Structural Characterization
of Recombinant Proteins
by MALDI-TOF/TOF MS
Including Top-Down Analysis

The full characterisation of the covalent protein structure requires a number of complementary methods that can be entirely based on mass spectrometry.
A Maldi-Tof/Tof mass spectrometer provides the instrument platform for MS, MS/MS and pseudo-MS3 experiments to achieve this.
While protein identification has become a routine task in many laboratories, the detailed characterisation of proteins remains difficult. Standard methods such as LC-separation of digested proteins, although indispensable, may not provide the full information about the protein structure such as the termini, since the achieved sequence coverage is typically lower than 100%. Top-down analysis of the undigested protein on a Maldi-Tof/Tof mass spectrometer was used here to overcome this limitation. The combined approach using intact MW determination, protein mass fingerprinting, MS/MS of the proteolytic peptides and the novel top-down method T3-sequencing allowed identification of an unknown protein, a point mutation and an N-terminal modification.

Sample
The sample was a purified recombinant polypeptide named penaeidin that has been produced in yeast Saccharomices cerevisiae, provided by Yannick Gueguen (IFREMER-CNRS, France). The analysis was performed as an analysis exercise at the 2nd Swiss Proteomics Society Meeting in Lausanne, Dec. 2002 [1]. Penaeidin is an antimicrobial peptide from the shrimp Penaeus vanname that contains an N-terminal proline rich domain and a C-terminal cysteine rich domain, where the six cysteine residues are forming three internal disulphide bonds in a chitin-binding motif. The recombinant sample is a mutated form of the wild type penaeidin 3a (Wt-Pen3a), with a mutation at position 8 (T8A-Pen3a). The N-terminus is blocked with a pyroglutamic acid. The sequence of Wt-Pen3a and of T8A-Pen3a is:
Mass spectrometry
A Bruker ultraflexTof/Tof has been used for all measurements including LIFT-Tof/Tof spectra of proteolytic peptides and reflector in-source decay spectra of the intact protein (reISD) as described in detail [2]. ReISD spectra provide predominant c- and a-ions by a fast fragmentation in the ion source. Downstream MS/MS analysis of selected ISD ions allows the detailed characterization of the protein N-terminus even in the case of terminal modification – this terminal Tof/Tof analysis has been termed T3-sequencing [6].
Results
The average molecular weight of the protein was determined to be 6643.1 Da in a linear Maldi spectrum (not shown). Later in the course of the analysis, any structure suggestion has to match the molecular mass of the intact protein. This small protein could not be identified by a protein mass fingerprint (PMF) due to largely unspecific tryptic digestion. However, Pen3a was identified based on 2 Lift-Tof/Tof spectra. In addition, MS/MS spectra not matching the Pen3a sequence were used to screen for modifications and mutations, revealing a T8A mutation (Thr-8 was mutated into Ala-8), e.g., in the MS/MS spectrum of m/z 1680.04 (Fig. 1).
The disulfide cross linked C-terminal domain was difficult to digest by trypsin, therefore the protein was reduced and alkylated using vinylpyridine. The N- and C-terminal structures were not revealed in the resulting PMF (Fig. 2), since the tryptic peptides were too small. The database entry (SWISS-PROT P81058), however, suggested that the N-terminus is pyroglutamic acid and the C-terminus amidated at the expense of C-terminal glycine. Protein termini were further characterized by “top down” sequence analysis using reISD. Proteins in the MW range 5-90 kDa have been successfully identified and further characterized with reISD without proteolytic digestion [2-6].
The reISD spectrum of the intact protein was matched to the T8A-Pen3a structure (Fig. 3) and allowed the characterization of the N-terminus. The
a-ion sequence tag stretching across 18 residues was in agreement with the presence of N-terminal pyroglutamylation. In contrast to previous results [2-6] the sequence was established based on a-ion tags due to the proline-rich N-terminal domain of Pen3a. The c-ions frequently observed in ISD spectra were strongly suppressed since the cleavage of two bonds in the ring structure of proline would be required. The tag suggested the N-terminal structure, although the N-terminus itself was not part of it. This is a striking feature of top-down analyses of undigested proteins in contrast to MS/MS analysis of tryptic peptides: The position of a tag in the mass spectrum on the m/z axis relates that sequence to the protein terminus.
In order to additionally obtain direct information about the N-terminal structure, the company applied a new technique to the analysis of T8A-Pen3a, which was termed T3-sequencing [6].
Fragment ions provided by ISD can be further analyzed by Lift-Tof/Tof methods, thus allowing the specific targeting of the protein N- and C-termini for MS/MS sequencing. The T3-spectrum of the a11-ion was obtained from the underivatized and undigested T8A-Pen3a (Fig. 4). The spectrum confirmed the N-terminal modification as well as the T8A mutation in a N-terminal peptide specifically “prepared” inside the mass spectrometer that could not be obtained by tryptic digestion. The intact calculated mass 6643.6 Da of the best matching structure was in great agreement with the experimental value of 6643.1. The matching molecular mass provided clear evidence that the C-terminus was in fact not amidated at the expense of glycine unlike wt-Pen3a.

Conclusion
All analytical approaches that fully relied on tryptic digests and MS/MS or LC-MS/MS experiments were sufficient to identify the protein but largely failed to get a deeper insight into the detailed structure even in this case of a rather small protein. T3-sequencing as a new top down method available on Maldi-Tof/Tof instrumentation appears to be well suited to provide terminal sequences even in the case of N-terminal modification. The full structure characterization of proteins including modifications and sequence variants will most likely require a combination of methods such as PMFs (“bottom up”), LC separations and top down methods. The exercise documented that detailed structure analysis of proteins can be efficiently achieved by the combination of classical "bottom up" protein mass fingerprints and novel "top-down" technologies.

References
[1] Binz, P.-A. et al. 2003 Proteomics, 3, 1562-1566.
[2] Suckau, D.; Resemann, A.; Schuerenberg, M.; Hufnagel, P.; Franzen, J.; Holle, A. 2003 Anal. Bioanal. Chem., 376, 952-965.
[3] Resemann, A.; Suckau, D. 2003 Anal. Chem., in press.
[4] Resemann, A.; Suckau, D. 2003 Proceedings of the 51th ASMS Conference on Mass Spectrom. and Allied Topics, Montreal, Canada, June 8 - 12, WOB pm 4:20.
[5] Suckau, D.; Cornett, D.S. 1998 Analusis, 26, N° 10, M19-M20.
[6] Resemann, A.; Schweiger-Hufnagel, U.; Witt, M.; Suckau, D. 2001, Proceedings of the 49th ASMS Conference on Mass Spectrom. and Allied Topics, Chicago, May 27-31, TPE-121.