Hi! Kindly check the attached file. I also attached some of the journal references I used.

Biochimica et Biophysica Acta 1587 (2002) 107 – 117
www.bba-direct.com

Review

Development of transition state analogues of purine nucleoside
phosphorylase as anti-T-cell agents
Vern L. Schramm*
Department of Biochemistry, Albert Einstein College of Medicine, 1300 Morris Park Avenue, Forch. 308, Bronx, NY 10461, USA
Received 24 January 2002; accepted 24 January 2002

Abstract
Newborns with a genetic deficiency of purine nucleoside phosphorylase (PNP) are normal, but exhibit a specific T-cell immunodeficiency
during the first years of development. All other cell and organ systems remain functional. The biological significance of human PNP is
degradation of deoxyguanosine, and apoptosis of T-cells occurs as a consequence of the accumulation of deoxyguanosine in the circulation,
and dGTP in the cells. Control of T-cell proliferation is desirable in T-cell cancers, autoimmune diseases, and tissue transplant rejection. The
search for powerful inhibitors of PNP as anti-T-cell agents has culminated in the immucillins. These inhibitors have been developed from
knowledge of the transition state structure for the reactions catalyzed by PNP, and inhibit with picomolar dissociation constants. ImmucillinH (Imm-H) causes deoxyguanosine-dependent apoptosis of rapidly dividing human T-cells, but not other cell types. Human T-cell leukemia
cells, and stimulated normal T-cells are both highly sensitive to the combination of Imm-H to block PNP and deoxyguanosine.
Deoxyguanosine is the cytotoxin, and Imm-H alone has low toxicity. Single doses of Imm-H to mice cause accumulation of deoxyguanosine
in the blood, and its administration prolongs the life of immunodeficient mice in a human T-cell tissue xenograft model. Immucillins are
capable of providing complete control of in vivo PNP levels and hold promise for treatment of proliferative T-cell disorders. D 2002 Elsevier
Science B.V. All rights reserved.
Keywords: Immucillin; Purine nucleoside phosphorylase; T-cell leukemia; Autoimmunity; Transition state; Apoptosis; Deoxyguanosine toxicity

1. Introduction
Purine nucleoside phosphorylase (PNP) catalyzes the
reversible reactions [1]:
ðdeoxyÞguanosine
guanine
ðdeoxyÞinosine þ PO4 X hypoxanthine þ a  D

 ðdeoxyÞribose 1  PO4
Prepared by her earlier discovery of severe combined immunodeficiency in adenosine deaminase deficiency, Eloise

Abbreviations: PNP, purine nucleoside phosphorylase; Imm-H, Immucillin-H [(1S)-1-(9-deazahypoxanthin-9-yl)-1,4-dideoxy-1,4-imino-D-ribitol]; Imm-G, Immucillin-G [(1S)-1-(9-deazaguanin-9-yl)-1,4-dideoxy-1,4imino-D-ribitol]; dNTP, 2V-deoxynucleoside 5V-triphosphates; Ki*, the
equilibrium dissociation constant between an enzyme and inhibitor that
includes a slow onset step; IL-2, interleukin-2; SCID mice, a mouse strain
with genetically derived severe combined immunodeficiency; BCX1777uImmucillin-H
*
Tel.: +1-718-430-2813; fax: +1-718-430-8565.
E-mail address: vern@aecom.yu.edu (V.L. Schramm).

Giblett discovered that infants with a rare T-cell immunodeficiency lacked PNP [2]. Subsequent studies indicated
that T-cell deficiency resulted from altered pathways of
purine metabolism. Deoxyguanosine accumulates in the
blood as a result of PNP deficiency, and is transported
and phosphorylated by T-cell deoxynucleoside kinases to
form pathologically elevated levels of dGTP specifically in
these cells [3 –6]. Inhibition of PNP was soon identified as a
target for the regulation of undesirable T-cell proliferation, a
campaign was launched for the discovery of powerful
inhibitors [7]. Thirty-three patents for PNP inhibitors were
listed by 1998, but clinical trials with the best of these
inhibitors failed to show adequate inhibition to cause
regulation of activated T-cells [8]. Type IV autoimmune
disorders are a primary disease target for PNP inhibitors,
and are caused by inappropriate activation of T-cells by selfantigens [9]. These disorders include rheumatoid arthritis,
psoriasis, inflammatory bowel disorders and multiple sclerosis. In addition, T-cell leukemias and lymphomas would
be primary proliferative targets for PNP inhibitors.
Inhibitor design for patented PNP inhibitors have used
structure – activity relationships and structure-based design,

0925-4439/02/$ - see front matter D 2002 Elsevier Science B.V. All rights reserved.
PII: S 0 9 2 5 - 4 4 3 9 ( 0 2 ) 0 0 0 7 3 - X

108

V.L. Schramm / Biochimica et Biophysica Acta 1587 (2002) 107–117

in which the catalytic site of PNP containing substrate or
product analogues or weak inhibitors was sequentially filled
with newly designed analogues followed by subsequent
structural and kinetic characterization and refinement of
catalytic site contacts [10,11]. The most powerful inhibitors
obtained by these methods are in the low nanomolar range
for inhibitory dissociation constants [8].
During this period, methods were also being developed for
the experimental analysis of enzymatic transition states by
kinetic isotope effect analysis [12 – 15]. Analogues that
resemble enzymatic transition states capture the enzymatic
forces used for catalysis and convert catalytic energy into
binding energy, resulting in powerful inhibition [15 – 17]. The
inversion of configuration at the anomeric carbon of PNP
substrates suggested a nucleophilic displacement reaction,
but compounds synthesized to resemble such transition states
bound no better than substrates. The possibility of transition
state inhibitor design for PNP became a reality in 1995, when
the transition state structure was resolved from studies of the
chemical mechanism and kinetic isotope effects [18 – 20].
This review provides a summary of the biology, enzymology
and chemistry that led to the development of the immucillins,
PNP inhibitors of sufficient specificity and affinity to cause
deoxguanosine accumulation in mammals.

2. PNP in human nucleotide metabolism
The (deoxy)nucleoside substrates of PNP are normally
absent from the blood as a result of robust PNP catalytic
activity in the intestine, liver, erythrocytes, lymphocytes,
spleen and kidney [21,22]. The activity of PNP in the blood
alone causes injected nucleoside substrates to undergo phosphorolysis with a half-life of a few seconds. However, PNP
substrates accumulate in the blood and urine of PNP-deficient patients, replacing uric acid, which is dramatically
reduced in both blood and urine (summarized in Ref. [23]).
This experiment of nature confirms that the pathway of
purine nucleoside degradation requires PNP, and without it,
nucleosides accumulate, purine bases for salvage pathways
and degradation are depleted, and the rate of de novo purine
synthesis increases [22]. Overproduction of purines is not
accompanied by purine precipitation disorders in PNP deficiency since the (deoxy)nucleosides are more soluble than
uric acid. Blood levels of these metabolites are significant
and the primary purine excretion products are urinary nucleosides. In addition, (deoxy)nucleoside salvage increases
because of excess substrate availability. Dividing T-cells
express an active deoxycytidine kinase, whose normal role
is the salvage of deoxycytidine to form dCMP ! ! dCTP
for DNA synthesis in activated T-cells [24]. When deoxyguanosine accumulates beyond normal levels, deoxycytidine
kinase accepts deoxguanosine to form dGMP ! ! dGTP.
The allosteric inhibition site for dGTP on ribonucleotide
diphosphate reductate inhibits cellular formation of dCDP
and dUDP [25], thereby preventing DNA synthesis (Fig. 1).

Fig. 1. Pathways of deoxyguanosine (dGuo) metabolism in human T-cells.
The normal function of deoxycytidine (dCyd) kinase is salvage of dCyd
arising from apoptosis of other T-cells. It is regulated by dCMP product
inhibition. Excess dGuo is phosphorylated to dGMP by the same enzyme,
but dGMP is not a good product inhibitor. Increased dGTP allosterically
inhibits ribonucleotide reductase, preventing DNA synthesis and T-cell
division.

Factors that have been proposed to make the disorder specific
for T-cells include high capacity for deoxynucleoside transport, high expression levels of deoxycytidine kinase, and low
levels of phosphatases for dGMP. The normal response to
antigenic T-cell stimulation requires clonal expansion from a
few cells with the appropriate receptors to the T-cell mass
required for an activated T-cell response. This response
requires rapid T-cell proliferation involving relatively large
quantities of DNA synthesis. Activation of T-cells under
conditions of unbalanced dNTPs induces apoptosis, and
instead of T-cell proliferation, depletion of T-cells occurs
[26]. T-cell populations are exquisitely poised for apoptotic
responses, since f 99% of developing thymocytes do not
receive an antigenic stimulatory response and undergo apoptosis under conditions of normal development [27].
Knowledge of human purine metabolism is required to
understand PNP deficiency since the metabolism of deoxyguanosine differs between humans and mice. Mice made
genetically deficient in PNP do not undergo T-cell depletion, and some of the cellular changes observed in the mice
are attributed to the action of a mitochondrial deoxyguanosine kinase [28 –30]. A useful outcome of this finding is that
xenografts of human T-cells into mice allow analysis of the
effects of PNP inhibitors with only modest effects on the
host T-cell profile [31,32].

3. Catalytic properties of PNP
Mammalian PNPs catalyze the phosphorolysis of the
natural 6-oxypurine (deoxy)nucleosides and are inactive
against (deoxy)adenosine or the pyrimidine (deoxy)nucleosides [33]. The catalytic efficiency is high for the deoxy-

V.L. Schramm / Biochimica et Biophysica Acta 1587 (2002) 107–117

guanosine nucleoside that characterizes its biological function [1]. The homotrimer exhibits Michaelis– Menten initial
rate kinetics and has no known physiologic regulatory sites.
PNP is present at micromolar concentrations in blood cells
and is coupled to a substrate trapping phenomenon known
as catalytic commitment [14,19]. Every collision of a substrate (deoxy)nucleoside with the catalytic site leads to its
trapping and conversion to product. With the high concentrations of enzyme, catalytic commitment, and low Km
value, cells containing high concentrations of PNP are assured of the virtual absence of free deoxyguanosine. Mitochondrial deoxyguanosine metabolism is exempt from this
degree of deoxyguanosine removal since PNP is absent and
a deoxyguanosine kinase is present [34]. Repair and recycling of mitochondrial DNA generates deoxguanosine that
is proposed to remain in this compartment.
Under nonphysiological conditions, in the absence of
phosphate, PNP catalyzes a slow hydrolysis of inosine in
which the first catalytic site releases ribose, but binds tightly
(1 pM) to the hypoxanthine product [18]. The enzyme stalls
in a complex with hypoxanthine bound at one of the three
catalytic sites, demonstrating sequential catalytic site action
now made familiar from the action of F1F0 ATPase [35].
The relevance of sequential site catalytic action for cancer
therapy is the prediction that inhibition of any single subunit
of the PNP homotrimer will lead to full inactivation of
catalytic activity. Immucillin-H (Imm-H) has been shown to
act by this mechanism and one-third-the-sites inhibition is
discussed in more detail in Section 5.

109

PNP demonstrated no kinetic isotope effects with
inosine and phosphate as substrates, because of catalytic
commitment, and bound products reforming substrates on
the enzyme prior to product release [19]. Intrinsic isotope
effects were established using the hydrolytic reaction,
where catalysis is slow and no back-reaction occurs, and
also with arsenate (AsO4) as a phosphate analogue [20].
Arsenate reacts to form a-D-ribose 1-arsenate, an unstable
intermediate that hydrolyzes rapidly and prevents the
products of the reaction from reforming as substrates
prior to release from the enzyme. The intrinsic isotope
effects were used to establish the transition state features

4. Transition state analysis and PNP
Chemical transition states occur within the lifetime of a
single bond vibration, approximately 10  13 s, and direct
observations have only been successful in gas-phase studies
by laser spectroscopy [36]. Enzymatic transition states have
similar lifetimes and can be established from intrinsic
kinetic isotope effects by the experimental steps: (1) synthesis of substrates with specific atomic labels surrounding
the bonds being made and broken in the transition state; (2)
establish chemical or kinetic conditions where the catalytic
step (transition state formation) is the first irreversible step
in the catalytic cycle, or where intrinsic isotope effects can
be obtained; (3) measure a family of kinetic isotope effects
for the atoms whose bonding patterns are perturbed at the
transition state; and (4) use bond vibrational analysis
coupled to quantum chemistry predictions to match the
kinetic isotope effects to the transition state structure. Geometric and electrostatic properties of the transition state can
be used as an atomic blueprint to design stable analogues.
Chemical synthesis of the transition state analogues provides the desired inhibitors for enzymatic and biological
testing against the target. The experimental implementation
of these steps has been outlined in reviews [14,37 – 39], and
will be exemplified here with the example of PNP.

Fig. 2. Substrate (inosine), transition state and transition state inhibitor
(Imm-H) for PNP. The geometry of the transition state is shown looking at
C1V of the ribosyl ring. Features of the transition state are highlighted in
bold, and also appear in the inhibitor [51].

110

V.L. Schramm / Biochimica et Biophysica Acta 1587 (2002) 107–117

Fig. 3. Molecular electrostatic potential surfaces for inosine, the transition state of PNP and Imm-H. The geometry around the 5V-hydroxymethyl group and the
ribosidic-base torsion angles were fixed at the values established from the X-ray crystallographic studies. In the transition state, N7 was protonated and the
N9 – C1Vbond was fixed at 1.8 Å. The electron distribution was calculated using Gaussian 98 with the STO-3G basis set.

of 0.38 Pauling bond order to the hypoxanthine leaving
group, protonation or hydrogen-bond stabilization to N7
of the purine, van der Waals contact to the attacking
phosphate nucleophile, and conversion of the ribosyl
group to a partially charged ribooxacarbenium ion [20].
The results established that the chemical property of the
transition state is dissociative, rather than the associative
property expected for a symmetric nucleophilic displacement (Fig. 2). The results provided the first sufficiently
complete structure of the transition state to permit the
design of transition state inhibitors. These features were
also consistent with the transition states established for
other purine N-ribohydrolases [40,41], and with an earlier
isotope effect measured for the PNP from Escherichia
coli [42].

5. Imm-H design and synthesis
The features of the transition state structure for PNP were
used to design a chemically stable isologue (same molecular
shape and volume) to act as transition state analogue inhibitor
(Fig. 2). Inosine was the substrate for transition state analysis,
and the transition state inhibitor was designed to mimic this
transition state. The ribosyl group at the transition state is a
partially positively charged ribooxacarbenium ion, and was
mimicked in Imm-H with an iminoribitol structure. In Imm-

H, the imino group1 is protonated with a pKa of 6.5, providing
a partial positive charge at physiological pH values. The bond
to the leaving group purine is more than 60% dissociated at
the transition state, causing the pKa at N7 of the purine to
increase. Immucillins were designed with a carbon – carbon
ribosidic link to provide chemical stability relative to the
carbon – nitrogen link in normal substrates. This carbon –
ribosidic bond also changes the bond conjugation pattern of
the inhibitor, increasing the pKa of N7 from f 2 in inosine to
f 9 in Imm-H. The pKa of this group at the transition state
has not been measured, but is elevated toward 7 or above as
the ribosidic bond is broken. Nitrogen-7 is proposed to be a
site of protonation or hydrogen bonding by the enzyme to
form the transition state, since the bonding electrons must be
accommodated in the leaving group and protonation at N7 or
a favorable H-bond at this site assists in electron departure
from the N-ribosidic bond. The elevated pKa at this site
constitutes one of the transition state features. The ability of
N7-methyl substituted inosine and guanosine molecules to
act as substrates of PNP [43] indicates that the Asn243 that Hbonds to this site is sufficiently flexible to accommodate an
N7-methyl substituent. The methyl group provides a similar

1

The IUPAC nomenclature for the ring nitrogen in sugar analogues
accepts amino or imino. The original description of the deoxyiminoribitols
[64] used the imino nomenclature, which we maintain here.

V.L. Schramm / Biochimica et Biophysica Acta 1587 (2002) 107–117

111

Fig. 4. Chemical synthesis of Imm-H by the linear and convergent pathways. These methods have been published and details of synthesis are provided in Refs.
[48,50].

electronic effect in assisting the purine to accept bonding
electrons.
The atomic replacements between inosine and Imm-H
make an insignificant change in atomic size, but a dramatic
change in the molecular electrostatic potential surface2
(Fig. 3) [44]. Analysis of the molecular electrostatic
potential surface similarity between transition state and
transition state inhibitors for several enzymes have established that this and the atomic size correlate with the
affinity between enzyme and inhibitor [45,46]. A departure
from other inhibitor design programs was to eliminate
features of the phosphate anion. Transition state analysis
revealed that phosphate is not bonded at the transition
state, but is in van der Waals contact, with less than 2%
covalent bond order [20]. This finding established that the
nucleoside contacts dominate transition state interactions,
and that the phosphate binding site will fill with inorganic
phosphate, an abundant component of cells.

Chemical synthesis of Imm-H was first accomplished
in a linear sequence of over 20 steps, beginning with
D-gulonolactone, and based on the precedents of aryl and
alkyl 1-substituted iminoribitols [47,48] and for synthesis of
9-deazapurine nucleosides [49] (Fig. 4). This synthetic
procedure builds the protected iminoribitol from D-gulonolactone, followed by the stepwise addition of the components of the 9-deazahypoxanthine and deprotection. The
chemical stability of the compound is revealed in the final
deprotection step, reflux in concentrated HCl. Although this
route provided the first access to Imm-H and Imm-G (the
9-deazaguanine analogue), it was not suitable for largescale synthesis. An improved synthetic route produced
activated and protected iminoribitol and 9-deazahypoxanthine in separate procedures, followed by reaction of the
two halves of the molecule [50] (Fig. 4).

6. Inhibition of PNPs by Imm-H
2

The molecular electrostatic potential surface is the force observed by
a point charge at every location on the van der Waals surface of the
molecule [44].

Imm-H and Imm-G inhibited both bovine and human
PNP, and exhibited the characteristics of slow-onset tight-

112

V.L. Schramm / Biochimica et Biophysica Acta 1587 (2002) 107–117

binding inhibitors [51]. Slow-onset inhibitors bind as reversible competitive ligands, followed by an isomerization of
the enzyme that causes increased inhibitor binding affinity
[52]. The initial binding phase for the bovine enzyme
exhibits a Ki of 41 nM, followed by a time-dependent, slow
onset (k5 = 0.06 s  1) increase of affinity by a factor of 2000,
to give a Ki* of 23 pM (Fig. 5). For the human enzyme, the
initial binding phase is rapid, and the Ki* value is 72 pM

(Table 1). The affinity of Imm-H is approximately one
million times that for the inosine substrate. The size and
molecular electrostatic potential properties of the inhibitor
are similar to the transition state [51]. Binding affinity is
dictated by on and off rates, in the case of tight-binding
inhibitors, that for k5 and k6, the conformational changes
that provide entry and departure from the tightly bound
transition state analogue complex (Fig. 5). Release of Imm-

Fig. 5. The kinetics of PNP inhibition by Imm-G and Imm-H. E, A, P and I represent PNP, inosine, products and Imm-H or Imm-G, respectively. The rate
constants k5 and k6 are the rate constants for formation of E * I and its conversion to EI. EI is the rapidly reversible, weakly bound inhibitor and E * I
represents a conformational change that accompanies tight-binding of the inhibitor. The center panel demonstrates the slow-onset of inhibition, and the bottom
panel demonstrates the post-slow onset rate (vs) as a function of inhibitor concentration using inosine at a concentration of 200 times its Km value. The reactions
were coupled to xanthine oxidase to yield uric acid as product. Adapted from reference [51].

V.L. Schramm / Biochimica et Biophysica Acta 1587 (2002) 107–117
Table 1
Inhibition of PNPs by Imm-H
PNP source
Human
Bovine
Mycobacterium
tuberculosis
Plasmodium
falciparum
a
b

Ki* (pM)

Km (AM)

Km /Ki

73
23
28

a

40
17a
26b

550,000
740,000
930,000

600

5a

8000

The Km values for inosine.
The Km values for 6-thio-7-methylguanosine.

H from the bovine enzyme has a t1/2 of 5 h, compared to that
from the human enzyme of 8 min. Inhibition was correlated
with moles of Imm-H bound per trimeric enzyme, and the
results established that binding of one Imm-H molecule per
trimer, with the affinity described above, is responsible for
the inhibition [51]. These inhibitory and binding studies
established that Imm-H was the most powerful PNP inhibitor yet described. It acts by complete inhibition of PNP
catalytic activity by filling only one of the three catalytic
sites of the trimer of PNP [51]. This result was unexpected
based on earlier kinetic studies and the previous X-ray
crystal structures of PNP, all of which indicated equivalent,
noninteracting sites in a symmetric trimer with identical
subunit occupancy with substrate and inhibitor analogues
[53]. Binding studies of Imm-H to bovine PNP indicated
that filling of the first subunit is responsible for the inhibition, and occurs with the highest affinity, followed by
saturation of the second and third sites with affinity orders
of magnitude lower than the first subunit [54]. However, at
large Imm-H excess, all three sites can be occupied. The
impact of the partial-sites inhibition for cancer therapy is
that physiological inhibition of PNP activity can be achieved
with only one molecule of inhibitor bound for every three
PNP subunits.
Although other applications of PNP inhibitors are
beyond the scope of this review, Imm-H is also a powerful
inhibitor of PNP from M. tuberculosis and P. falciparum
[55,56] (Table 1). The crystal structure of M. tuberculosis
has been solved with Imm-H and phosphate binding,
revealing that this enzyme is similar to the mammalism
enzyme [57]. In contrast, the PNP from P. falciparum is
more closely related to the hexameric E. coli PNP, and has a
lower affinity for Imm-H. Despite the lowered affinity, the
addition of Imm-H to cultures of human erythrocytes
infected with P. falciparum causes purine-less death of the
parasites under physiological culture conditions [58].

113

at all three sites, and permitted the collection of the highest
resolution data (1.5 Å) obtained for a complex of the bovine
enzyme. The structure was compared to structures of bovine
PNP solved earlier [53], to reveal the structural changes as
the enzyme progresses from the Michaelis complex to the
transition state and product complexes. Imm-H is situated in
the catalytic sites similar to the position of inosine, except
that contacts to the enzyme are closer to most parts of the
complex (Fig. 6). Comparison of substrate, Imm-H and
product complexes to the actual transition state structure
docked into the catalytic site permitted analysis of the
atomic motion that occurs as substrates are converted to
products, and the analysis of the similarity between Imm-H
and the docked transition state.
One surprising result is that the reaction occurs by
unprecedented atomic motion in the reaction coordinate.
These steps include; (a) enzymatic immobilization of the
purine ring and phosphate, (b) generation of the ribooxacarbenium ion transition state with the participation of
neighboring group oxygens from phosphate and the 5Vhydroxyl to stabilize the ribooxacarbenium ion and (c)
migration of the C1V anomeric carbon over the relatively
long distance of 1.7 Å between the enzymatically stabilized
nucleophiles, while the 5V-region of the ribosyl group
remains immobile. This mechanism is a departure from
solution chemistry, and also occurs in other ribosyltransferases. This mechanism has been called a nucleophilic
displacement by electrophilic migration, and reflects the
ability of enzymes to accomplish atomic motion of reactive
groups in the protected environment of the catalytic site,
(exemplified here by the ribooxacarbenium ion) [59,60]. A
second important lesson of the study was that the transition
state analogue and the actual transition state occupy the
same positions, except for the 0.4 Å difference in structures
resulting from the C– C covalent bond in the transition state
analogue [59]. A third surprise from the structure with ImmH involves the nature of the enzymatic activation of the
leaving group purine and the formation of the ribooxacarbenium ion. Enzyme-immobilized water oxygens form a
proton transfer bridge to solvent and provide the protons
required for a H-bond and a protonation of the leaving
group hypoxanthine. Activation of the leaving group occurs
by; (a) a hydrogen bond to N7 shared to the side-chain
carbonyl oxygen of Asn243, where the H-bonded proton is
provided by the water bridge, and (b) protonation of O6 by
an immobilized water that is sandwiched between O6 and
Glu201. Near-optimal contacts at every H-bond donor/
acceptor pair on Imm-H and phosphate at the catalytic site
indicated that improvement on the binding affinity by
changing the structure of the immucillins might be difficult.

7. Structure of PNPImm-HPO4
Crystals of bovine PNP with Imm-H and phosphate at the
catalytic sites were grown in the presence of large excesses
of phosphate and of Imm-H to saturate the catalytic sites
[59]. The crystals, demonstrated uniform catalytic site filling

8. Effect of Imm-H on transformed and activated T-cells
Human T-cell immunodeficiency from PNP loss is completely dependent on the presence of deoxyguanosine and its

114

V.L. Schramm / Biochimica et Biophysica Acta 1587 (2002) 107–117

conversion to dGTP [31,32]. Cultures of human T-cell
leukemia lines were tested for growth in the presence of
Imm-H with and without deoxyguanosine (Fig. 7). The
selective inhibition of T-cells, only in the presence of deoxyguanosine, established that Imm-H is biologically available,
and is dependent on deoxyguanosine. Analysis of treated
cells established that Imm-H with deoxyguanosine caused the
accumulation of dGTP [31]. Under conditions of 20 AM
deoxyguanosine in the medium, the IC50 for Imm-H was 0.4
to 5 nM for reduction of thymidine incorporation in cells
cultured for 3 days (Fig. 7). The toxicity of Imm-H and
deoxyguanosine is rescued by deoxycytidine, a metabolite

known to prevent dGTP accumulation by substrate competition for deoxycytidine kinase and by dCMP product inhibition of the enzyme [61] (Fig. 7C). Seventeen non-T-cell lines
were unaffected by Imm-H at 10 AM, several thousand-fold
above the effective dose for human leukemic T-cells [31]. Tcells isolated from normal human volunteers are not affected
by Imm-H unless they are stimulated to divide, and show
increasing sensitivity with increased stimulation [31,32] (Fig.
8). Peripheral human T-cells stimulated to rapid division by
excess interleukin-2 (IL-2) and mononuclear cells were
inhibited by Imm-H and deoxguanosine with an IC50 value
of 5 nM. The results demonstrate that inhibition of PNP by

Fig. 6. The structures of bovine PNP with substrate analogues (inosine + SO4), transition-state complex (Imm-H + PO4) and products (hypoxanthine + ribose 1PO4) bound at the catalytic sites. Substrate and product complexes are from Ref. [53], and this figure is reproduced from Ref. [59]. Hydrogen bond distances
are shown in angstroms. Red indicates bonds that shorten significantly and blue indicates bonds that lengthen significantly in the conversion of (a) to (b), and of
(b) to (c).

V.L. Schramm / Biochimica et Biophysica Acta 1587 (2002) 107–117

115

Fig. 8. The effect of Imm-H on [3H]thymidine incorporation by human Tlymphocytes. Reproduced from Ref. [31]. In the top panel, peripheral Tcells from four donors were incubated for 3 days with no stimulation (solid
squares) or with physiological levels of IL-2. In the bottom panel, cells
were treated with excess IL-2 and mitomycin-treated mononuclear cells,
followed by 6-day incubation. Error bars represent the results averaged for
the T-cells of four donors.

blood. Both humans and mice have high concentrations of
PNP in erythrocytes; therefore, deoxyguanosine is not present at detectable levels in the blood. A single oral dose of
10 mg/kg Imm-H to mice increased the plasma concenFig. 7. The inhibition of human T-cell leukemia lines by Imm-H and
deoxyguanosine. MOLT-4 and CEM are human T-cell leukemia cell lines,
while Geo is a human colon carcinoma cell line and BL-2 is a human B-cell
leukemia cell line. Panel A demonstrates the response of cell lines to 20 AM
deoxyguanosine and the indicated concentrations of Imm-H after 3 days of
culture and analysis of cell activity by the viability dye WST-1. Panel B is
the same experiment, but with cell viability tested by tritium – thymidine
incorporation. Panel C demonstrates that Imm-H alone has no effects on
CEM cell viability, and that deoxycytidine (dCR) rescues against the effects
of Imm-H and deoxyguanosine by the mechanism shown in Fig. 1. This
figure is reproduced from Ref. [31].

Imm-H is sufficient to cause dGTP accumulation and to
induce apoptosis specifically in rapidly dividing T-cells.

9. Imm-H in mice and effects on T-cell xenografts
Bioavailability of Imm-H was tested in mice by a comparison of oral and injected doses, and established that oral
availability is high, at 63% of the injected dose [32]. A key
test of the biological effectiveness of the inhibitor is the
ability to increase the concentration of deoxyguanosine in

Fig. 9. The effect of Imm-H on the human lymphocyte – mouse xenograft
tissue rejection model. The results are reproduced from Ref. [32]. SCID
(hu-PBL) mice were treated with 20 mg/kg Imm-H for 5 days, and the natural
killer lymphocytes depleted with anti-ASGMI antibodies. Mice were irradiated and human buffy coat cells were injected. The group with Imm-H treatment (triangles) survived approximately twice as long as the control group.

116

V.L. Schramm / Biochimica et Biophysica Acta 1587 (2002) 107–117

tration of deoxyguanosine to a peak concentration of 5 AM
over a period of 3 h, establishing in vivo efficacy for
whole-body PNP inhibition in the mouse [32]. Elevated
deoxyguanosine also occurs in the human genetic deficiency of PNP, and plasma levels are reported to be in the
range of 3 to 17 AM [23]. Thus, a single oral dose of ImmH to mice can achieve a deoxyguanosine concentration that
is adequate to cause T-cell deficiency in humans. These
experiments were a prelude to the use of a mouse model of
human immune transplantation rejection that has been
developed and extended to SCID mice [62,63]. In this
protocol, human peripheral blood lymphocytes (PBLs) are
engrafted into SCID mice, and are stimulated to divide by
the host antigens. Expansion of the human PBL cells
causes the SCID mice to die from xenogenic graft vs. host
disease, approximately 4 weeks after engraftment. Death is
caused by infiltration of human lymphocytes into the
tissues of the host mouse. The effects of Imm-H were
tested by pretreating mice with Imm-H to elevate the deoxyguanosine levels, followed by engraftment and continued treatment with Imm-H to determine effects on life span.
Control animals were treated in the same way, except that
no Imm-H was given. The results demonstrated a two-fold
increase in life span compared to control mice (Fig. 9) [32].
These results establish the ability of Imm-H to influence
human T-cell proliferation in the mouse model of host vs.
graft disease, and therefore the ability to obtain sufficient
inhibition of whole body PNP to sustain elevated deoxguanosine levels.

10. Prospectus for applications of Imm-H
Based on the enzymology, cell culture and animal model
work published to date, the prospects for Imm-H to function
as an in vivo anti-T-cell agent are the most encouraging yet
obtained in the 26 years since Eloise Giblett discovered the
genetic deficiency of PNP [2]. Recent studies on the mouse
model of Imm-H immunosuppression of human T-cells
concluded with the statement that this inhibitor is the first
known example of a PNP inhibitor that elevates deoxyguanosine in mice similar to the levels observed in PNPdeficient patients [32]. Phase I/II clinical trials of Imm-H
against human T-cell leukemia have been initiated under the
trade name of BCX-1777.

Acknowledgements
Research in this laboratory is supported by research
grants from the National Institutes of Health. Greg Kicska
and Robert Miles have been instrumental in the research on
PNP and Imm-H in this laboratory. Alexander Fedorov,
Wuxian Shi and Steven Almo are responsible for the X-ray
crystallography. Richard Furneaux and Peter Tyler of
Industrial Research, New Zealand, pioneered the synthesis

of Imm-H, and Shanta Bantia of BioCryst has had a central
role in cellular and animal studies. The author thanks Brett
Lewis for help in preparing figures.

References
[1] J.D. Stoeckler, C. Cambor, R.E. Parks Jr., Biochemistry 19 (1980)
102 – 107.
[2] E.R. Giblett, A.J. Ammann, D.W. Wara, R. Sandman, L.K. Diamond,
Lancet 1 (1975) 1010 – 1013.
[3] T.A. Krenitsky, J.V. Tuttle, G.W. Koszalka, I.S. Chen, L.M. Beachman,
J.L. Rideout, G.B. Elion, J. Biol. Chem. 251 (1976) 4055 – 4061.
[4] D.A. Carson, J. Kaye, J.E. Seegmiller, Proc. Natl. Acad. Sci. U. S. A.
74 (1977) 5677 – 5681.
[5] B.S. Mitchell, E. Mejias, P.E. Daddona, W.N. Kelly, Proc. Natl. Acad.
Sci. U. S. A. 75 (1978) 5011 – 5014.
[6] B. Ullman, L.J. Gudas, S.M. Clift, D.W. Martin Jr., Proc. Natl. Acad.
Sci. U. S. A. (1979) 1074 – 1078.
[7] S.E. Ealick, Y.D. Babu, C.E. Bugg, M.D. Erion, W.C. Guida, J.A.
Montgomery , J.A. Secrist III, Proc. Natl. Acad. Sci. U. S. A. 88 (1991)
11540 – 11544.
[8] P.E. Morris, J.A. Montgomery, Expert Opin. Ther. Pat. 8 (1998) 283 –
299.
[9] C.A. Janeway Jr., P. Travers, Immunobiology: The Immune System in
Health and Disease, 3rd edn., Current Biology/Garland Publishing,
London and New York, 1997.
[10] M.D. Erion, S. Niwas, J.D. Rose, S. Ananthan, M. Allen, J.A. Secrist
III, Y.S. Babu, et al., J. Med. Chem. 36 (1993) 3771 – 3783.
[11] W.C. Guida, R.D. Elliott, H.J. Thomas, J.A. Secrist III, Y.S. Babu,
C.E. Bugg, M.D. Erion, S.E. Ealick, J.A. Montgomery, J. Med. Chem.
37 (1994) 1109 – 1114.
[12] J. Rodgers, D.A. Femec, R.J. Schowen, J. Am. Chem. Soc. 104 (1982)
3263 – 3268.
[13] W.W. Cleland, Methods Enzymol. 87 (1982) 625 – 641.
[14] D.B. Northrop, Biochemistry 14 (1975) 2644 – 2651.
[15] V.L. Schramm, Ann. Rev. Biochem. 67 (1998) 693 – 720.
[16] W.P. Jencks, Adv. Enzymol. 43 (1975) 219 – 410.
[17] R. Wolfenden, Nature 223 (1969) 704 – 705.
[18] P.C. Kline, V.L. Schramm, Biochemistry 31 (1992) 5964 – 5973.
[19] P.C. Kline, V.L. Schramm, Biochemistry 32 (1993) 13212 – 13219.
[20] P.C. Kline, V.L. Schramm, Biochemistry 34 (1995) 1153 – 1162.
[21] H.A. Simmonds, L.D. Fairbanks, G.S. Morris, G. Morgan, A.R.
Watson, P. Timms, B. Singh, Arch. Dis. Child. 62 (1987) 385 – 391.
[22] A. Cohen, D. Doyle, D.W. Martin Jr., A.J. Ammann, N. Engl. J. Med.
295 (1976) 1449 – 1454.
[23] M.S. Hershfield, B.S. Mitchell, in: C.R. Scriber, A.L. Beaudet, W.S.
Sly, D. Valle (Eds.), The Metabolic Basis of Inherited Disease, 8th
edn., McGraw-Hill, NY, 2001, pp. 2585 – 2624, Chapter 109.
[24] N.S. Datta, D.S. Shewach, B.S. Mitchell, I.H. Fox, J. Biol. Chem. 264
(1989) 9359 – 9364.
[25] L. Thelander, P. Reichard, Ann. Rev. Biochem. 48 (1979) 133 – 158.
[26] M.L. Agarwal, A. Agarwal, W.R. Taylor, O. Chernova, Y. Sharma,
G.R. Stark, Proc. Natl. Acad. Sci. U. S. A. 95 (1998) 14775 – 14780.
[27] C.D. Surh, J. Sprent, Nature 372 (1994) 100 – 103.
[28] E.R. Mably, E. Fung, F.F. Snyder, Genome 32 (1989) 1026 – 1032.
[29] F.F. Synder, J.P. Jenuth, E.R. Mably, R.K. Mangat, Proc. Natl. Acad.
Sci. U. S. A. 94 (1997) 2522 – 2527.
[30] E. Arpaia, P. Benveniste, A. Di Cristofano, Y. Gu, I. Dalal, S. Kelly,
M. Hershfield, et al., J. Exp. Med. 191 (2000) 2197 – 2208.
[31] G.A. Kicska, K. Long, H. Hörig, C. Fairchild, P.C. Tyler, R.H. Furneaux, V.L. Schramm, H.L. Kaufman, Proc. Natl. Acad. Sci. U. S. A.
98 (2001) 4593 – 4598.
[32] S. Bantia, P.J. Miller, C.D. Parker, S.L. Ananth, L.L. Horn, J.M.
Kilpatrick, P.E. Morris, T.L. Hutchison, J.A. Montgomery, J.S.
Sanduh, Int. Immunopharmacol. 1 (2001) 1199 – 1210.

V.L. Schramm / Biochimica et Biophysica Acta 1587 (2002) 107–117
[33] B.K. Kim, S. Cha, R.E. Parks Jr., J. Biol. Chem. 243 (1968) 1763 –
1770.
[34] E. Arpaia, P. Benveniste, A. Di Cristofano, Y. Gu, I. Dalal, S. Kelly,
M. Hershfield, P.P. Pandolfi, C.M. Roifman, A. Cohen, J. Exp. Med.
191 (2000) 2197 – 2207.
[35] P.D. Boyer, Ann. Rev. Biochem. 66 (1997) 717 – 749.
[36] D. Zhong, S. Ahmad, P.Y. Cheng, A.H. Zewail, J. Am. Chem. Soc.
119 (1997) 2305 – 2306.
[37] V.L. Schramm, Methods Enzymol. 308 (1999) 301 – 355.
[38] P.J. Berti, Methods Enzymol. 308 (1999) 355 – 397.
[39] B.B. Braunheim, S.D. Schwartz, Methods Enzymol. 308 (1999) 398 –
426.
[40] B.A. Horenstein, D.W. Parkin, B. Estupinan, V.L. Schramm, Biochemistry 30 (1991) 10788 – 10795.
[41] F. Mentch, D.W. Parkin, V.L. Schramm, Biochemistry 26 (1987)
921 – 930.
[42] R.L. Stein, E.H. Cordes, J. Biol. Chem. 256 (1981) 767 – 772.
[43] A. Bzowaka, E. Kulikowska, E. Darzynkiewicz, D. Shugar, J. Biol.
Chem. 263 (1988) 9212 – 9217.
[44] B.A. Horenstein, V.L. Schramm, Biochemistry 32 (1993) 9917 – 9925.
[45] C.K. Bagdassarian, V.L. Schramm, S.D. Schwartz, J. Am. Chem. Soc.
118 (1996) 8825 – 8836.
[46] B.B. Braunheim, R.W. Miles, V.L. Schramm, S.D. Schwartz, Biochemistry 38 (1999) 16076 – 16083.
[47] B.A. Horenstein, R.F. Zabinski, V.L. Schramm, Tetrahedron Lett. 34
(1993) 7213 – 7216.
[48] G.B. Evans, R.H. Furneaux, G.J. Gainsford, V.L. Schramm, P.C. Tyler, Tetrahedron 56 (2000) 3053 – 3062.
[49] M.I. Lin, W.Y. Ren, B.A. Otter, R.S. Kline, J. Org. Chem. 48 (1983)
780 – 788.

117

[50] G.B. Evans, R.H. Furneaux, T.L. Hutchison, H.S. Kezar, P.E. Morris,
V.L. Schramm, P.C. Tyler, J. Org. Chem. 66 (2001) 5723 – 5730.
[51] R.W. Miles, P.C. Tyler, R.H. Furneaux, C.K. Bagdassarian, V.L.
Schramm, Biochemistry 37 (1998) 8615 – 8621.
[52] J.F. Morrison, C.T. Walsh, Adv. Enzymol. Relat. Areas Mol. Biol. 61
(1988) 201 – 301.
[53] C. Mao, W.J. Cook, M. Zhou, A.A. Fedorov, S.C. Almo, S.E. Ealick,
Biochemistry 37 (1998) 7135 – 7146.
[54] F. Wang, R.W. Miles, G.A. Kicska, E. Nieves, V.L. Schramm, R.H.
Angeletti, Protein Sci. 9 (2000) 1660 – 1668.
[55] L.A. Basso, D.S. Santos, W. Shi, R.H. Furneaux, P.C. Tyler, V.L.
Schramm, J.S. Blanchard, Biochemistry 40 (2001) 8196 – 8203.
[56] G.A. Kicska, P.C. Tyler, G.B. Evans, R.H. Furneaux, K. Kim, V.L.
Schramm, J. Biol. Chem. 277 (2002) 3219 – 3225.
[57] W. Shi, L.A. Basso, D.S. Santos, P.C. Tyler, R.H. Furneaux, J.S.
Blanchard, S.C. Almo, V.L. Schramm, Biochemistry 40 (2001)
8204 – 8215.
[58] G.A. Kicska, P.C. Tyler, G.B. Evans, R.H. Furneaux, V.L. Schramm,
K. Kim, J. Biol. Chem. 277 (2002) 3226 – 3231.
[59] A. Fedorov, W. Shi, G.A. Kicska, E. Fedorov, P.C. Tyler, R.H.
Furneaux, J.C. Hanson, G.J. Gainsford, J.Z. Larese, V.L. Schramm,
S.C. Almo, Biochemistry 40 (2001) 853 – 860.
[60] V.L. Schramm, W. Shi, Curr. Opin. Struct. Biol. 11 (2001) 657 – 665.
[61] W.R. Osborne, C.R. Scott, Biochem. J. 214 (1983) 711 – 718.
[62] D.E. Mosier, R.J. Guilizia, S.M. Baird, D.B. Wilson, Nature 335
(1988) 256 – 259.
[63] J.S. Sandhu, R. Gorczynski, B. Shpitz, S. Gallinger, H. Nguyen, N.A.
Hozumi, Transplantation 60 (1995) 179 – 184.
[64] G.W.J. Fleet, J.C. Son, Tetrahedron 44 (1988) 2637 – 2647.


Review
Combating Resistance: Infectious Diseases
For reprint orders, please contact reprints@future-science.com

Transition-state inhibitors of purine salvage
and other prospective enzyme targets in
malaria
Malaria is a leading cause of human death within the tropics. The gradual generation of drug resistance imposes an
urgent need for the development of new and selective antimalarial agents. Kinetic isotope effects coupled to
computational chemistry have provided the relevant details on geometry and charge of enzymatic transition states
to facilitate the design of transition-state analogs. These features have been reproduced into chemically stable
mimics through synthetic chemistry, generating inhibitors with dissociation constants in the pico- to femto-molar
range. Transition-state analogs are expected to contribute to the control of malaria.

Malaria is the most deadly of the human protozoan infections. This tropical disease can be
caused by five different species of the Plasmodium
genus: Plasmodium falciparum, Plasmodium
vivax, Plasmodium ovale, Plasmodium knowlesi
and Plasmodium malariae [1]. Its name arose
from the Italian term for ‘bad air’ (mala aria),
used to describe fevers associated with marsh
air. The disease is endemic to warm and humid
tropical areas worldwide, where mosquito larva
can grow and multiply appropriately. Individuals infected with these species develop a common set of nonspecific clinical symptoms, which
include multiple shaking chills, headaches,
high fever, muscular aches, tachycardia, diarrhea, nausea and mild anemia, usually 2 weeks
following infection [2,3]. Non-treated malaria
caused by P. falciparum, the predominant species in Africa and the most severe or complicated
malaria-causing species, progresses to a severe
condition, where patients can no longer swallow
medicines or food, and symptoms are characterized by severe anemia, kidney failure, hepatic
failure, cerebral ischemia, altered consciousness
and coma, ultimately leading to death.
Although the human disease can be triggered through infection by different species of
the genus, all parasites display a common and
complex life cycle (Figure 1) featuring three distinct stages: mosquito; human liver; and human
blood [201]. The host infection cycle is initiated
when a small number of Plasmodium sporozoites
are injected into the human skin by an infected
female Anopheles mosquito taking a blood meal.
Once reaching the bloodstream, the parasites are
transported throughout the body, gaining access

to the liver, where they infect hepatocytes and
undergo asymptomatic exoerythrocytic asexual
multiplication (the liver stage). The infected
hepatocytes rupture and several thousand merozoites are released into the circulatory system to
infect red blood cells (RBCs), the site of major
parasite expansion. The infecting parasites
remodel the cell to support their intraerythrocytic asexual multiplication (the blood stage).
The majority of antimalarial agents target the
intraerythrocytic phase of the parasite, as pathology associated with RBC infection is responsible
for the clinical disease. Although most merozoites
released upon cell rupture re-infect RBCs and
remain as merozoites, some divert from asexual
replication and develop into gametocytes. The
ingestion of gametocytes by a female Anopheles
mosquito taking a blood meal re-establishes the
Plasmodium life cycle into the mosquito stage [4].
Despite its complexity, the Plasmodium infection cycle is sufficiently robust to have survived
for countless millennia. A robust infection
cycle is supported by a genetic mechanism of
Plasmodium causing rapid cell surface epitope
switching to evade the human immune system.
Decades of effort to develop vaccines have not yet
led to a successful immunization program (see
below). The complex life style of Plasmodium
infections permits, in theory, therapeutic interruption at many key developmental steps, including: gametocyte development and multiplication
in mosquitoes, transmission to humans during
the blood meal, hepatocyte attachment, multiplication of merozoites in hepatocytes and attachment or multiplication of merozoites in erythrocytes. Public health eradication of mosquitoes

Rodrigo G Ducati, Hilda
A Namanja-Magliano
& Vern L Schramm*

10.4155/FMC.13.51 © 2013 Future Science Ltd

Future Med. Chem. (2013) 5(11), 1341–1360

ISSN 1756-8919

Department of Biochemistry, Albert
Einstein College of Medicine, Yeshiva
University, Bronx, NY 10461, USA
*Author for correspondence:
Tel.: +1 718 430 2813
Fax: +1 718 430 8565
E-mail: vern.schramm@einstein.yu.edu

1341

Review | Ducati, Namanja-Magliano & Schramm
Human liver stages
Liver cell

Infected
liver cell

Mosquito stages

2

12 Ruptured
oocyst

1

i

Mosquito takes a blood meal
(injects sporozoites)

Exo-erythrocytic cycle

Release of
i sporozoites

11 Oocyst

4

Ruptured schizont

Sporogonic cycle

3
Schizont

Human blood stages
5

10 Ookinete

Macrogametocyte

8
Mosquito takes
a blood meal
(ingests gametocytes)

d

Erythrocytic cycle
Microgamete entering
macrogamete
Exflagellated
microgametocyte
i = Infective stage
d = Diagnostic stage

Immature
trophozoite
(ring stage)

Plasmodium
falciparum

6
Ruptured
schizont

7
Gametocytes d
Plasmodium vivax
Plasmodium ovale
Plasmodium malariae

Schizont

Mature d
trophozoite

d
7
Gametocytes

Figure 1. Life cycle of the Plasmodium species that cause human malaria.
Reproduced from [201] .

with insecticides and bed nets has also had success (see below), but in cases of active disease,
more direct intervention is essential.
Key Term
Plasmodium falciparum:

The predominant Plasmodium
species in Africa and the most
severe or complicated malaria
parasite causing disease in
humans.

Enzymatic transition state:
When chemical reactants
participate in an enzymatic
reaction, an activated complex
or a transition state is formed.
In this chemical reaction, the
transition state exists
momentarily and is less stable
than both reactants or
products.

1342

Antimalarial agents & its limitations
Epidemiologists estimate that over 2 billion
people are currently at risk of malaria infection
worldwide within the tropics. Annual statistical
data from the WHO have accounted for over
200 million new cases reported worldwide, 80%
of which are in Africa, and almost 1 million
deaths, 86% of which occur in children below
the age of five [202]. These numbers have dire
health and economic consequences for tropical
nations.
The history on strategies to treat and control
malaria infections is nearly as old as the disease itself. The first class of antimalarial drugs,
the quinolines, include quinine, mefloquine,
Future Med. Chem. (2013) 5(11)

amidoquinine, chloroquine and halofantrine [5],
which are among the earliest drugs discovered
for malaria treatment. Quinine, was isolated
from cinchona tree bark in the early 19th century, and was one of the first agents to be used
on standardized dose treatments for malaria. Its
use is still recommended, particularly on severe
malaria treatment [6]. The mechanism of action
of quinoline drugs has been thought to be on
the inhibition of heme polymerization in the
Plasmodium acidic food vacuole [7–9]. Resistance
to quinolines has been attributed to mutations
in membrane proteins involved in transport of
antimalarial drugs into the Plasmodium acidic
vacuole. In the early and mid-20th century,
chloroquine was chemically synthesized and
reported to be as effective as the natural product [6]. The P. falciparum chloroquine-resistant
transporter is one putative protein at the parasite
future science group

Transition-state inhibitors of purine salvage & other prospective enzyme targets in malaria
food vacuole membrane that has been reported
to be responsible for Plasmodium resistance
towards the quinolines. Critical mutations, such
as K76T, confer resistance by reducing drug
accumulation in the digestive vacuole [10,11].
The antifolate class of drugs targets enzymes
found in the parasite cytosol involved in folate
metabolism. This class includes sulfadoxine
(inhibits DHPS; EC 2.5.1.15) and pyrimethamine (inhibits DHFR; EC 1.5.1.3), which in
combination (Fansidar™) are more effective
than chloroquine against P. falciparum [12].
Resistance towards antifolates has been attributed to mutations in DHPS and DHFR, which
result in reduced binding affinity for the drugs,
but with retention of catalytic function. Resistance to Fansidar is prevalent in east and central
Africa, as well as some parts of Asia and South
America.
The artemisinin combination therapies (ACTs)
are now the first-line malaria treatment in endemic
regions. The origin of the artemisinins comes
from traditional medicine in China, where leaves
of wormwood plants were used for thousands of
years to treat fevers [13]. These sesquiterpene lactones contain a peroxide bridge that is thought to
decompose into damaging free radicals. Its exact
mechanism of action is uncertain, but is thought
to be the inhibition of hemezoin production in the
parasite vacuole [14,15]. In 2001, the WHO recommended the use of four ACTs to treat malaria:
artemether-lamifantrine; artesunate–mefloquine;
artesunate–amodiaquine; and artesunate–
sulfadoxine/pyrimethamine. Although resistance
towards ACTs has already begun to evolve, it still
is the recommended and most effective class of
antimalarial treatment [16].
Other approaches to combat malaria include
disease prevention methods, such as vector control using indoor insecticide spraying (dichlorodiphenyl-trichloroethane) and the use of longlasting insecticidal nets. Both methods reduce
infection incidence but have not been successful in eradicating malaria [17]. A preventative
method for malaria currently in use provides
chemoprophylaxis for travelers, as infections
can be prevented by pre-exposure administration of chloroquine (for nonresistant areas),
atovaquone–proguanil, doxycycline and mefloquine [18]. P. falciparum is the predominant
malaria parasite in endemic African regions,
and numerous epidemiology studies associate a
prevalence of the sickle cell gene as a protective
mutation [19–21]. Other protective human genetic
mutations for malaria include thalassemia and
future science group

G6PDH (EC 1.1.1.49) deficiency, adaptations commonly found in malaria-endemic
populations.
Hope for the future is that a vaccination
strategy will be developed to prevent and eradicate malaria. The need for a vaccine is clear;
however, this field has been very challenging.
Although there is no current US FDA-approved
malaria vaccine, the RTS,S vaccine developed by
GlaxoSmithKline and others has entered clinical trials [22,23]. Until then, chemotherapeutic
treatment and vector control approaches must be
improved to control current disease. Enduring
problems are the development of resistance to
drug management approaches and insecticides
by the parasites and vectors. These public health
problems emphasize the need for new drugs to
become available for malaria treatment.

| Review

Key Terms
Kinetic isotope effect: The
ratio of reaction rates of labeled
(K labeled) and unlabeled (Kunlabeled)
reactants.
Transition-state analog:
Chemically stable compound
that mimics the unstable
transition-state structure.

Transition-state theory & inhibitor
design
Enzymatic transition-state ana­lysis provides
an excellent tool for developing powerful
drugs. During enzymatic catalysis, reactants
go through an unstable structure between the
structures of substrates and products, called
the transition state [24,25]. The lifetime of the
transition state is estimated to be approximately
10 -14 s, time for a single bond vibration [24,26].
The lifetime of enzymatic transition states prevents formation of tightly bound transition-state
complexes, thus, making it difficult to develop
methods that can be used to directly observe
their structures. The best approach to determine
the transition states of enzymatic reactions is
by measuring intrinsic kinetic isotope effects
(KIEs), which compare reaction rates for the
chemical step within enzymatic reactions using
isotopically labeled and unlabeled substrates.
This information is provided when making or
breaking covalent bonds are slow steps in the
reaction, or when corrections can be made for
interfering steps. Using quantum chemistry, a
theoretical transition-state structure matching
the intrinsic KIEs can be obtained. Chemically stable transition-state analogs developed
from electrostatic maps have been shown to
bind tighter than the actual substrates with dissociation constants into the femtomolar range
(10 -14 M). Upon binding of transition-state
analogs to the enzyme, the dynamic motions
involved in forming the enzymatic transition
state convert to stable thermodynamic interactions and, thus, the enzyme undergoes conformational changes. As a result, the release of the
www.future-science.com

1343

Review | Ducati, Namanja-Magliano & Schramm
chemical transition-state analog is very slow and
energetically unfavored, with release rates ranging from minutes to days [27]. Transition-state
analogs have proven to be powerful and specific enzyme inhibitors with favorable biological
properties based on their chemical similarity to
natural metabolites [24,27].
Purine salvage enzymes &
transition-state analogs
Extensive DNA replication occurs during blood
and liver stages of the malarial parasite, when
there is a high demand for purine nucleotides.
P. falciparum, as with most parasitic protozoa,
lacks a de novo purine nucleotide biosynthetic
pathway [28], relying exclusively on the salvage
of preformed purines from its human host to
supply the requirement for purine nucleotides
[29,30]. The component enzymes of purine
salvage pathways have been targeted for the
development of antimalarials since the discovery that Plasmodium parasites are purine
auxotrophs [31]. Although the purine salvage

pathway is present both in the parasite and its
human host, most temporary interruptions of
human purine salvage have no adverse health
effects. A thorough understanding of the component enzymes and enzyme-catalyzed reactions from the pathogen can reveal features to
be exploited in rational target-selective drug
design to inhibit Plasmodium viability. Based
on the transition-state analog approach, the
new class of compounds should also be active
against Plasmodium strains that display resistance to the anti­malarial agents currently used
to treat this disease.
Several homologs to enzymes of the purine
salvage pathway have been identified in
P. falciparum based on the published genome
sequence from the human malarial parasite
[32]. Within this pathway (F igure 2) , purine
bases, nucleosides and nucleotides can be
interconverted through the activities of ADA
(EC 3.5.4.4), PNP (EC 2.4.2.1), HGXPRT
(EC 2.4.2.8), IMPDH (EC 1.1.1.205), AdSS
(EC 6.3.4.4), AdSL (EC 4.3.2.2), and GMPS

Erythrocyte
Adenosine

Adenosine

hADA
Inosine

PfADA
Inosine

Hypoxanthine

Hypoxanthine
PfHGXPRT

hHGPRT

ATP

hA
MP

K

DA

hA

PfIMPDH

IMP

IMP

Adenine

MTI
PfPNP

hPNP

Adenosine

Polyamine
synthesis

MTA

hAPRT

ADP
AMP

Xanthine

PfAdSS
SAMP

Nucleic acids

PfAdSL
AMP
Xanthine
PfHGXPRT

Guanine

Guanine

XMP
PfGMPS
GMP

PfPNP
Guanosine

Guanosine
Plasmodium falciparum

Figure 2. Purine salvage and polyamine pathways in an erythrocyte infected by
Plasmodium falciparum. Bold arrows on reversible steps indicate the metabolically favored
direction.
MTA: 5’-methylthioadenosine; MTI: 5’-methylthioinosine; SAMP: Adenylosuccinate.
Adapted from [33] .

1344

Future Med. Chem. (2013) 5(11)

future science group

Transition-state inhibitors of purine salvage & other prospective enzyme targets in malaria
(EC 6.3.5.2) [33]. In this review, examples of
transition-state analog inhibitors of enzymes
involved in different pathways in P. falciparum
are provided, highlighting transition-state
analog design and lead compounds targeting
enzymes involved in the purine salvage pathway.

O
N
N

HO

future science group

NH

OH

O

N

O

+ HPO42-

HO

N

PfPNP

OH

O

NH

N
H

Inosine

„„
PNP

PNP plays a fundamental role in the P. falciparum
purine salvage pathway, catalyzing the reversible phosphorolysis of the N-glycosidic bond
of b-purine (deoxy)ribonucleosides to generate a-(deoxy)ribose 1-phosphate and the corresponding purine bases (Figure 3) [34]. This
enzyme is essential for the formation of hypoxanthine in humans and in the protozoan, and
hypoxanthine is the primary purine needed as a
purine precursor in the parasite.
The most powerful inhibitors known for PNP
have all been developed from KIE-based transition-state ana­lysis. Some of the first transitionstate analogs were based on the transition-state
structure of bovine PNP, and were derived from
(1S)-1-(9-deazahypoxanthin-9-yl)-1,4-dideoxy1,4-imino-d-ribitol (Immucillin H; ImmH;
F igure 4), displaying IC50 values within the
sub-micromolar range in mammalian cells, and
demonstrated to be effective for T-cell-selective
immunosuppression [35]. These slow-onset
tight-binding inhibitors were modeled to
exhibit an elevated pk a of N7 of the purine ring
and the formation of an oxocarbenium ion on
the ribosyl ring, and displayed equilibrium
dissociation constants in the picomolar range
[36–39]. As the KIEs and transition-state structures became available for human and malarial
PNP [40], second-generation immucillins were
syn...