Chemistry and catalysis advances in organometallic chemistry and catalysis
Download 11.05 Mb. Pdf ko'rish
|
- Bu sahifa navigatsiya:
- Scheme 41.2
- 41.3.4 Making Pharmaceutically Acceptable or Druglike CORMS
|
554 METAL CARBONYLS FOR CO-BASED THERAPIES: CHALLENGES AND SUCCESSES Fe N
N N CO N N N N N N N4Py N N N N O N N N N N N N N N N N Tpmen SBPy3 PaPy3
− [Fe
II (N5)CO)]
z CORM scaffold Figure 41.4 Monocarbonyl complexes of the [Fe II (CO)(N5)] z + type. are problematic in this respect, because once the CO is lost and the complex starts to oxidize, its chemistry becomes totally different from that of the original species and its scaffold is also rather different. Even the [RuCl 2 (DMSO)
4 ] complex is a false negative control for CORM-2 because the chemistry of the Ru(CO) 3 Cl 2 (DMSO) (the active component of CORM-2; Eq.1) is started by its water–gas shift reaction with water at the physiological pH. The oldest example (1933) of a reversible CO carrier, [Fe(l-cysteinate) 2 (CO)
2 ] 2 − , was reexamined and the complex mixture of isomers reinterpreted in favor of the predominance of the cis-Fe(CO) 2 species [102]. Since it is easily assembled from Fe 2 + , CO, and cysteine, it could be formed close to sites of HO activity. Its toxicity is very low, yet it is totally inactive in the NO inhibition anti-inflammatory tests. Breaking the usual dominance of 18-electron complexes in middle group transition metal organometallic chemistry, Zobi and coworkers [103] developed a remarkable series of 17-electron cis-trans-[Re II (CO)
2 Br 2 L 2 ] complexes that are very efficient and competent CORMs and protect neonatal rat ventricular cardiomyocytes in an ischemia/reperfusion model in vitro. Only monodentate pyridine and imidazole-type ligands elicit the carbonylation of deoxy-Mb. The t 1/2
value is sensitive to ligand substitution and pH, the pH being below the physiologic pH value (e.g., at pH 5.8). The ischemia-reperfusion tests cleanly reveal two crucial findings: (i) the rate of CO release in the Mb assay (t 1/2
) does not determine the cytoprotective effect and (ii) the compounds that offer best protection are those that do not permeate cell membranes. In this way, the most effective compound was the one with L = benzimidazole (Fig. 41.3), which is also the slowest releaser in the Mb assay. The compound with L = 4-picoline (Fig. 41.3) is a faster releaser but enters the cells and kills them. The next big step from here was the synthesis of B 12 -ReCORM-2 (Fig. 41.3), which is the adduct of cyanocobalamin (vitamin B12) with the 17-electron fragment [Re(CO) 2 (H 2 O)Br
2 ] [104].
When administered at the “onset of reperfusion,” B 12 -ReCORM-2 is strongly cytoprotective in the ischemia-reperfusion of neonatal rat cardiomyocytes (NRC) in culture. The introduction of the biological ligand improved not only the solubility but also the activity and stability of the original [Re II (CO)
2 ] fragment. B 12 -ReCORM-2 does not penetrate cells. It does not impair the oxygen consumption of the NRC under the conditions of cytoprotection but shows acute antioxidant properties at the “onset of reperfusion,” which is attributed to the oxidation of the low valent Re(II) to the final ReO 4 −
This system represents a major advance toward pharmaceutically acceptable CORMs because it provides important insights into the relationship between the CORM and the cell, which are necessary for the appearance of therapeutic activity, and also because it shows how the presence of biocompatible coordination spheres can improve the activity of CORMs. B
12 -ReCORM-2 undergoes rapid hydrolysis losing the two Br − ligands before releasing CO. A Re I (CO) species was identified by ESI-MS, showing that a reduction of the Re(II) takes place in the process. A study on the mechanism of the decomposition of the parent [Re II (CO)
2 Br 4 ] 2 − proposes that such reduction is carried out by Br - as summarized in Scheme 41.1. Interestingly, the CO seems to be released from the stable Re(I) species upon oxidation, forming perrhenate. This oxidation parallels the formation of polyoxomolybdates from the oxidation of the Mo(0) complexes such as ALF186.
THERAPEUTIC DELIVERY OF CO 555 [Re(CO)
2 Br 4 ] 2 − [Re(CO) 2 (H 2 O) 3 Br] + [Re(CO) 2 (H 2 O) 3 Br] ReO 4 − + CO +H 2 O −Br
− +Br
− −BrO
− O 2 Scheme 41.1 Decomposition of the 17-electron [Re(CO) 2 Br
] 2 − in aqueous medium. Enzyme-Triggered CO Release: ET-CORMs In 2011, the first example of enzymatically triggered CO releasing molecules (ET-CORM) was published [105]. These acyloxycyclohexadiene– iron tricarbonyl complexes are activated by esterase cleavage, producing the unstable diene-alcohol intermediate, which isomerizes to the ene-one complex (Scheme 41.2). Loss of this labile ligand allows rapid access of O 2 to the coordination sphere of the Fe(CO) 3 fragment, which irreversibly decomposes to release CO, the ene-one ligand, and Fe
+ ions. This trigger mechanism was substantiated by CO-release experiments (Mb assay) and by the inhibition of NO production in the LPS-induced RAW246.7 macrophage cellular assay. Complexes such as ET ac -CORM-1 only released CO to deoxy-Mb and only inhibited NO production in the LPS-induced macrophage assay in the presence of an esterase. The best complex studied, ET ac -CORM-1 (Fig. 41.5), is a very potent inhibitor of iNOS in this macrophage test (30% NO inhibition at 5 μM), although it still has some toxicity. The (acyloxydiene)Fe(CO) 3 complexes prepared do not respond equally to all esterases or lipases. Instead, each one of them seems to have a preference for a given type of esterase unless they are not very pure. This characteristic is very important because it brings an additional level of selectivity to the activation process. An extended variation of the substituents at the acyloxydiene ligand enabled the first structure–activity relationships: (i) high biological activity requires the acyl group to lie outside the diene system (dashed acyl in Scheme 41.2); (ii) increasing bulk of the acyl group decreases both NO inhibition activity and toxicity; (iii) introduction of a second acyloxy ester increases both NO inhibition activity and toxicity; and (iv) methylation of the ring (dashed methyl substituents in Scheme 41.2) strongly decreases toxicity only at the cost of a small decrease in activity. This led to the selection of the two complexes ET ac ,-CORM-1 and ET ac -CORM-2 (Fig. 41.5) as the best compounds, although the higher toxicity of ET ac -CORM-2 makes ET ac -CORM-1 the overall best choice of the compounds studied. The possibility to control biological activity through the ligand sphere had been mentioned by several authors before, but this study was the first to be published with a rational framework for equipping CORMs with tissue or cell-specific active targeting. This approach has now been extended to a phosphatase-triggered CORM by decorating the diene ligand with phosphoryloxy groups as depicted in Scheme 41.2 [106]. Although the screening of phosphatases is still limited and unsatisfactory, the proof-of-concept was obtained based on the same platform of tests, that is CO release and iNOS inhibition promoted by the (phosphoryloxydiene)Fe(CO) 3 CORMs in the presence of phosphatases. Interestingly, the more favored position of the phosphoryloxy substituent is now in the “inner” position (ET P -CORM-1) (Fig. 41.5) of the diene ligand instead of the “outer” position (ET P -CORM-2) that was favored for the esterase-triggered acyloxy complexes (ET ac -CORM-1). (OC) 3 Fe O O R (OC) 3 Fe OH
(OC) 3 Fe O CO Esterase (OC) 3 Fe O P H 3 CO O − Phosphatase O Oxidative O O R' Decomposition Scheme 41.2 Reaction pathways of enzyme-triggered CORMs. 556 METAL CARBONYLS FOR CO-BASED THERAPIES: CHALLENGES AND SUCCESSES (OC) 3
P H 3 CO O − O (OC)
3 Fe O P OCH 3 O − O ET P -CORM-1
ET P -CORM-2 ET ac -CORM-1 (OC) 3 Fe O O (OC) 3 Fe O O O O ET ac -CORM-2 Figure 41.5 Enzyme-triggered CORMs. Light-Triggered CO Release: PhotoCORMs Cleavage of the M–CO bonds by light is a general property of MCCs and its first application to CO release used Mn 2 (CO) 10 irradiation to deliver CO to the mice kidneys in vivo [107]. In 2008, Schatzschneider [108] reported that the photolysis of [(tpm)Mn(CO) 3 ] + (tpm
= tris(pyrazolyl)methane) releases 2 equiv of CO and results in a potent cytotoxic agent toward HT29 human colon cancer cells comparable to 5-fluorouracil (5-FU), an established anticancer drug. The biological action remains to be definitely assigned to CO, the decarbonylated metal fragment, or both. Importantly, Raman microscopy revealed that [(tpm)Mn(CO) 3 ]
accumulates preferably at the cell wall and nucleus [109]. Active targeting of [(tpm)Mn(CO) 3 ]
to the tumor tissue was approached by appending specific peptides to the tpm ligand [110]. Passive targeting through the enhanced permeability and retention effect of tumoral tissues was approached by grafting the [Mn I (CO) 3 ] + motif on the surface of SiO 2 nanoparticles [111] or polymeric methacrylate-based backbones [112]. The complex [Fe II (CO)(N4Py)] 2 + (Fig. 41.4) provides a similar platform for photodynamic therapy of cancer [113]. More in tune with the usual biological activity of CO, photoCORMs based on the [Mn I (CO) 3 ] + fragment and tripodal polypyridine amine ligands effectively relax isolated aortic rings [114], whereas cis-[Fe(CO) 2 (SCH
2 CH 2 NH 2 ) 2 ] activates BK Ca
+ channels [115]. The photoCORM field was recently reviewed and attention called to the need to define the therapeutic targets to treat with CO or the decarbonylated metal fragment [116, 117]. PhotoCORMs have the great advantage that they have a “clean” triggering mode where an “onset” of therapy can be accurately defined and literally “switched on” by the use of a light source. However, their use will certainly remain limited to administration within hospital settings under strict medical/surgical control. In this way, photoCORMs share common limitations with CO inhalation.
At Alfama the ultimate goal is to develop CORMs that can be taken by any patient in the absence of direct medical assistance, that is, in ambulatory regimen. Therefore, CORMs must be pharmaceutically acceptable and behave like other regular drugs or prodrugs. They must be stable to air, water and blood, soluble in water, toxicologically safe, therapeutically effective and potent, and exhibit a recognizable and acceptable pharmacokinetic profile. Ideally, each one of them must be equipped to target a specific tissue where it decomposes and delivers the CO. Although producing many arguments in favor of CORM therapy, the experimental CORMs lacked such druglike properties. ALF062 and CORM-2 are not water soluble, whereas CORM-3 and ALF186 are decomposed by aqueous, aerobic, biologically compatible solutions at physiological pH, as well as by plasma or blood, thereby losing any kind of controllable pharmacokinetic profile or tissue specificity. In early 2009, Alfama put together a methodology to produce druglike CORMs by selecting and assembling the building units of the model CORM depicted below (Fig. 41.6) in order to enable the final prodrug to target and specifically release therapeutic CO at the disease site [118]. This model CORM has three layers: the central metal ion, the first coordination sphere (inner sphere) formed by the ligands directly attached to the metal including the active principle CO, and an outer sphere (drug sphere) formed by the distal functional groups appended to the ligands, which largely control the solubility, biocompatibility, targeting,
THERAPEUTIC DELIVERY OF CO 557 Water-solubilizing group e.g., sugar, phosphate, carboxylate Aminoacid or peptide Targeting vector Ionizable or labile ligand Pharmacophore L4 L5
L3 L2 c M Outer coordination sphere ADME and PK control Inner coordination sphere CO release control
Conceptual model of a druglike CORM. The number of CO ligands can be greater than 1 and the other ancillary ligands (L2–L6) can be the same or different. Distal substituents are just given as possible examples. ADME, administration, distribution, metabolism, and excretion; PK, pharmacokinetics. pharmacological [absorption, distribution, metabolism, and excretion (ADME), pharmacokinetics (PK)], and toxicological properties of the prodrug. A wealth of data obtained from the broad screening of several hundred MCCs helped defining the inner sphere, whereas available tools of medicinal chemistry, metal-based drugs [119, 120], and radiopharmaceutical carbonyls, [M I (CO)
3 L 3 ] z ± (M = 99m
Tc and Re) [64, 121], provide much of the information necessary to build the drug sphere.
Most importantly, the process requires the a priori identification of the disease to be treated by CO. The nature of the tissue, the specificities of the disease, and even of the type of drug administration are fundamental to define the types of functions of the outer sphere that guarantee adequate targeting of the drug. We have already mentioned the liver targeting obtained with ALF492 (Fig. 41.2) [82]. The choice of the tissues to treat helps in selecting the best local triggers that promote the cleavage of the inner sphere and initiate the release of CO. Among other factors, this chemistry is dependent on the metal, the choice of which is of paramount importance. The toxicity of a complex does not depend solely on the metal, rather on the whole molecule. However, CORMs decompose, generating new metal fragments that may accumulate and/or become toxic. Our contention is that the choices are reduced to Mn(I), Re(I–II), Fe(0), Fe(II), Ru(II), and Mo(0–IV) [118]. If one accounts for nonchemical arguments and prejudice against metals often met even in the drug-development industry, this choice may be even narrower. In general, inner spheres should obey the 18- electron rule although the paramagnetic, 17-electron Re(II) CORMs developed by Zobi et al. [104] really shattered this paradigm. Any ancillary ligand is allowed, in principle, unless open toxicity forbids. The combination of such ligands in the inner sphere dictates CORM stability, trigger type, and CO-release profile. Here is where the organometallic chemist has the most exciting playground because the number of options is very large, not necessarily obvious, and eventually redundant. At Alfama, the medical indication chosen was the treatment (rescue) of acute liver failure induced by acetaminophen (Tylenol ® , paracetamol), where the activity of CO gas had already been demonstrated in animal models (Fig. 41.7) [51]. The treatment is restricted to the critical care room where injectable drugs are preferred [122]. The goal was to obtain a CORM providing total liver rescue within a wide time window, enabling later onset of treatment after intoxication challenge, thus outperforming the present treatment with N-acetylcysteine. The metal chosen was Mo(0), the ligand sphere was tested for a very large variety of combinations of CO with C, N, O, S, and P ligands in the Mo(CO) 5 L, Mo(CO) 4 L 2 , and Mo(CO) 3 L 3 stoichiometries. The ligands were decorated with functions that provide for enhanced solubility in water at physiological pH but had also tunable lipophilicity to gain liver specificity and avoid rapid excretion. 558 METAL CARBONYLS FOR CO-BASED THERAPIES: CHALLENGES AND SUCCESSES Mo CO
N C HO 2 C C CO OC C N C HO 2 C N C CO 2 H R 1 R 2 R 2 R 1 R 2 R 1 Figure 41.7 General structure of the family of CORMs selected for treatment of acute liver failure, induced by acetaminophen (Tylenol ® ,
1 , R
2 = CH
3 [123].
New compounds were pushed along a pass/fail multistep testing protocol that followed the order: water solubility; stability in water, plasma, and whole blood; and hemolysis. Each failure resulted in the compound being discarded. Those passing the hemolysis test entered cytotoxicity tests in several cell lines, and, when positively cleared, entered the in vivo animal model test. The best readouts of acute liver failure rescue were obtained for the family of isocyanide derivatives Mo(CO) 3
1 R 2 COOH) 3 . Such readouts include the inhibition of liver cell death measured by the dose-dependent reduction of the levels of the transaminase (ALT) and confirmed by liver histology, and total survival in a lethal model of acetaminophen poisoning [123]. Variation of the R 1 and R 2 substituents allowed the fine tuning of the biological activity and ADME properties. None of these compounds carbonylated whole blood in vitro. However, in vivo, low, sustained levels of COHb appeared in circulation, suggesting metabolic activation. Indeed, both rat and human liver microsomes accelerate the release of CO from the best compounds, relative to release in buffer. The distribution of CO in the tissues revealed a very high level of liver specificity that correlated directly with the biological activity in vivo. The PK profile of the more active compound Mo(CO) 3
2 COOH)
3 (ALF794) was very much like that of a regular drug and straightforward to determine by standard HPLC methods. Last, but not least, in vivo acute toxicity is unusually low [maximum tolerated dose (MTD) > 1000 mg/kg]. ALF794 has, therefore, a full pharmacologically acceptable, druglike profile and its development is being pursued. 41.4 FINAL REMARKS AND PERSPECTIVES The biology and chemistry that was described in the sections above shows that CO is a very powerful therapeutic principle and that metal carbonyls can mediate the delivery of CO in a pharmacologically useful manner. However, to meet this goal, the design of CORMs requires inputs beyond simple CO-release kinetics and cell toxicity assays. The pharmaceutical CORMs of the future must have properties that are similar to those common to any other drug or prodrug in terms of physical–chemical and biocompatibility properties, toxicological safety, efficacy, potency, and general pharmacological ADME and PK properties. If we can prepare CORMs that specifically reach the disease targets and release CO there, we will certainly make many useful drugs. Moreover, the production of CORMs should be much more cost effective than that of most modern therapeutic approaches— such as recombinant proteins, antibodies, gene therapy, or, more recently, cell therapy. However, in order to achieve this end, it is necessary that the organometallic chemists learn to speak beyond their usual jargon and collaborate with other specialists, namely, pharmacologists, drug developers, physiologists, and medical doctors who can show them the profiles and properties of the drugs that are needed. These 11 years of CORM development have taken us over a number of barriers and it is now clear that making metal carbonyl complexes with real pharmacological application is not only possible but will be part of the future of organometallic chemistry. Download 11.05 Mb. Do'stlaringiz bilan baham: 
|