ERYTHRO-HYDROXYNONYLADENINE ANALOGS WITH ENHANCED LIPOPHILIC AND ANTI-ISCHEMIA TRAITS
BACKGROUND OF THE INVENTION
This invention is in the fields of chemistry and pharmacology, and relates to drugs that can inhibit an enzyme called adenosine deaminase (ADA, also known as adenosine aminohydrolase) . ADA-inhibiting drugs can be used to reduce the enzymatic degradation of chemotherapeutic and anti-viral drugs, thereby increasing the therapeutic utility of such drugs. As disclosed herein, ADA-inhibiting drugs can also be used to protect heart muscle and brain tissue against damage caused by ischemia (inadequate blood flow) or hypoxia (inadequate oxygen supply) , as occurs during stroke, cardiac arrest, heart attack, asphyxiation, and various other crises.
The mammalian enzyme called adenosine deaminase (ADA) , which is designated E.C.3.5.4.4 under the international enzyme classification system, converts adenosine into inosine by removing an amine group from the #6 carbon in the two-ring adenyl structure of adenosine. ADA can also degrade a number of other molecules, including several nucleoside analogs that are used in cancer chemotherapy or for anti-viral therapy. Since ADA is known to reduce the therapeutic utility of various drugs used to treat cancer and viral infections, a substantial amount of work has been done to develop drugs which function as ADA inhibitors. The ADA inhibitor drugs can be used as adjuncts (i.e., as secondary agents to increase the effectiveness of a primary drug) to prolong the metabolic half-lives of therapeutic drugs during cancer or anti-viral chemotherapy. ADA inhibitors can also be used to artificially create ADA deficiencies, whic are of interest to some researchers.
A compound called erythro-hydroxynonyladenine (abbreviated as EHNA, usually pronounced as "eenah") is a relatively mild ADA inhibitor, and is of particular interest herein. EHNA is a stereoisomer with the following chemical structure, which shows the numbering of the carbon atoms in the nonyl "side chain" (i.e., in the erythro-hydroxy-nonyl straight chain which is attached to the double-ringed adenyl group) :
The "erythro-" pref x nd cates a certain stereoisomeric arrangement of the atoms attached to the #2 and #3 carbon atoms in the nonyl side chain. Both the #2 and #3 carbon atoms are chiral atoms (i.e., carbon atoms with four different groups attached to them, so that the spatial arrangement of the four groups will have either a dextrorotatory (or right, or +) or levorotatory (or sinister, or -) configuration, depending on how they rotate polarized light passing through an aqueous solution of a purified stereoisomer. These rotations are abbreviated as D/L, R/S, or +/-. Other purified steroeisomers having the same atoms as erythro compounds, but in a different stereoisomeric arrangement, are referred to as "threo-" compounds.
A "racemic" mixture of EHNA (i.e., a mixture containing both D/R/+ and L/S/- isomers) was identified as an ADA inhibitor in Schaeffer and Schwender 1974. Subsequent reports, including Bastian et al 1981 and Baker and Hawkins 1982, identified the (+)-2S,3R isomer (the erythro isomer) as the most potent ADA inhibitor from among the various hydroxynonyladenine isomers. Various analogs and derivatives of EHNA have been described in reports such as Harriman et al 1992. Those other analogs are not related to the EHNA analogs described herein.
EHNA apparently is metabolized and cleared from the mammalian bloodstream fairly rapidly (McConnell et al 1980; Lambe and Nelson 1982). In addition, EHNA's activity as an ADA inhibitor drug is not as strong as various other ADA inhibitor drugs, including deoxycoformycin (dCF, also known as Pentostatin) . The so-called "Ki" value of dCF (i.e., the negative log value of a molar concentration of dCF required to inactivate a standardized quantity of ADA) is very low, about
2.5 x 10"12, which indicates that dCF binds to ADA very tightly; dCF is sometimes called a "suicide inhibitor," which indicates that the binding between dCF and ADA is effectively irreversible, and neither molecule can be regenerated. This
process of irreversible binding is also referred to as "poisoning" an enzyme.
Because of its potency as an ADA inhibitor, dCF was tested by several research teams to determine whether it can be used therapeutically. Although dCF reportedly provided some beneficial activity in cardiovascular models (e.g., Dorheim et al 1991), neuroprotection (e.g., Phillis and O'Regan 1989), and cancer therapy, it was found to cause serious toxic side effects (e.g., O'Dwyer et al 1986). Therefore, attention subsequently returned to EHNA and various other milder or "softer" ADA inhibitors, in the hope that the milder ADA inhibitors would have fewer side effects and would be less toxic. The Ki value of (±)-EHNA is about 6 x 10"9, which indicates that EHNA binds to ADA about a thousand times less tightly than dCF. This invention discloses a class of compounds in which a hydrogen atom coupled to one of the "far end" carbon atoms (i.e., the #8 or #9 carbon atoms) is replaced by a hydroxyl group, to create a #8 or #9 hydroxylated EHNA, or by various other types of moieties to create other #8 or #9 analogs (including analogs which are more soluble in lipids than the hydroxylated analogs, and which have shown better therapeutic utility against ischemia) . As discussed below, the analogs that are of interest herein have both (1) a binding affinity for the ADA enzyme which is in the desired range, with a Ki value between about 10"7 and about 10'10, and (2) additional properties which render them substantially more useful and beneficial than unmodified EHNA in protecting heart tissue and/or brain tissue against damage caused by ischemia (inadequate blood flow) or hypoxia (inadequate oxygen supply), as occurs during stroke, heart attack, cardiac arrest, asphyxiation, and various other types of crises or conditions.
The utility of ADA-inhibiting drugs in protecting heart muscle or brain tissue against ischemic or hypoxic damage has not been widely recognized prior to this invention. Instead, nearly all research on ADA inhibitors has focused on their potential ability, as adjuncts, to slow the degradation of anti¬ cancer or anti-viral drugs by the ADA enzyme, in order to increase the efficacy of such anti-cancer or anti-viral drugs.
However, as disclosed herein, various EHNA analogs have
also been shown to provide substantial protection for the heart against ischemic or hypoxic damage, as would occur during a heart attack, cardiac arrest, or surgery requiring cardiopulmonary bypass. It is also believed that at least some of these analogs may also provide substantial protection for brain tissue against ischemic or hypoxic damage due to stroke, cardiac arrest, asphyxiation, etc.
It should be noted that the ADA enzyme acts inside cells. Despite the fact that this enzyme activity can affect the quantity of adenosine released by a cell, which will react with adenosine receptors on other cells, the fact remains that the ADA enzyme, itself, functions almost exclusively inside cells. Therefore, an ADA inhibitor drug must enter mammalian cells in order to function properly, and its efficacy will depend to a large extent on how readily it can be taken into cells.
One object of this invention is to disclose a class of analogs of EHNA which have been modified at the #8 or #9 carbon atoms on the side chain, in a manner which substantially improves the therapeutic efficacy of these analogs against ischemic or hypoxic damage to heart muscle or brain tissue, compared to either unmodified or hydroxylated EHNA.
Another object of this invention is to disclose a class of analogs of EHNA which have been modified at the #8 or #9 carbon atoms on the side chain, in a manner which provides various therapeutic advantages for these analogs while retaining a binding affinity for the ADA enzyme which is in the desired range (preferably with a Ki value between about 10'7 and about 10'10) . This status as a relatively mild and reversible ADA inhibitor allows such analogs to inhibit ADA activity at therapeutically effective levels, without irreversibly inactivating (poisoning) the ADA enzyme and increasing the risk of toxic side effects.
Another object of this invention is to disclose synthetic reagents and methods that can be used to create pharmacologically valuable analogs of EHNA which contain hydroxyl, halide, acid, ester, ether, amine, amide, imide, azide, nitrile, or other moieties at various controllable locations in the nonyl side chain, and particularly containing novel moieties coupled to the #8 or #9 carbon atoms in the side
chain .
Another object of this invention is to disclose a new set of EHNA analogs which can be used to slow down the degradation by the ADA enzyme of certain types of anti-cancer, anti-viral, or other therapeutic drugs.
These and other objects of the invention will become more clear and apparent from the following summary, detailed description, and examples.
SUMMARY OF THE INVENTION
Analogs of erythro-hydroxynonyladenine (EHNA) are disclosed which have been modified by bonding various types of moieties to the #8 or #9 carbon atoms in the "side chain" portion of the molecule (i.e., the 9-carbon erythro-hydroxynonyl straight chain portion, which is attached to an adenosine ring structure) . Analogs of EHNA with various moieties coupled to the #8 or #9 positions on the side chain have been discovered to have new and unexpected value as therapeutic drugs, as described below. In an early set of tests, one of the hydroxylated EHNA analogs described below (designated herein as 9-OH-EHNA, where the hydroxyl moiety was coupled to the #9 carbon atom on the EHNA side chain) was discovered to have a significant advantage in protecting heart muscle against ischemic damage, compared to unmodified EHNA. This assay is described in Example 4.
Based on that early finding, subsequent research on other newly-synthesized analogs of EHNA showed that several analogs had major therapeutic advantages, not just over unmodified EHNA, but also compared to the hydroxylated analogs. This subsequent research, which used a somewhat different heart perfusion technique as described in Example 6, indicated that the 9-OH- EHNA analog did not perform substantially better than unmodified EHNA in most parameters that were used to measure heart protection; however, by the time that was confirmed, it had been discovered that several other EHNA analogs showed a marked improvement over either the unmodified or the hydroxylated form of EHNA, and subsequent research was devoted to those other analogs while 9-OH-EHNA was disregarded and not tested further. The analogs which have showed the best combinations of
traits in the research done to date are believed to be more lipophilic (i.e., more soluble in lipids and other fatty, non- polar fluids, and less soluble in water) than the hydroxylated analogs. These lipophilic analogs are disclosed below, along with chemical methods for synthesizing them and any other desired analog of EHNA which has been modified at the #8 or #9 carbon atom of the side chain.
Any such analog which is synthesized as described herein can be screened, using assays as described below or otherwise known to those skilled in the art, to determine whether a particular analog has a desired combination of traits as described herein. These traits primarily involve (1) non-toxic potency as a mild, reversible ADA inhibitor; (2) a desirable level of lipophilicity, to increase cellular uptake; and (3) therapeutic utility in reducing ischemic or hypoxic tissue damage, or in reducing ADA degradation of anti-cancer, anti¬ viral, or other drugs.
BRIEF DESCRIPTION OF THE DRAWINGS FIGURE 1 depicts a series of chemical reactions used to create 9 '-hydroxy(+)-EHNA, designated as Compound [10], an intermediate compound that was used to create other subsequent analogs of EHNA that are more lipophilic.
FIGURE 2 depicts the reactions that were used to create 8 ■- hydroxy(+) -EHNA, designated as Compound [23].
FIGURE 3 depicts the reactions that were used to create 8 • ,9 '-dihydroxy(+)-EHNA, designated as Compound [14] .
FIGURE 4 depicts the reactions that were used (see Example 5) to create analogs of EHNA that contained various non-hydroxy moieties bonded to the #9 carbon atom.
FIGURES 5, 6, AND 7 are bar graphs showing that 9-chloro- EHNA (compound [29]) and 9-phthalimido-EHNA (compound [27] provided better protection for heart muscle against ischemic damage than either unmodified EHNA or 9-hydroxy-EHNA, in the tests described in Example 6.
DESCRIPTION OF THE PREFERRED EMBODIMENTS
This invention describeε analogs of EHNA in which the side chain (i.e., the straight chain erythro-hydroxynonyl portion,
which is attached to an adenyl ring structure) has been chemically modified by bonding certain types of chemical groups (moieties) to it. The preferred moieties provide the resulting analogs with certain pharmacological and therapeutic activities, which are substantially improved compared to unmodified EHNA. In one embodiment of this invention, a hydroxyl group can be bonded to either the #8 or #9 carbon atom on the EHNA side chain, to generate hydroxylated analogs referred to herein as 8- OH-EHNA or 9-OH-EHNA (or a di-hydroxylated analog with hydroxyl moieties bonded to both carbon atoms) . These hydroxylated analogs were synthesized as described in Examples 1 and 2, and were tested and discovered to have a slight but significant advantage, compared to unmodified EHNA, in protecting heart muscle against ischemic damage, using a laboratory model with intact perfused hearts taken from rats. The limited tests carried using these hydroxylated analogs indicated that the 9- OH-EHNA analog was preferable to the 8-OH-EHNA analog.
Based on those initial findings, other analogs of EHNA were synthesized and tested in various ways. Some of these analogs, synthesized in the laboratories of Prof. Elie Abushanab at the
University of Rhode Island, used a benzyl-protected precursor of the 9-OH-EHNA analog (compound [9] in the examples) as a starting reagent. Other analogs, synthesized in the laboratories of cypros Pharmaceutical Corporation in Carlsbad, California, used surplus quantities of the de-protected 9-OH-EHNA analog as a starting reagent. Research on these analogs showed that some of them had major therapeutic advantages over both unmodified EHNA, and over the 9-OH-EHNA analog as well.
The analogs which have showed the best combinations of traits in the research completed to date are believed to be more lipophilic (i.e., more soluble in lipids and in other fatty, non-polar fluids) and less soluble in water than the hydroxylated 9-OH-EHNA analog. These analogs, which include 9- chloro-EHNA and 9-phthalimido-EHNA, are disclosed below. Alternately, the chemical synthesis methods disclosed herein
(and other methods known to those skilled in the art of chemical synthesis) can be used to create other analogs of EHNA which have been modified by the addition of nearly any type of moiety at the #8 or #9 carbon atom of the side chain. After synthesis,
any such analog can be screened and tested, using the assays described herein or other assays known to those skilled in the art of biomedical testing, to determine whether any particular analog has a desired combination of traits. Three traits are of primary interest herein. One such trait is, an EHNA analog intended for therapeutic use in humans should have suitable potency as a reversible ADA inhibitor, preferably with a binding affinity for ADA that provides a Ki value in the range of about 10'7 to about 10'10. EHNA analogs with Ki values in this desired range can inhibit the ADA enzyme reversibly, without permanently poisoning enzyme molecules, and without causing the types of toxic side effects that have been caused in some patients or test animals by highly potent "suicide inhibitors" such as deoxycoformycin. The Ki value for any analog of EHNA can be determined by assays such as the spectrophotometric assay described in Harriman et al 1992 (described in more detail in Example 3) .
A second desirable trait for an EHNA analog intended for therapeutic use in humans involves a suitable level of lipid solubility (also called lipophilicity). In general, lipophilic drugs tend to be taken into cells more readily and in greater quantities than hydrophilic drugs, because of two chemical factors. First, like droplets of oil in water, lipophilic drugs generate surface tension between themselves and water molecules; they are not "comfortable" floating in the watery liquid that surrounds cells, and they seek configurations that minimize the area of their surface contact with water. This surface tension causes lipophilic molecules to attach and adhere to any lipophilic surfaces they encounter, including the surfaces of cells, to minimize the area of their surface contact with water.
And second, since the membranes of mammalian cells are themselves made of lipid bilayers, molecules that are soluble in lipids tend to dissolve in and move into cell membranes. This is a major step in cellular uptake, and it does not require an active transport mechanism to help these drugs cross a cell membrane and enter a cell.
Both of these factors tend to promote and increase cellular intake of lipophilic drugs, compared to drugs which are hydrophilic and/or highly charged. This is assumed and believed
to be a potentially important factor for ADA-inhibiting EHNA analogs, since EHNA and its analogs act via molecular mechanisms that occur inside cells.
However, the desired trait of high intake into cells does not increase in an unlimited manner as hydrophobicity increases; extremely hydrophobic drugs can be difficult to administer to a patient via conventional routes such as injection or ingestion, and once inside the body, they often tend to sequester themselves in lipid vesicles or globules, or they tend to cling to various membranes, plaque deposits, or particulates, either in the intestines or inside blood vessels.
For these reasons, a moderately high but not extreme level of lipophilicity (hydrophobicity) is often preferred for therapeutic drugs which must be taken inside cells in order to be fully effective. Accordingly, now that it has been discovered that other, more lipophilic analogs provide better protection against ischemic damage than the hydroxylated 9-OH-EHNA analog, other EHNA analogs with other lipophilic moieties have been and will be synthesized and tested, using the methods and assays disclosed herein, to determine the optimal lipophilic values and moieties that provide the best therapeutic benefits as described herein.
The lipophilic level of an EHNA analog with any candidate moiety can be assessed using a dual-solvent assay, such as the widely used octanol-water partition assay. The partition coefficient is usually referred to as P , where "o/w" refers to oil and water; this value is usually referred to by a base 10 logarithm, comparable to pH values; a high log P0/w value indicates a high degree of oil solubility, and a low (or negative) value refers to a high degree of water solubility.
Since lipophilicity involves chemical reactions rather than complex biological functions, octanol-water partition coefficients can be estimated using commercially available computer software (such as the ACD/LogP software program, sold by Advanced Chemistry Development, Inc. of Toronto, Canada) . This software was used to calculate the octanol-water partition coefficients listed in Table 1. A description of the methods used to calculate and estimate partition coefficients, based on their chemical structures, is described in Bodor et al 1989.
Accordingly, without tying or limiting this invention to any particular theory, it is generally believed that EHNA analogs which have a level of lipophilic solubility close to or greater than the chloro- or phthalimido- analogs described below are likely to be preferable to hydroxylated analogs and other more hydrophilic (water-soluble) analogs, for protecting against ischemic or hypoxic damage or for increasing the half-lives of drugs that are degraded by the ADA enzyme.
The third primary desirable trait for an EHNA analog intended for use as described herein involves therapeutic utility, either in mammalian patients or in laboratory studies which provide good models of therapeutic utility against certain types of cell damage or drug degradation. Currently, the two primary and most urgent uses for the EHNA analogs described herein are: (a) for reducing the amount of damage caused by ischemia or hypoxia in vulnerable tissues, especially in heart muscle or brain tissue; and (b) for prolonging the half-lives and increasing the therapeutic benefits of drugs that are being used to treat patients suffering from cancer, viral infection, or other disease conditions, by reducing the rate of degradation of such drugs by the ADA enzyme. Various other uses may also be currently available, or they may be discovered in the future after these compounds are announced and made available to scientific and medical researchers. This invention also discloses a method of synthesizing analogs of EHNA in which a moiety (such as a hydroxy group or halide atom) has been added to the side chain. This method comprises the following steps: a. reacting an epoxide reagent having a desired chiral orientation with an alkyl halide reagent having an unsaturated bond between two selected carbon atoms, under conditions which cause said reagents to create an unsaturated aliphatic compound comprising a first portion having a desired chiral orientation and a second portion having an unsaturated bond; b. reacting the unsaturated aliphatic compound with at least one third reagent, under conditions which cause the third reagent to modify the unsaturated aliphatic compound by adding at least one hydroxyl group (or other desired group or atom) to at least one of the carbon atoms involved in the unsaturated
bond, thereby creating a hydroxylated (or otherwise modified) saturated aliphatic compound; c. if a hydroxyl group was bonded to the EHNA side chain, the hydroxyl group can be replaced by or converted into a different moiety, as discussed herein.
When all of the initial steps have been completed, any additional processing is carried out to complete the synthesis of the desired analog, such as removal of benzyl or other protective groups; such groups are commonly used during synthesis to prevent undesired reactions involving a protected constituent. The final de-protected analog is then purified by any suitable means, such as chromatography, gel electrophoresis, or isoelectric focusing.
The particular processing and purification steps used to create a specific analog will depend on the exact molecular structure of the desired analog. Such steps are within the ordinary skill in the art, and various examples of suitable reagents and reactions which can be used for such purposes are described below. In the Examples and figures, each major starting reagent or intermediate is referred to by a bracketed number. For convenience, that bracketed number is then used to refer to that compound in subsequent processing steps.
Example 1 (below) describes in detail the reagents and reactions used to synthesize the EHNA analog which has a hydroxyl group bonded to the #9 carbon atom on the side chain. This compound, referred to herein as 9-hydroxy-EHNA or as 9-0H- EHNA, is designated as Compound [10]. Its full chemical name is 9-[2 (S) ,9-dihydroxy-3(R)-nonyl]adenine, and its synthesis is depicted in FIG. 1. The full chemical name includes the term
"dihydroxy" because this analog has two hydroxy groups on the side chain. One hydroxy group is attached to the #2 chiral carbon atom, in the same "S" orientation that occurs in unmodified EHNA; the other hydroxy group was added to the #9 carbon atom, to create the 9-OH-EHNA analog.
Example 2 describes the reagents and reactions used to synthesize the EHNA analog which has a hydroxyl group bonded to the #8 carbon atom on the side chain. This compound, referred to herein as 8-hydroxy-EHNA or as 8-OH-EHNA, is designated as
Compound [23]. Its full chemical name is 9-[2(S) ,8-dihydroxy- 3(R)-nonyl]adenine, and its synthesis is depicted in FIG. 2.
Example 2 also describes the synthesis of Compound [14], which is a di-hydroxylated EHNA analog with hydroxy groups added to both the 8' and 9' carbon atoms (in addition to the standard hydroxy group on the #2 carbon atom) . Its synthesis is depicted in FIG. 3.
All three of these hydroxylated analogs were shown to inhibit ADA activity in a reversible manner, with Ki values in the desired range, as described in Example 3. They were then tested in certain types of in vitro tissue tests involving perfused hearts, as described in Example 4. In both the ADA inhibition tests and the heart muscle tests, the 9-OH analog performed slightly better than either the 8-OH analog or the 8,9-dihydroxy analog. The 9-OH analog was the strongest binding agent of the three, with a Ki value of 3.8 x 10*9; the 8-OH analog was the weakest, with a Ki value of 15.8 x 10"9, while the 8,9-dihydroxy analog had an intermediate strength, with a Ki value of 6.4 x 10"9. In addition, the 9-OH analog also displayed a useful protective effect in the heart muscle reperfusion assays that was not shown at a significant level by the 8-OH analog, involving a reduction of undesired muscle stiffness.
Accordingly, the 9-OH analog was identified as the preferred candidate, and it was used as a starting reagent for synthesizing other analogs (more precisely, a benzyl-protected precursor of the 9-OH analog, designated as Compound [9] in Example 1, was used; the benzyl group protected the hydroxy group attached to the #2 carbon atom) . If desired, either the 8- hydroxy or the 8,9-dihydroxy analogs could be used instead, to create comparable lipophilic analogs with any desired moieties coupled to the #8 carbon atom instead of (or in addition to) the #9 carbon atom, using the same general procedures and reagents described herein.
This general procedure, using hydroxylated EHNA analogs as intermediates for synthesizing other analogs, can be used, if desired, to provide a general method for synthesizing any other desired type of EHNA analog, with any suitable type of moiety bonded to the side chain. However, since that is a relatively circuitous route which arose only because of the stepwise nature
of the research described herein, synthetic chemists will recognize that other, non-hydroxyl moieties can be added directly to the side chain, without having to go through hydroxylated intermediates, by using suitable alternate reagents in one or more of the reactions that were used to create the hydroxyl analogs. Such more direct methods of synthesis will likely provide better yields and require fewer purification steps, and will be generally be preferable to the indirect-via- hydroxyl method used during the initial research described herein.
Nevertheless, the hydroxyl route should be recognized as a potentially useful route for synthesizing a large number of analogs that can be generated by substituting or derivatizing hydroxyl groups, such as carboxylic acid groups, esters, and ethers, all of which can be created using techniques such as disclosed in the examples, or other techniques known to those skilled in the art of chemical synthesis. In addition, hydroxide groups can be converted to numerous other groups by known methods. As one example, a hydroxyl group can be converted into an azide group by reacting the hydroxyl with p-toluenesulfonyl chloride (TsCl) to create an 0-tosyl group (abbreviated as OTs in the figures; tosyl refers to toluenesulfonyl) , then reacting the O-tosyl compound with sodium azide (NaN3) , which displaces the O-tosyl group and leaves an N3 group attached to the carbon chain. As a second example, a hydroxyl group can be converted to a halide group (such as a chlorine, fluorine, bromine, or iodine atom) by methods such as in Example 5 relating to Compounds [28] and [29].
The analogs that can be synthesized as described herein include, but are not limited to, analogs in which the chemical moiety bonded to the #8 or #9 carbon atom on the nonyl side chain consists of a halide; a nitrogen-containing moiety such aε an amine, amide, azide, imide, or lactam; a carboxylic acid or salt thereof; or a moiety which is coupled to the #8 or #9 carbon atom via an ester or ether linkage. In order to be covered by the claims herein, any such analog must display the traits that can make such analogs therapeutically useful as disclosed herein (i.e., the resulting analog should have a Ki within the desired range of about 10'7 to about 10"10; it must be
pharmacologically acceptable; and it must be therapeutically useful in protecting tissue against ischemic or hypoxic damage, or as an adjunct with one or more anti-cancer, anti-viral, or other drugs) . Epoxide [1] was synthesized as described in Abushanab et al 1984 and 1988. It controls the orientation of the substituents on the two chiral carbon atoms in the final EHNA analog, which are provided by the #3 and #4 carbons in the epoxide. To synthesize different stereoisomers of any of the EHNA analogs discussed herein, different epoxide stereoisomers having any desired chiral configuration can be used as the starting reagent.
The benzyl group (-CH2C6H5) which was attached via an oxygen atom to the #3 carbon in the starting epoxide served as a protective group for the oxygen atom. In the final step of synthesis of each of the hydroxylated EHNA analogs, the benzyl group was displaced by hydrogen to create a hydroxyl group on the #2 carbon of the side chain. That #2 hydroxyl group is part of the normal EHNA molecule. If desired, that hydroxyl group can be eliminated by using a starting epoxide without a protected oxygen atom, or it can be modified during synthesis to provide a halide, carboxylic, ester, ether, azide, or other group, as described above. If a moiety is desired at the #1 carbon atom in the final EHNA analog, it can be provided by using a starting epoxide having the desired moiety or a precursor at the #4 carbon atom of the epoxide.
The synthesis reactions described herein also offer a method of derivatizing (i.e., bonding moieties to) the #4, #5, #6, or #7 carbon atoms on the nonyl side chain. In the synthetic method used herein, those carbon atoms were provided by the reagent 1-pentenylmagnesium bromide, which has a structure as shown in FIG. 1 in the reaction that converts epoxide [1] into compound [2] . The 1-pentenyl notation indicates that the unsaturated double bond is positioned between the #1 and #2 carbon atoms in 1-pentenylmagnesium bromide; those carbon atoms ultimately become the #8 and #9 carbon atoms in the EHNA analogs of this invention. The unsaturated carbon atoms in the double- bonded pentenyl compound became attachment points for hydroxyl groups during the reaction which converted compound [8] into
compound [9]. Hydroxyl groups were added to both of the unsaturated carbons, and the compound having the hydroxyl moiety at the desired location was subsequently purified. In an alternate approach, the double bond supplied by the pentenyl compound was converted into an epoxide intermediate, as shown in FIG. 2 in the reaction which generated compound [15].
Using either of these approaches, the location of a hydroxyl (or other desired) group on the side chain of an EHNA analog can be controlled by using a pentenylmagnesium bromide (or similar) compound having a double bond in any desired location. A 2-pentenyl compound would have a double bond between its #2 and #3 carbon atoms, which become the #8 and #7 carbon atoms in the final EHNA analog. A 3-pentenyl reagent (having a double bond between its #3 and #4 carbon atoms) would generate hydroxyl groups attached to the #7 or #6 carbons in the EHNA analog.
FIG. 2 also depicts a halogenated analog, Compound [21] . In Compound [21], the halogen (chlorine) atom was substituted into the adenine ring structure. Although that chlorine atom was substituted by an amine group during the synthesis of compound [22], that particular reaction could be omitted if desired, so that the halogen moiety would remain after removal of the benzyl protective group.
The method used to create the adenyl structure in the EHNA analogs described herein offers a general method for making various changes in the adenine group. The adenyl structure was provided by supplying and then manipulating a heterocyclic compound, 5-amino-4, 6-dichloropyrimidine (ADCP) , which is shown in FIG. 1 in the reaction that generated compound [6]; this same reagent was also used to generate compound [20] shown in FIG. 2.
The ADCP was coupled to the side chain by displacing one of the chlorine atoms on the ADCP with an amine group that was coupled to the side chain. The five-member ring in the adenine structure was then closed by forming a carbon bond between two proximal nitrogen atoms.
If desired, alternate heterocyclic reagents could be used instead of ADCP, to create analogs of EHNA with modified adenine structures, either as moieties attached to one of the rings, or as differing atoms incorporated into either of the rings.
Cristalli et al 1988 and 1991 report that certain analogues of EHNA with modified adenine structures (such as a 3-deaza-EHNA derivative) are active as ADA inhibitors. Such modifications to the adenyl structure could be incorporated into the analogs of this invention, which have modified side chains.
As mentioned above, all three of the hydroxylated EHNA analogs which were tested for ADA inhibition (as described in Example 4) were shown to be active. The 9-hydroxy analog (compound [10]) was the strongest binding agent of the three, with a Ki value of 3.8 x 10'9; the 8,9-dihydroxy analog
(compound [14]) was the weakest, with a Ki value of 15.8 x 10'9, while the 8-OH analog (compound [23]) had an intermediate strength, with a Ki value of 6.4 x 10"9.
All three of these Ki values are within a desired range, which covers about 10'7 to about lO'10. At one end of the desired range, ADA inhibitors having Ki values lower than about 10'10 run the risk of "poisoning" the enzyme by binding to it so tightly that the reaction is, for all practical purposes, irreversible. At the other end of the desired range, ADA inhibitors having Ki values higher than about 10'7 tend to be insufficiently potent to accomplish the desired level of ADA inhibition; they would need to be administered in relatively large quantities, and even in large quantities they might not be adequately potent.
The desired range of Ki values is relatively broad, since candidate compounds can be administered to a patient in any desired quantity, by various routes. An analog having a Ki value in the range of about 10'9 should be administered in relatively low dosages, such as up to about 10 milligrams per kilogram of body weight per day if injected intravenously, and up to about 50 mg/kg/day if administered orally. A less potent analog having a Ki value in the range of about 10"7 could be administered in higher dosages, such as up to about 25 mg/kg/day if administered orally or injected in response to a major crisis, or up to 20 mg/kg/day if injected intravenously. Since the metabolic problems caused by ADA deficiency tend to accumulate slowly, short-term dosages can be rather large.
After testing for ADA inhibition, the hydroxylated EHNA analogs were tested for protection against ischemic damage to hearts, using procedures described in Example 4. Briefly, these
tests involved hearts that were removed from laboratory rats, hooked up to perfusion equipment and given electrical stimulation to sustain the heartbeat, treated with the candidate drugs, subjected to a period of ischemia, and then reperfused, to evaluate how well the hearts could recover their pumping functions. In these initial assays, 9-OH-EHNA provided a higher level of protection than unmodified EHNA in a particular parameter involving reduction of unwanted muscle stiffness after ischemia. In subsequent assays using somewhat different heart preparations, described in Example 6, the advantages of 9-OH- EHNA were not as significant compared to unmodified EHNA; however, other analogs had been created by the time those subsequent assays were carried out, and the results of those other assays clearly indicated that the other preferred analogs were substantially better than either 9-OH-EHNA or unmodified EHNA, in protecting hearts against ischemic or hypoxic damage.
Example 5 describes, and Figure 4 depicts, the synthesis of several other analogs, using the benzyl-protected precursor (compound [9]) of the 9-OH analog as a starting reagent. These analogs include two relatively lipophilic analogs, referred to herein as 9-chloro-EHNA (Compound [29]) and 9-phthalimido-EHNA (Compound [26]). These two analogs have shown the best therapeutic results observed to date, in protecting both heart muscle and brain tissue against ischemic damage. Some additional analogs were also created by Cypros
Pharmaceutical Corporation, using the de-protected 9-OH-EHNA analog as a starting reagent, since a quantity was still available after completion of the initial biological testing. One such analog is the silicon-containing analog described in Example 5 (compound [33]). The silicon-containing moiety was chosen for two reasons: (1) calculations indicated that it had a very high lipophilicity, and could provide a potentially useful test compound to help evaluate that factor; and (2) it could be added to the #9 atom in a de-protected 9-OH-EHNA molecule, without disturbing the hydroxyl group on the #2 carbon atom of the side chain.
For convenience, the Ki values and oil/water solubility values that were gathered or calculated on the final (deprotected) analogs listed in Examples 1, 2, or 5 are compiled
in Table 1. In this table, these analogs are provided with simple names that indicate what type of modifying group was added to the side chain, and which carbon atom it was bonded to. Complete chemical names are provided in Examples 1, 2, and 5, correlated with bracketed compound numbers. It should be noted that the Ki values in Table 1 used extra-cellular ADA enzyme, and did not reflect the apparent ability of lipophilic analogs to enter cells more readily and in greater quantities.
TABLE 1
CHEMICAL DATA FOR VARIOUS EHNA ANALOGS
Compound Modifying Ki value number σroup x 10"9 (calculated)
— unmodified (±)-EHNA 6 2.60 ± 0.41
[10] 9-hydroxy-EHNA 3.8 ± 0.4 0.59 ± 0.41
[23] 8-hydroxy-EHNA 6.4 0.41 ± 0.41
[14] 8,9-dihydroxy-EHNA 15.8 ± 0.4 -1.10 ± 0.42
[25] 9-benzoyloxy-EHNA 0.2 3.50 ± 0.41
[27] 9-phthalimido-EHNA 2.3 2.86 ± 0.44
[29] 9-chloro-EHNA 3.7 2.28 ± 0.41
[31] 9-carboxymethyl-EHNA 5.0 0.95 ± 0.41
[32] 8, 9-unsaturated EHNA 2.5 2.06 ± 0.41
[33] 9-tert-BDPSi-EHNA ND 9.46 ± 0.69
Example 6 describes the testing of various analogs to evaluate their ability to protect heart muscle against ischemia. The results indicated that relatively lipophilic analogs (including 9-chloro-EHNA and 9-phthalimido-EHNA) provided substantially better protection against ischemic damage to heart muscle than either unmodified EHNA or hydroxylated EHNA.
Example 7 describes the results of cell culture tests to evaluate the ability of EHNA and several analogs both (1) to enter human cells, and (2) inhibit ADA activity inside the cells. These tests did not stress the cells, or test any analogs against ischemic damage; instead, they evaluated the ability of various analogs to reach the intracellular enzyme molecules and inhibit their activity. These tests used both red blood cells (which are easy to work with) , and human astrocytoma cells
(which are brain cells that can reproduce in cell culture; these
were used to provide an indication of whether EHNA analogs can help reduce ischemic damage in brain tissue) . The results indicated that several EHNA analogs which are more lipophilic than unmodified EHNA or 9-OH-EHNA were substantially more potent than unmodified EHNA or 9-OH-EHNA in inhibiting ADA activity inside cells, as indicated by lower IC50 values.
Example 8 describes the results of cell culture tests which used several different methods to generate ischemic damage in either brain cells or blood cells. Some of these tests used toxins such as 2-deoxyglucose or sodium azide to interfere with respiration and glycolysis. Other tests used culture media containing no free oxygen, obtained by bubbling nitrogen gas rather than oxygen gas through the cell culture media. In all of these tests, the cells were subjected to a period of oxygen deprivation (usually lasting several minutes) , then the oxygen supply was reestablished. After a brief period to allow the cells to reestablish equilibrium, selected metabolic indicators were evaluated to determine how close the cells had come to regaining their proper metabolic rates. The results indicated that some EHNA analogs (especially the lipophilic analogs) can indeed protect brain cells against ischemic damage.
Example 9 describes several assays that can be used to test EHNA analogs to quantify their ability to protect intact mammalian brain tissue against ischemia. Rather than using isolated cultured brain cells, as in Example 7, these tests use intact slices of brain tissue, from the hippocampal regions of sacrificed rats. The hippocampal region is used because it is highly vulnerable to ischemic damage, and the use of intact hippocampal slices that can still generate brain waves in response to electrical stimulation offers a better assurance of overall tissue functioning than the metabolic rates of isolated cells. These tests are currently underway. Although the final results are not yet available, it is believed (based on protection levels provided in other tests, including perfused heart tests and cultured brain cell tests) that at least some of the EHNA analogs described herein will provide a significant and therapeutic reduction in ischemic or hypoxic damage in brain tissue.
Analogs that show promising results in the hippocampal
slice tests described in Example 9 will be tested further, in in vivo tests on intact animals. These tests can use artery clamping, neck tourniquets, or other methods to induce either local or global ischemia in the brains of test animals, as described in articles such as Nellgard and Wieloch 1992, Buchan and Pulsinelli 1990, Michenfelder et al 1989, and Lanier et al 1988.
In summary, the examples, tables, and figures show that certain EHNA analogs described herein have a useful and previously unknown therapeutic benefit in protecting heart muscle and brain cells against ischemic damage. The benefits provided by the relatively lipophilic analogs exceed and surpass the benefits provided by unmodified EHNA or hydroxylated EHNA analogs. Included within the family of agents useful for the purposes described herein are any isomers (including "threo" isomers) , analogs, or salts of the compounds described herein, provided that such isomers, analogs, and salts are functionally effective as ADA inhibitors, are pharmacologically acceptable, and are therapeutically effective in either reducing ischemic or hypoxic damage or in slowing the degradation of anti-cancer, anti-viral, or other drugs. The potency of any candidate isomer, analog, or salt in inhibiting ADA activity can be tested using methods such as described in Example 3. The therapeutic efficacy of any candidate isomer, analog, or salt against ischemic or hypoxic damage can be tested using methods such as described in Examples 4, 6, and 7. The efficacy of any candidate isomer, analog, or salt in slowing the degradation of anti-cancer, anti¬ viral, or other drugs can be measured by methods known to those skilled in the art, such as by administering an EHNA analog to animals (or humans) that have received the drug of interest, and after an appropriate period (which will usually be in the range of 2 to 24 hours later, depending on the drug) , measuring the quantities of the drug that are present in the blood or tissue of the test animals (or humans, if blood tests are used) , and comparing that quantity to the quantity of the same drug in animals or humans that have not been treated with an EHNA analog.
The term "pharmacologically acceptable" embraces those
characteristics which make a drug suitable and practical for administration to humans; such compounds must be sufficiently chemically stable to have an adequate shelf life under reasonable storage conditions, and they must be physiologically acceptable when introduced into the body by a suitable route of administration. Acceptable salts can include alkali metal salts as well as addition salts of free acids or free bases. Examples of acids which are widely used to form pharmacologically acceptable acid-addition salts include inorganic acids such as hydrochloric acid, sulfuric acid and phosphoric acid, and organic acids such as maleic acid, succinic acid and citric acid. Alkali metal salts or alkaline earth metal salts could include, for example, sodium, potassium, calcium or magnesium salts. All of these salts may be prepared by conventional means. The nature of the salt is not critical, provided that it is non-toxic and does not substantially interfere with the desired activity.
The term "analog" is used herein in the conventional pharmaceutical sense, to refer to a molecule that structurally resembles a referent molecule (EHNA, 9-OH-EHNA, or 8-OH-EHNA, in this case) but which has been modified in a targeted and controlled manner to replace a specific substituent of the referent molecule with an alternate substituent, other than hydrogen (since replacement of the #9 hydroxyl group on 9-OH- EHNA with a hydrogen atom would give unmodified EHNA rather than a true analog of 9-OH-EHNA. A chemical analog requires an "offspring" type of relationship, wherein an analog is created by chemical modification of a known compound (often called a parent or referent compound) . Accordingly, the hydroxylated compounds [10], [14], and [23] are analogs of EHNA, but EHNA is not regarded as an analog of those hydroxylated compounds. A substitution which converts a known molecule into a new analog may be inserted into or coupled to any location in the molecule, such as in one of the rings in the adenyl structure, in the attached side chain, or in one of the pendant groups attached to the ring structure or side chain.
It should also be noted that analogs of 8-OH-EHNA or 9-OH- EHNA are covered by the claims herein only if they satisfy the requirements of pharmacological acceptability, ADA-inhibiting
efficacy, and therapeutic utility as disclosed herein, and only if such analogs retain an ADA inhibiting potency with a Ki value which is within the desired range of about 10'7 to about 10*10.
Administration of the compounds of this invention to humans or animals can be by any technique capable of introducing the compounds into the bloodstream, including oral administration or via intravenous or intramuscular injections. The active compound is usually administered in a pharmaceutical formulation, such as in a liquid carrier for injection, or in a capsule, tablet, or liquid form for oral ingestion. Such formulations may comprise a mixture of one or more active compounds mixed with one or more pharmaceutically acceptable carriers or diluents. When lipophilic drugs are formulated for injection, they are usually mixed with water, a buffer compound (such as a mixture of a carboxylic acid and a salt thereof) , and an organic compound having a plurality of hydroxyl groups; propylene glycol, dextran compounds, and cyclodextrin compounds are often used for such purposes.
If desired, other therapeutic agents (such as anti-cancer or anti-viral nucleoside analogs) may also be present in an injectable ingestible formulation which contains a suitable EHNA analog as described herein. A mixture of an anti-cancer or anti¬ viral nucleoside analog, with an EHNA analog, can be very useful, since the EHNA analog can prolong the half-life and efficacy of the nucleoside analog in the blood, by suppressing degradation of the nucleoside analog by ADA enzymes.
The tests completed to date indicate that different analogs showed different efficacy levels in different types of cells or tissues. For example, although the differences between 9-chloro- EHNA and 9-phthalimido-EHNA in heart muscle tests were relatively slight, the phthalimido analog tended to show significantly better efficacy in cell culture tests on brain cells. Accordingly, it is anticipated that one type of analog may be preferred for protection of heart muscle, while a different analog may be preferred for protection of brain tissue.
It should also be noted that the tissue-protecting efficacy of any analog will depend on a combination of factors, rather than on any factor in isolation. For example, as indicated in
Table 1, unmodified EHNA has a log P0/H coefficient that is roughly the same as for 9-chloro-EHNA or 9-phthalimido-EHNA, while 9-hydroxy-EHNA has a Ki value which is roughly the same as for 9-chloro-EHNA or 9-phthalimido-EHNA. However, neither unmodified EHNA nor 9-hydroxy-EHNA have the combined desirable traits of a low Ki value and a high P0/H value, as shown by 9- chloro-EHNA and 9-phthalimido-EHNA, and neither unmodified EHNA nor 9-hydroxy-EHNA show the same level of tissue protection shown by 9-chloro-EHNA and 9-phthalimido-EHNA. Accordingly, unless and until other parameters are determined to be of even greater use in predicting protective utility in intact tissue tests or in vivo tests, a combination of low Ki value and high Po/H value should be regarded as a better indicator than either trait considered by itself. At the current time, based on the tests completed to date, it is believed that an analog should have both (a) a Ki value for adenosine deaminase inhibition which is less than about 5 x 10*9, and (b) an octanol/water partition coefficient of at least about 2. Neither unmodified EHNA nor any of the hydroxyated analogs created to date have this combination of traits.
It was also noted that the 9-benzoyloxy-EHNA analog had the best combination of low Ki value and high P0/w value out of all the analogs listed in Table 1. In the future, it will be tested in both cell culture and intact tissue tests. In the assays carried out to date, it was not tested, due to concerns that the benzoyloxy group would likely be cleaved off from the EHNA molecule by various mammalian enzymes, thereby converting it into 9-OH-EHNA, which has relatively low efficacy for tissue protection.
EXAMPLES EXAMPLE 1: SYNTHESIS OF 9-OH-EHNA ANALOG This example describes how various intermediate and final compounds were synthesized. For convenience, bracketed numbers which correspond to the subheadings below and to the callout numbers in the drawings are used to refer to each compound. Unless otherwise indicated, all organic solutions were dried over anhydrous sodium sulfate, filtered, and evaporated to dryness under reduced pressure. Ratios of chromatography
solvents are expressed in v/v. Silica gel (Davison, grade H, 230-425 mesh) suitable for flash column chromatography was purchased from Fisher Scientific. A Chromatotron (centrifugally accelerated, preparative thin-layer, radial chromatograph) Model 5 7924T was used to complete various separations. The 1.0 and 2.0 mm plates used were coated with silica gel PF254 containing CaS04.
The molecular structures were confirmed in a number of ways. Elemental analyses (to compare measured values against 0 calculated values) were performed by MHW Laboratories (Phoenix, AZ) . Melting points were determined on a Buchi 535 melting point apparatus, and 1H NMR spectra were recorded on a Varian EM-390, Bruker AM-300, or Bruker AM-500 spectrometer. Optical rotations were obtained with a Perkin Elmer Model 141 digital readout 5 polarimeter. All of these analyses provided data that were within the expected limits.
COMPOUND 1: (2S.3S)-3-(benzyloxy)-1.2-epoxybutane
The starting epoxide [1] was synthesized as described in 0 Abushanab et al 1984 and 1988. This epoxide determines the orientation of the substituents on the two chiral carbon atoms in the final EHNA analog, which are provided by the #3 and #4 carbons in the epoxide. To synthesize different stereoisomers of any of the EHNA analogs discussed herein, different epoxide 5 stereoisomers having any desired chiral configuration can be used as the starting reagent. A benzyl group attached to the #3 carbon atom via an oxygen atom was used to protect the oxygen atom during synthesis.
0 COMPOUND 2: f2S.3Si-2-0-Benzyl-2.3-non-8-en-diol
A solution of epoxide [1] (2g, 11.24 mmol) in ether (50 mL) was added to a cold (-78°C) ether solution containing 1- pentenylmagnesium bromide [(22.5 mmol), prepared by reacting magnesium (0.66g, 22.5 mmol) and 5-bromo-pentene (4.11g, 22.5 5 mmol)] and 0.1 mmol of lithium tetrachlorocuprate, while stirring was continued for 1 h. The reaction was quenched with a saturated solution of NHC1 (100 mL) and extracted with ether. Pure [2] (2.67g, 96%) was obtained by silica gel column chromatography eluting with a mixture of ethyl acetate (EtOAc)
and hexane (5:95).
COMPOUND 3: (2S.3S) -2-0-Benzyl-3-Q-tosyl-2.3-nona-8-en-diol
To a stirred solution of the alcohol [2] (5.4g, 21.7 mmol) in pyridine (10 mL) was added p-toluenesulfonyl chloride (TsCl; 4.5 g, 23.9 mmol) and stirring was continued for 12 h at room temperature (RT) . The mixture was poured into water (100 mL) and extracted with CH2Cl2 (3 x 100 mL) . The combined organic solutions were then washed with cold HCl (2x50 mL) and water (2 x 100 mL) , dried (MgS04) and filtered. Removal of solvent left an oil, which was purified by silica gel chromatography eluting with EtOAc-hexane (1:50) to afford [3] (8.35g, 97%).
COMPOUND 4: (2S.3Rι-3-Azido-2-Q-benzγl-2-nona-8-en-ol Sodium azide (1.027g, 15.8 mmol) was added to a stirred solution of [2] (4.0g, 13.2 mmol) in anhydrous DMF (20 mL) .
After refluxing for 45 min, DMF was removed and pure [4] (2.61g,
96%) was obtained by silica gel column chromatography eluting with EtOAc-hexanes (5:95).
COMPOUND 5: (2S.3R)-3-Amino-2-0-Benzyl-2-nona-8-en-0l
To a stirred solution of lithium aluminum hydride (LAH, o.25g, 5.2 mmol) in anhydrous ether (50 mL) was added, dropwise, a solution of azide [4] (1 g, 3. 66 mmol) in anhydrous ether (50 L) . The reaction mixture was then heated at reflux for 2 h, cooled to RT, and excess LAH was decomposed by the careful successive dropwise addition of water (0.25 mL) , 15% NAOH (0.24 mL) , and water (0.5 mL) . Filtration, drying and evaporation of the solvent gave a pure colorless liquid [5] (1.06g, 96%) .
COMPOUND 6: 5-Amino-6-chloro-4[2 (S)-O-benzyl-3 CR) -nona-8- enyl]aminopyrimidine
5-Amino-4,6-dichloropyrimidine (ADCP, 0.39g, 2.336 mmol), N-tributylamine (n-Bu3N; 0.433g, 2.336 mmol) and [5] (0.5907g, 2.336 mmol) in anhydrous pentanol (10 mL) was heated at reflux for 48 h under an N2 atmosphere. Pentanol and n-Bu3N were removed and the residue was chromatographed over silica gel (EtOAc-hexanes 1:10) to give [6] (0.65g, 74%).
COMPOUND 7: 6-Chloro-9- \2 (S)-O-benzyl-3 (R) -nona-8-en-enyl1purine
An acidified (cone. HCl 0.3 mL) solution of [6] (0.58 g, 1.55 mmol) in triethyl orthoformate (TEOF; 15 mL) was stirred at RT for 24 h. The yellow oil obtained after removal of TEOF, was purified by silica gel column chromatography eluting with EtOAc- hexanes (1:10) to provide [7] (0.5 g, 85%).
COMPOUND 8: 9-[2 (S)-O-Benzyl-3 (R)-nona-8-enyl]adenine
Compound [7] (0.3 g, 0.78 mmol) was dissolved in liquid ammonia (15 mL) and heated at 90°C in a steel bomb for 24 h. After cooling, excess ammonia was allowed to evaporate. The residue was taken up in CH2C12 (25 mL) and washed with water (10 mL) . The organic layer was dried and pure [8] (0.245 g, 85%) was obtained as a white solid.
COMPOUND 9: 9-[2 ( S)-0-Benzyl-9-hvdroxy-3 fR)-nonyl]adenine
To a solution of the olefin [8] (0.365g, 1 mmol) in dry tetrahydrofuran (THF; 1 L) , placed in a three-necked flask fitted with a condenser and a septum, was added a 1 M solution of a diborane-THF complex (BH3.THF; 0.5 L, 0.5 mmol) at 0°C.
The reaction was done under nitrogen atmosphere. The mixture was permitted to stir for additional hours at RT to continue the completion of the reaction. Water (0.05 mL) was added slowly and the mixture was allowed to stir at RT until hydrogen no longer evolved. The flask was immersed in an ice bath and 3 molar NaOH
(0.17 L) was rapidly added to the reaction mixture. The organoboronic acid intermediate was oxidized by the slow addition of 30% hydrogen peroxide (0.11 mL) . The reaction mixture was then allowed to stir for 3 h at 50°C to ensure completion of the oxidation. The mixture was brought to RT and
NaCl was added to saturate the lower aqueous phase. The THF phase was separated and dried (MgSO ) . The crude compound obtained after solvent removal was purified by silica gel column chromatography eluting with EtOAc to provide 9 (0.268g, 70%).
COMPOUND 10: 9-T2 (S) .9-dihvdroxγ-3 (Ri-nonylladenine
In order to remove the benzyl protective group, a solution of compound [9] (0.227g, 0.593 mmol) in EtOH (25 mL) and cyclohexene (10 L) was treated with 20% palladium hydroxide on
charcoal [Pd(OH)2/C, often referred to as Pearlmann's reagent, 0.05g]. The resulting suspension was stirred at reflux for 12 h. After cooling to RT, the mixture was filtered and the filtrate was concentrated at reduced pressure. The residue was chromatographed over silica gel (EtOAc-MeOH, 9:1) to give pure [10] (0.156g, 90%) .
This compound is the 9-hydroxy analog of EHNA, which was tested as described in Examples 4 and 6.
EXAMPLE 2: SYNTHESIS OF 8-OH-EHNA AND 8.9-DIHYDROXY ANALOGS
COMPOUND 11: 6-Chloro-9-r2 fS)-O-benzγl-8.9-epoxy-3 (R) -nona-8- enyl]purine
To an ice cold solution of the olefin [7] (0.769g, 2 mmol) in CH2C12 (15 L) was added 85% m-chloroperbenzoic acid (0.488g, 2.4 mmol) . After stirring the reaction mixture at RT overnight, it was diluted with ether (50 mL) and washed successively with saturated NaHC03 (15 mL) , 10% NaHS03 (15 mL) , saturated NaHC03 (15 L) , and brine and dried (MgS04) . The residue, obtained after evaporation of ether, was purified by silica gel column chromatography eluting with EtOAc-hexanes (1:10) to afford pure epoxide [11] (0.721g, 90%).
COMPOUND 12: 6-Chloro-9 [ 2 (S) -O-benzyl-8.9-dihvdroxy-3 fR) -nonyll purine
Compound [11] (0.3g, 0.75 mmol), 5% HC104 (2 mL) , in acetonitrile (6 mL) was stirred at RT for 2 h. The reaction mixture was neutralized with solid NaHC03 and the mixture was filtered. The filtrate was diluted with CH2C12 (25 mL) and dried over MgS04. The residue obtained after solvent evaporation was purified by silica gel column chromatography using EtOAc-hexanes (1:1) to provide diol [12] (0.3g, 95%).
COMPOUND 13: 9-f2 (S) -O-Benzyl-8.9-dihvdroxy-3 fRΪ-nonylladenine Compound [12] was obtained from [11] according to the procedure described for the preparation of [8], in 90% yield.
COMPOUND 14: 9-T2fSi . 8.9-trihvdroxy-3 fR) -nonylladenine
Triol [14] was prepared in 90% yield by debenzylating [13]
using the procedure described for [10].
This compound is the 8,9-dihydroxy analog of EHNA, which was tested as described in Example 3. Although it was shown to be a moderately effective inhibitor of ADA activity, its potency was lower than the 9-OH-EHNA analog, and it was not tested further.
COMPOUND 15: 2S■3S)-2-0-Benzyl-3. i -O-tosyl-8.9-ePOXV-2■3- nonanediol Epoxidation of [3] was carried out as described for compound [11]. After workup, the residue was chromatographed using hexane-EtOAc (7:3) to give the epoxide [15] (99%) as an oil.
COMPOUND 16: 2S.3S)-2-0-Benzyl-3-0-tosyl-2.3.8-nonenetriol A solution of aluminum hydride (30 mL, 15 mmol, 1 M solution in THF), was added to epoxide [15] (3.143g, 7.52 mmol) in dry ether (100 mL) at 0°C. The mixture was stirred at RT for 1 h and decomposed slowly by the addition of water (25 mL) and the aqueous solution was extracted with ether (4 x 50 mL) . The combined ether extracts were dried (MgS04) and concentrated to furnish [16] (3g, 95%) . An analytical sample was obtained by silica gel column chromatography with hexane-EtOAc (7:3) as eluent.
COMPOUND 17: f2S.3S> -2-0-Benzyl-3-0-tosyl-8-0-tetrahvdropyranyl- 2.3.8-nonanetriol
A solution of alcohol [16] (2.8g, 6.6667 mmol), and dihydropyran (1.68g, 13,33 mmol) in dry CH2C12 (50 mL) containing pyridinium p-toluenesulfonate (PPTS, 0.33g, 1.33 mmol) was stirred at RT for 4 h. The reaction mixture was diluted with ether and washed once with half saturated brine to remove the catalyst PPTS and dried (MgS04) . The residue obtained after evaporation of the solvent was purified by silica gel column chromatography using hexane-EtOAc (95:5) as eluent, to afford [17] (3.3g, 99%) .
COMPOUND 18: 2S,3R) -3-Azido-2-0-benzyl-8-0-tetrahydrσpyranyl- 2.8-nonanediol
Compound [18] was prepared from [17] following the procedure described for the formation of [4]. The crude product, after column chromatography (hexane-EtOAc, 95:5), gave pure [18] (82%) .
COMPOUND 19: 2S.3R)-3-Amino-2-0—benzyl-8-Q-tetrahvdropyranyl- 2-8-nonanediol
The azide [18] was reduced by a procedure similar to that described for the preparation of [5], to afford the amine [19] quantitatively.
COMPOUND 20: 5-Amino-6-chloro-4r2fS)-O-benzyl-8-O- tetrahvdropγranγl-3fR)-2.8-dihydroxynonγllaminopγrimidine Compound [20] was prepared from 19, by following the procedure described for the formation of [6]. The residue, obtained after solvent removal, was chromatographed over silica gel (EtOAc-hexanes 1:5) to give [20] (28%).
COMPOUND 21: 6-Chloro-9-f2fS)-O-benzyl-2.8-dihvdroxy-3fR)- nonyll urine
An acidified (cone. HCl, 0.15 mL) solution of [20] (0.3g, 0.63 mmol) in TEOF (15 mL) was stirred at RT for 24 h. The residue obtained, after removal of TEOF, was dissolved in absolute ethanol (10 mL) , containing PPTS (0.025g) and refluxed for 1 h. Solvent was then removed under reduced pressure and the crude product was purified by silica gel column chromatography, using EtOAc-hexanes (3:1) to give [21] (0.15g, 60%).
COMPOUND 22: 9-r2 S)-O-Benzyl-2.8-dihvdroxy-3fR)-nonylladenine Amination of chloropurine derivative [21] was carried out as for the synthesis of [8] from [7]. Crude product, obtained after workup, was purified by silica gel column chromatography, eluting with MeOH-EtOAc (5:95) to give [22] (90%).
COMPOUND 23: 9-r2 (S) .8-Dihvdroxy-3fR)-nonyl1adenine
Diol [23] was prepared in 90% yield by debenzylating [22] using the procedure described for [10] .
This compound is the 8-hydroxy analog of EHNA which was tested as described in Example 3. It was shown to inhibit ADA
activity in the desired range, but its potency was lower than the 9-OH-EHNA analog, so it was not tested further.
EXAMPLE 3: TESTING FOR ADA INHIBITION Compound [10] (the 9-hydroxy analog) , compound [23] (the 8- hydroxy analog), and compound [14] (the 8,9-dihydroxy analog) were tested for ADA inhibition activity, using calf intestinal ADA (Type III, Sigma Chemical Company) measured at 30°C by direct spectrophoto etric assays at 265 nm, as described in Harriman et al 1992. These tests used extra-cellular enzyme preparations, and did not require any of the analogs to enter cells in order to reach the enzyme. The Ki values, listed in Table 1, indicated that the 9-OH analog was more potent than either of the other hydroxylated analogs, since a smaller quantity was required to achieve 50% inactivation of a standardized quantity of the ADA enzyme. Since it was more potent, the 9-hydroxy analog was used as a starting compound for synthesizing other analogs.
Table 1 compiles ADA inhibition data and octanol-water partitiion coefficients (an index of lipophilicity) for all of the final (deprotected) analogs listed in Examples 1, 2, or 5.
EXAMPLE 4: TESTING OF 9-OH-EHNA FOR PROTECTION AGAINST ISCHEMIC DAMAGE TO TISSUE After synthesis of the 9-hydroxy and 8-hydroxy analogs of
EHNA, samples were provided by the Applicant to Dr. Robert Rodgers of the Department of Pharmacology and Toxicology at the University of Rhode Island. There were sufficient quantities of 9-OH-EHNA, while quantities of 8-OH-EHNA were smaller. Accordingly, most tests used 9-OH-EHNA and compared it to unmodified EHNA and to disulfiram, an unrelated compound that is known to have certain protective anti-ischemic effects in cardiovascular tissue.
The tests carried out by Dr. Rodgers used a widely-used protocol known as a "working heart" preparation. These tests involved removing intact hearts from sacrificed lab animals (male Sprague-Dawley rats were used) , and perfusing the hearts with liquids containing controlled quantities (or deficits) of oxygen and glucose for fixed periods of time. The procedures
used in these experiments are described in detail in Davidoff and Rodgers, Hypertension 15: 633-642 (1990), with certain minor modifications. The left atrium was filled at 15 cm H20 pressure, and the left ventricle ejected into a buffer-filled column against a pressure which equated to 72 mm Hg, except during ischemic periods. The perfusate was Krebs-Henseleit buffer with HC03 ' (25 mM) , Ca++ (2.2 mM) , and glucose (10 mM) . When gassed with 95% 02 and 5% C02, the pH of the perfusate was 7.4 ± 0.2. Perfusate and ambient temperatures were held at 37°C, and the hearts were allowed to beat spontaneously.
After perfusion began, the isolated hearts were allowed to stabilize for 10 minutes, then they were treated for 10 minutes with one of the test drugs (unmodified EHNA, 9-OH-EHNA, or disulfiram) or buffered saline containing either dilute ethyl alcohol (used to increase the solubility of EHNA or 9-hydroxy- EHNA) or dilute dimethyl sulfoxide (used to increase the solubility of disulfiram) .
Following stabilization and treatment, the hearts were subjected to simulated ischemia for 20 minutes; no oxygen was added to the perfusate during this period. When the ischemic period ended, oxygen was again added to the perfusion buffer, and the following parameters were measured over a period of 10 minutes:
LVPP - left ventricular pulse pressures (time- dependent pressures, calculated as peak pressure minus diastolic pressure, in mm Hg, millimeters of mercury column)
LVEDP - left ventricular end diastolic pressure (i.e, time-dependent pressures as the ventricle relaxed during diastolic filling, in mm Hg)
CFR - coronary flow rate (mL/min)
HR - spontaneous heartbeat rate (beats/min)
ECG - electrocardiogram (surface potential, in mV)
The results indicated that both EHNA and 9-hydroxy-EHNA reduced the incidence of fibrillation; the difference between them was not significant. Both EHNA and 9-OH-EHNA also caused moderate increases in both LVPP and coronary flow rate after
ischemia; again, their effects were not different from each other at a significant level.
The most important difference observed between EHNA and 9-OH-EHNA in these initial tests appeared in measurements of LVEDP (left ventricular end diastolic pressure) . This parameter indicates whether the muscles of the left ventricular wall are able to relax promptly following contraction. Prompt relaxation is essential, since it allows the heart's pumping chambers to fill with blood during diastolic relaxation, between contractions. A high LVEDP level is very undesirable, since it indicates that the heart muscle has become stiffened by ischemic damage and no longer has sufficient flexibility and elasticity to properly fill the pumping chambers with blood during diastolic relaxation. In these tests, 9-OH-EHNA provided a higher level of protection against muscle stiffening than unmodified EHNA.
Testing of the 8-OH-EHNA analog in these perfused heart preparations was limited, due to the small quantity that was available. However, those limited tests indicated that 8-OH-EHNA analog was not as potent as 9-OH-EHNA in reducing left ventricular stiffness and LVEDP.
Because 9-OH-EHNA was more potent as an ADA inhibitor than 8-OH-EHNA, and because 9-OH-EHNA appeared to have an additional useful effect in reducing heart muscle stiffness, subsequent research used 9-OH-EHNA, compound [10], or its benzyl-protected precursor, compound [9], as starting compounds for synthesizing other analogs.
The methods used by Prof. Rodgers were slightly different in some respects from the heart muscle tests described in Example 6, which were carried out subsequently at a contract research firm called Coromed, Inc. (Troy, New York) . In the tests carried out by Coromed, under contract to the Applicant, Cypros Pharmaceutical Corporation, 9-OH-EHNA showed little or no improvement over unmodified EHNA in protecting heart muscle, and in some tests it did not perform as well as EHNA. However, by then, other more lipophilic analogs (including 9-chloro-EHNA and 9-phthaliraido-EHNA) had been synthesized and were being tested. As described in Example 6, those other analogs showed major advantages in protecting heart muscle against ischemia, compared
to either EHNA or 9-OH-EHNA. Accordingly, subsequent research has been devoted to those other analogs, while 9-OH-EHNA and 8- OH-EHNA are not being actively tested further.
EXAMPLE 5: SYNTHESIS OF OTHER ANALOGS OF EHNA
This example and Fig. 4 depict the synthesis of several additional analogs of EHNA. Except as noted, the synthetic reactions described below used the benzyl-protected compound [9] (described in Example 1) as the starting reagent. Elemental and NMR analyses confirmed that each compound was created in analytically pure form, except as noted for compounds [26] and [28]; these were intermediates rather than completed analogs, and they were not fully purified.
The final (de-protected) analogs (Compounds [25], [27], [29], [31], and [32]) described herein were tested for inhibition of the ADA enzyme, using the procedures described in Example 3, and all were found to have a potency in the desired range, indicating that they can inhibit the ADA enzyme without irreversibly poisoning it. These Ki values are provided in Table 1.
COMPOUND 24: 9-r9-Benzoyloxγ-2fS)-O-benzyl-3 fR)-nonyl]adenine
This analog, which has benzoyloxy groups coupled to the #2 carbon and a benzyloxy group at the #9 carbon atom, was prepared by adding n,n-diisopropylazo-dicarboxylate (DIAD, 202 mg, 1 mmol) to a stirred solution of compound [9] (314 mg, 0.82 mmol), benzoic acid (BzOH, 122 mg, 1 mmol) , and triphenyl phosphine (PPh3, 262 mg, 1 mmol) in THF (5 ml) . The mixture was stirred at room temperature for 24 hr and precipitated triphenyl phosphine oxide was filtered out. The filtrate was concentrated and the residue was chromatographed on silica gel using ethyl acetate (EtOAc) to provide [24], which was further purified by preparative thin layer chromatography (TLC) using EtOAc and methanol (MeOH) at 9:1. Yield was 250 mg (63%).
COMPOUND 25: 9-f9-Benzoyloxy-2 S)-hydroxy-3 R)-nonyl]adenine
This alcohol, which has a benzoyloxy group coupled to the #9 carbon atom, was created by treating compound [24] (200 mg) in ethanol (EtOH, 18 ml) and cyclohexene (6 ml) with 20%
palladium hydroxide on charcoal (PdOH2/C, 0.15 g) . The suspension was stirred at reflux for 12 hours. After cooling to room temperature, the mixture was filtered and the filtrate was concentrated at reduced pressure. The residue was chromatographed over silica (EtOAc and MeOH at 9:1) to give pure [25] , 130 mg (80%) .
COMPOUND 26: 9-[2 fS)-0-benzyl-9-phthalimido-3 R)-nonyl]adenine Compound [26] was prepared from compound [9] in 87% yield in the same way as compound [24], except that phthalimide (1 mmol, 147 mg) was used in place of benzoic acid. Analytically pure compound could not be obtained as it always contained traces of triphenyl phosphine oxide and dihydro-DIAD. It was used as a reagent in the following step, to create compound [26].
COMPOUND 27: 9-[2fS)-hvdroxy-9-phthalimido-3 fR)-nonyl]adenine
This alcohol was created in 85% yield, with [26] as the starting reagent (255 mg, 0.5 mmol) using the same palladium on charcoal (PdOH2/C) catalytic procedure used to create [25].
COMPOUND 28: 9-T2 S)-Q-benzyl-9-chloro-3 R)-nonylladenine
Analog [28], with a benzyl ring on the #2 carbon atom and a chlorine atom coupled to the #9 carbon atom, was created by adding PPh3 (400 mg, 1.5 mmol) to a stirred solution of [9] (500 mg, 1.3 mmol) and NaHC03 (50 mg) in anhydrous CC14 (5 ml). The mixture was treated at reflux for 12 hours, then filtered, and the filtrate was concentrated. The residue was chromatographed over silica gel (EtOAc) to provide 380 mg (75%) . Analytically pure compound could not be obtained, since it contained traces of triphenyl phosphine oxide. It was used as a reagent in the following step, to create compound [29].
COMPOUND 29: 9-r9-chloro-2(S)-hvdroxy-3 R)-nonyl1adenine This analog, with a halide moiety at the 9 carbon and an alcohol group at the #2 carbon atom, was prepared in 82% yield using [28] as the starting reagent and using the same palladium on charcoal (PdOH2/C) catalytic procedure used to create [25]. Other halide analogs with fluorine, bromine, or iodine
atoms can be created, if desired, by proper selection of reagents containing such atoms. For example, the synthetic procedure described for Compound [28] could be modified by using CBr4 or CI4 instead of CC14. As another example, a 9-fluoro-EHNA analog could be produced by reacting compound [9] (the benzyl- protected 9-OH-EHNA analog) with the well known fluorination agent diethylaminosulfur trifluoride (DAST) .
COMPOUND 30: Methyl-7fR)-adenine-9-yl)-8fS)-O-benzyl-nonoate Analog [30], with an ester group at the #9 carbon and a benzyl group at the #2 carbon, was created by adding pyridinium dichromate (PDC, 2.755 g, 7.3 mmol) to a solution of [9] (1.369 g, 3.4 mmol) in dimethyl formamide (DMF, 2 ml) . The mixture was stirred at room temperature for 24 hours, then diluted with ethyl acetate and passed through a mixture of silica gel and Na2S04 (1:1) to give the corresponding acid (220 mg, 15.7% yield) .
COMPOUND 31: Methγl-7fR)-adenine-9-yl)-8 S)-hvdroxy-nonoate Analog [31], with an ester group at the #9 carbon and a hydroxy group at the #2 carbon, was created in 82% yield using the same palladium on charcoal (PdOH2/C) catalytic procedure used to create [25], using the benzyl analog [30] as the starting reagent. Ester 31 can also be referred to as 9-[2(s)- hydroxy-9-carboxymethyl-3(R)-nonyl]adenine.
COMPOUND 32: 9-T2fS)-hvdroxy-3fR)-non-8-enylladenine
Unsaturated analog [32] was created by using the unsaturated benzyl-protected analog [7], shown in FIG. 3 and described in Example 1, as the starting reagent. Compound [7] (200 mg, 0.55 mmol) in toluene (10 ml) was cooled with dry ice/acetone and ammonia was bubbled through the solution until the volume of the mixture reached 40 ml. Sodium metal was added in portions with vigorous stirring until the mixture was neutralized with NH4C1 and methanol and evaporated to dryness. The compound was then extracted with CH2C12 and the extracts were dried over Na2S04 and evaporated. The product was purified by preparative thin layer chromatography using ethyl acetate to give a 70% yield of [32].
Compound 33: 9-r9-tert-Butγldiphenylsilyloxy-2fS)-hydroxy-3fR)- nonylladenine
An additional analog that deserves note was synthesized by cypros Pharmaceutical Corporation, using de-protected 9-OH-EHNA as a starting reagent, since a surplus of that compound was available after completion of the initial biological testing of the 9-OH-EHNA. The silicon-containing moiety was chosen for two reasons: (1) calculations indicated that it had a very high level of lipophilicity (with a log P0/w value in the range of 9, compared to 2 or 3 for the chloro and phthalimido analogs) and could provide a potentially useful compound to help evaluate high-level lipophilicity; and, (2) this moiety could be added to the #9 atom in a de-protected 9-OH-EHNA molecule without disturbing the unprotected hydroxyl group on the #2 carbon atom of the side chain.
Accordingly, compound [10] (30 mg, 0.102 mmol) was dissolved in 0.5 ml of DMF and added to a solution of 50 μl of diisopropylethylamine and 15 mg of dimethylaminopyridine contained in 1 ml of CH2C12. To this solution was added 40 mg (0.14 mmol) of t-butylchlorodiphenyl-silane and the mixture was stirred at room temperature for 16 hours. The solvent was evaporated and the product was purified by chromatography over silica gel (10% CH30H/CHC13) to give 26 mg (48%) of 9-t-BDPSi-EHNA [compound 33] .
EXAMPLE 6; PROTECTION OF HEART MUSCLE AGAINST ISCHEMIA Several of the analogs described in Example 5 were tested to evaluate their ability to protect heart muscle against ischemia, using a so-called "Langendorf heart preparation". Briefly, male Sprague-Dawley rats weighing between 250 and 350 grams were anaesthetized with sodium heparin and sacrificed with C02. The heart was rapidly excised via thoracotomy and placed in physiological salt solution (PSS) un contraction ceased. The heart was then mounted via the aortic aot to a cannula and retrogradely perfused with PSS containing (in mM) : NaCl (118) , KCl (4.7), CaCl2 (2.2), KH2P04 (1.18), MgS04 (1.17), NaHC03 (25), dextrose (11) at 80 mm Hg at 37 degrees C. The perfusion solution was aerated with 95% 02/5% C02 to maintain pH at 7.4. Hearts were allowed to equilibrate for 15 min, during which time a balloon-
tipped catheter was introduced into the lumen of the left ventricle via a small incision in the left atrium. The catheter was connected to a pressure transducer and was used to measure left ventricular hemodynamic performance, i.e. left ventricular systolic pressure (LVSP) , left ventricular end-diastolic pressure (LVEDP), left ventricular developed pressure (LVDP) , +dP/dtmaχ (the maximum rate at which pressure developed in the left ventricle during each contraction) , -dP/dt|raχ (the maximum rate at which left ventricular pressure declined following each contraction) , and heart rate. Following placement of the balloon- tipped catheter, the pulmonary artery was cannulated to collect coronary effluent for measurement of coronary flow.
At the conclusion of the stabilization period, measurements of left ventricular hemodynamic performance, heart rate and coronary flow were made. The hearts were then perfused for 10 minutes with PSS containing vehicle, 9-hydroxy-EHNA, 9-chloro- EHNA, or 9-phthalimido-EHNA, and measurements were repeated. Global ischemia was produced by clamping the aortic cannula, and measurements of these parameters were made at 5-minute intervals. After 35 min of global ischemia, the hearts were reperfused with
PSS for 20 min at a pressure of 80 mm Hg. Measurements were again taken at 5 minute intervals during the reperfusion period.
The results, in Tables 2-4 and Figures 5-7, indicated that the more lipophilic analogs provided substantially better protection against ischemic damage to heart muscle than either unmodified EHNA or 9-OH-EHNA.
KYAMPT.TC 7: CELL CULTURE TESTS. UNST ESSED CKT.T.g
In a first set of cell culture tests, EHNA and its analogs were evaluated for their ability to inhibit ADA activity in two different types of cells: human red blood cells (which are relatively easy to work with) , and human astrocytoma cells (which are brain cells that can reproduce in cell culture; these were used to provide an indication of whether EHNA analogs can help reduce ischemic damage in brain tissue) .
A first set of tests was carried out, using "normoxic" conditions (i.e., the cells had not been stressed by hypoxia or by simulated ischemia, using deoxyglucose or sodium azide) , to determine an IC50 value for EHNA and several analogs. The IC50
TABLE 4 VENTRICDLAR PRESSURE INCREASE RATES DURING CONTRACTIONS (maximum dP/ t)
(mean ± SD)
Time after vehicle EHNA 9-OH-EHNA 9-C1-EHNA 9-phtalimido EHNA reperfusion 10 ttM 10 uM 10 uM 10 un
10 minutes 180 ± 128 127 ± 314 407 i 529 465 ± 356 337 ± 260 15 minutes 284 ± 196 515 ± 246 161 ± 178 815 ± 360 703 ± 249 20 minutes 533 ± 304 993 ± 390 877 ± 464 1123 ± 360 1070 ± 218
Time after vehicle EHNA 9-chloro-EHNA 9-phtalimido EHNA reperfusion 30 UM 30 uM 30 UM
10 minutes 180 ± 128 199 ± 164 821 ± 424 1051 ± 653 ω 15 minutes 284 ± 196 343 ± 300 1141 ± 774 1249 ± 583 vo 20 minutes 533 ± 304 701 ± 365 1269 ± 427 1279 ± 566
value indicates the concentration of drug which was required to inhibit half of the ADA activity in the cells; this value reflects both the ability of a drug to enter cells and reach the ADA enzyme, and the potency of a drug in binding to and inhibiting the ADA enzyme inside the cells. A low IC50 value indicates that a drug is a potent ADA inhibitor and can enter cells readily.
To carry out these tests, cell populations were preincubated with EHNA or an EHNA analog, in varying concentrations, for l hour. The cells were then incubated for 30 minutes with 10 uM 5- iodotubercidin, which inhibits the activity of a different enzyme called adenosine kinase, which adds phosphate groups to adenosine. This step ensured that adenosine levels would not be altered by a phosphorylation pathway which can consume adenosine in intact cells. The cells were then incubated for 30 minutes with 100 uM radiolabelled adenosine. After 30 minutes, concentrations of radiolabelled inosine and hypoxanthine (the molecules that are created when the ADA enzyme degrades adenosine) were measured in cell medium after separation using cellulose thin layer chromatography. Several concentrations of EHNA or an EHNA analog were used in each set of tests, and an IC50 for each compound was calculated based on the dose-response curve for that compound.
The results, in Table 5, indicate that 9-chloro-EHNA was substantially more potent than unmodified EHNA or 9-OH-EHNA in inhibiting ADA activity inside red blood cells cells (as indicated by lower inhibitory concentration values) , and that 9- phthalimido-EHNA was substantially more potent than unmodified EHNA or 9-OH-EHNA in inhibiting ADA activity inside brain (astrocytoma) cells. ICS0 values in this table are averages followed by standard deviations.
TABLE 5 INHIBITION OF ADA ACTIVITY IN CELL CULTURE TESTS
Cell type/Compound IC50 ( iM)
Red Blood Cells EHNA 1.200 ± 0.70
9-OH-EHNA [cpd. 10] 2.100 ± 0.90
9-chloro-EHNA [cpd. 29] 0.220 ± 0.14
Astrocytoma cells
EHNA 3.0 ± 2.5 9-OH-EHNA [10] 3.5 ± 0.8
9-chloro-EHNA [cpd. 29] 4.8 ± 1.9
9-phthalimido-EHNA [cpd. 27] 1.1 ± 0.9
8,9-unsaturated-EHNA [cpd. 32] 2.5 ± 0.8
EXAMPLE 8: CELL CULTURE TESTS FOR ISCHEMIC PROTECTION
In a second set of cell culture tests, cells were stressed by either of two methods which simulate hypoxic or ischemic damage. In these tests, a population of astrocytoma cells containing radiolabelled ATP was incubated with EHNA or an analog for 60 minutes. Then, to disrupt glycolysis and respiration, 2- deoxyglucose or sodium azide (5.5 mM final concentration for either toxin) was added, and the cells were incubated for 60 minutes. After incubation, the cultures were tested to determine how much radiolabelled adenosine they released into medium. Adenosine release is a normal and proper metabolic function of cells during ischemia or hypoxia, and the quantity of adenosine released by EHNA-treated or analog-treated cells was compared to the quantity of adenosine released by control cells, which had been identically stressed by the same toxin without any treatment by EHNA or an analog. The results, in Tables 6 and 7, are expressed in percentages of adenosine release by treated cells, compared to untreated control cells.
TABLE 6 ADENOSINE RELEASE BY ASTRQCYTΠ ^ r?BI,fT..q DURING SIMULATED HYPOXIA
■STRESSED BY DEOXYGLUCOSEΪ
Protective drug Percent of control values
Untreated = 100% (baseline)
EHNA 339 ± 25
9-OH-EHNA [10] 289 ± 12
9-chloro-EHNA [cpd. 29] 375 ± 25 9-phthalimido-EHNA [cpd. 27] 559 ± 19 8,9-unsaturated-EHNA [cpd. 32] 300 ± 48 9-butyldiphenylsilyloxy-EHNA [33] 244 ± 21
TABLE 7
ADENOSINE PFT-Fa«;E BY ASTPOPYTOMA lφTj[& IN NORMOXIC CONDITIONS AND SIMULATED HYPOXIA
Protective drug (iM; conditions) Percent of control values Untreated = 100% (baseline) EHNA
0.01 normoxic 99 ± 10
0.1 normoxic 92 ± 8
0.01 stressed 139 ± 7 0.1 stressed 253 ± 21
9-phthalimido-EHNA [cpd. 27]
0.01 normoxic 102 ± 11
0.1 normoxic 111 ± 13 0.01 stressed 242 ± 9
0.1 stressed 425 ± 40
8,9-unsaturated-EHNA [cpd. 32]
0.01 normoxic 115 ± 22 0.1 normoxic 119 ± 12
0.01 stressed 159 ± 11
0.1 stressed 225 ± 31
These results indicate that the 9-chloro and 9-phthalimido
EHNA analogs provided better adenosine release during hypoxic damage, in brain cell cultures, than unmodified EHNA.
In addition, the absence of any significant effects in cells that were cultured under normal oxygen ("normoxic") conditions is important, because it indicates that EHNA does not disrupt the normal metabolic activities of cells that are not being stressed. It only becomes active in cells that are being stressed.
In a third set of cell culture tests, astrocytoma cells were subjected to actual hypoxia for 2 hours, in an anaerobic chamber while nitrogen gas (rather than oxygen) was bubbled through the cell culture medium. After the hypoxic period release of radiolabelled adenosine into the culture medium (TABLE XXX) .
TABLE 8 ADENOSINE PTϋf.KASE BY BRATW rr:τ.τ.s AFTER 2 HOURS HYPOXIA
Protective drug fuMi Percent of control value
Untreated = 100% (baseline)
EHNA 0.05 257 ± 34
0.5 769 ± 67
9-phthalimido-EHNA
0.05 386 ± 51
0.5 866 ± 83
These results indicate that EHNA and its 9-phthalimido analog elevate adenosine release in both simulated and unsimulated hypoxia, and that the phthalimido analog offers a higher level of protection. This ability to help sustain a normal metabolic function, in brain cells, in the face of ischemic insult indicates that the phthalimido analog can help protect brain tissue against ischemic damage.
EXAMPLE 9: ADDITIONAL ASSAYS FOR PROTECTION OF BRAIN TISSUE Various assays are currently underway to quantify the ability of several EHNA analogs to reduce hypoxic or ischemic damage in intact brain tissue, as distinct from cultured cells. These assays are being sponsored and funded by the Applicant, Cypros Pharmaceutical Corporation, and are being carried out at
the UCLA Medical Center in Los Angeles by an independent investigator. Based upon the results of other tests already completed (including the heart tests and the cultured brain cell tests) , it is anticipated that at least some of the EHNA analogs will display a substantial ability to reduce hypoxic or ischemic damage in intact brain tissue. Accordingly, this example describes experimental protocols that can be used to carry out such evaluative tests, using intact tissue slices in relatively simple, rapid and inexpensive tests. The in vitro assays currently being carried out use intact slices of hippocampal tissue from the brains of sacrificed rats, for two reasons. First, tissue from the hippocampal region of the brain is known to be highly vulnerable to ischemic damage. And second, tests using perfused intact tissue slices are often preferred in neurological research, since intact cohesive tissue often provides a better model of in vivo cell and tissue behavior than either broken cell fragments, or isolated cells cultured under artificial conditions (especially when such cells are cancerous or have been genetically altered to increase their replication rates in artificial cell culture conditions) .
If properly perfused, so-called "CAl° neurons in hippocampal tissue slices will generate electrophysiological responses (comparable to brain waves in living animals) which can be measured quickly and easily, using a device comparable to an electroenσephalograph, for several hours. Accordingly, a candidate neuroprotective drug can be tested to find out whether it can prolong, restore, or otherwise increase the ability of neurons in perfused hippocampal tissue slices to generate electrical spikes having normal and desirable amplitudes and frequencies (as distinct from seizure-inducing convulsant drugs, which induce spikes having abnormal amplitudes or frequencies) . If a candidate drug allows hippocampal slices to regain or prolong their normal electrophysiological responses despite ischemic insult, this provides strong and direct evidence that the candidate drug does indeed have substantial protective activity against ischemic damage to brain tissue. Such tests are widely used and accepted, and are described in various references such as Whittingham et al 1984, and Schuur and Rigor 1992.
The protocols described below are described in more detail
in Wallis et al 1992. Briefly, Sprague-Dawley rats are anesthetized with halothane, then decapitated. The brain is quickly removed and placed in cold artificial cerebrospinal fluid (CSF) for one minute. This fluid contains (in mM) NaCl, 126; KCl, 5 4.0; KH2P04, 1.4; MgS04, 1.3; CaCl2, 2.4; NaHC03, 26; and glucose, 4.0, with a pH of 7.4, saturated with a gas mixture of 95% 02 and 5% C02. The chilled brain tissue is then sliced to provide hippocampal tissue slices, which are placed in recording wells with the temperature of the surrounding bath thermostatically
10 controlled to 34°C.
One hour after placement of slices into the recording wells, the orthodromic CAl population spike (PS) is measured for each tissue slice. This output, which is an indicator of cell function, is elicited as a response to electrical stimulation
15 using a twisted bipolar electrode placed over the CA3 Schaffer collaterals. Responses are recorded using tungsten electrodes inserted into the pyramidal layer of CAl. Current strengths (in the stimulating electrodes) and electrode depth (for the measuring electrodes) are adjusted to obtain maximal amplitude of
20 the CAl spikes. Only slices showing an orthodromic CAl PS of 3 V or greater on initial assessment are used for further testing.
In control samples, tissue slices are submerged in artificial CSF fluid that dose not contain any EHNA analogs or other neuroprotective drugs. Test samples are treated
25 identically, but the CSF fluid contains a known concentration of
EHNA, or an EHNA analog as described herein.
The assays involve subjecting hippocampal tissue slices to hypoxic conditions for limited periods of time, and then measuring the ability of the CAl cells to respond to electrical
30 stimulation after that period of hypoxia. To initiate hypoxia, paired hippocampal slices (i.e., two tissue slices from the same animal) are placed in two recording wells, and the perfusion fluid to both wells is changed to artificial CSF containing no free oxygen; the CSF fluid is saturated with 95% N2 (instead of 35 02) and 5% CO...
One slice in each pair additionally receives exposure to EHNA or an EHNA analog, beginning 30 minutes prior to the onset of hypoxia, and continued until 15 minutes after the termination of hypoxia. The period of hypoxic deprivation is variable for
different slices; during hypoxia, each control tissue slice (i.e., each slice that had not been treated by EHNA or an EHNA analog) is monitored to ensure that a "hypoxic injury potential" (HIP; see Fairchild et al 1988) is still present. The hypoxic deprivation for paired slices is terminated by adding oxygen to the perfusion medium 5 minutes after the disappearance of the HIP in the untreated slice. The two paired slices are then monitored for an additional 90 minutes, and brain wave amplitudes are recorded. Final CAl orthodromic PS amplitude is then compared to the original CAl orthodromic PS amplitude which was measured prior to treatment. Antidromic PS amplitude is also assessed before hypoxia, and after 90 minutes of recovery.
In some tests, hippocampal tissue slices are given electrical stimulation every 30 seconds throughout the entire perfusion period, including the hypoxic period. In these tests, ongoing periodic stimulation imposes additional metabolic demands, which aggravates the excitotoxic injury and provides an even more rigorous test.
The data gathered from stimulated slices, which are monitored for mean PS recoveries throughout hypoxia, are analyzed using Student's t-test. Data from unstimulated slices, which are assessed for CAl orthodromic and antidromic PS amplitude only at the beginning and end of experiments, are analyzed using the Wilcoxon rank-sum test. Due to the design of these test protocols, the CAl injury produced by hypoxia in this assay is severe, and usually quite reliable, in unmedicated, unstimulated slices, and it provides a useful baseline for comparative purposes.
As noted above, these assays have not yet been completed, but they are currently underway, and they are expected to show that at least some of the EHNA analogs described herein offer substantially improved levels of neuroprotection against ischemia, compared to unmodified EHNA or hydroxylated EHNA. Any specific EHNA analog which substantially reduces hypoxic or ischemic damage in these relatively simple and inexpensive in vitro tests can then be tested in intact animals, using in vivo studies as described in articles such as Nellgard and Wieloch 1992 (surgically-induced ischemia in rats) , Buchan and Pulsinelli 1990 (surgical ischemia in gerbils) , Michenfelder et al 1989
W
(surgical ischemia in dogs) , and Lanier et al 1988 (neck tourniquets on primates) , or using other protocols known to those skilled in the art.
Thus, there has been shown and described a new class of EHNA analogs which are useful for reducing ischemic or hypoxic damage and for slowing the metabolic degradation of useful anti-cancer and anti-viral drugs. Also disclosed herein are methods of synthesizing these compounds, and methods of using these compounds to treat patients in need of such treatment. Although this invention has been exemplified for purposes of illustration and description by reference to certain specific embodiments, it will be apparent to those skilled in the art that various modifications and alterations of the illustrated examples are possible. Any such changes which derive directly from the teachings herein, and which do not depart from the spirit and scope of the invention, are deemed to be covered by this invention.
REFERENCES Abushanab, E. , et al, "Practical Enantiospecific Synthesis of (+)-erythro-9-(2-Hydroxy-3-nonyl)adenine," Tetrahedron Lett. 25: 3841 (1984)
Abushanab, E., et al, "The Chemistry of L-Ascorbic and D- isoascorbic Acids: 1. The Preparation of Chiral Butanetriols and -tetrols" J. Org. Chem. 53: 2598-2602 (1988)
Bastian, G. , et al, "Adenosine Deaminase Inhibitors: Conversion of a Single Chiral Synthon into erythro- and threo-9- (2-hydroxy-3-nonyl)adenines," J. Med. Chem. 24:1383-1385 (1981) Baker, D.C, and Hawkins, L.D., "Synthesis of Inhibitors of Adenosine Deaminase: A Total Synthesis of erythro-3-(Adenin-9- yl)-2-nonanol and Its Isomers from Chiral Precursors," J. Org. Chem. 47: 2179-2184 (1982)
Bodor, N. , et al., "A New Method for the Estimation of Partition Coefficients," J. Am. Chem. Soc. Ill: 3783-3786 (1989) Buchan, A. and Pulsinelli, W.A. , "Hypothermia, but not the NMDA Antagonist MK-801, attenuates the neuronal damage in gerbils subjected to transient global ischemia," J. Neuroscience 10: 311- 316 (1990)
Cristalli, G. , et al, "Adenosine Deaminase Inhibitors:
Synthesis and Biological Activity of Deaza Analogues of erythro- 9-c-2-Hvdroxv-3-nonylιadenine. " J. Med. Chem. 31: 390-397 (1988)
Cristalli, G. , et al, "Adenosine Deaminase Inhibitors: Synthesis and Structure-Activity Relationships of Imidazole Analogues of erythro-9-(2-Hydroxy-3nonyl)adenine, " J. Med. Chem. 34.: 1187-1192 (1991)
Davidoff, A.J., and Rodgers, R.L., "Insulin, thyroid hormone, and heart function of diabetic spontaneously hypertensive rat," Hypertension 15: 633-642 (1990) Dorheim, T.A. , et al, "Enhanced interstitial fluid adenosine attenuates myocardial stunning," Surgery 110: 136-145 (1991)
Harriman, G.C.B., et al, "Adenosine Deaminase Inhibitors: Synthesis and Biological Evaluation of CI* and Nor-Cl' Derivatives of (+)-erythro-9-(2(S)-Hydroxy-3(R)-nonyl)adenine, " J. Med. Chem. 35: 4180-4184 (1992)
Lambe, CU. , and Nelson, D.J. , "Pharmacokinetics of Inhibition of Adenosine Deaminase by erythro-9-(2-hydroxy-3- nonyl)adenine in CBA Mice," Biochem. Pharmacol. 31: 535-539 (1982) Lanier, W. , et al, "The effects of dizocilpine maleate (MK- 801) , an antagonist of the NMDA receptor, on neurological recovery and histopathology following complete cerebral ischemia in primates," Anesthes. Rev. 15: 36-37 (1988)
McConnell, W.R. , et al, "Metabolism and Disposition of erythro-9-(2-Hydroxy-3-nonyl) [UC]adenine in the Rhesus Monkey," Drug Metab. Disp. 8: 5-7 (1980)
Michenfelder, J. , et al, "Evaluation of the glutamate antagonist dizocilpine maleate (MK-801) on neurological outcome in a canine model of complete cerebral ischemia: correlation with hippocampal histopathology," Brain Research 481: 228-234 (1989)
Nellgard, B. and Wieloch, T. , "Postischemic blockage of AMPA but not NMDA receptors mitigates neuronal damage in the rat brain following transient severe cerebral ischemia," J Cereb Blood Flow Metab 12: 2-11 (1992) O'Dwyer, P.J. , et al, "Association of severe fatal infections and treatment with Pentostatin," Cancer Treat. Rep. 7_0: 1117-1120 (1986)
Phillis, J.W. and O'Regan, M.H., "Deoxycoformycin antagonizes ischemia-induced neuronal degeneration," Brain Res.
Bul l . 22 : 537 -40 ( 1989 )
Schaeffer, H.J. and Schwender, C.F., "Enzyme Inhibitors XXVI: Bridging Hydrophobic and Hydrophilic Regions on Adenosine Deaminase with some 9-(2-Hydroxy-3-alkyl)adenines, " . Med. Chem■ 17: 6-8 (1974)
Vargeese, C , et al, "Adenosine Deaminase Inhibitors: Synthesis and Biological Evaluation of Putative Metabolites of EHNA," J. Med Chem. 37: 3844 (1994)