Synthesis of Methylene Isosteres of Glycosyl Amino Acids and Glycosyl Phosphates
|
Glycosylation of proteins and lipids is a post-translational
modification important for numerous physiological processes (1).
These glycoconjugates act in protein folding, altering the physico-chemical
properties and activity, and adorn the landscape of the cell surface,
influencing cell–cell and cell–matrix interactions. Our attention has been directed toward isosteric mimics, which entail methylene substitution for the anomeric oxygen, the so-called C-linked glycosyl amino acids, C-linked glycosyl phosphonates, and C-linked glycosyl lipids, Figure 1. While such an exchange (CH2 for O) does slightly alter the bond angle, bond lengths, and the dihedral angles, there is ample precedent to suggest the conformational change is not insurmountable for binding (2) The methylene isostere offers a great deal of stability, in that degradation by glycosidases, reaction with glycosyltransferases, acid hydrolysis of the anomeric acetal, and b-elimination from the serine are all prohibited. |
![]() |
Our goal is to
provide convenient synthetic routes to C-linked glycosyl amino acids and
C-linked glycolipids, as well as their biosynthetic precursors. Previous synthetic work on C-linkages has explored a wide variety of C-C
bond forming reactions (see suggested reading).
Despite the
significant effort, many of the synthetic approaches require expensive starting
materials, are limited in scope to the preparation of only one anomeric
stereochemistry, and lose the isosteric relationship by excessively lengthening
the intervening carbon chain. While
our work does not remedy all these shortcomings, it does offer complementary
approaches to these biologically significant materials and is well-suited for
serious undergraduate research. The
proposed syntheses are short, are adaptable to a variety of biomimetics, and
utilitize chemistry that is well precedented and not experimentally tedious.
Synthesis of C-Glycosyl
Serines
Our previous work
in this area involved alkylation of iodides 3
with the enolate derived from Williams’ Boc-protected diphenyloxazinone 4,
Scheme 1. The iodides originated from a-C-allyl
glycosides, which are readily prepared and well-known in the literature. Although this reaction was stereoselective and provided access to the
isosteric glycosyl serine 5, and to
the b-anomers
as well, it was not amenable to acetyl glycosides or N-acetyl precursors (3).

A more functional group-tolerant reaction methodology for C-glycosyl amino acid synthesis was utilized in our lab by employing the Grubbs’ second generation catalyst (4) for the olefin cross-metathesis (CM) between C-allyl glycoside 6 and vinyl glycines 8-10, prepared from methionine or glutamic acid (5), Scheme 2. A explosion of reports have appeared since the catalyst has become commercially available and these have been recently reviewed (6). This methodology has provided a concise entry into glycosyl amino acids in which the intervening carbon chain contains three or more methylenes. In our hands and in accordance with others, the linking carbon chain between the glycoside and the amino acid cannot be made shorter than three carbons (e.g., serine mimics require only two methylenes) without seriously diminishing the yield, as evidenced by >70% recovered C-vinyl glycoside 7 in all attempted reactions utilizing this cross-metathesis partner and our vinyl glycines 8-10.

Students working in my lab have enjoyed recent success, attested to by the five undergraduate co-authors on publications in the past two years. As a mentor, the PI works alongside the students as they gain confidence learning the techniques involved in running reactions, purifying products on our new Isco Flash Chromatography station, and identifying structures. The acquisition of a new 400 MHz NMR has greatly simplified training on how to acquire NMR data, and has focused much more attention on the interpretation of data with students. The recent purchase of an ion-trap GC-MS and the soon to be acquired MALDI-TOF mass spectrometer facilitate molecular weight determination for characterization of products.
![]() |
![]() |
| Jennifer Potter ('04) writing up the results of a reaction in her laboratory notebook. (Spring 2004) | Michael Orlando ('04) loading a sample into Colgate's Bruker 400 MHz NMR spectrometer. (Spring 2004) |
References
Suggested Reading
The preparation of C-glycosyl amino acids has been reviewed by Dondoni (1). Since that review there have been several reports toward the development of concise and broadly applicable routes to C-glycosyl amino acids, including our own (see publication list), and others utilizing the Ramberg-Bäcklund rearrangement (2), acetylide couplings (3), and an asymmetric Strecker reaction (4) The C-phosphonate analogs have been prepared as mimics for the UDP-glycosides and glycolipids (5)
General approaches to form the C-glycosidic bond to phosphonates include anomeric radical additons to unsaturated phosphonates (6), Wittig chemistry with concomitant ring closures (7) elaboration from C-allyl glycosides (8) and phosphonate anion addition to glyconolactones (9) Application of some of these approaches has led to the preparation of some UDP-glycosyl mimics (10), decaprenolphosphoarabinose analogues (11), muramic acid derivatives (12) mannosyl phosphate mimics (13), and a bacterial peptidoglycan analogue (14).
Citations most closely related to our work include Dondoni (15) McGarvey (16) and Postema (17).
Top of Page - Back to Nolen Profile - Back to Faculty Research Interests - Chemistry Home Page - Colgate Home Page
![]() |
Design: Al Rehm and Roger S. Rowlett Updated February 24, 2004 Questions to: rrowlett@mail.colgate.edu © 1997 Colgate University. All Rights Reserved. |