Revisiting Amadori and Heyns synthesis: Critical percentage of acyclic form play the trick in addition to catalyst

Article abstract of DOI:10.1016/j.tetlet.2018.06.050

Amadori and Heyns reaction are landmark reactions of carbohydrate chemistry. Synthesis of simple Amadori and Heyns compounds are complicated by various factors and require tedious column chromatographic or ion chromatographic separations. Herein, we report an improved catalytic method based on classical synthetic method of Amadori and Heyns compounds in light of new understanding of a factor governing the reaction. By utilizing the improved catalytic method, we have accomplished several Amadori compounds of D-tagatose and also numerous other Amadori and Heyns compounds.

Full text of DOI:10.1016/j.tetlet.2018.06.050

Accepted Manuscript
Revisiting Amadori and Heyns Synthesis: Critical Percentage of Acyclic Form
Play the Trick in Addition to Catalyst
Debasree Chanda, Nanishankar V. Harohally
PII:
S0040-4039(18)30816-5
https://doi.org/10.1016/j.tetlet.2018.06.050
TETL 50088
DOI:
Reference:
Tetrahedron Letters
To appear in:
Received Date:
Revised Date:
Accepted Date:
26 April 2018
15 June 2018
21 June 2018
Please cite this article as: Chanda, D., Harohally, N.V., Revisiting Amadori and Heyns Synthesis: Critical Percentage
of Acyclic Form Play the Trick in Addition to Catalyst, Tetrahedron Letters (2018), doi: https://doi.org/10.1016/
j.tetlet.2018.06.050
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Title:
Revisiting Amadori and Heyns Synthesis: Critical Percentage of Acyclic Form Play the
Trick in Addition to Catalyst
Debasree Chanda1&2 and Nanishankar V. Harohally1&2
*
Author Names and affiliations
1. Department of Spice and Flavour Science
CSIR-CFTRI
KRS Road,
Karnataka
India
phone: +91 821 2512352
Fax: +91 821 2517233
email:nani.shankar@gmail.com
nani@cftri.res.in
2. Academy of Scientific and Innovative Research
CSIR-CFTRI
Mysore 570020
Corresponding Author
Nanishankar V. Harohally
1

Graphical Abstract
2

Abstract:
Amadori and Heyns reaction are landmark reactions of carbohydrate chemistry.
Synthesis of simple Amadori and Heyns compounds are complicated by various factors and
require tedious column chromatographic or ion chromatographic separations. Herein, we
report an improved catalytic method based on classical synthetic method of Amadori and
Heyns compounds in light of new understanding of a factor governing the reaction. By
utilizing the improved catalytic method, we have accomplished several Amadori compounds
of D-tagatose and also numerous other Amadori and Heyns compounds.
Keywords:
Amadori reaction, Heyns reaction, Acyclic form, Catalysis, Zinc acetate
3

Revisiting Amadori and Heyns Synthesis: Critical Percentage of Acyclic Form Play the
Trick in Addition to Catalyst
Debasree Chanda1,2 and Nanishankar V. Harohally1,2*
Amadori rearrangement is a cornerstone reaction of carbohydrates that has received
meagre interest from mainstream organic chemists.1,2 It has attracted the major attention from
food scientists and biochemists from the perspective of Maillard reaction related to flavours
and glycation respectively.3,4 Amadori reaction allows formation of 1-aminodeoxyketose
from aldose and amine/amino acid whereas, Heyns reaction results in the formation of 2-
aminodeoxyaldoses from the ketose and amine/amino acid without requirement of protection
and deprotection steps.1 Amadori and Heyns reaction forms the first step of complex Maillard
reaction and they are stable intermediates. The classical synthetic method for accomplishing
Amadori and Heyns compounds derived from amino acids employ methanol as a solvent
with acetic acid as a catalyst and pH regulator and necessitates purification by either column
or ion exchange chromatographic techniques.5 Recently, Lewis acid ZnCl2 has been
6-8
demonstrated as catalyst for the synthesis of Amadori/Heyns compounds
However, it
works only in aprotic solvent combination of pyridine and acetic acid and also requires
substantial time for completion. We were intrigued by lack of catalytic protocols in polar
solvents including methanol-acetic acid for these compounds and also driven by curiosity to
understand the factors hindering development of catalytic version of classical synthetic
reaction in protic solvents. Towards this step, we looked carefully in the literature pertaining
to mechanism of these rearrangement reactions for critical factors. The Amadori and Heyns
reaction has the first step, formation of Schiff base. Subsequently, Schiff base undergoes
rearrangement because of presence of α-hydroxy group (probably driven by pH) wherein
imine to keto-enol transformation occurs. Further, it has been shown previously that Zn2+
drives the rearrangement via formation of complex with Schiff base.6-8 Keeping this fact, We
4

hypothesized that formation of Schiff base could be having some hindrance which
subsequently results in moderate conversion to Amadori/Heyns product that necessitate
employment of separation techniques. We further, hypothesized that formation of Schiff base
is dictated by the acyclic ratio of the reducing sugar and that is the bottleneck controlling fate
of rearrangement reaction leading to only moderate conversion. This hypothesis has some
grounds supported by previous observations that reducing sugars consisting of hexoses or
ketoses occur as ring structures and only tiny fraction or trace amount occur as acyclic
structures and acyclic form is the most reactive form.5,9
Table 1: Percentage of acyclic forms of reducing sugars
Reducing sugar
Percentage
acyclic form
0.01*
of
D-glucose
D-mannose
D-galactose
D-fructose
0.026*
0.052*
0.5 #
D-xylose
0.075+
* ref10,#ref11,+ref 12
To validate our hypothesis, we carried out reaction of D-glucose with L-valine in
MeOH-AcOH combination in presence of ZnBr2. This resulted in majority of the Amadori
product N-(1-deoxy-fructos-1-yl)-L-valine but still some starting material was remaining.
Then, we did reaction of D-mannose with L-valine under similar conditions and found that,
the desired compound N-(1-deoxy-fructos-1-yl)-L- valine had indeed formed and no starting
material was left behind. The acyclic form of D-mannose occur nearly threefold higher than
D-glucose (Table 1).
5

Scheme 1: Amadori compounds derived from D-Mannose
We reasoned that, D-mannose with acyclic form of 0.026 % compared to only 0.01 %
of glucose was driving force for the formation of Schiff base and further with assistance from
the Lewis acid catalyst ZnBr2 lead to the accomplishment of the pure Amadori compound.
Enlightened by this new observation, we switched to D-galactose which has about 0.052 % of
acyclic form. From the earlier experiments, we anticipated neat reaction and the story turned
out to as expected. We observed fast reaction and the reaction time was only about 6 h
(Scheme 1). We then set out for the optimization of zinc based Lewis acid catalyst to figure
out effect of changing anions of zinc salts on the Amadori synthesis. We expected that in the
solvent mixture of MeOH-AcOH, which is relatively polar environment, ZnBr2 would
undergo dissociation to Zn2+ and Br- and the catalytic activity is due to solvated Zn2+. We
expected that reaction may work irrespective of nature of anion. The primary trouble is
solubility issues of reaction associated with the presence of different anions which alter the
catalytic reaction due to non-availability of catalyst. In this context, we screened several zinc
salts as shown in the table 2, and to our surprise, we found that Zn(OAc)2 was the best
catalyst. We believe that Zn(OAc)2 with acetate anion is the best choice as solvent mixture
6

has also acetic acid due to this the solubility of Zn(OAc)2 is much better than other catalyst
including ZnBr2, ZnCl2 or Zn(OTf)2. Utilizing Zn(OAc)2, we accomplished twelve tagatose
derivatives as shown in the Scheme 2. Only three tagatose derivatives have been reported
earlier.13-14 Our developed method works for all sorts of amino acids consisting of aromatic,
aliphatic and even sulphur containing amino acid D- methionine.
Table 2: Optimization of catalyst
Catalyst
Mol %
Product yield
(%)
Entry
ZnBr2
4.44
0.84
0.44
0.73
85
80
80
83
ZnCl2
Zn(OTf)2
Zn(OAc)2
0.27
80
5.45
2.72
1.63
0.81
0.54
85
86
84
86
85
7

Scheme 2: Amadori compounds derived from D-Galactose
Then, we switched to Heyns reaction and carried out reaction of D-fructose with L-
valine under the similar conditions of MeOH and AcOH, wherein, we observed the formation
of Heyns compounds (Scheme 3). The fructose has acyclic form of 0.5 % and it is 20 fold
higher than D-mannose. In total, we made seven compounds to demonstrate the concept. All
these Heyns compounds are already reported.
8

Scheme 3: Heyns compounds derived from D-Fructose
Then we also tried the same reaction with one of the pentose sugar D-xylose. D-
Xylose has acyclic form to the tune of 0.075 %. We reacted D-xylose with few amino acids
using Zn(OAc)2 catalyst. We accomplished the Amadori compounds in pure form and the
reaction was completed only in just 2h (Scheme 4). This short duration of reaction was
anticipated due to higher availability of acyclic form.
9

Scheme 4.:Amadori compounds derived from D-Xylose
We also carried out a reaction of D-maltose with L-valine under similar conditions to
exemplify the applicability of the concept to disaccharides. The reaction was facile and gave
the expected Amadori product N-(1-deoxy-maltos-1-yl)-L-valine in the pure form (Scheme
5). D-maltose is a glucose derived disaccharide, the amount of aldehyde form (acyclic form)
of one of the glucose is the key for the Amadori reaction to occur, however, the quantity of
partial acyclic form is not reported in the literature. As the D-maltose gave facile reaction, we
expect that the amount of partial acyclic form is higher than that of D-glucose. Probably, 13C
NMR investigations on 13C labelled D-maltose may shed light on this regard.
10

Scheme 5: Amadori compound derived from D-Maltose
The plausible mechanism for reaction is depicted in the scheme 6. The initial step is
formation of acyclic form and this is governed by thermal equilibrium. In the next step
acyclic sugar reacts with amino acid resulting in Schiff base. As a consequence, the formation
Schiff base is dictated by the availability of acyclic form. Once the Schiff base is formed,
then it reacts with Zn2+ to form Zinc-imine complex. Further, due to electron pull of Zn2+,
rearrangement of double bond occurs resulting in a Zinc-amine complex. Subsequently,
elimination of Zn2+ coupled with thermodynamically preferential enol to keto transformation
furnishes acyclic Amadori compound, which further upon ring closure leads to cyclic
Amadori compound.
11

Scheme 6: Plausible mechanism for the formation of Amadori compound
In conclusion, by revisiting classical method of Amadori and Heyns synthesis, we
transformed the classical method to catalytic method in light of newer understanding of
critical role of quantity of acyclic form assisted by Lewis acid catalyst. We also optimized
newer catalyst and found Zn(OAc)2 a cheaper and better catalyst. By utilizing optimized
derivatives, we accomplished several Amadori and Heyns compounds.
Experimental section:
Synthesis of N-(1-Deoxy-D-tagatos-1-yl)-L-valine
L-Valine(1.17 g, 10 mmol) was loaded into Erlenmeyer flask having 70 mL MeOH
and 5 mL of acetic acid, subsequently D-galactose (1.80 g, 10 mmol) was added followed by
12

Zn(OAc)2(10 mg, 0.54 mol%) and the reaction mixture was refluxed for about 6 h, during
which solution turned reddish brown colour. The resultant solution was filtered through celite
while hot by applying partial vacuum and the filtrate was concentrated to about 3 mL by
subjecting to vacuum. Further, 50 mL of acetone was added to cause the precipitation.
Furthermore, acetone was decanted carefully and again three times acetone (50mL) was
added, stirred and decanted to get off white precipitate and later the precipitate was dried
under vaccum. Yield = 2.37 g, (85%). HRMS (ESI) calcd for C11H21NO7Na [M+Na]+
302.1216, found, 302.1244. [α]D20 = (c =1, water) IR spectra(cm-1): 1701(C=O of acyclic
compound), 1648 (C=O), 1416 (O-H In plane bend), 1043 (C-N stretch), 786(out of plane N-
H bend).
4
O
HO
CH3
9
6
HO
HO
2
5
8
NH
HO
7
1
CH3
10
3
OH
11
O
α-pyranose form
1H NMR (D2O, 400 MHz): δ 0 9 (d, J = 6.8 Hz, 3H, C(10)H3), 1.01 (d, J = 6.7 Hz, 3H,
C(9)H3), 2.23 (m. 1H), 3.20 (d, J = 13.1 Hz, 1H, C(6)H), 3.38 (d, J = 13.1 Hz, 1H, C(6)H),
3.58 (d, J = 4.3 Hz, 1H, C(7)H), 3.58 (d, J = 10.0 Hz, 1H, C(5)H), 3.82 (d, J = 10.0 Hz, 1H,
C(5)H), 3.83 (m, J = 9.5 Hz, 1H, C(3)H), 3.84 (m, J = 10.1 Hz, 1H, C(4)H), 3.93 (d, J = 3.3
Hz, 1H, C(3)H).
13C NMR (D2O, 125 MHz): δ 1 95(C10), 18.16(C9), 28.95(C8),53.20(C6), 62.52(C5),
65.69(C4), 69.04(C7),70.63(C3), 72.41(C2), 95.04(C1), 172.42(C11).
β-pyranose form
13

13C NMR (D2O, 125 MHz): δ 1 85(C10), 18.32(C9), 30.20(C8),52.71(C6), 60.09(C5),
65.88(C4),68.85(C7),70.53(C3), 71.18(C2), 95.66(C1), 172.8(C11).
Supporting Information
The supporting information consisting of detailed synthetic procedures and spectra is
available in the journal website.
ORCID
Nanishankar V.Harohally:0000-0003-2306-5897
Present Addresses
† Spice and Flavour Science Department, CSIR-Central Food Technological research
Institute. KRS Road, Mysuru-570020, Karnataka, India
Notes
The authors declare no competing financial interest.
Author Contributions
The manuscript was written through contributions of all authors.
ACKNOWLEDGMENT
We thank Prof.Ram Rajsekharan for the support and encouragement. We acknowledge
Gangothri M.V. for helping in recording spectra. This work was supported by DST-WOS A
scheme grant awarded by DST, New Delhi.
REFERENCES
1. Wrodnigg TM, Eder B. Topics in current chemistry 2001;215:115–152.
2. Horvat S, Jakas A. J. Pept. Sci. 2004;10:119–137.
14

3. Mauron J. Prog. Food. Sci. 1981;5:5–35.
4. Bunn HF, Higgins PJ. Science 1981;213:222-223.
5. Rosen L, Woods JW, Pigman W. J. Am. Chem. Soc. 1958;80:4697–4702
6. Piispanen PS, Norin T. J. Org. Chem. 2003;68:628–630.
7. Srinivas SM, Harohally NV. J. Agric. Food Chem. 2012; 60:1522–1527.
8. Harohally NV, Srinivas SM, Umesh S. Food Chem. 2014;158:340–344.
9. DeWinter HL, Von Itzstein M. Biochemistry 1995;34:8299–8308.
10. Zhu Y, Jajicek J, Serianni AS. J. Org. Chem. 2001; 66:6244–6251.
11. Barclay T, Ginic-Markovic M, Johnston MR, Copper P, Petrovsky N. Carbohydr. Res.
2012;347:136–141.
12. Drew KN, Jajicek J, Bondo G, Bose B, Serianni AS, Carbohydr. Res. 1998;307:199–209.
13. Kaufmann M, Missner PM, Pelke D, Mugge C, Kroh LW. Carbohydr. Res. 2016; 428:
87–99.
14. Onodera K, Uehara T, Kitaoka S. Bul.Agr.Chem.Soc.Japan. 1960;24:703–710.
15

Highlights
 Acyclic form of reducing sugar is crucial for catalytic Amadori and Heyns synthesis
 Zn(OAc)2 is newer Lewis acid catalyst for Amadori and Heyns Synthesis
 Amadori compounds of D-tagatose accomplished
16