RECENT ADVANCES IN PRION CHEMOTHERAPEUTICS

Polyanions

Sulphated glycans and other polyanions are some of the oldest known inhibitors of PrP^res formation [30,31]. Their mechanism of action is likely competitive inhibition of physiological cofactors that can bind PrP and promote conversion, including RNA [2,32,33,34], or endogenous sulfated glycosaminoglycans (GAGs) [19,30,31,32,35,36]. The usefulness of these compounds is limited by potential anti-coagulant activity and/or poor blood brain barrier permeability. However, this latter drawback can be circumvented through intraventricular infusion in animal models [37].

Pentosan polysulphate (PPS) is a classic sulphated glycan which prevents new PrP^res accumulation [30,31] and is regularly employed to cure cell culture models of prion infection. It prolongs incubation times in rodent-adapted scrapie models [38,39,40,41], and has significant effect when administered via an intraventricular cannula and osmotic pump [37]. Its efficacy has recently been demonstrated in a CWD-infected deer-cell model [42], as well as in a mouse model of BSE [43]. There is also a significant amount of data from case reports of intraventricular PPS used in human prion diseases, with mixed results.

Several other polyanions have been studied. Fucoidan, a complex sulphated fucosylated polysaccharide, is an edible component of seaweed which has in vitro and in vivo anti-prion effects [44]. Phosphorothioate oligonucleotides, particularly those of 17 or more bases, have striking prophylactic effects in mice inoculated subcutaneously or intraperitoneally [45], and have reduced anti-coagulant effect compared with pentosan polysulfate. Copaxone, a polymer of alanine, glutamate, lysine, and tyrosine, is an approved treatment for multiple sclerosis, but can also prolong the incubation periods in rodent-adapted scrapie if it is mixed with the inoculum prior to infection, or if it is given at the time of inoculation [46].

Sulphonated dyes and similar compounds (Congo red, suramin, curcumin)

Congo red was the first identified inhibitor of PrP^res accumulation [47]. For a review of its role in neurodegenerative diseases see reference [48]. Its ability to stack and mimic larger polyanions [16,49,50], likely accounts for its ability to inhibit PrP^res formation through competition with sulphated glycans for PrP^c binding [35]. Inhibition may also result from an overstabilization of PrP^res [51]. In vitro, Congo red and analogues thereof decrease PrP^res and scrapie infectivity in cells [30,47,51,52,53,54,55,56], block cell-free conversion reactions [53,54,57,58], and decrease surface PrP^c in normal cells [19]. In vivo, Congo red prolongs incubation times in hamsters [59,60], but its clinical use is limited by the potential teratogenic and/or carcinogenic activity of the benzidine moiety, and by its poor blood brain barrier permeability.

Polysulphonated naphthyl urea (suramin) is a symmetrical aromatic which is structurally similar to Congo red and can decrease PrP^res, decrease surface PrP^c, and cause aggregation of recombinant PrP [61]. It has a modest early treatment effect on incubation times in hamsters [41], [61]. Suramin analogues with naphthalene- or benzene-sulphonic acid substitutions also decrease PrP^res accumulation and induce PrP aggregation, but have no effect on PrP^c expression [62].

Diferuloylmethane (the spice curcumin), has a related structure which lacks sulphonates and the benzidine moiety, is non-toxic, has good blood brain barrier permeability, and can inhibit PrP^res formation in vitro [63]. It binds a β-sheet rich form of PrP 90-231, as well as α-helical intermediates, but not the native α-helical form of this N-terminally truncated PrP fragment [64]. Whether curcumin can bind to native full-length PrP^C remains to be determined. While high dose treatment was ineffective in vivo [63,65], low doses (50 mg/kg/day) have recently been shown to prolong survival in intracerebrally inoculated mice that were treated starting at 100 days post-inoculation [65].

Cyclic tetrapyrroles (porphyrins, phthalocyanines)

These compounds have highly conjugated planar aromatic ring systems that bind transition metal ions and can be circumscribed by anionic, cationic, or uncharged peripheral substituent groups. Examples include phthalocyanine tetrasulphonate (PcTS), deuteroporphyrin IX 2,4-bis-(ethylene glycol) iron(III) (DPG₂-Fe³⁺), meso-tetra(2-N-methylpyridyl)porphine iron(III) (TMPP-Fe³⁺), tetra (4-N,N,N-trimethylanilinium)porphine (Fe-TAP), and tetra (4-sulphonatophenyl)porphine (Fe-TSP). They inhibit PrP^res accumulation in vitro [66] and significantly increase survival times in vivo when given early in the disease [17,42,67,68], probably by stacking and directly interacting with the flexible amino-terminal domain of PrP molecules [16,69]. Of note, Fe-TSP was more effective in treating mice inoculated intracerebrally when it was used in combination with pentosan polysulphate [17], thus highlighting a possible role for combination therapy. Tetrapyrroles may also be useful in non-rodent models; the compound indium(III) meso-tetra(4-sulphonatophenyl)porphine chloride (In-TSP) has recently been shown to affect CWD-infected deer cell cultures [42].

Another new player on the porphyrin stage is the natural cyclic tetrapyrrole, hemin, which inhibits PrP^res formation in cell culture (D. A. Kocisko and B. Caughey, unpublished data). The finding that hemin binds PrP^c and causes its clustering and internalization in cell culture [70], raises the possibility that there are physiologically significant hemin-PrP^c interactions, and suggests that prion inhibition may result from sequestration of PrP^c into a non-convertible state.

Lysosomotropic factors (quinacrine, quinoline, acridines, and phenothiazines)

Quinacrine, chlorpromazine, quinine and related molecules have inhibitory effects on PrP^res formation in vitro cells [71,72,73], and variable effects in vivo [37,74,75,76]. Quinacrine has received attention in human clinical trials, but a retrospective analysis of human CJD patients treated with quinacrine, compared to untreated control cases, demonstrated no benefit [77]. Interestingly, the combined use of quinacrine plus desipramine or the HMG-CoA reductase inhibitor simvastatin [78] or the recombinant mutant PrP peptide rPrP-Q218K [79], may be more effective than quinacrine alone (see below). Also, intraventricular cannula treatment of mice with quinine and biquinoline partially reduced pathology in the hemisphere with the cannula, suggesting that this route of administration might enhance efficacy for some of these drugs [80]. One compound, the anti-malarial mefloquine, looked promising in cell culture, but lacked efficacy in tg7 mice that were inoculated intraperitoneally [81].

Tetracyclic compounds

The experiments published to date on tetracycline and doxycycline look promising in that pre-incubation with inocula reduces infectivity [82,83] and detectable PrP^res [82,84], and the results of in vivo studies are pending. However, the related molecule minocycline did not demonstrate therapeutic effect [65].

Other amyloidophilic compounds

Recently, anti-prion effects have been studied in a new group of amyloidophilic compounds [85]. Prion-strain dependent effects were seen in cell culture and in intracerebrally inoculated mouse models, where oral drug administration, started early in the disease course, extended incubation times.

Pyridine dicarbonitrile compounds

Certain PrP mutants resist conversion, and a computational search for molecules that spatially resembled these mutants led to the discovery of several compounds with inhibitory effects in cell culture [24]. A pyridine dicarbonitrile substructure was found to be common to these agents [25], and further screening of these types of compounds discovered two which, at 50 nM concentration, could reduce PrP^res in cell culture to < 30% of controls [86]. No in vivo data is currently available.

Peptide aptamers and β-sheet breakers

In cell culture models and cell-free conversion systems, peptides of PrP residues 106-141, 119-136, 166-179, and 200-223 can block PrP^res formation [87,88,89]. Subsequently, searches have been made for other peptides which might interfere with conversion. One group placed random 16-mer sequences in a constant scaffold of thioredoxin A from E. coli, and selected those which recognized PrP sequence 23-231. Cell culture studies demonstrated that inhibition of PrP^res formation occurred in those which targeted PrP^c in the endoplasmic reticulum or lysosome [90]. Another group sought peptides which represented a portion of the target protein with prolines inserted to block incorporation into a β-sheet structure. iPrP13, one such β-sheet breaker, had some prophylactic effect in vivo [91]. However, there may be some strain or model dependence, as there was effect in a Chinese Hamster Ovarian cell model expressing mutant PrP, but no effect on infected N2a cells [92].

Cholesterol-depleting agents

PrP^c is primarily located in areas of membrane rich in sphingolipids and cholesterol, called lipid rafts, and conversion of membrane associated PrP^c depends on interactions with PrP^res in the same membrane [3]. Therefore, agents which affect membrane cholesterol levels could alter PrP^c distribution and thereby affect PrP^res formation. Mevinolin, an HMG CoA reductase inhibitor which blocks cholesterol synthesis, does reduce cell surface PrP^c and leads to the accumulation of PrP^c within the Golgi apparatus [93]. Amiodarone and progesterone cause cholesterol redistribution and can clear PrP^res from cells [78]. Lovastatin and squalestatin, statin drugs which inhibit cholesterol synthesis, reduce PrP^res accumulation in cell culture, an effect which is lost upon addition of cholesterol [94,95]. Inhibitors of cholesterol esterification have also been shown to block PrP^res accumulation, consistent with a correlation between cholesterol ester levels and susceptibility of cells and sheep to scrapie infection [96]. However, such in vitro effects of cholesterol modulators do not always translate into in vivo results; inhibitors of 7-dehydrocholesterol reductase and 24-dehydrocholesterol reductase prevented PrP^res formation in cell culture, but had no effect on scrapie incubation times, or on brain cholesterol levels in vivo, regardless of whether treatment was initiated prior to or at the time of inoculation [97].

In contrast, the statin drug simvastatin can cross the blood-brain barrier and give some benefit in vivo, despite its short half-life of 1.5 hours. While low doses (1 mg/kg/day) were only effective if begun at the time of intracerebral inoculation [98], high doses (100 mg/kg/day) started 100 days after intracerebral inoculation prolonged survival in mice [99]. Interestingly, in both cases there were no effects on PrP^res levels, and cholesterol levels were only reduced in the liver, not the brain. Thus, it may be that the benefits seen were from anti-inflammatory effects more than cholesterol lowering [99]. Regardless of the statin mechanism of action, its usefulness may be enhanced in combination therapy, as using simvastatin with quinacrine can produce a synergistic inhibitory effect [78].

Polyene antibiotics

The antifungal drug amphotericin B reduces PrP^res formation in cell culture, but is not curative [100]. It and its less toxic analog MS-8209 offer some early treatment benefits in hamsters [101,102,103,104,105,106,107,108], and some late treatment effects in C57BL/6 mice that have been inoculated intracerebrally [109,110], likely by perturbing the raft membrane domains with which PrP^c is associated [100,111]. However, the effect may be somewhat strain-specific [104,105,112], and amphotericin B has been unsuccessful in the treatment of humans with CJD [113]. Another analogue, mepartricin, has efficacy only against intraperitoneal inoculations. Most recently, filipin, a relative of amphotericin B, has been shown to reduce PrP^res in cell culture [111], but no in vivo data is available at this time.

Copper, chelators

The role of copper in the functioning and conversion of normal PrP, which has several copper binding sites, has been much debated [121,122,123,124,125,126,127,128]. The proteinase K resistance of PrP^Sc is enhanced in vitro as copper concentrations are increased, suggesting that decreasing copper may be beneficial therapeutically. Copper chelators have had promising effects in cell culture, and examples such as D(-)penicillamine [129] and clioquinol [130] have shown modest effects in vivo. It should be noted, though, that their effectiveness may correlate more with superoxide dismutase ability than chelating properties [131]. Also, clioquinol has structural similarity to quinacrine, raising questions regarding its actual mechanism of action. The chelator chrysoidine inhibits PrP^res in cell culture, without reducing PrP^c levels or destabilizing PrP^res [132], but again, chelation may not be its sole mechanism of action. Further confusing the issue is the recent observation that providing infected mice with copper in their water was sufficient to prolong their incubation periods [133].

Dimethylsulfoxide (DMSO)

This organic solvent inhibits PrP^res accumulation in cell culture [134] and lowers the infectivity titre in brain homogenate [135], but has significant toxicity issues in long term use [136]. Also, studies vary in their support for any treatment effect in vivo [136,137].

Antivirals

Early attempts to treat human CJD patients with a wide variety of antiviral agents were unsuccessful, except for some temporary modest improvements reported in two of three CJD patients intermittently treated with vidarabine (adenine arabinoside) [138]. Subsequent studies of rifampicin, cytosine arabinoside, adenosine arabinoside, isoprinosine, amantadine, methisazone, phosphonoacetic acid, virazole, methisoprinol, β-propiolactone, and thiamphenicol have failed to show any effect in mice [101,139,140,141,142].

Laminin receptor precursor protein (LRP/LR)

The 67 kDa laminin receptor and its 37 kDa precursor bind PrP^c [143,144], and possibly PrP^res [145], and play roles in PrP^res propagation [146]. Several techniques have been used to target LRP/LR, including small interfering RNAs (siRNAs), antisense RNA, LRP antibodies [146], and more recently, recombinant single chain antibodies (scFvs) [147]. These approaches have been well reviewed in reference [148].

Cell signalling pathways

A variety of enzyme and cell signalling inhibitors have been screened and tested for their anti-prion abilities. The mitogen-activated protein kinase ½ (MEK½) inhibitor SL327 [149], the cysteine-protease inhibitor L-trans-epoxysuccinyl-leucylamido (4-guanidino) butane (E64d) [71], and inhibitors of phospholipase A₂ (PLA₂) and platelet activating factor (PAF) [150], all decrease PrP^res in cell culture. The p53 inhibitor pifitrin-α also reduces PrP^res in vitro, but has no anti-prion effect in vivo [137]. The tyrosine kinase inhibitor STI571, also known as imatinib mesylate, has efficacy in vitro and in vivo, likely via targeting lysosomal degradation pathways. It cured cells of PrP^res after 10 days of treatment [151,152], decreased spleen PrP^res, and delayed neuroinvasion if it was administered early after intraperitoneal inoculation. However, even intracerebral delivery of this drug could not clear PrP^res from the brain [152].

Polycations

Anti-prion tendencies in vitro have been reported in a group of cationic polyamines which may exert their effects, in part, via enhanced PrP^res clearance [153]. Polypropyleneimine generation 4.0, polyethyleneimine, and polyamidoamide generation 4.0, are components of Qiagen SuperFect transfection reagent which prevent PrP^res formation [153] and remove infectivity from cell culture [154], possibly acting at the level of the lysosome [154]. Phosphorus dendrimers, generation 4, are less toxic, more bioavailable, and can reduce the splenic content of PrP^res in mice inoculated intraperitoneally [155]. Spermine and spermidine reduce the tRNA-induced polymerization of alpha-PrP, suggesting they might influence prion infection [156]. DOSPA is a lipopolyamine which enhances PrP^res clearance in cells and also blocks de novo formation. Its activity is dependent on membrane association [157].