Degree Name

Doctor of Philosophy


School of Chemistry


Cosmomycin D (CosD) is an anthracycline that possesses two trisaccharide chains bound to its aglycone. This glycosylation pattern is substantially different from that of anthracyclines used in the clinic such as daunomycin (Dn) and doxorubicin (Dx). A purification method for CosD from culture supernatants of Streptomyces olindensis ICB20 previously genetically engineered to overproduce CosD was developed. When the bacteria cultures were propagated for several generations, in addition to CosD, different anthracyclines appeared in the crude extracts, most likely as a result of spontaneous mutations. Characterisation by mass spectrometry showed that these anthracyclines were related to CosD in that they had two trisaccharide chains. The sequences of sugars bound to the aglycones were determined by collisionally-activated dissociation using an electrospray ionisation (ESI) quadrupole ion trap mass spectrometer. Some of the anthracyclines had the same mass but had different HPLC elution times using a C4 column. For example, several compounds differed from CosD by the mass of an hydroxyl group. This project coincided with the commercial release of an ESI quadrupole time-of-flight mass spectrometer equipped with a travelling wave ion mobility cell (TWIMS). Some of the isobaric anthracyclines could be resolved by TWIMS providing another example of the types of molecules that can be separated by this ion mobility method. Furthermore, the sequence of the sugars was also determined by fragmentation of the anthracyclines using the TWIMS cell and the results compared to those obtained with the ion trap instrument.

Since only small amounts of CosD could be obtained, the sensitivity of mass spectrometry (MS) was exploited in several different ways to characterise its interaction with DNA. ESI mass spectrometry was used to screen the binding of CosD to doublestranded (ds) DNA. The DNA was saturated with CosD at lower concentrations than Dn and Dx. Fewer molecules of CosD were able to bind to DNA suggesting a larger footprint, and all complexes formed with CosD were more stable than those obtained for Dn and Dx. Nuclease footprinting experiments analysed by mass spectrometry showed that CosD was able to protect the DNA to a greater extent than Dn and Dx consistent with the binding profiles for the DNA sequences. Titration mixtures of dsDNA with increasing amounts of CosD were also analysed by agarose gel electrophoresis (plasmid DNA and 16 mer DNA) and CD spectroscopy (16 mer DNA) to complement the ESIMS studies. Both techniques revealed that a greater change was induced on the DNA conformation upon binding of CosD than for Dn and Dx; this change was most likely caused by partial unwinding of the DNA double-helix upon binding (intercalation) of the drug molecule. In vitro RNA transcription assays showed that CosD was more effective in inhibiting RNA production than Dn and Dx. All the experiments carried out here support that CosD binds more tightly to DNA, exerts a larger footprint on DNA, and has a greater ability to perturb the DNA conformation than Dn and Dx.

Since CosD was a purified natural product available only in milligram quantities, synthesis of its aglycone, b-rhodomycinone, was attempted using Dn as starting material owing to the similarity of the aglycone of CosD and Dn. Sugar removal and functional group modification led to the unexpected synthesis of a novel compound, necessitating the devising of an alternative synthetic route. Although the synthesis of the aglycone was not completed owing to time restraints, the DNA-binding studies of the synthesised intermediate compounds were carried out and confirmed the importance of the sugars in DNA-binding.

This is the first comprehensive DNA-binding study of CosD. Early studies of CosD showed that it induced differentiation in mouse leukemia cells. This effect was not observed for Dn and Dx. The different effects of CosD on cells related to its structure are yet to be fully investigated. The difference from Dn and Dx may lead to a new chemotherapeutic agent, or alternatively, if the effects on cell differentiation are shown to be reproducible, CosD may serve as a tool to be used to study the process of differentiation in cell culture.



Unless otherwise indicated, the views expressed in this thesis are those of the author and do not necessarily represent the views of the University of Wollongong.