MICROBORE REVERSED-PHASE HIGH-PERFORMANCE LIQUID-CHROMATOGRAPHIC PURIFICATION OF PEPTIDES FOR COMBINED CHEMICAL SEQUENCING-LASER-DESORPTIONMASS-SPECTROMETRIC ANALYSIS

An optimized microbore RP-HPLC system (1.0 mm I.D. columns) for the pu rification of low picomole amounts (<5 pmol) of peptides is described. It is comprised of commercially available columns, instrument compone nts and parts. These were selected on the basis of a comparative evalu ation and to yield the highest resolution and most efficient peak coll ection. The sensitivity of this system equals, probably surpasses, tha t of advanced chemical microsequencing for which 2-4 pmol of peptide a re minimally required. As an automated sequencer cannot be ''on-line'' connected with a micro-preparative HPLC system, fractions must be col lected and transferred. With a typical flow of 30 mu l, efficient manu al collection is possible and fractions (about 20 mu l in volume) can still be handled without unacceptable losses, albeit with great precau tion. Furthermore, major difficulties were encountered to efficiently and quantitatively load low- or sub-picomole amounts of peptide mixtur es onto the RP-HPLC column for separation. Discipline and rigorous adh erence to sample handling protocols are thus on order when working at those levels of sensitivity. With adequate instrumentation and handlin g procedures in place, we demonstrate that low picomole amounts of pep tides can now be routinely prepared for analysis by combined Edman-che mical sequencing-matrix-assisted laser-desorption mass spectrometry (M ALDI-MS). The integrated method was applied to covalent structural cha racterization of minute quantities of a gel-purified protein of known biological function but unknown identity. The results allowed unambigu ous identification and illustrated the power of MALDI-MS-aided interpr etation of chemical sequencing data: accurate peptide masses were cruc ial for (i) confirmation of the results, (ii) deconvolution of mixed s equences, (iii) proposal of complete structures on the basis of partia l sequences, and (iv) confirmation of protein identification (obtained by database search with a single, small stretch of peptide sequence) by ''mass matching'' of several more peptides with predicted proteolyt ic fragments.