Synthesis of a dendron and dendrimer consisting of MALDI matrix like branching units

Article abstract of DOI:10.1016/S0040-4039(02)01522-8

The synthesis of a novel dendron and dendrimer with potential application in matrix-assisted laser desorption/ionization mass spectrometry (MALDI MS) is described. The dendrimer branching unit was designed to have structural similarities to cinnamic acid based matrices used in MALDI MS. Branching was incorporated into the structure by introducing a second acrylic acid side chain. Some intermediates of the monomer synthesis were shown to function as MALDI matrices. A second-generation dendron was assembled on a solid support employing an N-to-C directed convergent synthetic strategy. Coupling of the dendrons with a trivalent core molecule resulted in a second generation dendrimer.

Full text of DOI:10.1016/S0040-4039(02)01522-8

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 43 (2002) 6723–6727
Synthesis of a dendron and dendrimer consisting of MALDI
matrix like branching units
Hendrik Neubert, Andrew T. Kicman, David A. Cowan and Sukhvinder S. Bansal*
Drug Control Centre and Department of Pharmacy, King’s College London, Franklin-Wilkins Building, 150 Stamford Street,
London SE1 9NN, UK
Received 16 May 2002; accepted 23 July 2002
Abstract—The synthesis of a novel dendron and dendrimer with potential application in matrix-assisted laser desorption/ioniza-
tion mass spectrometry (MALDI MS) is described. The dendrimer branching unit was designed to have structural similarities to
cinnamic acid based matrices used in MALDI MS. Branching was incorporated into the structure by introducing a second acrylic
acid side chain. Some intermediates of the monomer synthesis were shown to function as MALDI matrices. A second-generation
dendron was assembled on a solid support employing an N-to-C directed convergent synthetic strategy. Coupling of the dendrons
with a trivalent core molecule resulted in a second generation dendrimer. © 2002 Elsevier Science Ltd. All rights reserved.
1. Introduction
energy absorbing surface or as a photolabile tether
between a covalently immobilized biomolecule and
MALDI target.
Since the introduction of matrix-assisted laser desorp-
tion/ionization mass spectrometry (MALDI MS) in the
late 1980s,1 this techniques has become a valuable tool
for the analysis of biological molecules. In MALDI
analysis, the analyte is co-crystallized in a suitable
matrix of highly UV absorbing small organic molecules.
The excitation of the matrix by a laser pulse causes
desorption and ionization of the matrix and the analyte
molecules. Despite enormous research efforts, the
mechanisms of this phenomenon are still not com-
pletely understood. In an attempt to enhance the des-
orption and ionization process, matrix molecules such
as a-cyano-hydroxycinnamic acid2 or sulfhydryl con-
taining benzoic acid derivatives3 have been covalently
attached to the surface of a MALDI target. In contrast
to conventional MALDI, the sample is added to the
matrix coated target surface prior to MS analysis. This
simplifies the sample preparation and may be of impor-
tance to automated high throughput MALDI analysis.
However, the synthesis of dendrons and dendrimers
consisting of matrix-like molecules as aids for MALDI
mass spectrometry has not been investigated. A poten-
tially useful avenue of investigation is the covalent
attachment of a ‘matrix dendrimer’ to a MALDI target
surface in order to form a film that can be used as an
In recent years, dendrimers have emerged from syn-
thetic interest into tailor-made structures for specific
applications.4,5 The construction of dendrimers is car-
ried out in a stepwise manner; whereby two fundamen-
tally different synthetic routes can be employed. In a
divergent strategy,6 monomer units are successively
attached to a core molecule which results in dendrimer
growth. In a convergent approach,7 the synthesis starts
on the periphery and goes towards what will be the core
of the dendrimeric wedge followed by a subsequent
reaction with a core molecule. Traditionally, den-
drimers were synthesized in solution by both of these
strategies. In parallel with the development of den-
drimer chemistry, there have been vast advances in the
field of solid-phase and combinatorial chemistry which
resulted in considerable progress in the solid-phase
synthesis of dendrimers.8–10 The major advantages are
that unreacted soluble reagents can be removed by
simple filtration and washing without sample loss. Fur-
thermore, excess soluble reagents can be used to drive
reactions to completion. Herein, we describe the design
and synthesis of a dendrimer branching unit possessing
MALDI matrix like properties and the convergent
solid-phase synthesis to give a generation two dendron.
We also demonstrate the assembly of dendrons with a
core molecule in order to synthesize a second-genera-
tion dendrimer.
Keywords: MALDI matrix; dendrimer; solid phase synthesis; reverse
synthesis.
* Corresponding author. Fax: +44-(0)20-7848-4785; e-mail:
sukhi.bansal@kcl.ac.uk
0040-4039/02/$ - see front matter © 2002 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(02)01522-8

6724
H. Neubert et al. / Tetrahedron Letters 43 (2002) 6723–6727
2. Design features
3. Results and discussion
The building block for the dendrimer synthesis was
designed to have structural similarities to a family of
organic molecules used as matrices in MALDI MS.
Members of that family are analogues of cinnamic acid
such as 3,4-dihydroxycinnamic acid (caffeic acid), 3-
methoxy-4-hydroxycinnamic acid (ferulic acid) and 3,5-
dimethoxy-4-hydroxycinnamic acid (sinapinic acid).
The common cinnamic acid structure was the basis of
the design of a MALDI matrix like dendrimer branch-
ing unit. To obtain a branching multiplicity of two
(type AB2), a second arm was introduced into the
cinnamic acid structure by placing an additional acrylic
acid moiety in the meta position and the methoxy
group was replaced with an amino group to afford
3-amino-5-(carboxyvinyl)-phenyl acrylic acid 4. In the
design, it was envisaged that the nucleophilicity of the
aromatic amine and the reactivity of the acrylic acids
might present problems during the synthesis of the
dendron. Thus, in the design strategy a glycyl moiety
was attached to both the amino and carboxylate ter-
mini to afford the amino acid which would be prepared
as the 9-fluorenylmethyl ester for utilization in solid
phase dendron synthesis. The structure of the designed
building block dictates the synthetic route for the solid
phase dendron assembly, which would involve N to C
synthesis and the use of hyperacid labile attachment
(trityl based) to the support to produce the protected
dendron for attachment to the core molecule. Our
strategy involved the activation of the carboxylic acid,
which is immobilized on the solid support followed by
treatment with the free amine. This ‘reverse’ coupling is
not very widely studied and would require optimiza-
tion. As an adjunct, the extended conjugated system in
the monomeric unit was required for enhanced absorp-
tion of the laser light in UV-MALDI.
The synthesis of the 9-fluorenylmethyl protected build-
ing block 7 (GAGFm) is shown in Scheme 1. Isophthal-
aldehyde 1 was nitrated using nitric acid and sulphuric
acid to afford the 5-nitro-isophthalaldehyde 2.11 Con-
densation of 5-nitro-isophthalaldehyde with malonic
acid under basic conditions gave 2-carboxyvinyl-5-
nitrophenylacrylic acid 3. The nitro group was subse-
quently reduced with ferrous sulphate to yield
3-amino-5-(carboxyvinyl)-phenyl acrylic acid 4.12 4 was
then acylated with tert-butoxycarbonyl glycine anhy-
dride to afford the diacid 5, which was subsequently
treated with 2-(1H-benzotriazol-1-yl)-1,1,3,3-tetra-
methyl uronium tetra-fluoroborate (TBTU), diiso-
propylethylamine (DIPEA) and glycine fluorenylmethyl
ester to yield the orthogonally protected building block
6. Finally, the tert-butoxycarbonyl group was removed
with trifluoroacetic acid to give the monomeric unit 7.
Batch solid-phase dendrimer synthesis was carried out
in a manual sintered flow reactor (Scheme 2). A trityl
alcohol PEG-PS (polyethylene glycol–polystyrene) resin
(NovaSyn® TGT), was utilized for solid phase dendron
assembly. The trityl alcohol resin 8 was treated with
thionyl chloride at room temperature under a N2 atmo-
sphere to yield the chlorotrityl resin 9.13 As the chloro-
trityl resin is susceptible to hydrolysis it was used
immediately. The chlorotrityl resin 9 (0.26 mmol/g) was
treated with half an equivalent of GAGFm 7 in the
presence of DIPEA. This reaction was monitored by
HPLC analysis of the 7, which indicated that half of the
sites on the resin were acylated. The remaining unre-
acted sites on the resin were capped with methanol.
This procedure effectively reduced the resin loading to
0.013 mmol/g in order to prevent steric problems
Scheme 1. Synthetic pathway of MALDI matrix like dendrimer branching unit. Reagents: (a) (NH4)2SO4, conc. H2SO4, fuming
HNO3; (b) (i) malonic acid, pyridine, piperidine; (ii) glacial acetic acid; (c) ammonium hydroxide, H2O, FeSO4; (d) (Boc-Gly)2O,
DMF; (e) H-Gly-OFm, TBTU, DIPEA, dioxane; (f) TFA, CH2Cl2, H2O.

H. Neubert et al. / Tetrahedron Letters 43 (2002) 6723–6727
6725
acylation yields. Coupling reactions in solid-phase pep-
tide synthesis are usually undertaken at room temperature
in order to minimize side reactions and racemization.
Since our acylation reactions failed to reach completion
at room temperature we conducted the reactions at 37°C,
which resulted in successful and efficient acylation as
determined by reverse-phase HPLC (Fig. 1). The peak
area percentage of the total chromatogram was used to
plot the graph. The second largest HPLC peak was related
to deprotected 11b, i.e. only one branching unit being
coupled to the generation 1. The coupling reaction was
regarded as being complete after 4 h.
The standard procedure for final resin cleavage from
the trityl solid support with acetic acid (20% v/v) or
TFA (1% v/v) solution in trifluoroethanol (20% v/v)
15
and CH2Cl2 was slow and therefore unsuitable. How-
ever, successful cleavage was achieved with TFA/
CH2Cl2 (1:4). Final purification of 12 was performed by
size exclusion chromatography using an ammonium
bicarbonate buffer (purity 94.5%).
All intermediates of Scheme 1 were tested for their
potential to function as MALDI matrices. In fact, 3, 4
and 5 showed the ability to promote desorption and
ionization of cytochrome C (12.2 kDa) when co-
deposited on a MALDI target. MALDI spectra of
cytochrome C were successfully recorded after co-crys-
tallizing 0.8 ml of the intermediate 3, 4 or 5 (10 mg/ml
in 70% acetonitrile/water) with 2 pmol of the protein.
However, in initial experiments the observed MALDI
signal intensities for cytochrome C were weaker than
those obtained from conventional sinapinic acid/
cytochrome C co-crystals and higher laser power was
required for desorption. Expectedly, compounds 2, 6
and 7 did not exhibit any MALDI matrix like proper-
ties as they lack the ability to donate a proton for
protein ionization. The dendron, 12, was also co-
deposited with cytochrome C on a MALDI sample
Scheme 2. Solid phase synthesis of the G-2 dendron. Reagents:
(a) SOCl2, CH2Cl2; (b) 7, DIPEA, DMF; (c) (i) 20% piperidine;
(ii) 7, HOAt, DIPCDI, DMF; (d) (i) 20% piperidine; (ii) CH2Cl2,
TFA, H2O, TIPS.
during the solid-phase synthesis. To couple the next
generation of the building block 7, the fluorenylmethyl
esters were removed employing 20% piperidine in DMF14
andthepiperidiniumcounterionwasremovedbywashing
with diluted acetic acid (0.5%). The trityl linker was found
to be completely stable to the treatment with 0.5% acetic
acid. The parameters investigated to achieve optimized
coupling of the second generation of 7 were the reagent
concentration, the reaction temperature and time. It was
found that 15 equiv. of the building block 7, 20 equiv.
of 1-hydroxyazabenzotriazole (HOAt) and 40 equiv. of
diisopropylcarbodiimide (DIPCDI) gave acceptable
Figure 1. Reaction kinetic of acylation of resin bound branching unit GAGFm at 37°C, 11c=complete G-2 dendron, 11b=one
branching unit missing.

6726
H. Neubert et al. / Tetrahedron Letters 43 (2002) 6723–6727
plate. However, no signal for cytochrome C was
observed which was related to the exclusion of the
protein from crystals of 12 upon drying due to dissimi-
lar physico-chemical characteristics.
Treatment of the trivalent core tris-succinimidyl amino-
triacetate with the dendron 14 failed to produce the
final dendrimer 13. However, on addition of HOAt,
the dendrimer was successfully prepared. To drive
the reaction to completion and to overcome steric
problems during the assembly, the reaction was per-
formed at 60°C and 9 molar equiv. of 14 and HOAt
were utilized.
The dendrons were covalently attached to a suitable
core to give a complete dendrimer, a strategy known
from conventional convergent dendrimer growth.7
Figure 2. Negative ion mode ESI spectrum of G-2 dendron (M, 12).

H. Neubert et al. / Tetrahedron Letters 43 (2002) 6723–6727
2301.
6727
The products of all synthetic steps were analyzed by direct
infusion electrospray ionization iontrap and MALDI-
TOF mass spectrometry. Exemplarily, Fig. 2 shows the
negative ion mode ESI mass spectrum of the generation
2 dendron, 12.
2. Hutchens, T. W.; Yip, T. T. Rapid Commun. Mass
Spectrom. 1993, 7, 576–580.
3. Voivodov, K. I.; Ching, J.; Hutchens, T. W. Tetra-
hedron Lett. 1996, 37, 5669–5672.
4. Matthews, O. A.; Shipway, A. N.; Stoddart, J. F. Prog.
Polym. Sci. 1998, 23, 1–56.
5. Fischer, M.; Vo¨gtle, F. Angew. Chem., Int. Ed. Engl.
1999, 38, 884–905.
6. Buhleier, E.; Wehner, W.; Vo¨gtle, F. Synthesis 1978,
155–158.
7. Hawker, C. J.; Frechet, J. M. J. J. Am. Chem. Soc.
1990, 112, 7638–7647.
8. Lebreton, S.; Monaghan, S.; Bradley, M. Aldrichim.
Acta 2001, 34, 75–83.
9. Swali, V.; Wells, N. J.; Langley, G. J.; Bradley, M. J.
Org. Chem. 1997, 62, 4902–4903.
In conclusion, we have designed a novel dendrimer and
dendron, which is derived from MALDI matrices to
possess matrix properties for MALDI MS. A second
generation dendron has been successfully synthesized on
a solid support at elevated temperature and in the N-to-C
(reverse) direction. All products were characterized by
HPLC and by MALDI and ESI mass spectrometry. The
dendron has been attached to a core molecule to produce
a dendrimer. These novel derivatives have been shown to
possess matrix characteristics and further studies are
currently being carried out.
10. Mahajan, A.; Chhabra, S. R.; Chan, W. C. Tetrahedron
Lett. 1999, 40, 4909–4912.
Acknowledgements
11. Jennings, K. F. J. Chem. Soc. 1957, 1172–1175.
12. Ackerman, J. H. US Patent 3,856,853: Iodinated 5-sub-
stituted-1,3-benzenediacrylic and dipropionic acids; Ster-
ling Drug Inc.: USA, 1974.
13. Harre, M.; Nickisch, K.; Tilstam, U. React. Funct.
Polym. 1999, 41, 111–114.
We gratefully acknowledge A. Przyborowska for assis-
tance with recording electrospray spectra and the BBSRC
for a JREI grant for MS equipment. We also thank J.
Hawkes of the University of London Intercollegiate
Research Service for providing high-field NMR spectra.
14. Henkel, B.; Zhang, L. S.; Bayer, E. Liebigs Ann.-Recl.
1997, 2161–2168.
15. Barlos, K.; Chatzi, O.; Gatos, D.; Stavropoulos, G. Int.
J. Pept. Protein Res. 1991, 37, 513–520.
References
1. Karas, M.; Hillenkamp, F. Anal. Chem. 1988, 60, 2299–