Full length article
Negative mode MS/MS to read digital information encoded in sequence-defined oligo(urethane)s: A mechanistic study

https://doi.org/10.1016/j.ijms.2017.07.006Get rights and content

Highlights

  • Small polyurethanes (PUs) were studied to anticipate MS/MS behavior of long chains.

  • Three mechanisms were evidenced for Osingle bond(CO)NH bond cleavage in deprotonated PUs.

  • Main issue with long chains was not related to CID but to negative mode ESI yield.

  • A new design of PU end-groups allowed this sensitivity issue to be addressed.

Abstract

MS/MS sequencing is an unrivaled technique to decipher binary information chemically encoded in the backbone of sequence-controlled synthetic polymers constructed with two co-monomers of different mass, arbitrarily designated as the 0- and 1-bit of the ASCII alphabet. Efficiency of this “reading” step relies however on the simplicity of MS/MS patterns, which depends on both polymer chemistry and chain length. In this context, polyurethanes (PUs) were very promising candidates as dissociation of small deprotonated oligomers (n < 8) yielded a single fragment series. The carbamate bond cleavage reaction was hence studied in details to tentatively anticipate the CID behavior of longer chains prior to optimizing their synthesis. In spite of the simplicity of MS/MS spectra, three different mechanisms were evidenced; however, they were not expected to induce MS/MS complexity when activating longer chains, as verified for sequence-controlled PUs containing up to two bytes of information (i.e., 16 co-monomers). In contrast, the ionization step appeared to be an issue: deprotonation yield of the end-group in negative ion mode electrospray was observed to strongly decrease as PU chain length increases. This sensitivity issue was addressed by introducing a second acidic end-group to allow doubly deprotonated oligomers with no impact on their CID behavior.

Introduction

Synthetic digital polymers have recently emerged as an interesting new class of macromolecules that can be used to store information at the molecular level [1], [2]. By using a set of two co-monomers arbitrarily defined as the 0- and 1-bits of the ASCII alphabet [3], [4], our group has reported various synthesis methodologies involving different chemistries to produce so-called digitally encoded polymers such as poly(alkoxyamine amide)s [5], [6], poly(phosphodiester)s [7], [8], poly(alkoxyamine phosphodiester)s [9], poly(triazole amide)s [10], [11], and polyurethanes [12], [13]. In all these species, the chemical code was based on mass, with a simple H/CH3 molecular variation between the two co-monomers. As a result, MS/MS sequencing appeared as a key technique to “read” information that was “written” in the chains. Indeed, similarly to biopolymers such as peptides [14], oligonucleotides [15], or carbohydrates [16], collision-induced dissociation (CID) of sequence-defined synthetic polymers proceeds via backbone cleavages independently of co-monomer arrangement along the chain, yielding MS/MS data that allow their sequence to be re-constructed. However, complexity of CID spectra (which affects MS/MS readability of encoded information) highly depends on the chemistry of the polymer backbone, as commonly observed for synthetic polymers [17]. Most often, and similar to the case of biopolymers, multiple fragment series are generated upon activation, which leads to redundant information when used for sequencing purposes. Moreover, spectral complexity further increases with the size of dissociating chains: because longer chains promptly adopt high charge states upon electrospray ionization (ESI), their dissociation products also exhibit a charge state distribution. Typically, a chain-length of at least 100 residues is wished for such digital polymers, in order to prepare libraries of moderate size macromolecules to compose long messages rather than using very high molecular weight polymers, as learnt from studies involving DNA chains as a medium to store information [18], [19].

In this context, simplicity of MS/MS pattern obtained for small oligomers (n < 10) prepared in the earlier development stages of new sequence-defined polymers is a key requirement before dedicating further efforts in optimizing synthesis protocols to produce longer congeners with perfect monodispersity. So far, candidates that best fulfilled this requirement were polyurethanes (PUs) [12]. As depicted in Scheme 1, information-encoded PUs are built from two co-monomeric units composed of a carbamate function linked to an alkyl segment, the latter being used to implement a binary 0/1 code with R = H for the 0-bit and R = CH3 for the 1-bit. Co-oligomeric chains are capped by a carboxypentyl α end-group and a hydroxyl ω termination. Although their structure allowed coded PU oligomers to readily ionize in both polarity modes, dissociation of deprotonated molecules was found to offer an extremely simple MS/MS pattern, with a single series of fragments spaced by the mass of one or the other co-monomer (vide infra). This peculiar pattern allows any binary messages to be literally read from CID spectra. In order to figure out whether such a simple MS/MS sequencing could also be anticipated for longer chains, mechanisms underlying dissociation reactions observed for PU co-oligomers were studied in details.

Section snippets

Chemicals

Digitally-encoded polyurethanes studied in this work were synthesized using an orthogonal iterative solid-phase approach described in a recent publication [12]. The oligourethane with an ω-carboxylic acid moiety was prepared using ethyl 6-aminohexanoate in the last step of the iterative synthesis. The terminal ester function was then acidified during cleavage and the formed ethanol was evaporated along with cleavage reagents, thus affording the desired compound. The sample containing a

Results and discussion

Sequence-defined PUs were ionized in the negative ion mode ESI after deprotonation of the carboxylic acid group in the α moiety (Scheme 1). Single stage MS experiments confirmed that the requested uniform structure has been achieved for these sequence-defined polymers, with main detection of targeted species validating that all chains exhibit the same length (Supporting Fig. S1). As mentioned in the Introduction, MS/MS spectra obtained for deprotonated PU oligomers were extremely simple and

Conclusion

As shown in this mechanistic study, most simple MS/MS patterns do not necessarily imply that dissociating ions experienced a single reaction. Cleavage of Osingle bond(CO) bond in carbamate groups was indeed evidenced to proceed via three mechanisms of different efficiencies. The bottom-up approach implemented here to anticipate the CID behavior of long chains from that observed for very short oligomers showed the same dissociation reactions applied for PUs with a chain length up to 16 residues. However,

Acknowledgments

L.C. acknowledges support from Spectropole, the Analytical Facility of Aix-Marseille University, by allowing a special access to the instruments purchased with European Funding (FEDER OBJ2142-3341). J.F.L. thanks the H2020 program of the European Union (project Euro-Sequences, H2020-MSCA-ITN-2014, grant agreement n°642083), and the Cluster of Excellence Chemistry of Complex Systems (LabEx CSC) for financial support. The PhD positions of D. K. and B. E. P. are supported by the ITN Euro-Sequences

References (23)

  • A. Al Ouahabi et al.

    Synthesis of non-natural sequence-encoded polymers using phosphoramidite chemistry

    J. Am. Chem. Soc.

    (2015)
  • Cited by (15)

    • Promoting carboxylate salts in the ESI source to simplify positive mode MS/MS sequencing of acid-terminated encoded polyurethanes

      2020, International Journal of Mass Spectrometry
      Citation Excerpt :

      This approach is explored here for sequenced-defined polyurethanes (PUs). When subjected to collision-induced dissociation (CID), PU oligomers with DP = 4–12 exhibited very simple MS/MS pattern which revealed that only one reaction occurred in all repeating units, giving rise to a single fragment series due to a fixed negative charge in the oligomer α end-group [14]. As a result, coded sequences could be literally read from peak-to-peak distance in MS/MS spectra and a simple algorithm was created for automated extraction of digital data [15].

    • Easily encodable/decodable digital polymers linked by dithiosuccinimide motif

      2019, European Polymer Journal
      Citation Excerpt :

      This digital polymer with uniform (monodisperse) chain length and defined sequences of two monomer units (0-bit and 1-bit) was proposed and developed by Lutz et al. based on modern synthetic (encoding) strategies and sequencing (decoding) techniques [2–4,7–19]. So far, a variety of digital polymers, such as poly(alkoxyamine amide)s [16,20–23], poly(phosphodiester)s [7,14,17], poly(alkoxyamine phosphodiester)s [9], poly(triazole amide)s [24–26] and polyurethanes [8,10,11,27], have been created to encode information. In general, an ideal digital polymer should be easily accessed for synthesis and sequencing[22].

    View all citing articles on Scopus
    View full text