Polypeptide composition of an adenovirus type 5 used in cancer gene therapy.

For cancer gene therapy, a recombinant adenovirus serotype 5 named RPR/INGN201 has been constructed by susbtitution of the E1 region with human tumor suppressor gene p53. The protein components of RPR/INGN201 virions were separated by reversed-phase HPLC and were individually identified by electrospray time-of-flight mass spectrometry and N-terminal sequencing, both on intact proteins and on their proteolytic fragments after trypsin digestion. Twenty-five peptide components of the proteome (including fiber) with greater than 0.25-0.5% contribution to the protein content of the virus were identified and characterized. Fiber was confirmed to be partially glycosylated (both the non-glycosylated and the monoglycosylated states were identified), and two proteins were isolated and identified as phosphorylation derivatives, namely protein V (non-phosphorylated and monophosphorylated) and protein IIIa (mono- and diphosphorylated). This new analytical tool proved to be very useful not only for refining our current knowledge of the polypeptide repertoire of purified infectious virions but also for monitoring and very rapidly identifying structural modifications resulting from changes in the manufacturing process. It was also used successfully for the characterization of various adenoviral constructs.

Primary structure control of recombinant proteins using high-performance liquid chromatography, mass spectrometry and microsequencing.

The conformity of two recombinant proteins (a von Willbrand factor fragment and human serum albumin, consisting of respectively 289 and 585 amino acids) has been examined by HPLC combined with mass spectrometry and microsequencing, on both intact material and fragment peptides obtained by proteolytic cleavage. These studies confirmed that the primary structure of the recombinant proteins corresponds to that predicted from their gene, particularly the integrity of their N and C termini, and, in the case of albumin, the agreement between the observed disulfide bond pattern and the published model. Furthermore, the structure of an albumin-related compound could be elucidated. Application of LC-MS for batch-to-batch quality control is also under discussion.

The origin of the diversity of crotoxin isoforms in the venom of Crotalus durissus terrificus.

Crotoxin, the main toxin from the venom of the South American rattlesnake Crotalus durissus terrificus, is a beta-neurotoxin which consists of the non-covalent association of two subunits: a phospholipase A2 subunit B (CB), and a non-enzymic subunit A (CA). We have previously purified and characterized several isoforms of each subunit of crotoxin in the venom collected from numerous snakes. Furthermore, three cDNAs encoding two CB isoforms and the precursor, pro-CA, of subunit A have been isolated from a cDNA library prepared from a single venom gland of Crotalus durissus terrificus. The aim of this study is to analyse an individual snake venom from an animal that has been used to construct a cDNA library. Several isoforms of subunit A and two isoforms of subunit B were isolated and compared to purified and characterized subunit isoforms from pooled venom. The result of this study showed that the multiplicity and the diversity of crotoxin isoforms result from post-translational modifications occurring on a precursor and from the expression of different messenger RNAs present in an individual snake. It allowed for the identification of the two CB isoforms encoding cDNAs expressed in the individual venom with two isoforms from pooled venom, CBc and probably CBa2, that belong to two classes of crotoxin complexes which can be distinguished biochemically and pharmacologically.