Pulmonary bioavailability of a phosphorothioate oligonucleotide (CGP 64128A): comparison with other delivery routes.

PURPOSE: Phosphorothioate antisense oligodeoxynucleotides are promising therapeutic candidates. When given systemically in clinical trials they are administered via slow intravenous infusion to avoid their putative plasma concentration-dependent haemodynamic side-effects. In this study, we have evaluated alternative parenteral and non-parenteral administration routes which have the potential to enhance the therapeutic and commercial potential of these agents. METHODS: The delivery of CGP 64128A by intravenous, subcutaneous, intra-peritoneal, oral and intra-tracheal (pulmonary) routes was investigated in rats using radiolabelled compound and supported by more specific capillary gel electrophoretic analyses. RESULTS: Intravenously administered CGP 64128A exhibited the rapid blood clearance and distinctive tissue distribution which are typical for phosphorothioate oligodeoxynucleotides. Subcutaneous and intraperitoneal administration resulted in significant bioavailabilities (30.9% and 28.1% over 360 min, respectively) and reduced peak plasma levels when compared with intravenous dosing. Administration via the gastrointestinal tract gave negligible bioavailability (< 2%). Intra-tracheal administration resulted in significant but dose-dependent bioavailabilities of 3.2, 16.5 and 39.8% at 0.06, 0.6 and 6.0 mg/kg, respectively. CONCLUSIONS: Significant bioavailabilities of CGP 64128A were achieved following subcutaneous, intra-peritoneal and intra-tracheal administration. Pulmonary delivery represents a promising mode of non-parenteral dosing for antisense oligonucleotides. The dose-dependent increase in pulmonary bioavailability suggests that low doses may be retained in the lungs for local effects whereas higher doses may be suitable for the treatment of a broader spectrum of systemic diseases.

Studies on the uptake of low molecular weight monomeric tris-galactosyl conjugates by the rat liver.

We have attempted to direct low molecular weight compounds to the liver via the internalizing asialoglycoprotein receptor on parenchymal cells by conjugation to a monomeric triantennary galactosyl cluster. Acetate and a hypolipidaemic ansamycin were derivatized and the biodistribution of the conjugates was determined 250 sec and 30 min after administration to Wistar rats. The ansamycin conjugate (CGH46) was rapidly cleared from the circulation by the liver; after 250 sec, 64% of the radiolabelled dose was found in the liver compared to 18% in the blood. However, the distribution of the conjugate did not differ significantly from that of unconjugated ansamycin (CGH45). Tris-galactosyl acetate showed no capacity to localize in the liver, with only 2% recovered from that organ 250 sec after administration compared to 38% in the blood and 13-18% in the kidneys, skin and muscle. Extraction efficiency of CGH46 by isolated perfused rat livers was almost 20% of the administered dose and this value was not significantly changed by co-administration of specific inhibitors of the uptake process. It is concluded that derivatization of low molecular weight molecules with monomeric triantennary galactosyl residues is unlikely to increase their affinity for the liver.

Hypolipidaemic activity of a novel acyclic ansamycin derivative. Synthesis and pharmacokinetic and pharmacodynamic studies in rodents.

Certain classes of the antibacterial agent rifamycin SV have recently been shown to possess marked hypolipidaemic activity. An acyclic oxazolylrifamycin has been prepared and its hypolipidaemic properties evaluated; it was found to be significantly more potent when administered orally to Wistar and Sprague-Dawley rats than other previously described rifamycin hypolipidaemics. The plasma decay rate of a bolus of intravenously administered 125I-LDL (low density lipoprotein) was significantly greater in treated rats than in rats receiving vehicle alone, compatible with a drug-induced increase in LDL catabolism. A bolus of radiolabelled drug was rapidly removed from the circulation by the liver, the presumed target organ. Compound 3 constitutes the first example of a new class of acyclic hypolipidaemic ansamycins.

The transport and metabolism of the uridine mononucleotides by rat jejunum in vitro.

1. Both uridine 3'-monophosphate (3'-UMP) and uridine 5'-monophosphate (5'-UMP) when perfused through the lumen of isolated rat jejunum gave rise to uracil as the only transported pyrimidine appearing in the serosal medium; neither the nucleotide nor the nucleoside could be detected in the serosal fluid. 2. There was a low level of the nucleoside, uridine, in the luminal fluid after the nucleotide had passed through the jejunal segment. Luminal nucleoside appearance was more marked from the 3' form of the nucleotide. 3. The hydrolysis of the nucleotides to the nucleoside form occurred via a brush-border membrane enzyme, which had the same maximal velocity (Vmax) for the two nucleotides (699 +/- 35 and 747 +/- 10 nmol min-1 (mg protein)-1 for 3'-UMP and 5'-UMP, respectively) but a different Michaelis constant (Km) so that 3'-UMP (Km = 58 +/- 3 microM) hydrolysis is favoured over 5'-UMP hydrolysis (Km = 108 +/- microM) at lower concentrations. 4. At 0.05 mM, luminal 3'-UMP gave rise to a higher rate of serosal uracil appearance than luminal 5'-UMP, but at higher luminal concentrations (0.1-0.2 mM) the rate of serosal uracil appearance was the same from both nucleotides. 5. The transmural transport of uracil from the uridine mononucleotides is discussed with reference to the metabolism and compartmentalization of the small intestine responsible for the appearance of the free pyrimidine in the serosal fluid.

The specificity of pyrimidine nucleoside transport and metabolism by rat jejunum in vitro.

1. 5-Methyluridine perfused through the lumen of isolated loops of rat jejunum gave rise to serosal thymine as the major species transported across the epithelial layer; no serosal 5-methyluridine was detected. 2. At 0.1 mM-luminal 5-methyluridine there was enhanced transmural transport of thymine (P less than 0.001) when compared with either 0.1 mM-thymine or 0.1 mM-thymidine as the luminal substrate for thymine transport. 3. At low luminal concentrations (0.025 mM) thymine was a significantly better substrate (P less than 0.001) for the transport of the free pyrimidine than either of the two nucleosides (thymidine or 5-methyluridine). 4. Luminal deoxyuridine gave rise to uracil as the major species appearing in the serosal secretion. High luminal concentrations of deoxyuridine (0.5 and 1.0 mM) gave rise to low levels of the nucleoside in the serosal fluid. 5. As a consequence of the specificity of the mucosal phosphorolysis the ribonucleosides are favoured over the deoxyribonucleosides as substrates for transmural pyrimidine transport.

The transport and metabolism of naturally occurring pyrimidine nucleosides by isolated rat jejunum.

1. Uridine perfused through the lumen of isolated loops of rat jejunum over a concentration range of 0.1-1.0 mM gave rise to higher serosal concentrations of uracil than the equivalent luminal concentration of uracil (P less than 0.001). No serosal uridine could be detected. 2. Luminal thymidine over a concentration range of 0.1-0.5 mM gave rise to the same serosal concentration of thymine as the equivalent luminal concentration of thymine (P greater than 0.1). Low concentrations of serosal thymidine were detected. Both luminal thymidine and thymine gave rise to elevated levels of serosal uracil. 3. Luminal cytidine at concentrations of 0.1-0.5 mM was poorly transported and yielded low serosal concentrations of cytidine. No serosal cytosine was detected, although elevated levels of uracil were found in the serosal secretions. 4. Cytosine over a luminal concentration range of 0.1-0.5 mM gave rise to low concentrations of cytosine in the serosal secretions. These results were consistent with a passive diffusion model for cytosine transport. No increase in serosal uracil was detected. 5. The cleavage of uridine and thymidine to their respective pyrimidine bases occurred via a cytoplasmic nucleoside phosphorylase, which had a similar Michaelis constant (Km), (61.0 +/- 4.4 and 97.1 +/- 5.7 microM for uridine and thymidine, respectively) but a maximal velocity (Vmax) for uridine cleavage (320 +/- 32 nmol min-1 (mg protein)-1) 13 times that for thymidine cleavage (24.7 +/- 1.4 nmol min-1 (mg protein)-1). 6. The differences between the three pyrimidine nucleosides are discussed with reference to the interactions between their epithelial transport and metabolism.

The transport of pyrimidines into tissue rings cut from rat small intestine.

1. At low concentrations (0.1 mM) the transport of uracil, 5-fluorouracil and thymine into jejunal tissue rings is an active process. 2. The transport of 5-fluorouracil into tissue rings cut from the duodenum and jejunum was greater than the transport into rings cut from the ileum. This difference was abolished by starving the rats for 48 h before the experiment. 3. The active transport can be abolished by replacing the Na+ in the incubation medium with either K+ or mannitol, or by increasing the concentration of the pyrimidine to 1.0 mM. 4. The accumulation of uracil or 5-fluorouracil into the jejunal rings was identical when determined by radioactive tracer or by high-performance liquid chromatography. 5. The apparent Michaelis constant (Km) for 5-fluorouracil transport into jejunal rings was 0.074 mM in the standard Na+ bicarbonate Krebs-Ringer solution and 0.394 mM in the K+-substituted bicarbonate Krebs-Ringer solution. 6. Both thymine and uracil inhibited the transport of 5-fluorouracil into jejunal tissue rings: however, cytosine and orotic acid did not.

D-galactose transport in rat intestinal brush border membrane vesicles studied with a molecular-sieve technique.

A technique using G-50 Sephadex molecular-sieve chromatography has been developed to study both the accumulation of transported solutes by rat intestinal brush border membrane vesicles and the wash-out of these solutes from pre-loaded vesicles. The extent to which D-galactose was taken up into the brush border membrane vesicles was dependent on the medium osmolarity and this dependence was significantly reduced (P less than 0.01) by pre-treating the vesicles with the non-ionic detergent Triton X-100. The transport of D-galactose into the brush border membrane vesicles was sodium dependent, electrogenic and phloridzin sensitive. It is, therefore, qualitatively similar to D-glucose transport into brush border membrane vesicles. The wash-out of D-galactose from pre-loaded brush border membrane vesicles was inhibited by phloridzin and accelerated by the presence of an outwardly directed sodium gradient. D-Galactose was released from the pre-loaded brush border membrane vesicles more slowly than D-glucose and this difference was consistent with the higher overshoot found for D-galactose during the transport of the sugars into the vesicles.