Design and Evaluation of Nucleoside Derivatives for Targeted Drug Delivery and Therapeutic Applications

2',3'-Dideoxynucleoside analogs are commonly used as anti-HIV, anti-HBV, and anti-cancer drugs. Despite of their potent activities, there are some major limitations in using 2',3'-dideoxynucleosides as therapeutic agents. The nucleosides have usually poor cellular uptake because of their hydrophilic nature. Some of the nucleoside \ analogs, such as anti-HIV agents, become ineffective after multiple administrations because of the development of the drug resistance, and therefore they must be administered in combination therapy. It is hard to deliver the nucleoside analogs to a particular tissue for site specific targeting. Furthermore, nucleoside analogs undergo three intracellular phosphorylation steps to become active. The first phosphorylation step is slow and a rate-limiting process for several compounds. Herein, we report the synthesis and evaluation of 2',3'-dideoxynucleoside conjugates with fatty acids, peptides, other nucleosides, fatty acyl phosphotriesters, or polymer derivatives. The primary hypothesis of this project was that conjugation of nucleosides with other compounds offers a novel strategy in designing compounds with enhanced anti-HIV activity. This combination may result in development of anti-HIV agents having enhanced lipophilicity, longer duration of action by sustained intracellular release of active substrates at adequate concentrations, higher uptake into infected cells, and/or site specificity. The development of viral resistance to the nucleosides would occur at a slower rate than to either compound alone. Furthermore, some of the compounds may be used to bypass first rate-limiting phosphorylation step. In the first two chapters, synthesis and anti-HIV activities of fatty acyl derivatives of Zidovudine (AZT), Allovudine (FLT), Emtricitabine (FTC), Lamivudine (3TC), and Stavudine (d4T) are discussed. Among all the compounds, 5'-0-myristoyl derivative of FTC (2.31, ECso = 70 nM against cell-free virus) exhibited the best antiHIV profile when compared with other fatty acyl derivatives of other nucleosides and the physical mixture of FTC and myristic acid. 5'-0-Fatty acyl derivatives of FLT, 5'~ 0-(12-azidododecanoyl) derivative of FLT (KP-1), and 5'-0-(12thioethyldodecanoyl)thymidine (KP-17), also displayed good activity against cell-free (EC50 values of <0.2 to 0.4 μM, respectively) and cell-associated (EC50 values of 0.9 to 1.0 μM , respectively) virus and minimal cellular toxicity. Cellular uptake studies for 5'0-fatty acyl derivatives of FLT and 3TC were conducted on CCRF-CEM cell line using a 5(6)-carboxyfluorescein derivative attached through 12-aminododecanoic acid as a linker to the nucleosides. The fluorescence-based studies indicated that the fatty acyl derivatives of FLT and 3TC have a higher cellular uptake versus that of the corresponding parent nucleoside substituted with a short alkyl group, such as ~-alanine. The cellular uptake was concentrationand time-dependent. In the third chapter, the synthesis and anti-HIV activities of succinate, suberate, and peptide derivatives of AZT, FLT, and 3TC are discussed. The compounds were designed in such a way to have 1 to 3 nucleosides. The hypothesis underlying this project is that the conjugates are able to deliver 1 to 3 nucleoside analogs to the HIVinfected cells. Some of the nucleoside-peptide conjugates were also substituted with the fatty acids. Peptides conjugated with fatty acids and nucleosides exhibited higher antiHIV activities when compared with those substituted only with nucleosides. Increasing the number of anti-HIV nucleosides to 2 or 3 on the peptide chain enhanced the antiHIV potency. A glutamic acid ester derivative, FLT-Succinate-AZT(glutamyl)-3TC, containing three different nucleosides was the most active compound among all the derivatives with an ECso value of 0.9 μM. Chapter 4 describes the synthesis of FLT from thymidine using a solid-phase method to circumvent some of the problems associated with the solution-phase methods, such as multiple protecting and deprotecting steps. Fifth chapter discusses the synthesis and anti-HIV activities of phosphotriesters of AZT and FLT. The conjugates were expected to get hydrolyzed inside the cell, to release nucleoside monophosphates, and to bypass first rate limiting phosphorylation step. The synthesized phosphotriester derivatives showed only modest anti-HIV activity, significantly lower than that of their parent nucleosides In chapter 6, synthesis and characterization of dextran prodrug (3TCSD) of the antiviral drug 3TC is discussed. Dextran-3TC conjugate was synthesized to localize 3TC selectively in the liver and provide sustained release of the drug by the action of liver lysosomes. Liver accumulation of conjugated 3TC was enhanced by 50 fold when compared to that of parent drug. In chapter 7 the synthesis and biological evaluation of double-barreled conjugates of sodium cellulose sulfate (CS) with 2',3'-dideoxynucleosides analogs (AZT, FLT and 3TC) using different linkers are described. Cellulose sulfate is a polyanionic polymer which blocks HIV entry into the cells by interacting with the positive charge of viral gp 120 protein. Nucleosides analogs act as reverse transcriptase inhibitors (RTis). Conjugates were expected to undergo enzymatic hydrolysis and \ thereby releasing nucleosides and cellulose sulfate targeting two different strains of virus. Cellulose sulfate conjugates of nucleosides containing an acetate linker showed good activity against both R5 and X4 strains of HIV. For example a CS-AZT conjugate (acetate linker; 1.73% loading) was more effective than CS, especially against the RS HIV-I lab-adapted strain BaL. Similarly, sodium cellulose sulfate-acetate-FLT and showed better anti-HIV profile than sodium cellulose sulfate and the mixture of sodium cellulose sulfate and FLT. Overall, the research described in this dissertation demonstrated that conjugation of anti-HIV nucleoside analogs with appropriate compounds (e.g., fatty acids, polymers, peptides groups, or other nucleosides) is an alternative strategy for designing more effective anti-HIV agents that can be further developed as therapeutic or preventative agents.

The cellular uptake was concentration-and time-dependent.
In the third chapter, the synthesis and anti-HIV activities of succinate, suberate, and peptide derivatives of AZT, FLT, and 3TC are discussed. The compounds were designed in such a way to have 1 to 3 nucleosides. The hypothesis underlying this project is that the conjugates are able to deliver 1 to 3 nucleoside analogs to the HIVinfected cells. Some of the nucleoside-peptide conjugates were also substituted with the fatty acids. Peptides conjugated with fatty acids and nucleosides exhibited higher anti-HIV activities when compared with those substituted only with nucleosides. Increasing the number of anti-HIV nucleosides to 2 or 3 on the peptide chain enhanced the anti-HIV potency. A glutamic acid ester derivative, FLT-Succinate-AZT(glutamyl)-3TC, containing three different nucleosides was the most active compound among all the derivatives with an ECso value of 0.9 µM.
Chapter 4 describes the synthesis of FLT from thymidine using a solid-phase method to circumvent some of the problems associated with the solution-phase methods, such as multiple protecting and deprotecting steps.
Fifth chapter discusses the synthesis and anti-HIV activities of phosphotriesters of AZT and FLT. The conjugates were expected to get hydrolyzed inside the cell, to release nucleoside monophosphates, and to bypass first rate limiting phosphorylation step. The synthesized phosphotriester derivatives showed only modest anti-HIV activity, significantly lower than that of their parent nucleosides In chapter 6, synthesis and characterization of dextran prodrug (3TCSD) of the antiviral drug 3TC is discussed. Dextran-3TC conjugate was synthesized to localize 3TC selectively in the liver and provide sustained release of the drug by the action of liver lysosomes. Liver accumulation of conjugated 3TC was enhanced by 50 fold when compared to that of parent drug.
In chapter 7 the synthesis and biological evaluation of double-barreled conjugates of sodium cellulose sulfate (CS) with 2',3'-dideoxynucleosides analogs (AZT,FLT and 3TC)  I would like to acknowledge USAID-CONRAD (HRN-A-00-98-00020-00) for providing me funding during first two years of Ph.D.

Preface
This thesis is written in the manuscript format. This work is dedicated to my beloved parents whose constant support and unmatched love guided me throughout the period of this study.
Chapter 1 and Chapter 2 discuss the synthesis and biological evaluations of fatty ' acyl derivatives of various ddNs including FLT, AZT, 3TC, FTC, and d4T. Nucleosides and myristic acid analogs act as RT inhibitors (RTis) and viral NMT inhibitors, respectively. It was expected that the conjugation of compounds to enhance the lipophilicity and thus the cellular uptake and to reduce the toxicity associated with nucleosides. Furthermore, development of viral resistance to two active drugs would occur at a slower rate than to either compound alone.
In Chapter 3, various peptide, succinate and suberate derivatives of nucleosides were synthesized and evaluated for anti-HIV activities. Derivatives were synthesized in such a way to allow the incorporation of several anti-HIV nucleosides in one compound for combinational therapy. Peptide derivatives were myristoylated at N-terminal to improve the cellular uptake. The derivatives were expected to release different nucleosides intracellularly, to provide synergic effect, and to reduce the viral drug resistance.
Chapter 4 deals with reported solution-phase methods for the synthesis of 3'fluoro-3 '-deoxythymidine (FLT) are cumbersome, require purification of intermediates, Vlll and include several protecting/deprotecting steps. To circumvent these problems, a solidphase strategy was designed for the synthesis of FLT.
In Chapter 5 nucleosides are converted into their monophosphate, diphosphate and finally to triphosphate by enzymatic phosphorylation. Conversion to nucleoside monophosphate is the rate-limiting step. Several phosphate triester derivatives of FLT \ and AZT with myristic acid analogues were synthesized using glycol as a linker in order to improve their cellular uptake and bypass rate-limiting monophosphorylation.
In Chapter 6, 3TC is used to treat hepatitis B viral infection. Treatment of HBV infection is significantly dependent on its distribution and accumulation in the liver.
Therefore, 3TC was conjugated with dextran (25 kD) by using succinate linker to synthesize 3TC-succinate-dextran conjugates. Since dextran (25 kD) has the capacity to accumulate in the liver, the conjugate was expected to get hydrolyzed inside the liver releasing free 3TC. Using this approach allowed a higher amount of 3TC to target the liver.
Capter 7 deals with cellulose sulfate that belongs to the category of sulfonate and sulfate polyanionic microbicides which inhibitors HIV entry and sperm-function.
Bifunctional conjugates containing AZT or FLT as RTis and cellulose sulfate as HIV entry blockers were synthesized. The conjugates were expected to provide additional bisubstrate compounds having synergistic and broad-spectrum activity against susceptible and AZT-resistant strains and sperm and STD-pathogen inhibiting properties.
lX Introduction           Human immunodeficiency virus (HIV) is a reterovirus, which mainly targets the immune cells, such as T-lymphocytes, monocytes, B lymphocytes, and macrophages that have CD4, a member of the immunoglobulin superfamily (Costin, 2007). The infection induces progressive loss of immune system, which ultimately results in the opportunistic infections and malignancies associated with acquired immunodeficiency syndrome (AIDS). According to the UNAIDS reports almost 33.2 million people were living with HIV at the end of 2007, a year in which 2.5 million people were newly infected with HIV infection and 2.1 million died of AIDS. Current antiretroviral drugs do not eliminate HIV and restore the immune system completely.
However, all combination therapy can reduce the viral replication to the minimum level to prevent the advance of the infection. Another problem is the continued development of drug-resistant virus to current antiretroviral drugs. Thus, there is an urgent need to discover new, safe, and potent anti-HIV agents and preventive strategies as existing therapies succumb to newly developed resistant virus.
HIV shares features common to all retroviruses and is able to route genetic information from RNA to DNA. This is accomplished by a unique enzyme, Reverse transcriptase (RT), which is encoded by a gene within the retroviral genome. HIV contains three different types of structural proteins named the external glycoprotein (Env), the capsid protein (Gag), and the viral enzymes necessary for replication (Pol) 1 proteins. Env proteins (gp 120 and gp41) are responsible for viral binding with the host cell membrane and for the infectivity of the viral particle by means of attachment to specific cellular receptors. Gag proteins are responsible for forming the reteroviral core (capsid). Pol proteins include pr integrases, RT, and protease which are responsible for viral replication (Cohen et al., 2008). receptor of the host cell (Dimitrov et al., 2005 andWeissenhom et al., 1997). This binding induces conformational changes in gp 120 molecule and exposes its other binding sites becoming suitable for attachment with coreceptors. Coreceptor binding 2 leads to another conformational change in the viral envelope leading to gp 120 dissociation from gp41. Exposure of hydrophobic gp41 domains results in gp41-cell membrane interaction. Finally, HRl and HR2 regions of gp41 form a six-helix hairpin like structure bringing the two membranes closer to each other initiating fusion process and release of viral contents in the host cell ( Fig. 1 ).  Wu et al., 2004).

RS and X4 Strains of HIV-1
Positively charged V3 loop of the viral protein gp 120 interacts with the negatively charged CD4 receptor, CCR5, and CXCR4 coreceptors , Cheng-Mayer et al., 1997. Transmembrane chemokine receptors belong to two different classes of receptors C-X-C (a-receptor) and C-C W-receptor) (Deng et al., 1996). The classification is on the basis of separation of first two cysteines by single amino acid in C-X-C class and adjacent in C-C class.
Depending on the type of coreceptors used for viral binding to the cell membrane, HIV can be classified in two categories; R5 and X4 strains of virus. These two strains show completely different interactions with the host cells and produce different pathogenic effects (Pollaskis et al., 2004, Fais et al., 1999. R5 strain of virus interacts with CCR5 chemokine coreceptors for cell-binding (monocytotropic strain, M-tropic) (Cheng-Mayer et al., 1997, Knox et al., 2004, Alkhatib et al., 1996. X4 strains of virus uses CXCR4 coreceptors to enter in the cells (lymphocytotropic strain, T-tropic) (Yi et al., 1999. X4-strains of HIV contain higher strength of positive charges at V3 loop than R5 virus . Therefore, X4 virus interacts much better with the cell-membrane than R5, but at the same time are more vulnerable to polyanionic entry blockers. deoxythymidine (zidovudine, AZT), (R)-9-(2-phosphonomethoxypropyl)adenine (tenofovir, TFV), and 5-fluoro-2'-deoxyuridine (Floxuridine, 5FU) are commercially used as in combination therapy with other drugs.
In order to produce their pharmacological effects, on entering the cells the majority of ddNs are phosphorylated intracellularly to monophosphates, diphosphates, and triphosphates in the presence of host cellular kinases. RT is a key enzyme in the replicative cycle of HIV and HBV. In case of anti-retroviral therapy, ddNs are called RT inhibitors. For example, anti-HIV ddNs are prodrugs that must enter the infected cell and then be phosphorylated to the active triphosphates by host cell kinases. As triphosphates, the ddNs act through inhibition of RT by means of substrate competition with natural deoxynucleosides and through chain termination of the nascent DNA being transcribed by the viral RT by means of incorporation of the ddN triphosphates that lacks the 3'-hydroxyl group (Lee et al., 2001, Nikolenko et al., 2005). Fig. 3 shows the activation of AZT as a representative example.  The major problems with ddNs are their high level of clinical toxicities such as bone marrow suppression and neuropathy For example, AZT triphosphate also inhibits mitochondrial DNA polymerase (Lewis et al., 2006, Lund et al., 2007. Thus, treatment with ddNs faces several challenges, such as a low therapeutic index caused in part by inhibition of cellular polymerases, absolute dependence on host cell kinasemediated activation (Fig. 1), limited brain uptake, short half-life in blood, low potential for metabolic activation, and the rapid development of resistance to drugs by HIV-I. Some important limitations are discussed briefly.
The hydrophilic nature of ddNs leads to limited cellular uptake and bioavailability. Extensive efforts have been carried out to synthesize lipophilic prodrugs of anti-HIV nucleosides . The lipophilic prodrugs 6 must have acceptable stability prior to cellular uptake and selective biotransformation to the active species.
The individuals being treated with the ddNs stops responding to the treatment due to drug resistance. The continual use of ddNs often results in emergence of drugresistant virus. For example, single point mutation at Met 184 with Val and Ile results in 3TC and FTC resistant HIV strains . HIV also produces resistance against d4T by K65R mutation (Garcia-Lerma et al., 2007). Similarly, mutation at Met 552 with Val and Ile results in 3TC and FTC resistant HBV strains . Viruses with resistance mutations accumulate, sometimes with complete replacement of wild-type virus by drug resistant mutants.
Combination therapy for controlling HIV-1 infections involving different classes of anti-HIV drugs provides several potential advantages to reduce the drug resistance (Zdanowicz, 2006). Two or more drugs may have additive or synergistic interactions that produce better efficacy than either drug alone. In highly active antireteroviral therapy (HAAR T) HIV is targeted by different classes of reverse transcriptase inhibitors along with protease inhibitors.
Furthermore, the first phosphorylation step of conversion of several ddNs to their monophosphates is a slow and rate-limiting process ( . In attempts to bypass the first rate-limiting phosphorylation step in the metabolic conversion of nucleoside analogs, numerous prodrugs of 5'-monophosphate types, such as neutral species of phosphotriester derivatives of nucleosides have been 7 proposed  which are readily taken by the infected cells. After the action of hydrolytic enzymes, phosphotriesters results are converted to active nucleoside monophosphate intracellularly.

Objectives of research
Various ddN conjugates with fatty acids, peptides, other nucleosides, and polymer derivatives were synthesized with an intention to develop multifunctional anti-HIV-1 agents. The hypothesis underlying this project was that safe, potent, and broad-spectrum multifunctional anti-HIV agents can be designed to deliver and release different active species intracellularly at the same time. Furthermore, development of viral resistance to several active drugs would occur at a much slower rate than to either compound alone. Subtype and mutant coverage will also be enhanced. Specific objectives for each class of compounds are discussed here briefly.

Chapter 1 and Chapter 2
First two chapters discuss the synthesis and biological evaluations of fatty acyl derivatives of various ddNs including 3'-fluoro-3'-deoxythymidine (FLT), AZT, 3TC, FTC, and d4T. Nucleosides and myristic acid analogs act as RT inhibitors (RTis) and viral NMT inhibitors, respectively. It was expected that the conjugation of compounds to enhance the lipophilicity and thus the cellular uptake and to reduce the toxicity associated with nucleosides. Furthermore, development of viral resistance to two active drugs would occur at a slower rate than to either compound alone.

Chapter 3
Various peptide, succinate and suberate derivatives of nucleosides were synthesized and evaluated for anti-HIV activities. Derivatives were synthesized in such a way to allow the incorporation of several anti-HIV nucleosides in one compound for combinational therapy. Peptide derivatives were myristoylated at Nterminal to improve the cellular uptake. The derivatives were expected to release different nucleosides intracellularly, to provide synergic effect, and to reduce the viral drug resistance.

Chapter 4
Reported solution-phase methods for the synthesis of FLT are cumbersome, require purification of intermediates, and include several protecting/deprotecting steps.
To circumvent these problems, a solid-phase strategy was designed for the synthesis of FLT  Chapter 5 Nucleosides are converted into their monophosphate, diphosphate and finally to triphosphate by enzymatic phosphorylation. Conversion to nucleoside monophosphate is the rate-limiting step. Several phosphate triester derivatives of FLT and AZT with myristic acid analogues were synthesized using glycol as a linker in order to improve their cellular uptake and bypass rate-limiting monophosphorylation .

Chapter 6
3TC is used to treat hepatitis B viral infection. Treatment of HBV infection is significantly dependent on its distribution and accumulation in the liver. Therefore, 3TC was conjugated with dextran (25 kD) by using succinate linker to synthesize 3TC-succinate-dextran conjugates (Chimalakonda et al., 2007). Since dextran (25 kD) has the capacity to accumulate in the liver, the conjugate was expected to get hydrolyzed inside the liver releasing free 3TC. Using this approach allowed a higher amount of 3TC to target the liver.

Chapter 7
Cellulose sulfate belongs to the category of sulfonate and sulfate polyanionic microbicides which inhibitors HIV entry ) and sperm-function . Bifunctional conjugates containing AZT or FLT as RTis and cellulose sulfate as HIV entry blockers were synthesized. The conjugates were expected to provide additional bisubstrate compounds having synergistic and broad-spectrum activity against susceptible and AZT-resistant strains and sperm and STD-pathogen inhibiting properties.
(K.P-17) derivatives of FLT with EC 50 values of 0.4 µM, 1.1 µM , and < 0.2 µM, respectively, against cell-free virus were found the most potent compounds with minimal cellular toxicity, and were selected for further studies. The tetradecanol ether analogs of FLT (1.7, EC 5 o = 176 µM) and AZT (1.8, EC 5 o = 27.6 µM) were found to be inactive under similar conditions because of the lack of hydrolysis to the parent compounds, nucleosides and myristic acid. The data suggest that the ester hydrolysis to FLT or AZT and fatty acids was critical for producing the anti-HIV activity. A number of FLT derivatives were further studied to determine their physicochemical properties (e.g., solubility, Log P, pKa) and cellular uptake. Cellular uptake studies were conducted on CCRF-CEM cell line using 5(6)-carboxyfluorescein derivatives of FLT attached through ~-alanine (1.5) or 12-aminododecanoic acid (1.6) as linkers.
Fluorescein-substituted analog of FLT with long chain length (1.6) showed > 12 times better cellular uptake profile than analog with short chain length (1.5). Cellular uptake studies revealed that the attachment of fatty acid improves the cellular uptake of the nucleoside conjugate. KP-1 and KP-17 are currently under evaluation m the preclinical studies.
Once enters the cell, FLT gets converted into FLT triphosphate by the action of host cellular kinases . FLT triphosphate is then incorporated into the DNA of HIV leading to chain termination at 3'-position. FLT is also a potent inhibitor of RT enzyme, which coverts viral RNA into proviral DNA   (Figure 1.1 ).
FLT was under clinical evaluation from 1990-1992. The studies were stopped after FLT failed phase II clinical trials because of the observed hematological toxicities including neutropenia, leucopenia, and anemia . The toxicity of FLT was suggested to be the result of DNA damage and apoptosis (Sundseth et al., 1996). In 2001, Medivir (Sweden) again started the phase II clinical trials of FLT . The trials were conducted on fifteen HIV infected patients with 7.5 mg/day alovudine and all the patients showed significant reduction in HIV load with no serious side effects. In the latest study, alovudine was used in doses of 0.5, 1.0 and 2.0 mg/day for four weeks to test the viral inhibition (Ghosn et al., 2007). The results indicated that FLT produced modest viral load reduction but could not produce the desired clinical anti-viral activity.
N-Myristoyl transferase (NMT) enzyme is involved in catalyzing the myristoylation of several proteins in HIV life cycle (e.g., capsid protein pl 7, Pr160gag-P01, Pr55gag, p27nef). At N-terminal glycine, viral proteins (gag and nef) are covalently attached to myristic acid in the presence of NMT. Myristic acid attachment makes the proteins more hydrophobic, which improves protein-protein and protein-membrane interactions . For example, after the N-myristoylation, p17 protein localizes itself towards the cell membrane, where new virus is produced (Wu et al., 2004) (Figure 1.1).
The replication of HIV-1 can be inhibited by heteroatom-containing analogs of myristic acid without accompanying cellular toxicity , Takamune et al. , 2002. It has been previously reported that several fatty acids, such as 2methoxydodecanoic acid, 4-oxatetradecanoic acid, and 12-thioethyldodecanoic acid, reduce HIV-1 replication in acutely infected T-lymphocytes. For example, 12thioethyldodecanoic acid was moderately active (EC 50 = 9 .4 µM) against HIV-infected T4 lymphocytes   It is hypothesized that the attachment of nucleoside analogs to the long chain myristic acid analogs enhances their lipophilicity and thus their cellular uptake. Once the ester conjugate enters the cells, it gets hydrolyzed by esterases and generates two active molecules, nucleoside analog and fatty acid, targeting reverse transcriptase (RT) and N-myristoyl transferase (NMT) enzymes, respectively (Figure 1.1 ).
A number of 5'-0-fatty acyl derivatives of FLT were previously reported to have better and wider activity profile than FLT . These compounds were designed to act as bifunctional anti-HIV agents targeting two important enzymes for viral reproduction. Herein, we report the synthesis of additional compounds, a more extensive evaluation of biological activities of 5'-0-fatty acyl derivatives of FLT in comparison with 5'-0-fatty acyl derivatives of AZT and parent nucleosides, cellular uptake, mechanistic studies, and their applications as anti-HIV agents and microbicides.
Microbicides are the compounds that can be applied inside the vagma or rectum topically to protect against sexually transmitted diseases including HIV. There is an urgent need to develop a safe over-the-counter intravaginal/intrarectal anti-HIV microbicide for prevention of HIV transmission.

Materials
FLT was synthesized in 5 g scale according to the previously reported method . FLT and AZT were purchased from Euro Asia Tran 20 Continental (Bombay, India) for large-scale synthesis of ester conjugates. 12-Bromododecanoic acid was purchased from Sigma Aldrich Chemical Co. 5(6)-Carboxyfluorescein (FAM) was purchased from Novabiochem. All the other reagents including solvents were purchased from Fisher scientific.
The products were purified on a Phenomenex®Gemini 10 µm ODS reversedphase column (2.1 x 25 cm) with Hitachi HPLC system using a gradient system at constant flow rate of 17 ml/min (Table 1.1 ).  The purity of the compounds was confirmed by usmg analytical Hitachi For cellular uptake studies, cells were analyzed by flow cytometry (F ACSCalibur: Becton Dickinson) using FITC channel and CellQuest software. Cell-viability studies were conducted using Cellometer Auto T.4 (Nexcelom Biosciences). The real time microscopy in live CCRF-CEM cell line with or with compounds were imaged using ZEISS Axioplan 2 light microscope equipped with transmitted light microscopy with a differential-interference contrast method and an Achroplan 40X objective.

pKa
The pKa of KP-1 was determined using the D-PAS (spectroscopic) technique.
The sample was initially titrated in a fast titration between pH 1.8 and pH 12.

Log P and Log D
It was not possible to measure the partition coefficient of the fatty acyl derivatives of FLT and AZT in standard n-octanol/water mixture, because of the insolubility of the compounds in water. The Log P of KP-1 was initially investigated by the pH-metric (potentiometric) method. The sample was titrated in three triple titrations from pH 2. 5 to pH 11 . 9 at concentrations of 0. 7-1.1 mM in various ratios of octanol/water. The results indicated high sample lipophilicity although the Log P could not be determined potentiometrically due to the apparent pKa in octanol shifting out of the measurable range.

Log D Determination by LDA-Liquid Liquid Distribution Chromatography
The Log D at pH 7.4 was measured as 5.04 using liquid chromatography. The ProfilerLDA is an isocratic chromatography system, which uses an octanol-coated column with octanol saturated mobile phases adjusted to pH 7.4.
A set of standard compounds with well known Log D octanol values are run through the column before the samples, the generated retention times are used as a calibration curve to relate retention times generated for sample compounds to Log D (Table 1.2., Figure 1.2).  Detection was carried out by using an UV diode array. A multi-wavelength peak location system is used to home on the largest peak present in the chromatogram.
This reduces interference from impurity peaks to a minimum (assuming that impurities are much smaller in size than the sample peak). It is assumed that the largest peak in the chromatogram is the compound of interest as there is no positive identification in this system (e.g. MS detection). A value of 5.04 was obtained, which should correspond to the neutral Log P, in consideration of the sample pKa ( Figure 1.3

Solubility
Difficulties were also encountered in our solubility analysis of KP-1 due to the low sample solubility in ionized form and suspected sample decomposition at high pH.
We were able to determine an upper limit for the sample solubility of 510 nM by shake-flask methodology, however.
The sample solubility was initially analyzed using the Sirius CheqSol method.
The sample was titrated under aqueous conditions from pH 12.1 to low pH at an initial concentration of 3.1 mM. As full sample dissolution was not evident at the start of the CheqSol study, the experiment was paused and the sample was sonicated in an ultrasonic bath containing hot water, for several minutes. When the experiment was resumed, sample precipitate was only observed below approximately pH 7, where a second sample ionization stage was apparent. We consider the sample to have decomposed to produce another species with a significantly lower acidic pKa during the "hot"-sonification at high pH. To avoid further complication due to sample decomposition, the sample solubility was subsequently investigated by shake-flask methodology with UV -spectroscopic sample detection.
To produce a saturated solution of KP-1, approximately 1 mg of solid material was added to 10 mL of aqueous solution, adjusted to pH 2.3 with 0.5 M HCI. The sample was then sonicated in an ultrasonic bath for several hours (at room temperature) before being left to equilibrate for a period of approximately three days.

Anti-HIV Assays
Anti-HIV activities of the compounds were evaluated according to the previously reported procedure . In summary, HeLa (Human

Cellular Uptake Study
All of the stock solutions for compounds FAM, 1.5, and 1.6 were prepared in DMSO. The human T lymphoblastoid cells CCRF-CEM (ATCC No. CCL-119) were grown on 25 cm 2 cell culture flasks with RMPI-1640 medium containing 10% fetal bovine serum. Upon reaching about 70% confluency, the cells were treated as described below and incubated for 1 h or longer at 37 °C.

Cellular Uptake of FAM, 1.5 and 1.6 at Different Time Points
When the cells reached about 70% confluency, FAM, 1.5, or 1.6 (1 mL, 20 µM) in RMPI-1640 medium were added to 1 mL of cells to make the final concentration as 10 µM. The cells were incubated for 0.5, 1, 2, 4 and 8 h at 37 °C.
Then the flow cytometry assays were performed as described below.

Cell Viability Assay
When the cells reached about 70% confluency, the cells were incubated with a solution of CCRF-CEM cell alone or 10 µM FAM, 1.5, or 1.6 for 24 hat 37 °C. Then 20 µL of the cells from each flask were treated with 2 µL of trypan blue (0.1 %) for 1 36 min. The cells were then transferred to a Cellometer® counting slide and analyzed using Cellometer® Auto T.4 (Nexcelom Bioscience). All the assays were performed in triplicate.

Real Time Fluorescence Microscopy in Live CCRF-CEM Cell Line
The cellular uptake studies and intracellular localization of CCRF-CEM cell alone, or incubated with 1.5 and 1.6 were imaged using a ZEISS Axioplan 2 light microscope equipped with transmitted light microscopy with a differentialinterference contrast method and an Achroplan 40X objective. The human T lymphoblastoid cells CCRF-CEM (ATCC No. CCL-119) were grown on 60 mm Petri Dishes with RPMI-1640 medium containing 10% fetal bovine serum. Upon reaching about 70% confluency, the cells were incubated with a solution of 10 µM 1.5 or 1.6 for 1 hat 37 °C. The cells were then observed under the fluorescent microscope under bright field and FITC channels ( 480/520 run).

5'-0-(Fatty acyl) Ester Derivatives of FLT and AZT
FLT was synthesized using thymidine as the starting material according to previously reported procedure . 5'-0-(Fatty acyl) ester derivatives of FLT and AZT (Table 1. 4) were synthesized from the reaction of FLT or AZT with the corresponding fatty acyl chloride in the presence of oxalyl chloride and DMAP as described previously  at the scale of 100 mg . KP-1, KP-l6, and KP-17 showed higher potency and minimal cellular toxicity when compared to the other compounds (Table 1.2). KP-1 and KP-17 were then synthesized in larger scale (25 g) for further biological evaluation, preclinical and formulation studies.
Compounds were purified first by using silica gel column chromatography and then HPLC to achieve >99% purity.

Physicochemical Properties
Physicochemical properties including pKa, LogD, and solubility were determined for KP-1 as a model compound

pKa
The pKa of KP-1 was determined using the D-PAS (spectroscopic) technique.
The sample was subsequently titrated under methanol-water cosolvent conditions in two triple titrations from pH 12.2 to pH 3. 7 at concentrations of 31-49 µM. The pKa was determined from the spectroscopic data with an aqueous value of 9.67± 0.02, obtained by Yasuda-Shedlovsky extrapolation of the individual results obtained.

Log P and Log D
The Log D was measured at pH 7.4 by liquid chromatography. The ProfilerLDA is an isocratic chromatography system, which uses an octanol-coated column with octanol saturated mobile phases adjusted to pH 7.4. A value of 5.04 was obtained, which should correspond to the neutral Log P, in consideration of the sample pKa.

Solubility
An upper limit for the sample solubility of 510 nM was determined by shakeflask methodology with UV-spectroscopic sample detection. To produce a saturated solution of KP-1, approximately 1 mg of solid material was added to 10 rnL of aqueous solution, adjusted to pH 2.3 with 0.5 M HCI. The sample was then sonicated in an ultrasonic bath for several hours at room temperature and then equilibrated for a period of approximately three days. The supernatant was then filtered, and the UV absorption spectrum of the sample was measured. The UV-absorption signal (0.0039) of the supernatant at the absorption maximum of KP-1 (264 nm, CJ= 7650 dm 3 cm-1Mor1) was close to the detection limit of the apparatus and it is considered appropriate to quote the solubility value determined as an upper limit. The intrinsic aqueous solubility of KP-1 is therefore determined to be <510 nM. KP-1 was completely soluble in ethanol (>30 mg/rnL) and the mixture of water/methanol (60:40). KP-1 was less soluble in DMSO (~4.1 mg/mL). concentrations that were not cytotoxic. As shown in Table 1 observed that the 5'-0-myristoyl ester conjugate of FLT (KP-16) was more consistent in displaying antiviral activity against cell-associated virus compared to FLT, AZT, and physical mixtures of FLT or AZT with fatty acids (50:50 in equimolar ratio; 1.9-1.11). All three physical mixtures, 1.9 (myristic acid and AZT (50:50)), 1.10 (myristic acid and FLT (50:50)), and 1.11 (12-bromododecanoic acid and AZT (50:50)), showed lower inhibitory potency against cell-associated HIV compared to KP-1 and KP-17. Compound 1.10 exhibited higher potency than that of FLT in cell-associated virus, but not as much as KP-1 and KP-17, suggesting the conjugation is required for efficient inhibition of cell-associated virus.  virus when compared to that of AZT. KP-1, KP-16 and KP-17 were found to be safe compounds for cell viability studies as their toxic concentration limits were >40. AZT and FLT on the other hand displayed toxicity even at half the concentration. KP-2 not only showed less activity but also demonstrated high toxicity towards the cells at very low concentration. The impact of the compounds was further studied by looking at their antiviral index which is the ratio of TC 5 o/IC 5 o. KP-1, KP-16 and KP-17 had >4 times better Also values than FLT and >30 times effective values than AZT.

Cytotoxicity and Proinflammatory Effects
Compounds cytotoxicity was evaluated using human vaginal cells (VK-2).
Contrary to N-9 (used as positive control), at 1 mg/mL, highest concentration tested, KP-1, KP-16, and KP-17 did not show significant cytotoxic effects during a 6 h incubation at multiple concentrations (Figure 1.5). Furthermore, although FLT is considerably more cytotoxic than AZT toward uninfected lymphocytes, 5'-fatty acyl derivatives of FLT did not exhibit higher toxicity in epithelial cell and vaginal cell cytotoxicity assays (Figure 1.5). All analogs of FLT demonstrated lower toxicity than FLT probably due to a sustained release of FLT upon the hydrolysis of the conjugates.   The spermicidal activities of several fatty acids have been previously reported (Brown-Woodman et al., 1985;Jianzhong et al., 1987). None of these derivatives showed significant spermicidal activity (Figure 1.11). In a dose-response study to evaluate spermicidal activity, compounds KP-1, KP-16, and KP-17 did not show significant sperm immobilizing or spermicidal activity, even at their maximum concentrations (1 mg/mL). KP-7, one of the FLT analogs, displayed spermimmobilizing activity, although it was comparatively weak (Table 1.8).  To confirm that the enhanced uptake of 5(6)-carboxyfluorescein derivative of FLT, 1.6, is not due to the absorption on the cell membrane surface, cells were incubated with 10 µM of DMSO, FAM, 1.5, and 1.6 for 1 h and then treated with trypsin for 5 min to wash the adsorbed molecules (if any) from the cell membrane. The cellular uptake studies after trypsin treatment showed that the cellular uptake of 1.6 was still much higher than those of contro 1 compounds, FAM and 1.5 (Figure 1.14 ), suggesting that the higher cellular uptake of 1.6 is not due to artificial absorption to the cell membrane.

Real Time Fluorescence Microscopy in Live CCRF-CEM Cells
CCRF-CEM cells were incubated with 10 µM of DMSO, FAM, 1.5 and 1.6 for 1 hand were imaged using light microscope (ZEISS Axioplan 2) equipped with transmitted light microscopy with a differential-interference contrast method and an Achroplan 40X objective. Cells showed no significant fluorescence when incubated with DMSO, FAM, and 1.5 (Figure 1. 16). On the other hand, cells incubated with 1.6 showed fluorescence.
The results further confirm the higher cellular uptake of 1.6, a fatty acyl derivative of FLT, in comparison to 1.5 and FAM alone. In general, these data indicate that the fatty acyl derivatives of nucleosides have better cellular uptake than their parent nucleosides.
59 The presence of long chain fatty acid at 5'-position enhanced the lipophilicity of FLT and the cellular uptake as was shown by cellular uptake studies of 5'-carboxyfluroscein derivatives of FLT containing short chain (1.5) and long chain (1.6) alkyl ester groups. F ACS experiments showed that 1.6 had at least 8-fold higher cellular uptake in CCRF-CEM cells than 1.5. Fluorescence microscopy of the cells incubated with these compounds further confirmed the F ACS results as cells incubated with 1.6 showed significantly higher fluorescence when compared with cells incubated with FAM and 1.5. These results suggest that the increased inhibition by KP-1, KP-16, and KP-17 may be due to a higher intracellular level of active nucleoside achieved by the conjugate.
The high activity of these compounds was possibly due to their increased rate of uptake and intracellular hydrolysis yielding two antiviral agents with different targets, FLT and fatty acid analog.
KP-1 and KP-17 are currently undergoing further preclinical studies, such as ADMET, animal toxicity, lactobacillus inhibition, and preformulation studies. These compounds may be used as topical microbicidal applications (such as vaginal insert or jellies) to prevent HIV infection during the sexual activity. These data provided insights for more rational design of additional potent and safe anti-HIV microbicides using the FLT as the parent nucleoside. When taken together, the results will have significant implications for the design of more potent and innovative anti-HIV agents. higher activity against HIV when compared with that of (+)-isomer and is nearly two times less cytotoxic, therefore it is used in clinic (Skalski et al., 1993). Although

Acknowledgments
Larnivudine has good activity against wild type HIV, a single point mutation at 184 residue results in 3TC-resistant mutant virus (M184V/I) . Several studies have provided different reasons for resistance development, such as cytidine deamination and the generation of steric hindrance at 184 amino acid residues. Similar to the HIV, mutation at Met5 52 with Val and Ile (M552V/I) results in 3TC and FTC resistant HBV strains .
Stavudine is a thymidine nucleoside analog and was approved in 1994 for treatment against HIV-1. d4 T is well absorbed orally and is metabolized intracellularly to FTC, a 5-fluoro derivative of 3TC, is 10-17 times more potent than 3TC. FTC and 3TC share common mechanism of action and drug resistance patterns (Masho et al., 2007). It is suggested that use of FTC with Tenofovir results in a higher barrier to drug resistance Pozniak et al., 2006).
N-Myristoyl transferase (NMT) enzyme is involved in catalyzing the myristoylation of several proteins in HIV life cycle (e.g., capsid protein pl 7, Prl60gag-pol, Pr55gag, p27ner) . At N-terminal glycine, viral proteins (gag and nef) are covalently attached to myristic acid in the presence of NMT. Myristic acid attachment makes the proteins more hydrophobic, which improves protein-protein and protein-membrane interactions . For example, after the N-myristoylation, pl 7 protein localizes itself towards the cell membrane, where new virus is produced (Wu et al., 2004) 69 The replication of HIV-I can be inhibited by heteroatom-containing analogs of myristic acid without accompanying cellular toxicity (Bryant et al., I993, Takamune et al., 2002). It has been previously reported that several fatty acids, such as 2methoxydodecanoic acid, 4-oxatetradecanoic acid, and I 2-thioethyldodecanoic acid, reduce HIV-I replication in acutely infected T-lymphocytes. For example, 12thioethyldodecanoic acid was moderately active (ECso = 9.4 µM) against HIV-infected T4 lymphocytes .
It is hypothesized that the attachment of nucleoside analogs to the long chain myristic acid analogs enhances their lipophilicity and thus their cellular uptake. Once the ester conjugate enters the cells; it gets hydrolyzed by esterases; and generates two active molecules, nucleoside analog and fatty acid targeting RT and NMT enzymes, respectively.
We previously reported the synthesis and evaluation of fatty acyl derivatives of AZT and FLT (chapter I). Herein, we report the synthesis of fatty acyl derivatives of The products were purified on a Phenomenex®Gemini 10 µm ODS reversedphase column (2.1 x 25 cm) with Hitachi HPLC system using a gradient system at constant flow rate of 17 ml/min (Table 2.1 ).  compounds were imaged using ZEISS Axioplan 2 light microscope equipped with transmitted light microscopy with a differential-interference contrast method and an Achroplan 40X objective.

.3.3. Anti-IDV Assays
Anti-HIV activities of the compounds were evaluated according to the previously reported procedure . In summary, HeLa (Human cervical carcinoma:

Cellular Uptake Study
All of the stock solutions for compounds FAM, 2.38, and 2.39 were prepared in DMSO. The human T lymphoblastoid cells CCRF-CEM (ATCC No. CCL-119) were grown on 25 cm 2 cell culture flasks with RMPI-1640 medium containing 10% fetal bovine serum. Upon reaching about 70% confluency, the cells were treated as described below and incubated for 1 h or longer at 3 7 °C.

. 3 .4.1. Cellular Uptake of FAM, 2.38 and 2.39 at Different Time Points
When the cells reached about 70% confluency, FAM, 2.38, or 2.39 (1 mL, 20 µM) in RMPI-1640 medium were added to 1 rnL of cells to make the final concentration as 10 µM. The cells were incubated for 0.5, 1, 2, 4, and 8 hat 37 °C. Then the flow cytometry assays were performed as described below.

Cellular Uptake of 2.39 at Different Concentrations
When the cells reached about 70% confluency, 1 rnL of graded concentrations ( Information.

Cellular Uptake of 2.38 and 2.39 with Trypsin Treatment.
The assays were performed as previously described in section 2.3.4.1 at 1 h time point with the exception that the cells used were incubated with 0.25% trypsin/0.53 mM EDTA for 5 min before washing with PBS (pH 7.4).

Flow Cytometry
The cells were washed twice with PBS (pH 7.4) at 2000 rpm for 5 min. Then the cells were analyzed by flow cytometry (F ACS Cali bur: Becton Dickinson) using FITC channel and CellQuest software. The data presented are based on the mean fluorescence signal for 10000 cells collected. All the assays were done in triplicate.

Cell Viability Assay
When the cells reached about 70% confluency, the cells were incubated with a solution of CCRF-CEM cell alone or 10 µM FAM, 2.38, or 2.39 for 24 hat 37 °C. Then 20 µL of the cells from each flask were treated with µl of trypan blue (0.1 %) for 1 min.
The cells were then transferred to a Cellometer® counting slide and analyzed using Cellometer Auto T.4 (Nexcelom Bioscience). All the assays were performed in triplicate.

Real Time Fluorescence Microscopy in Live CCRF-CEM Cell Line
The cellular uptake studies and intracellular localization of CCRE-CEM cell alone, or incubated with 2.38 and 2.39 were imaged using a ZEISS Axioplan 2 light microscope equipped with transmitted light microscopy with a differential-interference contrast method and an Achroplan 40X objective. The human T lymphoblastoid cells  All compounds were synthesized at 100 mg scale and were tested for the anti-HIV and cytotoxicity assays. Compounds 2.8, 2.16, and 2.17 were further synthesized in larger 99 scale (5 g) for further biological evaluations. These three compounds were first purified by using silica gel column chromatography (>90% purity) and then HPLC (>99% purity

.4.2. Biological Activities
Although 3TC, FTC, and d4T are less potent than FLT against cell-free virus assays, but they exhibited a higher anti-HIV activity against cell-associated virus. Table   2.2 illustrates the anti-HIV-1 activities of the fatty acyl ester derivatives of3TC, d4T, and FTC in comparison with 3TC, FTC, and d4T against cell-free and cell-associated virus.
The data provide structure-anti-HIV activity relationships for different fatty acyl

Cellular Uptake Study
Studies were performed to understand cellular uptake profile of 5'-0-fatty acyl derivatives in comparison with 3TC. 3TC attached to FAM through /)-alanine (2.38) was used as a control 3TC analog. 3TC attached to FAM through 12-aminododecanoic acid   For example, diversity in the structure and physicochemical properties of peptides allow their applications in targeted drug delivery, enzyme inhibitors, and scaffolds.
Peptides-based prodrugs are commonly used in drug delivery. Peptides have been used as linkers to deliver drugs at desired site where they undergo site specific enzymatic hydrolysis to deliver the active molecules. For example, Chau et al. have used a specific peptide sequence of matrix metalloproteinase, an enzyme overexpressed in cancer cells, to deliver anti-cancer drugs, such as methotrexate to the cancer cells , Chau et al., 2005. Peptides with different chain lengths have also been used as spacers to deliver active drugs after Jysosomal hydrolysis of the peptide conjugates (Penugonda et al., 2007, Subr et al., I 992, Soyez et al., 1996.
Peptide esters have been previously used to improve the bioavailability of the active drugs. Peptide prodrugs of lopinavir showed higher oral bioavailability than Jopinavir itself . Peptide conjugates of 5-aminolaevulinic acid showed improved pharmacological response as a result of better cellular uptake .
Furthermore, peptide derivatives are also being used to produce direct pharmacological activity against some targeted enzymes. Ramipril, enalapril, and captopril are peptide-based compounds that are used as angiotensin converting enzyme inhibitors . Enfuvirtide is a recently approved anti-HIV drug that acts as the anti-HIV entry blocker, and is a peptide structure based derivative . Several HIV protease inhibitors, such as lopinavir, saquinavir, are also peptide-based drugs .
Although the introduction of highly active antiretroviral therapy (HAART) in the mid-l 990s has resulted in a decrease of the morbidity and mortality in the HIV-1 patient population that has access to treatment, therapy failure still occurs. A combination of reverse transcriptase (RT) inhibitor nucleoside analogs is used m HAART to reduce the viral load. Each nucleoside analog has different cellular uptake rate and pharmacokinetics. Several of nucleoside analogs succumb to newly developed resistant virus. For example, Lamivudine is a (-)-2',3'-dideoxy-3'-thiacytidine analog that is used in the treatment of both HIV-1 and hepatitis disease (Skalski et al., 1993) ..
Although Lamivudine has good activity against wild type HIV, a single point mutation at 184 residue results in 3TC-resistant mutant virus (Ml84V/I) . Several studies have provided different reasons for resistance development, such as cytidine deamination and the generation of steric hindrance at 184 amino acid residues. Similar to the HIV, mutation at Met552 with Val and Ile (M552V/I) results in 3TC and FTC resistant HBV strains . Nucleoside analogs are also very polar and have limited cellular uptake.
Therefore it is logical to develop new and more potent multi-nucleoside conjugates, with major advantages to HAART therapy that display broad-spectrum activity against drug-resistant HIV, have higher cellular uptake, and can deliver several RT nucleoside inhibitors simultaneously to the HIV-infected cells.
The objective of this research was to design multi-nucleoside conjugates substituted on a multivalent scaffold. The conjugates may have application in delivery of several nucleosides to the infected cells, broad-spectrum activity, and a higher barrier to drug resistance. Herein, we report the synthesis and anti-HIV evaluation of, three classes of nucleoside analog (AZT, FLT, or 3TC) conjugates. In the first class, combinations of two similar or different nucleosides, (AZT, FLT, or 3TC) were attached to the carboxylic acid groups of succinic acid and suberic acid. Second class of compounds includes myristoyl or acetyl derivatives of di-or trinucleoside-glutamic acid conjugates containing more than one nucleoside. In the third class of compounds, peptide derivatives containing nucleosides and myristoyl group on the side chain were synthesized.
Nucleoside-scaffold conjugates were designed with the expectation that the attachment of more than one nucleoside analog to the scaffold will generate a prodrug capable of delivering different nucleosides to the HIV-infected cells . Myristic acid was attached to the scaffolds to improve the lipophilicity of the conjugates and their cellular uptake. It was expected that once the conjugate enters the cells, it will be hydrolyzed by esterase and/or peptidases to generate parent nucleoside analogs. The release of different nucleosides will help to increase the barrier to resistance to the individual compounds. The combined conjugates may have also the benefits of synergistic antiviral effects on HIV-1 and HIV-2, increased antiviral spectrum, dosing simplicity, and favorable pharmacokinetic properties.

126
The products were purified on a Phenomenex®Gemini 1 O µm ODS reversedphase column (2.1 x 25 cm) with Hitachi HPLC system using a gradient system at constant flow rate of 17 ml/min (Table 3.1).  nucleoside (AZT or FLT) (l.45 mmol) were dissolved in dry pyridine (15.0 mL). The reaction mixture was stirred at room temperature overnight. The solvent was evaporated under reduced pressure and the crude product was purified with reversed phase HPLC using a C 18 column and water/acetonitrile as solvents as described above in   (3.7). FLT or AZT (0.45 mmol) and DMAP (110 mg, 0.90 mmoL) were dissolved in dry benzene (10 mL). Succinyl chloride (22 µL, 0.2 mmoL) was added to the reaction mixture. The reaction mixture was stirred overnight at room temperature, concentrated at reduced pressure, and dried 132 under vacuum. The residue was purified with reversed phase HPLC using a C1s column and water/acetonitile as solvents as described above to yield 3.6 or 3.7. nunol) were dissolved in dry DMF (10 mL). The reaction mixture was stirred overnight at room temperature, concentrated, and dried under vacuum. The residue was purified with reversed phase HPLC using a C1s column and water/acetonitile as solvents as described above to yield 3.8 and 3.9.

3.8.
Yield ( and DIPEA (2 mL, 15 mmol) were dissolved in dry DMF (10 mL). The mixture was stirred overnight at room temperature and was concentrated at reduced pressure.
Acetic acid (80%, 10 mL) was added to the residue and the reaction mixture was stirred at 80 °C for 30 min. The reaction mixture was concentrated and dried under vacuum. The residue was purified with reversed phase HPLC using a C 18 column and water/acetonitile as solvents as described above to yield 3.10 (50 mg, 30%).

.3.2.2. Synthesis of Peptide-Nucleosides Conjugates (Peptides Containing one nucleoside and one Myristoyl Group)
Several peptide conjugates of AZT, FLT, 3TC, and myristic acid were synthesized employing a PS3 automated peptide synthesizer and Fmoc solid-phase peptide synthesis using Fmoc-L-amino acid building blocks. The peptide-nucleoside conjugates were assembled on Wang resin solid support at room temperature. The building blocks, Fmoc-Glu(nucleoside)-OH and Fmoc-Ser(myristoyl)-OH, were synthesized from Fmoc-Glu(OH)-tBu and Fmoc-Ser-OH, respectively, as described below:
To the resin was added TFA:DCM (5%, 10 mL). The mixture was shaken for 1 hat room temperature. The resin was washed with DCM (3 x 10 mL) and DMF (10 mL) and swelled in DMF (10 mL). Free amino group at lysine side chain was further myristoylated by adding myristic anhydride (100 mg, 1.08 mmol) and DIPEA (2 mL, 15 mmol) to the swelled resin. The mixture was shaken for 2 h at room temperature.
The resin was washed with DMF (3 x 10 mL). A mixture of TF A/anisole/water (95:2.5:2.5 v/v/v, 10 rnL) was added to the resin and the mixture was shaken for 1 h.
After filtration, the solution was concentrated and dried under reduced pressure. The crude peptide conjugates were purified with reversed phase HPLC using a C 18 column and water/acetonitile as solvents as described above and were lyophilized to yield 3.17 and3.18 .
HBTU (1.08 mmol) and NMM (1.08 mmol) in DMF were used as coupling and activating reagents, respectively. Fmoc deprotection at each step was carried out using piperidine in DMF (20%). NH2-Ser(My)-,BAla-Glu(FLT)Gly-Wang resin or NH2-Ser(My)-,BA1a-Glu(AZT)-Gly-W ang resin was transferred to the reaction vessel and swelled in DMF (2 mL) for 30 min. Acetic anhydride (2 mL) and DIPEA (2 mL, 15 mmol) were added to the mixture. The reaction was shaken at room temperature for 30 min to cap N-terminal with acetyl group. N-Acetylated resin was washed DMF (2 x 10 mL). To the resin was added the peptide was cleaved from the resin by a mixture of TFA/anisole/water (95:2.5:2.5 v/v/v, 10 mL) and the mixture was shaken for 1 h. After :filtration, the solution was concentrated and dried under reduced pressure. The crude peptide conjugates were purified with reversed phase HPLC using a C 18 column and water/acetonitile as solvents as described above and were lyophilized to yield 3.19 and 3.20. DIPEA (5 mL, 37 mmol) was added to the solution and the reaction mixture was stirred overnight at room temperature. The solvent was removed and the residue was dried under reduced pressure. The residue was purified with reversed phase HPLC using a C 18 column and water/acetonitile as solvents as described above and was lyophilized to yield 3.21 (820 mg, 75%). Compound 3.23 was dissolved in acetic acid (80% in water, 10 mL) and was heated at 80 °C for 30 min to remove DMTr protection. Acetic acid was removed under reduced pressure and the residue was purified with reversed phase HPLC using a C1s column and water/acetonitrile as solvents as described above and was lyophilized to yield 3.24 (40 mg, 45%).  26.34, 30.34, 30.21, 30.09, 29.91, 29.76, 29.69, 26.80, 23.16 29.90, 29.83, 29.66, 29.60, 29.51 , 29.46, 29.36, 29.31, 29.16, 27.27 25.59, 24.97, 22.70  (535 mg, 1 mmol), and HBTU (650 mg 1.7 mmol) were dissolved DMF (10 mL).

3.24.
DIPEA (5 mL, 37 mmol) was added to the solution and to the reaction mixture was stirred overnight at room temperature. The solvent was removed and the residue was dried reduced pressure. The residue purified with reversed phase HPLC using a C 18 column and water/acetonitile as solvents as described above and was lyophilized to yield 3.31 (840 mg, 87%).  c 13 H 2 ,-CONH-Glu(AZT)-3TC (3.34 Acetic acid (80% in water, 10 mL) was added to compound 3.33. The mixture was heated at 80 °C for 30 min to remove DMTr protection. Acetic acid was removed under reduced pressure and the residue was purified with reversed phase HPLC using a C 18 column and water/acetonitile as solvents as described above to yield 3.34 (35 mg, 50%). (CH 2 COO), 36.24 (3TC C-2'), 32.11, 30.11, 29.99, 29.83, 29.55, 29.47, 29.40, 25

Nucleoside and One Myristoyl Group).
Peptide-nucleoside derivatives (3.16-3.20) containing one nucleoside and one myristoyl moiety attached in the peptide side chains were synthesized in three steps: (i) Fmoc-building block synthesis; (ii) Peptide assembly on the resin using automated peptide synthesizer; and (iii) Deprotection followed by acylation of the free amino groups.
Step-by-step synthetic procedures are shown in the schemes 3.6-3.8.

Synthesis of Dinucleoside-and Trinucleoside Glutamic Acid Derivatives with or without Myristoyl Moiety
The synthesis of glutamate esters containing two to three nucleosides, with or without myristic acid, was accomplished by solution phase synthesis. The conjugates were synthesized by reaction of an appropriate building block, such as Fmoc-Glu(FLT)-OH (3.12) or Fmoc-Glu(AZT)-OH (3.14) as at free a-carboxylic acid with other nucleosides, such as 3TC-DMTr or AZT, in the presence of HBTU and DIPEA  acid (3.24, 3.26, 3.29, 3.30, 3.34,   and 3.36). Myristoyl and acetyl capping at N-terminal was carried out to provide peptides with high and low lypophilicity, respectively (Schemes 3.9 and 3.10).
Finally, a glutamate conjugate containing three different nucleosides attached was synthesized (Scheme 3.10). In this molecule, the first, second, and third nucleosides were attached to C-terminal, to the side chain, amino group, respectively.

Biological evaluation
All the compounds were tested for their anti-HIV activity and cytotoxicity profile.  As shown in Table 3.2, all the peptide-nucleoside conjugates (3.16-3.19) displayed poor anti-HIV activity (EC 50 > 10 µM when compared to FLT (EC 50 = 0.8 µM).
Myristoylated peptide-AZT conjugate 3.18 was almost two fold less active than AZT.
Peptide conjugates were nearly 13-79 folds less active than FLT. The results indicate that the peptides may not be an appropriate scaffold for the attachment of nucleosides and attachment of myristic acid did not improve the ant-HIV profile of the 168 nucleosides. This may be due to the low cellular uptake of compounds containing peptide backbone.
The anti-HIV activities for glutamate esters of two different nucleosides were evaluated against monocytotropic stain of virus and compared with the physical mixture of the corresponding nucleosides with or without myristic acid (Table 3.3).
The data showed improved anti-HIV activity for molecules containing myristic acid when compared to the molecules without the fatty acid. Glutamate ester of FLT and 3TC with myristic acid (3.24) exhibited 8.5-fold higher activity than the corresponding conjugate without myristic acid (3.26). Furthermore, compound 3.24 was the most active conjugate among glutamic ester of two nucleosides. However, the anti-HIV activity for all the compounds was less than that of FLT.
Surprisingly, the physical mixtures of nucleosides and glutamic acid with myristic acid showed higher potency than the corresponding conjugates. This result indicates that attachment of nucleosides to the amino acid backbone is deleterious, possibly because of poor cellular uptake.
In general, the presence of the myristic acid in the conjugates or physical mixtures improved the anti-HIV activity. The possible reason for the better activity could be the improved lypophilicity, which ultimately results in higher cellular uptake of the active compounds and higher solubility of the physical mixtures in the presence of myristic acid in tested solutions. More investigations are required to confirm these hypotheses.
The activity of 3.37 was also compared to the corresponding physical mixtures either with or without myristic acid (Table 3. This data is consistent with our earlier results suggesting that glutamate and peptides are not appropriate scaffolds for improving anti-HIV profile. Addition of myristic acid in equivalent ratio to the physical mixture (3.46) improved the activity by three fold in comparison with 3.37 and the physical mixture 3.45. These data were consistent with those obtained for the physical mixtures of two nucleosides and myristic acid described above.

Conclusions
Three classes of mono-di-, or trinucleoside conjugated on multivalent scaffolds (e.g., polycarboxylic acids, amino acids, and peptides) were synthesized with the expectation to improve the cellular uptake profile of the nucleosides, to exert synergistic effect by delivering different ant-HIV nucleosides at the same time to the infected cells, and to generate broad-spectrum anti-HIV agents with higher barrier to drug resistance.
Peptides or glutamate conjugated with myristic acid and nucleosides exhibited higher anti-HIV activities when compared with those substituted only with nucleosides. Increasing the number of anti-HIV nucleosides to 2 or 3 on the peptide chain enhanced the anti-HIV potency. Physical mixtures of nucleosides with amino acids and fatty acids used in the conjugation also showed significantly higher potency.
The presence of one myristic acid in the conjugates or physical mixtures improved the anti-HIV activity, but addition of two myristic acids to the conjugates was not beneficial.
Glutamate-nucleoside derivatives showed higher anti-HIV activity than dinucleoside succinate derivatives. The glutamate conjugate with three different nucleosides (3.37) was found to be the most potent compound in three classes of compounds evaluated here. Compound 3.37 had higher anti-HIV activity than AZT and 3TC, and showed comparable activity to FLT (EC 50 = 0.8 µM). Although glutamate conjugates containing two nucleosides exhibited higher activity than AZT and 3TC, but they were less active than FLT. Presence of myristic acid in the glutamic acid conjugates and their corresponding physical mixtures improved the anti-HIV activity.
The advantages of these compounds will be more clearly defined with further evaluation against multiple drug resistant strains. Selected compounds are currently under further biological evaluations for their broad-spectrum anti-HIV properties. J.6. Acknowledgments Support for this subproject (MSA-03-367) was provided by CONRAD, Eastern Virginia Medical School under a Cooperative Agreement (HRN-A-00-98-00020-00) with the United States Agency for International Development (USAID). The views expressed by the authors do not necessarily reflect the views ofUSAID or CONRAD.

Abstract
Reported solution-phase methods for the synthesis of 3'-fluoro-3'-deoxythymidine (FLT) are cumbersome, require purification of intermediates, and include several protecting/deprotecting steps. To circumvent these problems, a solid-phase strategy was designed for the synthesis of FLT. Thyrnidine was immobilized on trityl resin via the 5'-hydroxyl group. The subsequent mesylation of the free 3'-hydroxyl group in the presence of methanesulfonyl chloride afforded the polymer-bound 3'-0mesylthymidine. Nucleophilic substitution of the mesyl moiety by hydroxyl group in the presence of sodium hydroxide produced the polymer-bound threothymidine.
Fluorination with diethylarninosulfur trifluoride followed by acidic cleavage of the polymer-bound FLT in the presence of trifluoroacetic acid afforded FLT.

Introduction
3'-Fluoro-3'-deoxythymidine (FLT, alovudine) is a nucleoside analogue structurally related to 3'-azido-3'-deoxythymidine (AZT), a commercially available anti-human immunodeficiency virus type 1 (HIV-1) drug. FLT has a substitution of fluorine for the hydroxyl group at the 3' position of the ribose ring of thymidine, and has been reported to be one of the most active inhibitors of HIV in vitro. FLT is up to 10-fold more potent than AZT in vitro  and is at least 10 times more active than AZT in monkeys infected with suruan immunodeficiency virus . Further investigations of this compound  showed that FLT-5'-triphosphate (FLT-TP) is a potent and selective inhibitor ofHIV-1 reverse transcriptase. mv isolates with mutations resulting in multidrug resistance against all currently available reverse transcriptase inhibitors including AZT had no evidence of cross resistance to FLT (Kim et al., 1994). American Cyanamid Co discontinued the development of FLT in 1994 because of the observed hematological toxicity . However, Medivir continued to test FLT for the treatment of patients with multidrug-resistant HIV infection. The phase Ila clinical trials of FLT was successfully completed in July 2002. All patients underwent treatment without any serious side effects . FLT is currently undergoing further clinical tests.

Recently 3'-[ 18 F]fluoro-3'-deoxythymidine ([ 18 F]FLT) has been proposed as a
new marker for imaging tumor proliferation by positron emission tomography (PET) . The introduction of 18 F at the ribose rather than labelling the nucleotide with 18 F enhanced the metabolic stability of the marker . [ 18 F]FLT was predominantly taken up by proliferating cells.
The synthesis of FLT in solution phase has been carried out by using several protecting and deprotecting steps , Yun et al., 2003. These reactions are cumbersome and the intermediates need to be purified in each step. Furthermore, the overall yield is not satisfactory. Because of revival of research interest for using FLT as anti-HIV agent or as a marker in tumor imaging by PET, there is a need for an alternative facile and effective synthesis of this compound. We designed a solid-phase strategy for the synthesis of FLT using unprotected thymidine to circumvent some of the problems associated with the solution-phase methods.

Materials
All the chemicals and solvents were purchased from fisher scientific. Trityl resin was purchased from Novabiochem. All reactions were carried out in Bio-Rad polypropylene columns by shaking and mixing using a Glass-Col small tube rotator in dry conditions at room temperature unless otherwise stated. Trityl chloride resin (1.6 mmol/g) was purchased from Novabiochem. Other chemicals and reagents were purchased from Sigma-Aldrich Chemical Co. (Milwaukee, WI). The chemical structure of FLT was confirmed by nuclear magnetic resonance spectrometry ( 1 H NMR, 13 C NMR) on a NMR spectrometer ( 400 MHz) and a high-resolution PE Biosystems Mariner API time-of-flight mass spectrometer.

Synthesis
Polymer-bound thymidine (4.3). The reaction vessel containing trityl chloride resin  (4.7). Resin 4.6 (248 mg) was suspended in DCM containing 3% TF A (I 0 mL) and was shaken at room temperature for 1 h. The resin was collected by filtration. The filtrate was concentrated under reduced pressure and purified by silica-gel column chromatography using DCM and methanol as eluents (98:2, v/v) to afford FLT (39 mg, 55%). 1 H NMR, 13 C NMR, and high resolution timeof-flight electrospray mass spectrometry confirmed the structure of the compound.

184
The novelty of the method lies in its simplicity. Thymidine is mixed with trityl chloride and is thereby "captured" as an immobilized compound through the 5'hydroxyl group. Washing the support allows for the recovery of an excess of thymidine and removal of unreacted reagents, and guarantees that no unreacted starting materials remain. This makes the method very economical and cost-effective.
Trityl chloride resin has a hindered structure, thereby allowing for the regioselective reaction. The most reactive hydroxyl group (5' -hydroxyl group) of thymidine reacts selectively with hindered resin when an excess amount ofthymidine (4 eq) is used. Reported solution-phase approaches to the synthesis of FLT include a variety of protecting groups . For example, in a parallel modified solution-phase method, (Yun et al., 2003) 5'-hydroxyl group of thymidine was protected by 4,4'-dimethoxytrityl group to afford 4.8. Subsequent mesylation, basic hydrolysis, fluorination, and acidic cleavage reactions afforded FLT ( 4. 7). All the intermediates were purified by silica gel column chromatography in a time consuming process.
FLT was synthesized in a higher overall yield ( 45% overall yield) by the solidphase method when compared to that for the solution-phase method[1 8 J carried out in parallel (19% overall yield). The successful application of the solid-phase strategy for the synthesis of FLT provides insight for the synthesis of other 3'-substituted nucleosides using a similar methodology.
Additionally, a number of fatty alcohols were reacted with diisopropylphoramidous dichloride to produce the phosphitylating intermediates, which underwent coupling reactions with 3'-azido-2',3'-dideoxythymidine (AZT) and FLT followed by oxidation with t-BuOOH to yield fatty alcohol phosphate triester derivatives of AZT and FLT.

Introduction
2',3'-Dideoxynucleoside analogs are used clinically against the human immunodeficiency virus (HIV). There are numerous reasons to utilize nucleotide prodrug strategy in order to make anti-HIV nucleosides more effective against the virus. Bypassing the first rate-limiting phosphorylation step ( , increasing the lipophilicity, and enhancing the cellular uptake and half-life in blood are some of them . On entering the cell, the majority of anti-HIV nucleoside analogs, such as 3'fluoro-2 ',3 '-dideoxythymidine (FLT, alovudine) and 3 '-azido-2 ',3 '-dideoxythyrnidine (AZT, zidovudine), are phosphorylated to monophosphates, diphosphates, and triphosphates forms by host cellular kinases to show antiviral activity. Negativelycharged nucleoside monophosphates cannot be directly used because of their high polarity and poor cellular uptake. Furthermore, they are readily dephosphorylated on cell surfaces and in extracellular fluids by non-specific phosphohydrolases. In order to bypass the first rate-limiting phosphorylation step in the metabolic conversion of nucleoside analogs, numerous prodrugs of 5'-monophosphate types, such as neutral species of phosphotriester derivatives of nucleosides have been proposed , Thumann et al., 2996 with the hope that these prodrugs would release the corresponding nucleoside-monophosphates intracellularly. The phosphotriesters must have acceptable stability prior to cellular uptake and selective intracellular biotransformation of the active species. Furthermore, extensive efforts have been carried out to synthesize lipophilic prodrugs of anti-HIV nucleosides by esterification strategy . Both strategies have not yet provided an anti-HIV prodrug agent with a clear-cut therapeutic advantage for clinical use. The major challenge of developing nucleotide prodrugs has been in the selection of a suitable phosphate-masking group. By judicious choice of the alcohols used in triester formation, it may be possible to improve cellular uptake and to direct intracellular hydrolysis to nucleoside monophosphates. Thus, further research to identify prodrugs containing both phosphotriester and lipophilic groups with distinct advantages, relative to parent anti-HIV nucleosides is warranted.
Herein, we report the synthesis of uncharged fatty acyl and fatty alcohol phosphotriester derivatives of AZT and FLT ( Figure 5 .1 ). The lipophilic moieties, fatty acyls or fatty alcohols, were incorporated into the structure with the aim of improving interaction with membrane bilayers and cellular uptake of anti-HIV nucleoside phosphotriester derivatives and to release nucleoside monophosphates intracellularly, bypassing the first phosphorylation step (Figure 5.2).

Materials
Alovudine (FLT) was synthesized according to the previously reported method . Zidovudine (

2-
The solvent was evaporated using reduced pressure and the crude product was purified using silica gel column chromatography and hexane and dichloromethane (50:50) as eluting solvents to afford 5.4 (55% yield). Synthesis of fatty acyl-glycol ester conjugates 5.14-5.17.

Anti-HIV assays
Anti-HIV activities of the compounds were evaluated according to the previously reported procedure . In summary, HeLa (Human

Results and Discussion
In the first class of compounds, two identical fatty acids were linked through a glycol linker to a phosphate group, which was attached to 5'-0-position of FLT to afford bis(fatty acyl-glycol)phosphate triester derivatives. The selection of fatty acids was based on the previously reported moderate anti-HN activities of 12bromododecanoic acid, 12-azidododecanoic acid, and 12-thioethydodecanoic acid . In the second class of compounds, fatty alcohols and nucleosides, FLT and AZT, were directly attached to phosphotriester group.
The synthetic procedures used for the synthesis of phosphotriester derivatives of nucleosides were based on P(III) chemistry using phosphoramidite approach and Low temperature proved to be important for the success of this coupling reaction as shown by the failure of the reaction of 5.1 with diisopropylphosphoramidous dichloride in the presence of pyridine at room temperature. The intermediates 5.6-5.9 should be used immediately in the next reaction without purification because of the activity of the phosphorous in trivalent form in these compounds. Subsequent conversion of phosphoramidite intermediates to phosphotriesters was accomplished by treatment with FLT (1.5 mmol) in presence of 5-ethyl-lH-tetrazole (4.5 mmol) followed by in situ oxidation with tert-butyl hydroperoxide (t-BuOOH, 4.5 mmol) to obtain bis(fatty acyl-glycol)phosphotriester derivatives of FLT (5.14-5.17). The chemical structures of 5.14-5.17 were determined by 1 H NMR, 13 C NMR, 31 p NMR, and high-resolution ESI mass spectrometry (Table 5.1).  and 5-ethyl-lH-tetrazole (4.5 mmol) were added to the reaction mixture to yield 5.20-  (5.23-5.25).

Rl(O~OH
Using a single-round infection assay  with HIV-1 IIIB and transformed HeLa cells expressing HIV receptors (CD4) and coreceptors (CXCR4 and CCR5), the newly synthesized triester derivatives showed only modest antiviral activity, significantly lower than that of their parent nucleosides, AZT and FLT (IC 50 = 10 and 0.8 µM, respectively).
In summary, the results presented herein show that the synthesis of different classes of lipophilic phosphate triesters of FLT and AZT can be successfully accomplished by using P(III) chemistry. The extension of this methodology should prove to be useful for the development of lipophilic phosphotriester prodrugs of other nucleosides. The premature hydrolysis of the phosphate-masking group bond in the extracellular medium, however, may have yielded a negatively-charged diester with low cellular uptake and reduced antiviral potency. The phosphotriesters must have acceptable stability in cell culture prior to cellular uptake and selective intracellular transformation of the active species. We were not able to determine the stability of compounds because of their extremely low water solubility. The extracellular hydrolysis of phosphotriester derivatives of nucleosides has been previously reported.
For example,  reported that some of dialkyl and diaryl phosphotriester derivatives of AZT were inactive because of the rapid in vitro hydrolysis to release the nucleotide extracellularly. The utility of phosphotriester derivatives of nucleosides will be enhanced by a clearer understanding of the mechanisms pertaining to their bioconversion, uptake, and cellular incorporation.

Abstract
A liver-selective prodrug (3TCSD) of the antiviral drug lamivudine (3TC) was developed and characterized. 3TC was coupled to dextran (~25 kDa) using a succinate linker and the in vitro and in vivo behavior of the conjugate were studied using newlydeveloped size-exclusion and reversed-phase analytical methods. Synthesized 3TCSD had a purity of> 99% with a degree of substitution of 6.5 mg 3TC per 100 mg of the conjugate. Furthermore, the developed assays were precise and accurate in the concentration ranges of 0.125-20, 0.36-18, and 1-50 µg/mL for 3TC, 3TC succinate (3TCS), and 3TCSD, respectively. In vitro, the conjugate slowly released 3TC in the presence of rat liver lysosomes, whereas it was stable in the corresponding buffer. In vivo in rats, conjugation of 3TC to dextran resulted in forty-and seven-fold decreases in the clearance and volume of distribution of the drug, respectively. However, the accumulation of the conjugated 3TC in the liver was fifty-fold higher than that of the parent drug. The high accumulation of the conjugate in the liver was associated with a gradual and sustained release of 3TC in the liver. These studies indicate the feasibility of the synthesis of 3TC-succinate-dextran and its potential use for the selective delivery of 3TC to the liver.

Introduction
Lamivudine (~-L-2',3'-dideoxy-3'-thiacytidine, 3TC) is a deoxycytidine nucleoside analog that inhibits hepatitis B virus (HBV) replication and is used in the treatment of chronic hepatitis B infection . In addition to the efficacy of antiviral drugs against HBV, treatment of HBV infection is significantly dependent on the pharmacokinetics of these drugs, in particular their distribution and accumulation in the liver. To be effective, 3TC needs to enter liver cells and be phosphorylated to its active form 3TC 5'-triphosphate intracellularly before incorporation into DNA of HBV by DNA polymerase . Therefore, strategies that direct the drug to its site of action (liver) may increase the effectiveness of the drug and decrease its potential side effects · in other non-target organs. Indeed, several studies , DeVrueh et al. , 2000 have attempted to target antiviral drugs to the liver for the treatment of HBV infection.
Dextrans are glucose polymers, which are under investigation as macromolecular carriers for delivery of drugs and other therapeutic agents, such as proteins, after the systemic administration of the dextran-drug conjugates . Additionally, conjugates of dextrans with non-steroidal  and steroidal  anti-inflammatory drugs have been studied for local delivery of these drugs to the colon. In a series of studies , we previously showed that the plasma kinetics and tissue distribution of dextran carriers are dependent on the Mw of the polymer. Therefore, dextrans of different Mw's may be useful for the delivery of drugs to different tissues after the systemic administration of the conjugate. For example, dextrans with Mw' s of 20 kDa to 70 kDa showed a high degree of selectivity for the liver when liver:plasma area under the concentration-time curve (AUC) ratio was considered . The liver selectivity of these dextrans was attributed to their sizes, which restrict their passage through most vascular bed pores, while allowing unrestricted passage through the substantially larger pore sizes of the liver sinusoids , Mehvar et al., 1987.
In addition to the molecular weight, both the electric charge  and chemical modification (V ansteenkiste et al. , 1991(V ansteenkiste et al. , , Nishikawa et al., 1993 of dextrans may affect their organ and cellular distribution. For example, it has been reported that positively charged dextrans are taken up by the liver more rapidly and extensively than neutral or negatively charged dextrans . Additionally, whereas unmodified dextrans distribute to both parenchymal and non-parenchymal cells of the liver , galactosylation or mannosylation of dextrans results in selective delivery of the macromolecule to the parenchymal and non-parenchymal cells, respectively . Because electrical or chemical modifications of dextrans may alter their safety profiles (Y arnaoka et al. , 1995), we have recently used unmodified neutral dextrans for delivery of methylprednisolone to the liver for local immunosuppression . These studies showed that after systemic administration, the conjugate selectively accumulates in the liver, where it gradually releases the active drug, resulting in more intense and sustained pharrnacologic effects. Collectively, the above studies suggest that both unmodified and glycosylated dextrans may be suitable for systemic delivery of therapeutic agents to the liver.
The aim of the present study was to synthesize and characterize a conjugate of 3TC with dextran -25 kDa, intended for selective delivery of the anti-HBV drug to the liver. The conjugation of 3TC to dextran was achieved through a succinate linker, resulting in a macromolecular prodrug, potentially releasing 3TC and 3TC succinate (3TCS). Therefore, analytical methods were also developed for quantitation of the intact prodrug and simultaneous quantitation of its potential release products (3TC and 3TCS) in biological samples. In addition to in vitro characterization, the plasma pharrnacokinetics and tissue disposition of the prodrug and parent drug were also studied in rats, a species which has recently been used as a model for human HBV infection . To the best of our knowledge, this is the first report of designing a macromolecular prodrug of anti-HBV drug larnivudine for targeted delivery to the liver.

Materials
Dextran with an average M. were analytical grade and obtained through commercial sources.

Animals
Adult, male Sprague-Dawley rats were used in this study for in vitro blood and liver lysosome and in vivo disposition studies as outlined in the subsequent sections.
All the procedures involving animals in this study were consistent with the "Principles of Laboratory Animal Care" (NIH publication Vol. 25, No. 28, revised 1996) and approved by the Texas Tech University Health Sciences Center Institutional Animal Care and Use Committee.

Synthesis of 3TC-Succinate-Dextran (3TCSD) Conjugate
The complete procedure for the synthesis of 3TCSD conjugate is depicted in Schemes 6.1 and 6.2 and described in detail below. The chemical structures of final desalted products were characterized by nuclear magnetic resonance spectrometry eH NMR, 13 C NMR, 31 P NMR) determined on a Bruker DPX NMR spectrometer (

6.4)
The purity of the powder was determined using both the size exclusion (SEC) and reversed-phase (RPC) chromatographic methods. The degree of substitution of 3TC in 3TCSD was determined by hydrolysis of the conjugate under basic conditions.
To 1 mg of the conjugate were added 1 mL of 0.1 N NaOH and 0.6 mL of methanol.
After leaving at room temperature for 5 min, 30 min, and 24 h, 100 µL of the sample was micropipetted into a microcentrifuge tube containing 100 µL of 0.1 M HCl. An aliquot (50 µL) was then injected into HPLC. The concentration of the released 3TC was determined using a reversed-phase method based on 3TC standard solutions as described below.

High Performance Liquid Chromatography
Size-exclusion (SEC) and reversed-phase (RPC) chromatographic methods were developed and validated for quantitation of the conjugate 3TCSD and its potential hydrolysis products (3TC and 3TCS), respectively, in buffers or biological samples.

Reversed-Phase Liquid Chromatography (RPC)
A reversed-phase chromatographic method was developed to quantitate the

HPLC System
The HPLC instrument consisted of a 510 pump (Waters; Milford, MA), a 717 plus auto sampler (Waters; Milford, MA), and a 486 UV detector (Waters) set at a wavelength of 276 nm. The chromatographic data was managed using Empower version 2 software. Calibration curves were constructed by plotting the peak areas of 3TCSD or peak area ratios of 3TC or 3TCS over IS against the concentration in the sample using a weight of 1/concentration.

Sample Preparation
For the size exclusion chromatography, to 100 µL of plasma in microcentrifuge tubes were added 20 µL of methanol and 20 µL of 20% (w/v) trichloroacetic acid. After vortex mixing for 5 s, the samples were centrifuged in a microcentrifuge at 16,000 rpm for 5 min. A 100 µL aliquot of the supernatant was mixed with 50 µL of 0.5 M phosphate buffer (pH 6.0), and a 75-µL aliquot was injected into the HPLC.
To determine the concentrations of 3TCSD in the tissues by the SEC method, tissues were homogenized in 3 volumes of water using a homogenizer at a rate of 10,000 rpm. To 100 µL of the whole homogenate in siliconized microcentrifuge tubes were added 50 µL of 0.5 M phosphate buffer (pH 7.0) and 50 µL of cold methanol.
Samples were then briefly vortex-mixed and 40 µL of trichloroacetic acid (40%) was added to precipitate proteins. After vortex-mixing for 5 s, the samples were centrifuged in a microcentrifuge at 10,000 rpm for 3 min. A 100 µL aliquot of the supernatant was mixed with 50 µL of 1 M sodium acetate, and a 75-µL aliquot was injected into the HPLC.
The preparation of plasma samples for reversed-phase chromatography was similar to that for the SEC method with one exception; for the reversed-phase system, 50 µL of 50 µg/mL stavudine was added as internal standard to the plasma sample before protein precipitation. Similarly, the tissue samples for reversed-phase chromatography were prepared as described above for the SEC method, but without the addition of 0.5 M phosphate buffer.

Validation of Assays
The inter-run precision (%CV) and accuracy (%error) of the assays were detennined from the analysis of quality control samples (n = 5) based on reported guidelines . To determine the recovery of 3TCSD, 3TC, and 3TCS from plasma after protein precipitation, plasma samples (n = 5) containing 5 or 100 µg/mL 3TCSD, 5 or 50 µg/mL 3TC, or 1.8 or 18 µg/mL 3TCS were subjected to the above assays and the peak areas were determined. The peak areas of these samples were then compared with those containing an equal concentration of each analyte in distilled water, injected directly into the HPLC. Similarly, the recoveries of 3TCSD and 3TC from the liver samples (n = 5) containing 5 µg/mL 3TCSD or 0.5 µg/mL of 3TC were determined.

Release Characteristics in Rat Blood
Blood was obtained from rats by cardiac puncture. Approximately 4 ID of heparin was added to each rnL of blood to prevent coagulation. Immediately after the collection of blood, 3TCSD conjugate (in 10 rnM isotonic phosphate buffer at pH 7.4) was added to produce a blood concentration of 100 µg/rnL (3TC equivalent) (n = 3).
The solution was then incubated at 37 °C. Blood samples (1 rnL) were collected at 0, 3, 6, and 12 h in heparinized microcentrifuge tubes. After centrifugation of the blood, the plasma samples (100 µL) were subjected to the assays described above for the determination of 3TC, 3TCS, and/or intact 3TCSD.

Release Characteristics in Rat Liver Lysosomes
Crude lysosomal fractions were prepared from the liver of untreated rats according to the procedure described in the lysosome isolation kit (Sigma). Briefly, the rat livers were perfused with ice-cold PBS before removal of the livers. The livers were then homogenized in 4 volumes of the extraction buffer, followed by differential centrifugation for isolating the lysosomal fraction. The protein concentrations in Iysosomal preparations were determined by Bio-Rad protein assay (Bio-Rad, Herecules, CA, USA). The activity of acid phosphatase, a lysosomal marker, in the preparation was tested using a commercial kit (Sigma). The specific enzyme activity in the lysosomal fraction was >9-fold that in the liver homogenate.

In Vivo Disposition
A total of 42 adult male Sprague-Dawley rats were divided equally into two groups of 21 rats each, treated with 3TCSD or 3TC. The mean± S.D. of the body weights of rats were 241 ± 8 and 240 ± 6 g for the 3TCSD-and 3TC-injected groups, respectively. The animals had free access to drinking water and rat chow before and during the course of experiments.
Single iv bolus doses (5 mg/kg; 3TC equivalent) of 3TCSD or the parent drug 3TC were injected into the penile vein of the animals under isoflurane anesthesia.
Immediately after excision, the tissues were rinsed in ice-cold saline solution to remove excess blood. Afterwards, the tissues were blotted dry and kept frozen until analysis. After centrifugation of the blood in a pre-chilled and heparin-coated rnicrocentrifuge tube, the plasma was collected. To prevent in vitro hydrolysis of 3TCSD during storage, plasma (500 µL) was mixed with a 10% acetic acid solution (100 µL). Plasma, tissue, and urine samples were kept frozen at -80 °C until analysis.

Pharmacokinetic Analysis
Non concentrations of 3TCSD or 3TC in plasma (Co) after the injection of the conjugate or parent drug were estimated by the program. The concentrations of drugs in the tissues were corrected  for the residual blood using the blood volume fraction of different tissues.

Statistical Analysis
Because of destructive sampling procedure used for the collection of blood and tissues from different animals at each time point, the composite kinetic parameter AUC could not be obtained for individual rats . Therefore the variance of AUC was estimated by a reported  procedure based on the standard error of mean and number of samples at each time point. The pairwise comparison of AUCs was then carried out at an a level of 0.05 and a Bonferroni-adjusted a of 0.05 or 0.0167 for pairwise comparison of two (1 comparison) or three (3 comparison) means, respectively. The critical values of Z (Zcrit) for the two-sided test using the Bonferroni-adjusted a of 0.05 and 0.0167 were 1.96 and 2.39, respectively, and the observed Z (Zobs) was calculated as reported before . A Zobs value > Zcrit was used as an indication of a significant difference between the AU Cs.
The differences between groups in their kinetic parameters that could be estimated for individual rats (e.g., Cmax and CLR) were determined using a two-tailed unpaired t test at a significance level (a) of 0.05. When possible, data are presented as mean ± S.D.

Results and Discussion
Targeted delivery of anti-HBV drugs to the liver, for the treatment of viral hepatic diseases, has attracted the attention of scientists for several years. In one of the first publications on this subject, Balboni et al. showed that conjugation of cytosine arabinoside and 5-fluorodeoxyuridine to albumin resulted in accumulation of these antiviral drugs in the mouse liver cells, increasing the effectiveness of the drugs in comparison with the free drugs . Further works by Fiume and colleagues  modified this strategy by the use of lactosaminated albumin as the carrier, selectively targeting the asialoglycoprotein receptors on the hepatocytes with the galactose moiety of lactose. Additionally, conjugates of antiviral drugs with galactosylated poly-L-lysine, instead of albumin, have been .synthesized and tested by the same group . Others have used arabinogalactan  or glycosylated lipoproteins , DeVrueh et al., 2000, as liver-accumulating carriers, or prodrugs that release the active drug based on metabolism by hepatic cytochrome P450  for targeted delivery of antiviral drugs to the liver. These studies support the general concept that targeted delivery of antiviral drugs to the liver potentially increases the efficacy of these drugs in the treatment of viral liver infections while, at the same time, decreasing their toxic effects in other tissues. However, the choice of carrier and targeting moieties need optimization to reduce carrier-related side effects, such as increased alkaline phosphatase levels seen with lactosylated human serum albumin , or to improve drawbacks associated with variability in density and affinity seen with the asialoglycoprotein receptors .

Synthesis and Characterization of 3TCSD
In addition to our study, we are aware of only one other study that used dextran as a macromolecular prodrug of antiviral agents . In that study, acyclovir was conjugated to dextran with a Mw of 40 kDa. However, to synthesize the conjugate, dextran was first oxidized to dextran aldehyde before direct reaction of the aldehyde groups of dextran with the amine group of acyclovir to produce a Schiff s base, without any spacer between acyclovir and dextran. Modification of dextrans, including their oxidation, may alter the degradation and safety profiles of native dextrans, which have been used clinically for almost six decades as plasma volume expanders . Therefore, we designed this strategy for conjugation of 3TC to dextran to minimize changing the structure of native dextran molecule.
Consequently, 3TC was coupled to dextran through a succinate linker, thus avoiding a need for oxidation of dextran as previously reported .
In the present study, we used dextran with a Mw of -25 kDa as an alternative to the currently studied carriers for the selective delivery of 3TC to the liver in HBV treatment. The synthesis of the conjugate was challenging because of the presence of free N4-amino group in the structure of 3TC, which had to be protected before the reaction of 3TC with succinic anhydride for the synthesis of 3TCS. Lamivudine (3TC) was conjugated to dextran using a succinate linker in two major steps by synthesis of 232 5'-0-succinate ester of the drug (Scheme 6.1), followed by the reaction of the ester conjugate with dextran (Scheme 6.2).
Compound 6.1 was reacted with dextran in the presence of DIC and DMAP to afford N 4 -(DMTr)larnivudine-succinate-dextran conjugate (6.3) that was deprotected in the presence of acetic acid to afford larnivudine-succinate-dextran conjugate (3TCSD, 6.4) with a reasonable degree of 3TC substitution (6.5%, w/w) (Scheme 6.2). The purity of the conjugate was > 99% as determined by the reversed phase analysis of a 100-µg/mL solution of the conjugate (n = 3), which contained only 0.296 ± 0.041 µg/mL of 3TC without any measurable peak of 3TCS. The degree of substitution of the final product, which was obtained by the base hydrolysis of 3TCSD to 3TC followed by reversed-phase quantitation of the released 3TC, was 6.5 mg larnivudine in 100 mg of 3TCSD powder. The complete hydrolysis of 3TCSD was confirmed by a complete disappearance of the 3TCSD peak in the SEC method.
Additionally, the area of the unhydrolyzed 3TCSD peak in the SEC method was the same as the area of the released 3TC peak in the reversed-phase method after appropriate volume correction, indicating that the degree of substitution may be determined directly from the area of 3TCSD peak without hydrolysis.

6.3
Larnivudine-succinate-dextran (3TCSD) has two ester bonds in its structure (Scheme 6.2), whose hydrolysis potentially releases 3TC or 3TCS. Therefore, to completely determine the fate of the conjugate and its potential release products in both in vitro and in vivo experiments, analytical methods capable of measurement of all three moieties are needed. To quantitate 3TCSD in both aqueous and biological samples, a SEC assay was developed. The method is capable of separating the conjugate peak from the endogenous peaks (Figure 6.1) and accurately and precisely quantitating the conjugate (Table 6.2).

235
Chromatograms of plasma samples taken from a rat before (blank) and 180 min after the administration of a single 5-mg/kg dose (3TC equivalent) of 3TCSD are demonstrated in Figure 6.1. Dextran-lamivudine succinate eluted at -5.3 min and was well separated from the endogenous peaks in plasma, which eluted after the conjugate peak ( Figure 6.1 ). Additionally, the relationship between the peak area of 3TCSD and the detector response was linear (/ 2: 0.998) over the studied range 1 to 50 µg/mL .
The results of the assay validation in plasma are presented in Table 6.1.
Excellent accuracy of the assay is demonstrated by error values of < 1 % for all the concentrations (even at the lowest concentration of 1 µg/mL). The assay is also deemed precise because the C.V. values are < 13% for the inter-run data (Table 6.1).
Based on the data presented in Table 1, the limit of quantitation of the assay is at least equal to 1 µg/mL.  The recovery of 3TCSD from plasma was 88.0 ± 4.2% and 94.1 ± 4.6% at concentrations of 5 and 50 µg/mL, respectively. The recovery of 3TCSD from liver was 94.2 ± 9.5% at a concentration of 5 µg/mL.

Reversed-Phase Chromatographic Method.
Several HPLC assays are currently available for determination of 3 TC m plasma and/or tissues . These assays use either solid-phase extraction  or protein precipitation  for sample preparation. Additionally, all of them use reversed-phase chromatography with C1s , Cs , or phenyl  columns for separation of 3TC from the endogenous peaks. To quantitate both 3TC and 3TCS simultaneously in our samples, we had to modify these assays. We found that a protein precipitation method with a combination of methanol and trichloroacetic acid resulted in the highest recovery (2' .:. 86%) for both 3TC and 3TCS from biological samples. Additionally, our chromatographic conditions resulted in complete separation of 3TC, 3TCS, and IS from the endogenous peaks (Figure 6.2) and accurate and precise quantitation of 3TC and 3TCS (Table 6.2). Validation results for the reversed-phase assay of 3TC and 3TCS in plasma are presented in Table 6.2. Excellent accuracy of the assay is demonstrated by error values of< 11 % for all the concentrations. The assay is also deemed precise because the C.V.
values are< 9% for all the concentrations except for the lowest concentration of 3TC, which showed an acceptable C.V. of 16.8% (Table 6.2). Based on the data presented in Table 6.2, the limit of quantitation of the assay is 0.125 and 0.36 µg/mL for 3TC and 3TCS, respectively. Chromatograms of plasma samples taken from a rat before (A) and 15 min after (B) the administration of a single 5 mg/kg dose of 3TC to rats and at 3 h after in vitro incubation of 3TCSD with rat blood (C), subjected to the reversed-phase chromatographic method. Sample B contained 1.84 µg/rnL 3TC, and sample C contained 3.97 and 5.12 µg/rnL 3TC and 3TCS, respectively. The recovery of 3TC from the liver samples at a concentration of 0.5 µg/mL was 95.7 ± 5.25%.

Release Characteristics in Buffers
The SEC and reversed-phase HPLC assays described above were used to investigate the stability of 3TCSD and formation of 3TC and 3TCS at 37 °C in buffers at pH 4.4 (Figure 6. .5) suggests that the degradation of 3TCSD in blood is only due to a slow chemical hydrolysis of the conjugate. This is because the degradation half lives of 3TCSD in buffer at pH 7.4 and in blood are almost identical (108 and 110 h, respectively), indicating lack of enzymatic hydrolysis of the conjugate in blood. This data is consistent with our previous report on a dextran conjugate of methylprednisolone succinate, which similarly showed little, if any, enzymatic hydrolysis in blood . Nevertheless, the relatively long half life of 3TCSD in blood in vitro (110 h) indicates that the conjugate is stable in blood long enough for the uptake by the liver.   . In most cases, error bars are too small to be observable.

Release Characteristics in Rat Liver Lysosomes
Previous works  on other ester conjugates of dextrans have proposed that the intact conjugates are resistant to the effects of esterases because of steric hindrance of dextrans. However, after exposure of the conjugate to dextranases, which depolymerize dextrans, esterases can hydrolyze the ester bonds in the conjugate. It has been reported ) that dextranases are present in lysosomes, predominantly in the liver. Additionally, dextrans enter cells via endocytosis, which results in their accumulation in the lysosomal compartments . Therefore, lysosomal exposure of dextran conjugates to dextranases may facilitate their further enzymatic hydrolysis. In vitro data showed significant amounts of 3TC released at pH 4 only in the presence of rat liver lysosomes (Figure 6.6). It was found that the presence of lysosomes in the medium induced a slow release of 3TC (7.36 ± 0.30 µg/rnL after 24 h) without any detectable 3TCS. In contrast, the release of 3TC or 3TCS in the same medium in the absence of lysosomes was negligible (Figure 6.6). These results confirms our hypothesis of 3TC release in lysosomes

In Vivo Disposition
Plasma concentration-time courses of 3TCSD and 3TC after the injection of single 5-mg/kg (3TC equivalent) doses of the prodrug or the parent drug are depicted in Figure 6.7. After the injection of unconjugated 3TC, the drug was eliminated relatively rapidly and could not be detected at the last sampling time of 3 h ( Figure   6.7). In contrast, the concentrations of 3TCSD were several folds higher than those of 3TC and remained high even at the last sampling time after the injection of the conjugate. The plasma concentrations of both 3TC and 3TCSD declined multiexponentially ( Figure 6.7). Additionally, no quantifiable concentrations of 3TC or 3TCS were detected in plasma of 3TCSD-injected rats.
The plasma pharmacokinetic parameters of 3TC and 3TCSD after the injection of the parent drug or the prodrug are listed in Table 6.3. Conjugation of 3TC to dextran resulted in an almost forty-fold decrease in the total CL of the drug and a similar degree of increase in its plasma AUC. The decrease in total CL was associated with an eighty-fold decrease in the CLR of the drug when conjugated to dextran.
Consequently, the fraction of the drug excreted unchanged into urine decreased from 65% for the parent drug to 31 % for the conjugate (Table 6.3). Additionally, dextran conjugation decreased the extent of distribution of the drug as reflected in V ss and V z values (Table 6.3), although to a lesser degree than that seen with the CL values (7-  The literature information about the pharmacokinetics of 3TC in rats is limited. We are aware of only one study reporting the plasma concentration-time course and AUC of the drug after a 230 mg/kg iv dose . However, no other pharmacokinetic parameters were reported in that study. We estimated a CL value of 27. l mL/min/kg from the AUC and dose values reported in that study. This value is very close to our value of 32.6 mL/min/kg after a much smaller dose of 5 mg/kg (Table 3), suggesting linear pharmacokinetics of 3TC in the dosage range of 5-230 mg/kg. The hepatic concentration-time courses (top) and AUC values (bottom) of the conjugate and regenerated 3TC after the administration of the conjugate and those of 3TC after the injection of the parent drug itself are depicted in Figure 6.8. After the injection of 3TC, the hepatic concentrations of the drug were measurable only in the first two samples; no detectable 3TC levels were found in the liver beyond 15 min following the administration of the parent drug ( Figure 6.8, top). However, relatively high concentrations of 3 TCSD were detected until the last sampling point. This resulted in > fifty-fold higher (P<0.0001) AUCs for the conjugated 3TC, compared with the parent drug ( Figure 6. 8, bottom). Additionally, the conjugated 3TC slowly released 3TC in the liver (Figure 6.8, top) with an AUC that was approximately 2.5fold larger (P<0.0001) than that of the parent drug during the sampling time ( Figure   6. 8, bottom). No measurable concentrations of 3TCS were found in the liver.
Conjugation of 3TC to dextran substantially decreased both the CL and volume of distribution of 3TC (Table 6.3), while at the same time increasing the accumulation of the drug in the liver (Figure 8). A similar disposition pattern was also observed for a dextran prodrug of methylprednisolone with a succinate linker .
The simultaneous decrease in volume of distribution and increase in liver accumulation upon dextran conjugation is due to the relatively large pore sizes of the liver sinusoids in comparison with those in the vascular beds of most other organs, allowing unrestricted access of the conjugate to the space of Disse and internalization of the conjugate .
Because of limited number of liver samples with quantifiable 3TC concentrations after 3TC injection and slow apparent declines in the liver concentrations of the conjugate and regenerated 3TC after conjugate injection ( Figure   6.8, top), terminal half lives were not estimated for the liver. Consequently, liver AUC values depicted in Figure 6.8 (bottom panel) are related to the sampling period without extrapolation to infinity. However, the significant differences between the liver AUCs after 3TC and 3TCSD injections would be expected to be even larger if extrapolated AUCs were used. This is because the apparent decline in the 3TCSD or regenerated 3TC concentrations was substantially slower than that in the liver concentrations of the parent drug ( Figure 6.8, top).
The slow decline in the 3TCSD concentrations in the liver (Figure 6.8, top) is in agreement with previous ex vivo studies using isolated perfused rat livers  showing that, once in the liver, the rate of exocytosis of dextrans from the liver cells is negligible. Therefore, the main reason for the decline in the liver concentrations of the conjugate is its elimination by the liver, which is relatively slow based on the in vitro studies in lysosomes (Figure 6.6). Nevertheless, in contrast to undetectable concentrations of 3TC in the liver beyond 15 min after the parent drug injection, high concentrations of the conjugate in the liver were associated with a gradual and sustained generation of the parent drug in this tissue (Figure 6 Table 4. In addition to the liver (Figure 6.8 infection . Therefore, accumulation of the conjugate and release of 3TC in this organ, in addition to that in the liver, may be of potential therapeutic benefit in chronic HBV infection. In contrast to the liver (Figure 6.8) and kidneys ( Figure 6.9), the concentrations of the conjugate and/or regenerated 3TC were very low or undetectable in the lungs, spleen, and heart (brain) after 3TCSD injection (data not shown). Additionally, no released 3TC could be detected in these tissues. The concentrations of 3TC after 3TC injection in organs other than kidney were also low or below the limit of detection in most of the samples.
Although our conjugate released both 3 TC and 3 TCS in vitro in buffers (Figures 6.3 and 6.4) and rat blood ( Figure 6.5), only 3TC was observed in biological samples after in vivo administration of the conjugate (Figures 6. 8 and 6.9). This suggests rapid conversion of any released 3TCS to 3TC in vivo. Our previous work on a methylprednisolone-succinate-dextran conjugate  also showed similar results, in that no measurable concentrations of methylprednisolone succinate were found in biological samples after the administration of the conjugate to rats. The rapid conversion of 3TCS to 3TC in vivo is advantageous because 3TCS is not expected to have any biological effects by itself. This is due to the fact that to be effective, 3TC should undergo stepwise phosphorylation at the 5'-position to monophosphate, diphosphate, and triphosphate form before 5'-triphosphate 3TC is incorporated into the viral DNA . Therefore, 3TCS, which has a succinate moiety at the 5 '-position (Schemes 6.1, and 6.2), cannot be converted to the active 5' -triphosphate 3TC without first being converted to 3TC.

255
The tissue exposure to 3TC after the administrations of the equivalent doses of the conjugate and the parent drug is the most relevant parameter in terms of targeting index. The ratios of 3TC AUCs after the conjugate and parent drug administration were 2.41 for the liver, 0.721 for the kidneys, and zero for the remaining studied tissues. These data clearly show that the conjugation of 3TC to dextran only increased the targeting of the drug to the liver. On the other hand, conjugation decreased accumulation of 3TC in all the other tissues except the kidneys, where conjugation did not have any significant effect. The significantly higher accumulation of 3TC in the liver is in conformity with the hypothesis of the study, i.e., conjugation of 3TC with dextran allows preferential accumulation of the drug in the liver.

Conclusions
In conclusion, a method is presented for the synthesis of a conjugate of native dextran with the antiviral drug 3TC using a succinate linker. Additionally, validated size-exclusion and reversed phase assays are developed for the determination of purity, in vitro release characteristics, and in vivo disposition of 3TCSD. Using these methods, the parent conjugate and its degradation products, 3TC and 3TCS, may be quantitated in non-biological and biological samples. In vitro studies indicate an evidence of lysosomal degradation and relative stability of the conjugate in buffers and blood. Additionally, in vivo studies after the administration of the conjugate or the parent drug to rats suggest that the conjugation increases the delivery of the drug to the liver, resulting in higher exposure of the liver to the regenerated antiviral drug.

Abstract
Sulfonate and sulfate polyanionic microbicides, such as sodium cellulose sulfate (CS, 7.1), are inhibitors of HIV entry and sperm-function. CS was first reacted with 2-bromoacetic acid to yield CS-acetate (CSA, 7.2), which was further reacted with AZT, FLT, and 3TC to give AZT-CSA (7.3, 1.78% loading), FLT-CSA (7.4, 1.43% loading), and 3TC-CSA (7.5, 1.07% loading), respectively. Furthermore, CS was conjugated to nucleoside analogs (AZT, FLT, and 3TC) through succinate linker to produce bifunctional nucleoside-CS conjugates. For the synthesis of the conjugates containing succinate linker, CS was reacted with AZT-succinate and FLT-succinate to give AZT-succinate-CS (7.6, 18.48% loading) and FLT-succinate-CS (7.7, 7.87% loading), respectively. These conjugates were designed to improve the anti-HIV profile of parent compounds. It was expected that extracellular hydrolysis of conjugates will release nucleosides and CS providing broad-spectrum activity, a higher barrier to drug resistance, microbicidal, and contraceptive activity. The anti-HIV activities of the conjugates were evaluated and were compared with the corresponding physical mixtures of nucleoside analogs and anionic polymers and with different classes of polyanionic polymers, such as dextran acetate, cellulose phosphate, and cellulose acetate. Cellulose acetate, cellulose phosphate, and dextran acetate were found to have no anti-HIV activities, suggesting the importance of the presence of sulfate on the cellulose for generating anti-HIV activities. The conjugates containing CS-acetate were found to be more potent than CS and other derivatives. AZT-CSA (7.3) and FLT-CSA conjugates (7.4) exhibited higher anti-HIV activities than CS (7.1) and AZT-succinate-CS and FLT-succinate-CS-conjugates (7.6 and 7.7). The presence of both sulfate and the acetate groups on cellulose were critical for generating maximum anti-HIV activity, possibly by increasing the negative charge density that is essential for blocking HIV entry. However, the compounds were less potent than the corresponding physical mixtures of nucleoside analogs with CSA (7.2), due to incomplete hydrolysis.

Introduction
Sulfonate and sulfate polyanionic microbicides inhibit the HIV entry into the cell  and sperm-function . Sodium cellulose sulfate (Ushercell, CS) is a polyanionic non-cytotoxic microbicide and its 6% gel has been studied as vaginal contraceptive and a broad spectrum antimicrobial agent .
Polyanionic sulfates are the polymers such as dextran, cellulose, styrene etc with sulfate groups in their structure . The negatively-charged sulfate group interacts with positively-charged viral protein . Viral envelop protein gp120 is know to have positively charged residues in its V3 loop. Viral entry in the host cell depends on the interaction with the negatively-charged surface of the coreceptors CXCR4 and CCR5 . Polyanionic sulfates exhibit their inhibitory activity by blocking the interaction between negatively-charged coreceptors and 264 positively charged V3 loop of viral gp120 protein , Baba et al., 1998. according to the previously reported procedure . In summary, HeLa
For the synthesis of sodium cellulose sulfate conjugates linked to AZT or FLT through a succinate linker, AZT and FLT were first reacted with succinic anhydride to synthesize AZT succinate and FLT succinate, which were then reacted with cellulose sulfate to afford cellulose sulfate succinate conjugates of AZT (7.6, 18.48% loading) and FLT (7.7, 7.87% loading) (Scheme 7.2). The Purity and percentage of loading of the nucleosides in the conjugates were determined using the SEC method as described above. Scheme 7.2. Synthesis of AZT-succinate-CS (7.6) and FLT-succinate-CS (7.7) conjugates. Table 7 .1 shows the antiviral activities of the cellulose sulfate-nucleoside conjugates with different loading percentages compared to those of CS, AZT, and FLT. CS exhibited approximately I 0-fold higher activity against X4 virus (IIIB strain, ECso = 5.9 µg/ml)) than R5 virus (BaL strain, ECso = 62.5 µg/ml) ( Table 7.1). The data is consistent with the reported data  that X4 strains possess higher number of positive charges on their V3 loop of gp120

Anti-HIV Activities Against Cell-Free and Cell-Associated Strains
protein surface compared to those of R5 strains. Higher number of positive charges on V3 loop of X4 strains makes them more susceptible to interaction with anionic polymer CS. On the other hand, conjugation of CS with nucleosides in all conjugates made R5 strains susceptible more susceptible sometimes even more than X4 strains. sulfate increased the anti-HIV activity of CS, but cellulose acetate (7.14) was completely inactive (Table 7.2), suggesting that the presence of sulfate of cellulose is critical in maintaining the anti-HIV activity of the polymer.
3TC-CSA conjugate (7.5, 1.07% loading) showed significantly different anti-HIV profile compared with AZT-CSA and FLT-CSA conjugates. Conjugate 7.5 showed almost 37-fold less anti-HIV activity against X4 strain, and was not active against cell-associated virus. The poor anti-HIV activity of 3TC conjugate could be due to the interaction of free 4-arnino group of 3TC with the negatively charged groups on CSA that reduces the available free negative charge of the conjugate for binding to V3 loops of the virus.
To determine the contribution of sulfate group in generating anti-HIV activity of CS, cellulose phosphate and dextran acetate were studied as controls. These compounds were found to be totally inactive in viral inhibition assay (Ta~le 7.2), suggesting that negatively charged acetate and phosphate alone are not sufficient for efficient interactions with V3 loops of the virus.
Both conjugates demonstrated at least 6-fold higher anti-HIV activity against X4 strains than CS suggesting the contribution of nucleoside analog in anti-HIV activity (Table 7.1). However, 7.6 and 7.7 exhibited less anti-HIV activity against the cell associated virus (EC 50 = 75-88 µg/mL) than that of CS (EC 50 = 2.5 µg/mL). The conjugates also showed less anti-HIV activity against cell-free virus when compared with AZT and FLT. aCytotoxicity assay; b50% Effective concentration; 0 Viral entry inhibition assay (lymphocytotropic strain); dViral entry inhibition assay (monocytotropic strain); °Cell-to-cell transmission assay (UIB). Although 7.6 and 7.7 had 10 and 6 times higher loading values than the corresponding conjugates substituted with acetate (7.3 and 7.4), respectively, the cellulose sulfate succinate conjugates were generally less active than cellulose sulfate acetate conjugates (Table 7.1). The less anti-HIV activity of conjugates containing succinate linker, despite of their higher nucleoside loading, compared to those containing acetate linker could be due to incomplete hydrolysis of 7.6 and 7.7 to parent nucleosides or the hydrolysis of the conjugate to generate inactive nucleosidesuccinate derivatives instead of free nucleosides. Furthermore, upon hydrolysis of conjugates containing acetate linker, the acetate group will remain intact on the CS that contributes to overall negative charge of the anionic polymer. Table 7.2 shows the anti-HIV activities (in µg/mL) of the nucleoside-CS conjugates compared with the corresponding physical mixtures. The physical mixture of AZT ( 1. 78%) + CSA (7 .8) showed slightly better anti-HIV activity against cell free virus than the corresponding AZT-CSA conjugate (7.3). In comparison to 7.3, the anti-HIV activity of 7.8 was almost 1.5-and 3.5-fold higher against lymphocytotropic and monocytotropic strains of cell-free virus, respectively, but was 1.5-fold less active against cell associated virus.
When CSA was replaced with CS in the physical mixture with AZT in 7.11 (17.2%), the anti-HIV activity was reduced significantly (2-9 fold) against cell-free virus when compared with the corresponding CSA conjugate 7.8 containing even a lower loading of AZT (l.78%), suggesting the major contribution of CSA in overall activity. The physical mixture of AZT with CS (7.11, 17.2%) that was in a similar ratio to the AZT-succinate-CS conjugate (7.6), showed 2-7-fold less anti-HIV activity against cell-free virus when compared with AZT-CSA conjugate 7.3 (1.78%) and AZT-succinate-CS conjugate 7.6 (17.2%).
Similar results were found in the case of FLT-CSA conjugate (7.9) and 3TC-CSA (7.5) when compared with the corresponding physical mixtures, 7.4 and 7.10, respectively. Physical mixture of CSA and FLT (7.9, 1.43%) exhibited approximately 3-5 fold higher anti-HIV activity against VBI (cell-free virus) when compared with the corresponding FLT-CSA conjugate (7.4, 1.43%). In the case of 7.10, the free amino group was not able to reduce the anionic interactions of sulfate group as described above in the conjugate 7.5.
However, the overall anti-HIV activity of 7.9 was reduced by 8-22 times against cell-free virus when CSA in 7.9 (l.43%) was replaced with CS in 7.12 (1.43%), suggesting the importance of CSA in overall anti-HIV activity.
The anti-HIV activity in 7.12 was increased by 10-20 fold against cell-free virus when the FLT content was increased from 1.43% to 7.85% in 7.13. The data indicates that higher concentration of FLT in the physical mixture improves activity.
The physical mixture of FLT with CS (7.13%) still exhibited 6-fold lower anti-HIV activity against cell-associated virus when compared with FLT-CSA conjugate (7.9, 1.43%). These data confirm that CSA is a more appropriate polymer for conjugation with nucleosides or making physical mixtures since all CSA derivatives exhibited better anti-HIV profile when compared with CS derivatives.
The physical mixture of AZT + cellulose acetate (7.17, 1.78%) showed significantly less activity against cell-free virus when compared with AZT-CSA conjugate (7.3, 1.78%) and AZT + CSA (7.8, 1.78%) and was inactive against cellassociated virus. Similarly, the physical mixture of FLT + cellulose acetate (7.18, 1.43%) showed 9-21 fold less activity against cell-free virus when compared with FLT + CSA (7.9, 1.43%) and was inactive against cell-associated virus. The result was not surprising since cellulose acetate is inactive polymer and the percentage of AZT or FLT were low in the physical mixture.
In general, the conjugation of nucleosides with CS provided better anti-HIV profile against both X4 and RS strains of virus. The substitution of acetate group on CS improved the anti-HIV activity, possibly by creating new negative charges after hydrolysis or the presence of free acetate groups on the polymer. Succinate spacer was less optimal than the acetate group for linking of the nucleoside with CS. •cTS: Cytotoxicity assay; rEC(50) = 50% effective concentration; 8 VBl(IIIB): Viral entry inhibition assay (lymphocytotropic strain); hVBI(BaL): Viral entry inhibition assay (monocytotropic strain); iCTC: cell-to-cell transmission assay (IIIB).

Anti-HIV activities against Multi-Drug Resistant (MDR) Isolates
AZT-CSA (7.3) and FLT-CSA (7.4) conjugates were evaluated for their anti-HIV activities against MDR virus and the data were compared with the controls CS and dextran sulfate (Table 7.3). Both CS and dextran sulfate were active against MDR virus (ICso = 1.61-3.12 µg/mL), but showed less activity against R5 strain of virus (ICso > 15 µg/mL). the result was expected since R5 virus has less positively charged V3 loops in gp 120 protein required for interactions with anionic polymers.
On the other hand, AZT-CSA (7.3) showed almost similar anti-HIV activity against both R5 and MDR strains than CS. AZT-CSA (7.3) was more effective than CS, against the R5 HIV-1 lab-adapted strain BaL. The higher activity against R5 strain is a result of AZT attachment. Furthermore, AZT is not active against MDR strain and hence its conjugate with CS in AZT-CSA (7.3), is 2-fold less active than CS against MDR strain. Similarly FLT-CSA (7.4) showed significantly higher activity against R5 strain versus CS (7.1) due to the presence of FLT in the conjugate. The anti-HIV activity of 7.4 against MDR is 5-fold higher than R5 strains, since released FLT has anti-HIV activity against AZT-MDR resistant.

Contraceptive activity
AZT-CSA conjugate (7.3) was selected for in vivo testing in rats for contraceptive activity. Application of CS and 7.3 in female rats prevented the pregnancy to 100% (Table 7.4). CS is known to have contraceptive properties and this result indicated that contraceptive property of the CS is retained after the AZT conjugation to the polymer.  Furthermore, AZT-CSA and FLT-CSA conjugates were more effective than CS against both X4 and RS HIV-1 viruses. The above-described conjugates present the advantage of not displaying weaker activity against HIV RS strains. Although in weight the AZT-CSA and AZT were similarly potent against cell-free virus, in moles (based on CS -2 x 10 6 Daltons), the conjugate was 5 orders of magnitude more potent (from µM to subnanomolar). Furthermore, unlike AZT, the conjugate was consistently active against cell-associated HIV.
This study presents an alternative approach for designing more optimal anti-HIV agents that may have broad-spectrum anti-HIV activities against cell-free, cellassociated and MDR virus by targeting both HIV entry and reverse transcription in HIV life cycle.