
bring about a conformational change. As seen in Table 1 the
energy of 12 kJ/mol may be available from the conversion of
electrostatic repulsion to relative kinetic energy with a relative
motion of the phosphate ion from ADP of about 0.1 nm, or 1
, in a medium of relative permittivity of 10, in the presence
of Mg, line 3 in Table 1.
H. Examples Where the Suggestion Seems To Play No
Role. In an hypothesis for the use of ATP for pumping protons
across mitochondrial membranes, and in the reverse process of
ATP synthesis, Boyer14,15 proposed a series of conformational
changes in the F1F0-type ATP synthase to achieve these
functions. There seems to be no need for our suggestion.
However further studies are necessary to relate these conformational changes to the utilization of ATP energy, which
according to the experiments in ref 7, is near 100%. In these
further studies our suggestion may find an application.
V. Possible Ions for the Choice of a Source of Energy
What ions are available, and which are likely candidates, for
a physical source of energy based on establishing repulsion
between ions of like charge? The ions present in seawater must
surely be considered; a list of some such ions and their
concentrations are given in Table 2. What combinations of like
charges exist, either by themselves or in combination with other
molecular moieties, which have the property of an energy source
as suggested here? Among the positive ions there appears to
be none. Among the negative ions there are the phosphates

6990 J. Phys. Chem. B, Vol. 110, No. 13, 2006

Ross
E. Kool, V. Pande, J. Spudich, R. N. Zare, and Dr. Marcel Vlad.
This work was supported in part by the National Science
Foundation.

TABLE 2: Principle Constituents of Seawater

chloride
sodium
magnesium
sulfate
calcium
phosphate

0.54 mol/kg of seawater

0.46 mol/kg of seawater
0.053 mol/kg of seawater
0.028 mol/kg of seawater
0.010 mol/kg of seawater
0.1-0.3 mol/kg of seawater

References and Notes

PO4-3 , P2O7-3, and P3O10-4 , which can react analogously to

eq 1

P3O10-4 + H2O f P2O7-3 + HPO4-2 + H+

(5)

Downloaded by UNIV AUTONOMA DE BARCELONA on October 30, 2009 | http://pubs.acs.org

Publication Date (Web): March 16, 2006 | doi: 10.1021/jp0556862

This reaction occurs spontaneously at pH 7, but the rate is slow

in the absence of phosphatases.
There is another possibility, that of the sulfates

S2O7-2 + H2O f 2SO4-2 + 2H+

(6)

for which G ) -142. kJ/mol. However S2O7-2 is not stable

at neutral pH.
The conclusion reached here, on the basis of arguments not
considered by Westheimer,2 is the same as that given in ref 2,
see the quote cited in the Introduction: that phosphates appear
to be the only choice for the universal role of energy storage
and transfer into mechanical and electrical energy in living
systems at normal temperatures.
Acknowledgment. Helpful discussions are gratefully acknowledged with Professors S. Block, W. Huestis, C. Khosla,

(1) Stryer, L. Biochemistry, 4th ed.; W. H. Freeman and Co.: New

York, 1995.
(2) Westheimer, F. H. Science 1987, 235, 1173-1178.
(3) Lippmann, L. In Phosphorous Metabolism; McElroy, W. D., Glass,
H. B., Eds.; Johns Hopkins Press: Baltimore, MD, 1961; Vol. 1, p 521.
(4) Schutz, C. N.; Warshel, A. Proteins: Struct., Funct., Genet. 2001,
44, 400-427.
(5) Mehta, A. D.; Rock, R. S.; Rief, M.; Spudich, J. A.; Mooseker, M.
S.; Cheney, R. E. Nature 1999, 400, 590-593.
(6) Rief, M.; Rock, R. S.; Mehta, A. D.; Mooseker, M. S.; Cheney, R.
E.; Spudich, J. A. Proc. Natl. Acad. Sci. U.S.A. 2002, 97, 9482-9486.
(7) Kinosita, K. Jr.; Yasuda, R.; Noji, H. Essays Biochem. 2000, 35,
3-18.
(8) Graceffa, P.; Dominguez, R. J. Biol. Chem 2003, 278, 3417234180.
(9) Nyman, T.; Schuler, H.; Korenbaum, E.; Schutt, C. E.; Karlsson,
R.; Lindberg, U. J. Mol. Biol. 2002, 317, 577-589.
(10) Suzuki, Y.; Yasunaga, T.; Ohkura, R.; Wakabayashi, T.; Sutho, K.
Nature 1998, 396, 380-383.
(11) Mizukura, Y.; Maruta, S. J. Biochem. 2002, 132, 471-482.
(12) Rayment, I. J. Biol. Chem. 1996, 271, 15850-15853.
(13) Armstrong, M. R.; Ogilvie, J. P.; Cowan, Nagy, A. M.; Miller, R.
J. D. Proc. Natl. Acad. Sci. U.S.A. 2003, 100, 4990-4994.
(14) Boyer, P. D. Biochim. Biophys. Acta 1993, 1140, 215-250.
(15) Capaldi, R. A.; Aggeler, R. Trends Biochem. Sci. 2002, 27, 154160.