Gregory E. Garcia

Multitarget Drug Design Strategy: Quinone–Tacrine Hybrids Designed To Block Amyloid-β Aggregation and To Exert Anticholinesterase and Antioxidant Effects

We report the identification of multitarget anti-Alzheimer compounds designed by combining a naphthoquinone function and a tacrine fragment. In vitro, 15 compounds displayed excellent acetylcholinesterase (AChE) inhibitory potencies and interesting capabilities to block amyloid-β (Aβ) aggregation. The X-ray analysis of one of those compounds in complex with AChE allowed rationalizing the outstanding activity data (IC50 = 0.72 nM). Two of the compounds showed negligible toxicity in immortalized mouse cortical neurons Neuro2A and primary rat cerebellar granule neurons. However, only one of them was less hepatotoxic than tacrine in HepG2 cells. In T67 cells, both compounds showed antioxidant activity, following NQO1 induction. Furthermore, in Neuro2A, they were able to completely revert the decrease in viability induced by Aβ. Importantly, they crossed the blood-brain barrier, as demonstrated in ex vivo experiments with rats. When ex vivo results were combined with in vitro studies, these two compounds emerged to be promising multitarget lead candidates worthy of further pursuit.

Isolation of a substance from the mosquito that activates Plasmodium fertilization

We have isolated a small, heat stabile, hydrophilic molecule from the gut lumen of unfed, female Anopheles stephensi that is a potent inducer of gametogenesis in Plasmodium falciparum and P. gallinaceum at a hydrogen ion concentration, pH 7.4, that normally suppresses activation. This gamete activation factor (GAF) was purified using reverse phase high performance liquid chromatography and determined to have a major ion m/z of 206.1 by low resolution electrospray mass spectrometry. The molecule, which was also found in the heads of both female and male A. stephensi, absorbed light in the ultraviolet region at three maxima (lambda(max) = 213, 245 and 350 nm); the 245/350 nm absorbance ratio was 7.0. The structure of the molecule and its normal function in the mosquito are not yet known, but in a sample of diverse insect species, extracts from those that feed on blood were bioactive. We propose that GAF is the previously observed malaria exflagellation factor (MEF).

Hydrolysis potential of recombinant human skin and kidney prolidase against diisopropylfluorophosphate and sarin by in vitro analysis

Human prolidase (PROL), which has structural homology to bacterial organophosphate acid anhydrolase that hydrolyze organophosphates and nerve agents has been proposed recently as a potential catalytic bioscavenger. To develop PROL as a catalytic bioscavenger, we evaluated the in vitro hydrolysis efficiency of purified recombinant human PROL against organophosphates and nerve agents. Human liver PROL was purified by chromatographic procedures, whereas recombinant human skin and kidney PROL was expressed in Trichoplusia ni larvae, affinity purified and analyzed by gel electrophoresis. The catalytic efficiency of PROL against diisopropylfluorophosphate (DFP) and nerve agents was evaluated by acetylcholinesterase back-titration assay. Partially purified human liver PROL hydrolyzed DFP and various nerve agents, which was abolished by specific PROL inhibitor showing the specificity of hydrolysis. Both the recombinant human skin and kidney PROL expressed in T. ni larvae showed ∼99% purity and efficiently hydrolyzed DFP and sarin. In contrast to human liver PROL, both skin and kidney PROL showed significantly low hydrolyzing potential against nerve agents soman, tabun and VX. In conclusion, compared to human liver PROL, recombinant human skin and kidney PROL hydrolyze only DFP and sarin showing the substrate specificity of PROL from various tissue sources.

Buforin I, a natural peptide, inhibits botulinum neurotoxin B activity in vitro.

Botulinum neurotoxin B (BoNT/B) serotype specifically cleaves between the amino acids glutamine and phenylalanine (Q and F bond) in position 76–77 of synaptobrevin (VAMP2). We evaluated peptides that contain the QF cleavage site but are not identical in primary structure to the VAMP2 sequence surrounding the QF site for both inhibition of BoNT/B proteolytic activity and as substrates for BoNT/B. A reverse‐phase high‐performance liquid chromatography (RP‐HPLC) method was used to measure digested peptides. A dose as high as 600 μM of substance P, and 11‐amino acid peptide containing the QF bond, was neither a substrate nor inhibitor of BoNT/B in our assay, suggesting that more than the QF bond is required to be recognized by BoNT/B. Buforin I (B‐I, QF site 24–25) is 39 amino acids in length, and sequence comparison of B‐I and VAMP2 indicated a similarity of 18% for conserved amino acids around the QF site. Furthermore, computer‐aided secondary structure computations predict α‐helical structures flanking the QF site for VAMP2 and for the upstream sequence of B‐I. Although predictions for the downstream sequence give nearly equal tendencies for α‐helical and β‐sheet structures, Yi et al. showed that the downstream sequence is likely to be the α‐helix based on their examination of buforin II (B‐II, a 21‐amino acid subset of B‐I (16–36)), which includes the QF site and the downstream sequence of B‐I. Buforin I was found not to be a substrate for BoNT/B; however, B‐I dose dependently and competitively inhibited BoNT/B activity, yielding IC50 = 1 × 10−6 M. In contrast, B‐II was not a substrate for BoNT/B and exhibited only 25% of the B‐I inhibition of BoNT/B. Two additional B‐I deletion peptides were tested for inhibition of BoNT/B proteolysis: peptide 36 (36 mer; containing B‐I amino acids 1–36) and peptide 24 (24 mer; B‐I amino acids 16–39). Peptide 24 had a similar inhibitory effect to B‐II (ca. 25% of B‐I) but peptide 36 was almost 50% as potent as B‐I. These findings suggest that the buforin tertiary structure is important for the inhibitory activity of these peptides for BoNT/B.

Protection of red blood cell acetylcholinesterase by oral huperzine A against ex vivo soman exposure: next generation prophylaxis and sequestering of acetylcholinesterase over butyrylcholinesterase.

As part of a phase Ib clinical trial to determine the tolerability and safety of the highly specific acetylcholinesterase (AChE) inhibitor huperzine A, twelve (12) healthy elderly individuals received an escalating dose regimen of huperzine A (100, 200, 300, and 400 microg doses, twice daily for a week at each dose), with three (3) individuals as controls receiving a placebo. Using the WRAIR whole blood cholinesterase assay, red blood cell AChE and plasma butyrylcholinesterase (BChE) were measured in unprocessed whole blood samples from the volunteers following each dose, and then for up to 48h following the final and highest (400 microg) dose to monitor the profile of inhibition and recovery of AChE. Significant inhibition of AChE was observed, ranging from 30-40% after 100 microg to >50% at 400 microg, and peaking 1.5h after the last dose. Gradual recovery of AChE activity then occurs, but even 48 h after the last dose red blood cell AChE was about 10% below control (pre-dose) values. Huperzine A levels in plasma peaked 1.5h after the final 400 microg dose (5.47+/-2.15 ng/mL). Plasma BChE was unaffected by huperzine A treatment (as expected). Aliquots of huperzine A-containing (from three individuals) and placebo blood samples were exposed ex vivo to the irreversible nerve agent soman (GD) for 10 min, followed by removal of unbound huperzine and soman from the blood by passing through a small C(18) reverse phase spin column. Eluted blood was diluted in buffer, and aliquots taken at various time intervals for AChE and BChE activity measurement to determine the time taken to achieve full return in activity of the free enzyme (dissociation from the active site of AChE by huperzine A), and thus the proportion of AChE that can be protected from soman exposure. Huperzine A-inhibited red blood cell (RBC) AChE activity was restored almost to the level that was initially inhibited by the drug. The increased doses of huperzine A used were well tolerated by these patients and in this ex vivo study sequestered more red blood cell AChE than has been previously demonstrated for pyridostigmine bromide (PB), indicating the potential improved prophylaxis against organophosphate (OP) poisoning.

Apoptosis of L1210 leukemia cells induced by 3-deazaadenosine analogs: Differential expression of c-myc, NF-kappa B and molecular events

A new class of potent apogens (apoptosis-inducing agents) has been identified, consisting of 3-deazaadenosine (DZA), 3-deaza-(+/-)aristeromycin (DZAri) and 1-beta-D-arabinofuranosyl-1H-imidazo[4,5-&cumacr;]pyridine (ara-3-deazaadenine; DZAra-A). They are inhibitors of S-adenosylhomocysteine hydrolase and indirect inhibitors of methylation. Furthermore, they have also been found to form 3-deaza-nucleotide analogs. The DZA analogs, DZA, DZAri, and DZAra-A, induced DNA fragmentation in a dose- and time-dependent manner, reaching a maximum at 250 &mgr;M after 72 h. Cycloheximide at 0.5 &mgr;g/ml completely blocked the DNA fragmentation induced by 250 &mgr;M of each of the analogs. Interestingly, exogenous 100 &mgr;M L-homocysteine thiolactone abrogated the DNA fragmentation caused by DZAri and DZAra-A, but not by DZA. Flow cytometric analysis showed that DZA arrested the cells in the G(2)/M phase, whereas the S phase was arrested by DZAri. Correlated with the effect of DZA was a rapid decrease in the expression of c-myc, whereas nur77 and GAPDH were unaffected. In comparison, there was an elevated expression of IFN-gamma mRNA without apparent change in bax, p53 or GAPDH mRNA after 24 h. After treatment with DZA, there was an elevated expression of NF-kappaB DNA binding activity, which became more pronounced at 24 h. Simultaneously, there was an apparent disappearance of AP-1 activity. Thus, DZA most likely inhibited the RNA synthesis of c-myc, a reduction of which could trigger a cascade of gene transcription leading to apoptosis in L1210 cells. Copyright 1997 S. Karger AG, Basel

TCEP treatment reduces proteolytic activity of BoNT/B in human neuronal SHSY‐5Y cells

The light chain (LC) of botulinum neurotoxin B (BoNT/B) is unable to enter target neuronal cells by itself. It is brought into the cell in association with the BoNT/B heavy chain (HC) through endocytosis. The BoNT HC‐LC subunits are held together by a single disulfide bond. Intracellular reduction of this bond and separation of the two subunits activates the endopeptidase activity of the LC. This requirement suggests a strategy to prevent uptake by prophylactic reduction to disrupt the disulfide bond prior to endocytosis of the complex. We examined the utility of tris‐(2‐carboxyethyl)‐phosphine hydrochloride (TCEP), a relatively non‐toxic, non‐sulfur containing disulfide bond reducing agent that lacks the undesirable properties of mercapto‐containing reducing agents. We found that TCEP was as effective as DTT with maximal LC endopeptidase activation occurring at 1 mM, a concentration not toxic to the human neuronal cell line, SHSY‐5Y. In these cells, 1 mM TCEP maximally protected against BoNT/B inhibition of [3H]‐NA release, achieving 72% of the release from un‐intoxicated controls. This effect appears to be due to the sparing of SNARE proteins as the levels of VAMP‐2, the specific target of BoNT/B, were protected. These results show that TCEP disrupts the structure of BoNT/B by reduction of the LC and HC bridging disulfide bond and prevents neuronal intoxication. Since disulfide bond coupling between toxin subunits is a general motif for many toxins, e.g., ricin, snake venom, and all BoNT serotypes, this suggests that TCEP is a promising means to protect against these toxins by preventing cell penetration. J. Cell. Biochem. 107: 1021–1030, 2009. Published 2009 Wiley‐Liss, Inc.

Un-nicked BoNT/B activity in human SHSY-5Y neuronal cells

BoNT/B holotoxin (HT) from the native source is a mixture of nicked and un‐nicked forms. A previous study showed that while un‐nicked HT could be transcytosed by intestinal epithelial cells, they did not correlate this with proteolytic activity or biological effect(s). Un‐nicked HT is likely to be present in BoNT biological warfare agents (BWA), so it is important to investigate the relative toxicity of un‐nicked HT in this BWA. To address this issue, we purified un‐nicked HT from commercial sources and evaluated its ability to cleave substrates both in vitro and in vivo, and its effects on vesicle trafficking. The un‐nicked HT was unable to cleave VAMPTide™ substrate used for in vitro proteolytic assays. Brief digestion of the un‐nicked toxin with trypsin resulted in significant activation of the toxin proteolytic ability. SHSY‐5Y human neuroblastoma cells were used to examine HT uptake and activation in vivo. Vesicle trafficking can be measured following K+ stimulation of cells preloaded with [3H]‐noradrenaline (NA). We found that highly purified un‐nicked HT did inhibit NA release but at much reduced levels compared to the nicked toxin. That the reduction in NA release was due to BoNT effects on SNARE proteins was supported by the finding that VAMP‐2 protein levels in un‐nicked toxin treated cells was greater than those treated with nicked toxin. These results demonstrate that although un‐nicked HT has markedly reduced toxicity than the nicked form, due to the preponderance in BoNT/B preparations from the native bacteria, it is a major source of toxicity. J. Cell. Biochem. 105: 129–135, 2008.

Induction of apoptosis and c-myc in L1210 lymphocytic leukemia cells by adenosine

High concentrations of adenosine (Ado), when added to L1210 lymphocytic leukemia cells, resulted in apoptosis or programmed cell death. The apoptotic process was accompanied by distinct morphological changes including chromatin condensation and blebbing of plasma membranes. Extensive DNA fragmentation was correlated with Ado concentrations. Furthermore, apoptosis in these cells was preceded by an early but transient expression of c-myc proto-oncogene, and was not influenced by homocysteine thiolactone added to the cells. Since severe combined immunodeficiency (SCID) is associated with a deficiency of adenosine deaminase, leading to defects in both cellular and humoral immunity, Ado-induced apoptosis may thus be a contributing factor in the pathology of SCID. Copyright 1994 S. Karger AG, Basel

Amino Acid Sequence of Horse Serum Butyrylcholinesterase

Horse serum butyrylcholinesterase (Eq-BChE) has been shown to be an effective pretreatment drug (bio-scavenger) for organophosphate toxicity in mice and rhesus monkeys. In addition, it has the longest mean retention time in animals among all ChE preparations tested. We now report the complete amino acid sequence of this protein (Figure 1). The tetrameric form of Eq-BChE was purified and subjected to fragmentation by cyanogen bromide, trypsin, and endo Arg-C protease. Digests were fractionated by reverse phase HPLC, and isolated peptides were sequenced by automated Edman degradation. Overlapping peptides were used to determine sequence order. Like other BChEs it is composed of 574 amino acids. Eight carbohydrate modified sites were found at amino acids 57, 106, 241, 256, 341, 455, 481, and 486 as N-linked Asn residues. The active site charge-relay triad amino acids were Glu325, Ser198 and His438. The N-terminal amino acids were Glu-Glu-Asp-Ile, consistent with this being the mature form of the enzyme. Eq-BChE has a high degree of primary sequence identity to human BChE, 90.1%, but has one less N-glycosylation site and Cys residue. If conserved amino acid substitutions are included, the proteins share 93.4% homology.