Content uploaded by M Rosario Domingues
Author content
All content in this area was uploaded by M Rosario Domingues on Jul 09, 2014
Content may be subject to copyright.
Coffee melanoidins: structures, mechanisms of formation and potential health
impacts
Ana S. P. Moreira,
a
Fernando M. Nunes,
b
M. Ros
ario Domingues
a
and Manuel A. Coimbra*
a
Received 28th February 2012, Accepted 5th April 2012
DOI: 10.1039/c2fo30048f
During the roasting process, coffee bean components undergo structural changes leading to the
formation of melanoidins, which are defined as high molecular weight nitrogenous and brown-colored
compounds. As coffee brew is one of the main sources of melanoidins in the human diet, their health
implications are of great interest. In fact, several biological activities, such as antioxidant,
antimicrobial, anticariogenic, anti-inflammatory, antihypertensive, and antiglycative activities, have
been attributed to coffee melanoidins. To understand the potential of coffee melanoidin health benefits,
it is essential to know their chemical structures. The studies undertaken to date dealing with the
structural characterization of coffee melanoidins have shown that polysaccharides, proteins, and
chlorogenic acids are involved in coffee melanoidin formation. However, exact structures of coffee
melanoidins and mechanisms involved in their formation are far to be elucidated. This paper
systematizes the available information and provides a critical overview of the knowledge obtained so
far about the structure of coffee melanoidins, mechanisms of their formation, and their potential health
implications.
1. Introduction
The roasting of green coffee beans is an important step in coffee
processing. The characteristic aroma, taste, and color of the
coffee brew, prepared by hot water extraction from roasted and
ground coffee beans, are to a great extent determined by the
roasting process.
1–5
During this process, the chemical
composition of the beans is changed due to the degradation and/
or transformation of some of the compounds identified in green
coffee beans. Green coffee beans are, on a dry matter basis,
mainly composed by carbohydrates (59–61%), lipids (10–16%),
proteins (10%), and chlorogenic acids (7–10%), containing lower
amounts of minerals (4%), aliphatic acids (2%), caffeine (1–2%),
trigonelline (1%), and free amino acids (<1%). Upon roasting,
a decrease is observed in coffee bean carbohydrates (38–42%),
proteins (8%), chlorogenic acids (3–4%), and free amino acids,
whereas lipids (11–17%), minerals (5%), aliphatic acids (3%),
caffeine (1–2%), and trigonelline (1%) keep their relative
From left to right :Ana
S:P:Moreira;Fernando M:
Nunes;M:Ros
ario Domingues;Manuel A:Coimbra
Ana S. P. Moreira initiated in the present year her PhD project in the
Department of Chemistry at the University of Aveiro (Portugal),
under the scientific supervision of Dr M. Ros
ario Domingues and Prof.
Manuel A. Coimbra. Her work is focused on the study of structural
modifications induced by thermal and oxidative treatments in model
oligosaccharides and polysaccharides isolated from coffee, using mass
spectrometry as the main technique for structural analysis. She received
her BA in Biochemistry in 2009 and her Master in Analytical Chem-
istry and Quality in 2011 at the University of Aveiro.
a
Departamento de Qu
ımica, Universidade de Aveiro, 3810- 193 Aveiro,
Portugal. E-mail: mac@ua.pt; Fax: +351 234 370084; Tel: +351 234
370706
b
Departamento de Qu
ımica, Universidade de Tr
as-os-Montes e Alto Douro,
5001- 801 Vila Real, Portugal
This journal is ªThe Royal Society of Chemistry 2012 Food Funct., 2012, 3, 903–915 | 903
Dynamic Article LinksC
<
Food & Function
Cite this: Food Funct., 2012, 3, 903
www.rsc.org/foodfunction REVIEW
content.
1
Also, as a result of the transformations occurring
during the coffee roasting, melanoidins are formed.
6
Melanoidins are the final products of the Maillard reaction.
7–9
This non-enzymatic browning reaction encompasses a network
of various reactions between reducing sugars and compounds
with a free amino group forming a variety of products, which can
be classified as early stage products, intermediate stage products,
and last stage products, the melanoidins.
7
Melanoidins are
generically defined as high molecular weight nitrogenous brown-
colored compounds.
6,10
Scarce information is available about
their chemical structures, although they are formed during the
heat processing of a large range of food products beyond coffee,
such as bread,
11
malt,
12
meat,
13
and tomato sauce.
14
Because
melanoidins cannot be directly analyzed due to the uncertainty of
their structures, they are usually quantified by difference, sub-
tracting the total percentage of known compounds from 100
percent. Using this criterion, they were estimated to account for
up to around 25% (w/w) of the dry weight of roasted coffee
beans.
1,15
In coffee brew, melanoidins were estimated to account for up
to around 29% (w/w) of the dry matter, quantified by differ-
ence.
15
The melanoidin content in coffee brew (and in coffee brew
fractions) has also been evaluated based on their contribution to
the brown color of the brew using color dilution analysis,
16,17
or
by measuring the absorption near 400 nm,
18–26
particularly at 405
nm.
20–26
The absorption spectrum from 200 to 700 nm of a coffee
brew shows two absorption maxima, one at 280 nm and a second
at 325 nm. The absorption maximum at 280 nm can be explained
by the presence of the aromatic rings of proteins, caffeine,
chlorogenic acids, and caffeic acid. The absorption maximum at
325 nm can be explained by the presence of chlorogenic acids and
caffeic acid.
22
The melanoidin content in coffee brew has been
evaluated by measuring the absorption at 405 nm, since this is
a wavelength often chosen to measure the intensity of the brown
color. To be able to compare absorption values at 405 nm of
different coffee brews (and coffee brew fractions), Bekedam
et al.
22
proposed to express the absorption (A) at this wavelength
as K
mix 405nm
, which is the specific extinction coefficient at 405 nm
calculated using the law of Lambert–Beer, expressed as A¼K
mix
(L g
1
cm
1
)concentration (g L
1
)length of light path (cm).
Comparing coffee brews prepared from roasted coffee beans
with different degrees of roast (and their high molecular weight
fractions), an increase of the K
mix 405nm
value with increasing the
degree of roast was observed,
20
suggesting a direct relation
between the melanoidin content and the degree of roast.
Coffee brew is considered one of the main sources of mela-
noidins in the human diet, since it is consumed by millions of
people worldwide every day.
11
Several studies on coffee mela-
noidins have been performed since at least the 1960s.
27
These
studies, focused primarily on melanoidins structural character-
ization, have more recently been extended to their biological
activities and effects on human health. In this paper a critical
overview is presented about what has been published to date
about the structure of coffee melanoidins, mechanisms of their
formation, and their potential health implications.
2. Coffee melanoidin structures
Despite all efforts, the chemical structure of coffee melanoidins
remains largely unknown. On the basis of current knowledge, it
can be stated that the structural elucidation of coffee melanoidins
is hampered by the extreme complexity and diversity of their
structures and the consequent difficulty in isolation of pure
melanoidin fractions.
Different approaches have been employed for isolation and
purification (if any) of coffee brew melanoidins.
6
In all
approaches, coffee melanoidins are first isolated taking advan-
tage of their high molecular weight. The high molecular weight
material (HMWM) of coffee brews has been isolated by dialysis
using membranes with a molecular weight cut-off (MWCO) of
2 kDa
28
and 12–14 kDa,
16
by diafiltration using a hollow fiber
with a MWCO of 3 kDa,
20,22,23
and by stepwise tangential flow
ultrafiltration using membranes with a MWCO of 100, 50, 10,
and 3 kDa.
17
Also, the HMWM has often been isolated by
ultrafiltration using a membrane with a MWCO of 10 kDa.
29–36
Uncommonly, the high molecular weight coffee brew fractions
have been isolated by gel filtration chromatography, monitoring
Fernando M. Nunes is presently assistant professor at the Chem-
istry Research Center, Chemistry Department at the University of
Tr
as-os-Montes e Alto Douro, Portugal. He received his BA in
Food Chemistry at University of Aveiro in 1995, MSc in Natural
Products and Food Chemistry at University of Aveiro in 1998 and
PhD in Chemistry/Food Chemistry at University of Tr
as-os-
Montes e Alto Douro in 2005. His research is concerned with
structural characterization of polysaccharides and changes induced
by oxidative and thermal treatments. Main work has been devel-
oped in the characterization of coffee polysaccharides and their
involvement in coffee melanoidin formation.
M. Ros
ario Domingues is presently assistant professor at the Mass
Spectrometry Center, Chemistry Department at the University of
Aveiro, Portugal. She received her BA in Pharmaceutical Sciences
at University of Coimbra in 1990 and her PhD in Chemistry-Mass
Spectrometry at Aveiro University in 1998. Her research is con-
cerned with applications of mass spectrometry for the study of
biomolecules structural features, and in the identification of the
biomolecules modification induced by oxidative stress and thermal
treatments. Main work has been developed in the characterization
of oligo and polysaccharides and in the study of lipids/phospho-
lipids from natural sources by mass spectrometry.
Manuel A. Coimbra is associate professor with habilitation at the
Department of Chemistry at the University of Aveiro, Portugal.
He received his BA in Biochemistry at University of Porto and
PhD in Chemistry/Food Chemistry at University of Aveiro in 1993,
under the supervision of Prof. Ivonne Delgadillo (U Aveiro) and
Dr Robert R. Selvendran (Institute of Food Research, Norwich,
UK). He is Professor of Biochemistry and Food Chemistry. Since
2011, he is Associate Editor of Carbohydrate Polymers (Elsevier).
The principal research interests are polysaccharides chemistry,
food chemistry and biochemistry, and volatile compounds. Co-
author of 128 scientific papers.
904 | Food Funct., 2012, 3, 903–915 This journal is ªThe Royal Society of Chemistry 2012
the elution profile by using the absorbance at 405 nm.
21,26
The
diversity of separation techniques used, namely dialysis, diafil-
tration, and ultrafiltration using membranes with different
MWCOs, shows that the different authors did not yet establish
a consensus for the minimum molecular weight for melanoidins.
Because the HMWM of the coffee brew prepared from
roasted beans comprises proteins and polysaccharides beyond
melanoidins,
37–39
additional purification procedures are
required for their separation. Over the last decade, different
melanoidin populations have been purified from the HMWM
of the coffee brew.
16,17,20,23,25
The purification of different mel-
anoidin populations has been achieved exploiting different
physico-chemical properties such as solubility, charge, metal
chelating ability, and hydrophobicity. The HMWM has been
fractionated into fractions with different polysaccharide
composition using an ethanol fractionation procedure. The
fractions most enriched in galactomannans precipitate at an
ethanol concentration of 40–50% and the fractions most
enriched in arabinogalactans precipitate at an ethanol concen-
tration of 75–80% (% by volume or weight). The fractions that
remain soluble at an ethanol concentration of 75–80% present
the lowest carbohydrate content.
16,22,37–39
Based on the anionic
nature of coffee melanoidins,
19
anionic melanoidin fractions
have been isolated performing anion exchange chromatography
on the fractions obtained by ethanol fractionation,
16
or on the
HMWM isolated directly from the coffee brew.
23,25
Also, due to
the metal chelating capacity of coffee melanoidins,
40,41
a further
fractionation of the anionic melanoidin fractions performed by
copper affinity chromatography allows the isolation of anionic
and chelating melanoidin fractions.
16
Also, some studies have
shown that arabinogalactans in the form of arabinogalactan-
proteins (AGPs) are present in green and roasted coffee beans
as well as in green and roasted coffee brews, the latter including
instant coffee.
23,42–46
Melanoidin fractions containing intact
AGPs have been isolated from the HMWM of (roasted) coffee
brews by precipitation with the AGP-specific b-glucosyl Yariv
reagent.
20,23
In addition, hydrophobic melanoidin fractions
have been isolated from the HMWM by gel permeation chro-
matography on Sephadex LH-20 followed by hydrophobic
interaction chromatography.
17
In respect to the physico-chemical properties of the melanoi-
dins present in the HMWM of the coffee brew, new insights have
been gained during the last decade using the purification proce-
dures above described. It is now known that a fraction of coffee
melanoidins present anionic character whereas other fractions do
not present it or present it in a very low extent.
16,23
This hetero-
geneity is also extended to their metal chelating capacity, where
about half of the anionic melanoidin fractions have chelating
ability for immobilized copper ions.
16
Moreover, there are coffee
melanoidin fractions that present hydrophobic character.
17
It has been suggested that carbohydrates, amino acids, and
phenolic compounds are components of coffee melanoi-
dins.
27,47–49
Over the last decade, the chemical characterization of
isolated and purified melanoidin fractions by a wide range of
analyses, including analysis of sugars, glycosidic linkages, amino
acids, nitrogen content, and phenolic groups content, has given
increasing evidences that polysaccharides, proteins, and chloro-
genic acids are involved in the formation of coffee melanoi-
dins.
16,17,20,23,25
The following sections summarize what is known
about the involvement of polysaccharides, proteins, and
chlorogenic acids in the structures of coffee melanoidins.
2.1. The contribution of polysaccharides
Polysaccharides are the predominant carbohydrates present in
green coffee beans, comprising about 50% of the dry weight of
the beans. Galactomannans and type II arabinogalactans are
their most abundant polysaccharides.
1,50
The galactomannans
extracted with hot water from green coffee beans are composed
by a main backbone of b-(1 /4)-linked D-mannose residues,
some of them substituted at O-6 by single residues of a-D-
galactose or L-arabinose and at O-2 and/or O-3 by acetyl groups.
Regarding to the acetylated mannose residues, there are single
acetylated residues, di-acetylated residues and consecutively
acetylated residues. Also, b-(1 /4)-linked D-glucose residues are
components of the mannan backbone.
51
Fig. 1a shows the main
structural features of hot water extractable green coffee gal-
actomannans. The arabinogalactans extracted with hot water
from green coffee beans are composed by a main backbone of b-
(1 /3)-linked D-galactose residues, some of them substituted at
O-6 with short chains of b-(1 /6)-linked D-galactose residues.
The galactose residues of these short chains are substituted with
various combinations of a-L-arabinose, a-L-rhamnose, and b-
D-glucuronic acid residues.
43
Almost all arabinogalactans in
green coffee beans are covalently linked to proteins, so called
arabinogalactan–proteins (AGPs).
42
Fig. 1b shows the main
structural features of hot water extractable green coffee arabi-
nogalactans (the polysaccharide moiety of AGPs).
During coffee roasting, arabinogalactans and galactomannans
undergo several structural modifications. Based on sugar and
Fig. 1 Illustration of the main structural features of galactomannans (a)
and arabinogalactans (b) isolated by hot water extraction of green coffee
beans.
This journal is ªThe Royal Society of Chemistry 2012 Food Funct., 2012, 3, 903–915 | 905
methylation analyses of the high molecular weight material
(HMWM) isolated from green and roasted coffee brews, it has
been stated that the roasting of green coffee beans promotes the
decrease of the degree of polymerization as well as the
debranching of these polysaccharides. In particular, arabinose,
mainly present as side chains in the arabinogalactans, was found
to be the sugar most sensitive to degradation during roast-
ing.
37–39,52,53
However, dry thermal treatments on manno- and
galactomanno- oligosaccharides, structurally related with coffee
galactomannans, showed the occurrence of polymerization
reactions.
54
This finding raised the hypothesis that the roasting of
green coffee beans also promotes the polymerization of the gal-
actomannans. Also, a detailed study on roasted coffee brew
galactomannans showed that their reducing end is modified by
the occurrence of caramelization, isomerization, oxidation,
decarboxylation, and Maillard reactions.
55
Melanoidin populations with different polysaccharide
compositions have been isolated from the HMWM of coffee
brews.
16,20,23,28
Based on the sugar composition of the fractions
obtained from the HMWM of a roasted coffee brew by ethanol
fractionation, Bekedam et al.
22
suggested the involvement of
polysaccharides, especially arabinogalactans, in coffee melanoi-
din formation. The existence of polysaccharides covalently
linked in melanoidins was proven later on by Nunes and
Coimbra
16
by the observation of galactomannans and arabino-
galactans in anionic fractions isolated from the ethanol fractions
of roasted coffee brews by anion exchange chromatography,
which contrasted with the negligible retention observed for the
green coffee polysaccharides under the same chromatographic
conditions.
Arabinogalactans can be incorporated in coffee melanoidin
structures in the form of intact AGPs.
23
This was demonstrated
by the addition of the Yariv reagent (a reagent that selectively
precipitates AGPs) to the HMWM of a roasted coffee brew.
Instead of the white color of the AGP fraction previously
recovered from green coffee beans,
56
a brown colored precipitate
was obtained. Based on the K
mix 405nm
values, it was shown that
the AGP fraction accounts for approximately half of the mela-
noidins present in the HMWM of the coffee brew. However, as
the amount of the galactose and arabinose of the AGP fraction
obtained by precipitation with the Yariv reagent is smaller than
that observed in the HMWM, it can be inferred that some of the
green coffee bean AGPs are transformed during roasting by
losing their protein moiety, as suggested by Bekedam et al.
23
The studies carried out so far on coffee melanoidin poly-
saccharides showed that they are covalently-linked components
of these high molecular weight brown structures. However, the
types of linkages between the polysaccharides and the other
components are not yet known.
2.2. The contribution of proteins
Proteins of green coffee beans, accounting for about 10% of bean
dry weight, can be divided according to their solubility in water:
50% are water-soluble and 50% are water-insoluble proteins.
1
It
is also known that coffee beans contain 11 S-type storage
proteins,
57,58
representing almost 45% of total proteins.
59
The roasting process leads to protein denaturation with
degradation, the later inferred by the decrease in total amount of
amino acids identified in coffee beans upon roasting. Among the
amino acids that make up green coffee proteins, some of them,
such as arginine, cysteine, lysine, and serine, showed a high
decrease in their amount during roasting.
1,15,22,60
The changes in
green coffee proteins promoted by roasting have also been fol-
lowed by sodium dodecyl sulfate-polyacrylamide gel electropho-
resis (SDS-PAGE).
37–39,61
The SDS-PAGE patterns obtained
under nonreducing conditions of the HMWM isolated from green
and roasted coffee brews are clearly distinct. While green coffees
presented a major protein band at 58 kDa and a second one at
38 kDa, roasted coffees presented a defined band with #14 kDa
and a diffuse band with >200 kDa,
37–39
the latter possibly due to
the involvement of proteins in melanoidins formation.
Different melanoidin fractions have been isolated from the
HMWM of coffee brews containing variable amounts of protein-
like materials, quantified by amino acid analysis after acid
hydrolysis.
16,17,22
Comparing different melanoidin fractions
obtained from ethanol fractionation of the HMWM, it was
noticed that the fraction that remains soluble at a high ethanol
concentration (75–80%) has the highest protein content.
However, the amino acid composition (in terms of amino acid
relative abundances) is similar for all these melanoidin fractions.
Alanine, aspartic acid/asparagine, glutamic acid/glutamine, and
glycine are among the most abundant amino acids in all frac-
tions, whereas histidine, lysine, methionine, and tyrosine are the
least abundant. Arginine was not found.
16,22
The amino acid
composition of these melanoidin fractions is also similar to those
reported for roasted coffee beans and roasted coffee brews.
22
Moreover, various melanoidin fractions contain hydroxypro-
line,
16
an amino acid found in high amounts in AGPs isolated
from green coffee beans and green coffee brews.
42,43
The presence
of hydroxyproline in melanoidin fractions containing also
arabinose and galactose residues is more evidence for the exis-
tence of AGPs in coffee melanoidin structures, as discussed
earlier.
Amino acids account for the majority of the nitrogen present
in the HMWM of the roasted coffee brew (>70%) and in the
melanoidin fractions obtained from the HMWM by ethanol
fractionation. Also, the amount of non-amino acid nitrogen is
higher in the melanoidin fraction with highest ethanol solubility.
This distinction between nitrogen from intact amino acids and
non-amino acid nitrogen, also referred as non-protein nitrogen,
has been achieved based on the amino acid/protein content
determined by amino acid analysis after acid hydrolysis and the
total nitrogen content.
22
However, the degraded/modified amino
acids may account for the nitrogen of non-proteic origin deter-
mined using this approach. Effectively, Maillard reaction prod-
ucts derived from lysine, including N
3
-(fructosyl)lysine (FL,
detected as furosine after acid hydrolysis), N
3
-(carboxymethyl)
lysine (CML), and N
3
-(carboxyethyl)lysine (CEL) (Fig. 2), were
identified in the HMWM and melanoidin fractions.
16
The iden-
tification of these compounds, although at very low amounts,
suggests that amino acids modified during coffee roasting are
also incorporated in melanoidin structures.
2.3. The contribution of chlorogenic acids
Phenolic compounds of green coffee beans are predominantly
chlorogenic acids (CGAs), a family of esters formed between
906 | Food Funct., 2012, 3, 903–915 This journal is ªThe Royal Society of Chemistry 2012
quinic acid and trans-cinnamic acids, such as caffeic, p-coumaric,
and ferulic acids.
1,62,63
More than 60 CGAs were already identi-
fied in green coffee beans,
64
being the most abundant 5-O-caf-
feoylquinic acid (5-CQA, Fig. 3).
65,66
Total chlorogenic acid content found in green coffee beans
account for up to 10% of their dry weight.
1
Upon roasting, the
amount of hot water extractable chlorogenic acids decreases by
50% or more, depending on roasting intensity.
65
In respect to the
products derived from the thermal degradation of CGAs during
coffee roasting, it is already known that they are diverse and
range from simple phenols to condensation products of high
structural complexity, as recently reviewed by Nunes and
Coimbra.
6
Briefly, the coffee roasting process promotes the
isomerization of CGAs, as well as their hydrolysis yielding quinic
acid (non-phenolic moiety) and various cinnamic acids (phenolic
moieties).
63,67,68
Model studies on dry thermal treatment of 5-
CQA, caffeic acid, and ferulic acid suggest the occurrence of
decarboxylation of quinic and cinnamic acids during coffee
roasting, yielding a range of simple phenols.
69–71
For caffeic acid,
several condensation products were also identified.
71
Accord-
ingly, bitter compounds present in coffee brews seem to be
generated by oligomerization of a simple phenol (4-vinyl-
catechol) released from caffeic acid moieties upon roasting.
72
Furthermore, CGAs can be converted into chlorogenic acid
lactones by the loss of a water molecule from the quinic acid
moiety and formation of an intramolecular ester bond
(Fig. 4).
66,67,73
Various melanoidin fractions containing CGAs (or their
derivatives) have been isolated from the HMWM of coffee
brews.
16,20,25
The presence of CGAs in the HMWM and mela-
noidin fractions was suggested based on the quantification of
phenolic compounds using the Folin–Ciocalteu reagent.
22,23
However, this colorimetric assay is not specific for phenolic
compounds, but measures reducing compounds.
74
The incorpo-
ration of CGAs in coffee melanoidins was also suggested by the
detection of phenols, such as guaiacol, 4-ethylguaiacol, and 4-
vinylguaiacol, after thermal degradation of the HMWM isolated
from a coffee brew,
14
and also identified after thermal treatment
of ferulic acid.
69
However, this approach does not allow
evidencing the type of interaction (covalent or non-covalent)
between CGAs and other components. The presence of cova-
lently-linked CGAs (or their derivatives) in the HMWM and
melanoidin fractions was demonstrated using the alkaline fusion
method, known as an efficient method to release condensed
phenolic structures.
16
To ensure the absence of non-covalently-
linked CGAs, since non-covalently-linked CGAs were previously
shown to occur in melanoidin fractions isolated from instant
coffees,
29
the fractions subjected to alkaline fusion were previ-
ously submitted to a reversed phase high performance liquid
chromatography after overnight incubation in 2 M NaCl. For
both HMWM and melanoidin fractions, the most abundant of
the monomeric phenolic compounds recovered after alkaline
fusion was 3,4-dihydroxibenzoic acid. This compound was also
the most abundant upon the alkaline fusion of ferulic and caffeic
acid standards.
16
Alkaline fusion of a melanoidin fraction
obtained from instant coffee by zinc precipitation gave similar
monomeric phenolic compounds.
41
On the basis of these results,
it can be stated that CGAs and/or their derivatives are incor-
porated in coffee melanoidin structures. The incorporation of 5-
CQA, and caffeic, ferulic, and quinic acids was also evaluated in
two subsequent studies using both the HMWM and the inter-
mediate-MWM (IMWM) isolated from a coffee brew.
20,25
These
studies showed the presence of ester linked caffeic, ferulic, and
quinic acids, released after alkaline saponification. Also, these
ester linkages were observed in the melanoidin fractions isolated
from both HMWM and IMWM, with a more abundant presence
of quinic acid than caffeic and ferulic acids. The presence of
intact CGAs into the melanoidin structures, incorporated via
caffeic acid moiety through mainly non-ester linkages, was
inferred by the enzymatic treatment with chlorogenate esterase
(EC 3.1.1.42).
25
In other studies, two-dimensional nuclear
Fig. 2 Maillard reaction products identified in coffee melanoidin
structures: N
3
-(fructosyl)lysine (FL), N
3
-(carboxymethyl)lysine (CML),
and N
3
-(carboxyethyl)lysine (CEL).
Fig. 3 Structure of 5-O-caffeoylquinic acid (5-CQA).
Fig. 4 Formation of a 1,5-g-quinolactone from 3-O-caffeoylquinic acid
(3-CQA) occurring during roasting as proposed by Farah et al.
73
This journal is ªThe Royal Society of Chemistry 2012 Food Funct., 2012, 3, 903–915 | 907
magnetic resonance (2D NMR) analysis did not support the idea
of intact caffeic or ferulic acid moieties integrated into the mel-
anoidins of the high molecular weight coffee fraction.
17,28
The
different results reported about the integration of intact cinnamic
acid moieties into the coffee melanoidins point out the need of
future studies designed to investigate this point.
Although there is currently strong evidence that CGA deriv-
atives are components of coffee melanoidins, it is not yet known
how they are linked within the melanoidin structure. Nunes and
Coimbra
6
suggested that proteins can be a possible binding site
for the CGA derivatives. This hypothesis was proposed based on
studies demonstrating that the changes in green coffee protein
profiles observed during roasting are similar to what are
observed when oxidized CGAs are reacted with green coffee
proteins in model systems.
61,75
Also, carbohydrates, particularly
arabinose residues, seem to be a possible binding site for the
CGA derivatives, as suggested by Bekedam et al.
25
This
hypothesis was proposed based on studies demonstrating that
the arabinose residues are quite susceptible to degradation
during roasting,
46,53
and on a model study developed under
simulated roasting conditions in which the reaction of epi-
catechin with sugar fragments is reported to occur.
76
3. Formation mechanisms of coffee melanoidins
Although some structural features of coffee melanoidins have
already been elucidated, the mechanisms involved in their
formation are far to be completely understood.
Based on model studies (mostly in solution), three theories on
melanoidin formation have been described.
6,10
One theory is that
the melanoidins are formed by polymerization (via poly-
condensation reactions) of low molecular weight (LMW) Mail-
lard reaction products, such as furans and pyrroles, formed in the
advanced stages of the reaction.
77–79
Hofmann
80–82
has suggested
that the melanoidins are derived from cross-linking of LMW
Maillard reaction products to proteins via reactive side chains of
amino acids such as lysine, arginine, and cysteine. The third
theory is that the melanoidin skeleton is mainly built up of sugar
degradation products, formed in the early stages of the Maillard
reaction and polymerized through aldol-type condensation.
83–85
Analysis of the volatiles released upon the thermal degrada-
tion of the HMWM isolated from a coffee brew revealed the
presence of mostly furans (65%) followed by carbonyl
compounds (16%).
14
The protein-bound 1,4-bis-(5-amino-5-car-
boxy-1-pentyl)pyrazinium radical cation (CROSSPY) was
shown to be formed during roasting of coffee beans,
86
but also
during heating of aqueous solutions of bovine serum albumin
and glycolaldehyde used as model systems for the melanoidin
formation.
87
These reports suggest that the three theories
proposed for melanoidin formation may occur during coffee
melanoidin formation. Also, the identification of the CROSSPY
radical suggests that radical mechanisms may be involved in the
formation of coffee melanoidins.
Fig. 5 is a simplistic illustration of coffee melanoidin forma-
tion, since the exact mechanisms involved in their formation
remain unclear. This is an adaptation from the original produced
by Nunes and Coimbra,
6
aiming to include the possible occur-
rence of galactomannan polymerization, as was reviewed here. In
summary, polysaccharides, proteins, and chlorogenic acids
(phenolic compounds) are known to be involved in coffee mel-
anoidin formation. However, several questions remain open
about their structures as the nature of the unknown material that
can contribute up to 90% to the melanoidin weight, as well as the
type of linkages between polysaccharides, proteins, and chloro-
genic acids. Thus, the structural characterization of coffee mel-
anoidins remains a topic of great research interest in the near
future.
4. Potential health impacts of coffee melanoidins
As coffee brew is one of the main sources of melanoidins in the
human diet,
11
biological activities of coffee melanoidins and their
health implications are of great interest. As shown in Table 1 and
described below, different biological activities have been attrib-
uted to coffee melanoidins. However, their physiological rele-
vance is so far yet to be elucidated. First, studies on biological
activities of coffee melanoidins have often been developed using
the high molecular weight material (HMWM) isolated from
coffee brews without subsequent purification, simply denomi-
nated as melanoidins. This approach is limited in the sense that
the HMWM, as previously discussed, comprises other high
molecular weight compounds, hampering a definitive conclusion
about the active principle responsible for the biological activity.
On the other hand, little is known about the metabolic transit
and biotransformation of melanoidins (from coffee and other
foods), as was previously reviewed.
88,89
Regarding coffee, the
HMWM isolated from a coffee brew by ultrafiltration (MWCO
10 kDa) was digested in vitro by simulating gastrointestinal
enzymatic digestion. The low molecular weight fraction recov-
ered after digestion represented 14% of the HMWM, suggesting
that the coffee melanoidins are largely resistant to digestion in
the human gastrointestinal tract.
31
However, little else is known
about the metabolic transit and biotransformation of coffee
melanoidins. In fact, one question that remains unclear is
whether the biological activities attributed to coffee melanoidins
(mostly based on in vitro studies) have a significant impact on
human health. This question is particularly relevant for those
activities whose potential health effects rely on the presence of
melanoidins in the bloodstream or their transportation through
the blood to the organs, since melanoidins are high molecular
weight compounds by definition and their ability to cross the
intestinal epithelial barrier has not yet been reported.
The following sections are intended to present the studies
performed to date on biological activities of coffee melanoidins,
namely, their reported antioxidant activity and ability to inhibit
matrix metalloproteases, the antimicrobial activity and ability to
modulate the bacterial colon population, as well as the anti-
cariogenic, anti-inflammatory, antihypertensive, and anti-
glycative activities.
4.1. Antioxidant activity
Several studies have shown that coffee melanoidins (most refer-
ring to the HMWM isolated from coffee brews without subse-
quent purification) present in vitro antioxidant activity.
17,21,29–31
Based on the observation that low molecular weight compounds
released from the HMWM of coffee brews after overnight
incubation in 2 M NaCl showed higher antioxidant activity than
908 | Food Funct., 2012, 3, 903–915 This journal is ªThe Royal Society of Chemistry 2012
that of the remaining polymeric material, it has been suggested
that, although melanoidins present antioxidant activity, an
important contribution to the overall antioxidant activity of the
HMWM is given by the low molecular weight compounds linked
non-covalently to the melanoidin skeleton, such as chlorogenic
acids.
29–31
The antioxidant activity of melanoidins was reinforced
by the observation of in vitro antioxidant activity in melanoidin
fractions not containing chlorogenic acids, isolated from the
HMWM by gel permeation chromatography and characterized
by 2D NMR analysis.
17
The mechanism of the antioxidant action
of coffee melanoidins is still unclear. However, it has been
assumed that it is based on their radical scavenging activity and/
or their metal chelating capacity.
29,30
The antioxidant activity of coffee melanoidins has been eval-
uated by measuring their radical scavenging activity against
stable free radicals: 2,20-azinobis(3-ethylbenzothiazoline-6-
sulfonic acid) radical cation (ABTS_
+
), 1,1-diphenyl-2-picrylhy-
drazyl radical (DPPH_), and N,N-dymethyl-p-phenylenediamine
radical cation (DMPD_
+
). Among others, assays testing their
ability to prevent lipid peroxidation have also been used.
17,21,29–31
Borrelli et al.
21
showed a decrease of the radical scavenging
activity of melanoidins isolated by gel filtration chromatography
from coffee brews towards ABTS_
+
and DMPD_
+
with the
increasing of the degree of roast, but the ability to prevent
linoleic acid peroxidation was higher in the dark-roasted sample.
On the contrary, with respect to the results obtained by the
ABTS_
+
assay, Delgado-Andrade et al.
30
showed that the anti-
oxidant activity of melanoidins isolated by ultrafiltration from
instant coffees was lower in the light-roasted sample. As for the
melanoidin fractions isolated from coffee brews, different results
have also been reported on the effect of roasting on the antiox-
idant activity of whole coffee brews. For example, Richelle
et al.
90
reported that coffee antioxidant activity decreased with
roasting, whereas in other studies maximum antioxidant activity
was observed for the medium-roasted coffee.
91,92
The differences
in the assay procedures may contribute to the different results
reported by the different authors. However, the previously dis-
cussed differences in the composition of coffee melanoidin frac-
tions, in which the different melanoidin populations may exhibit
antioxidant capacities through different mechanisms, should also
be taken into account to explain the different results. Better
knowledge of melanoidin structures would benefit the under-
standing of their antioxidant properties.
In terms of possible implications on human health, the anti-
oxidant activity of coffee melanoidins has been associated with
protective effects against oxidative damages. The HMWM of
a coffee brew isolated by ultrafiltration and digested by simulating
gastrointestinal enzymatic digestion showed protective effects on
cultured human hepatoma HepG2 cells submitted to oxidative
stress, such as a decrease on the activities of antioxidant enzymes
Fig. 5 Illustration of coffee melanoidin formation (an adaptation of the formation originally produced by Nunes and Coimbra).
6
This journal is ªThe Royal Society of Chemistry 2012 Food Funct., 2012, 3, 903–915 | 909
Table 1 Synoptic table of the studies focused on biological activities of coffee brew melanoidins
Biological activity
HMWM
isolation
HMWM
fractionation
Assay(s) for
biological activity Findings Ref.
Antioxidant
activity
Gel filtration
chromatography
a
N/A
b
,
c
— ABTS, DMPD, DPPH,
inhibition of linoleic
acid peroxidation, and
redox potential
ABTS_
+
/DMPD_
+
scavenging ability
decreased with the
increasing of the
roasting degree. The
ability to prevent linoleic
acid peroxidation was
higher
in the dark-roasted sample
21
Ultrafiltration 10 kDa Ultrafiltration
(10 kDa) after
incubation in
2 M NaCl
DPPH, ABTS, FRAP,
and inhibition of
(AAPH)-induced
linoleic acid oxidation
>50% of the antioxidant
activity is due to low
molecular weight
compounds linked
non-covalently to the
melanoidin skeleton.
29,30
Ultrafiltration 10 kDa Ultrafiltration
(10 kDa) after
gastrointestinal
digestion and after
incubation in
2 M NaCl
DPPH, ABTS, FRAP,
ORAC, and HOSC
Low molecular
weight compounds
released after
gastrointestinal digestion
exerted the highest
antioxidant activity.
31
Dialysis 3.5 kDa Gel filtration
chromatography
Lipid peroxidation in
a rat liver microsome
biological system
HMWM exerted a
higher antioxidant activity
than that of low-MWM.
98
Inhibition of matrix
metalloproteases
Ultrafiltration 10 kDa — Inhibition of MMP-1,
MMP-2, and MMP-9
activities
IC
50
values for medium
roasting ranged
between 0.2–1.1 mg mL
1
of HMWM.
36
Antimicrobial
activity
Ultrafiltration 10 kDa — MIC values against
bacterial strains of
Gram-negative, Gram-
positive, and Gram-
negative that produce
siderophores
MIC values ranged
between 2–10 mg mL
1
of HMWM.
32,33
Modulation of the
bacterial colon
population
Stepwise
ultrafiltration
100, 50, 10,
and 3 kDa
Ultrafiltration
(1 kDa) after
incubation with
human fecal
bacteria
Microbial population
analysis by FISH and
DGGE
Growth of bacterial cells
belonging to the
Bacteroides-Prevotella
group
94
Anticariogenic
activity
Dialysis 3.5 kDa Gel filtration
chromatography
Streptococcus mutans
adhesion to and
detachment from saliva-
coated hydroxypatite
beads and biofilm
formation
HMWM (6 mg mL
1
)
exerted 91% of adherence
inhibition, 23% of
detached bacteria and
100% of biofilm
formation inhibition
117
Anti-inflammatory
activity
Dialysis 12–14 kDa — Analysis of inflammatory
markers in liver samples
from rats subjected to a
high-fat diet
Reduction of
proinflammatory
cytokines and increase
of anti-inflammatory
cytokines in
melanoidin-drinking rats
119
Antihypertensive
activity
Ultrafiltration 10 kDa Ultrafiltration
(10 kDa) after
overnight incubation
in 2 M NaCl
ACE-inhibitory activity HMWM (2 mg mL
1
)
exerted 37–45% of ACE
inhibition.
34
Antiglycative
activity
Ultrafiltration 10 kDa — Inhibition of bovine serum
albumin glycation with glucose
IC
50
¼274 mgmL
1
of
HMWM
35
a
Method used for HMWM isolation.
b
Molecular weight cut-off (of the membrane or fiber used).
c
N/A, Not applicable. Other abbreviations used:
ABTS, 2,20-azinobis(3-ethylbenzothiazoline-6-sulfonic acid); ACE, angiotensin-I converting enzyme; APPH, 2,20-azobis(2-amidinopropane)
dihydrochloride; DGGE, denaturing gradient gel electrophoresis; DMPD, N,N-dimethyl-p-phenylenediamine; DPPH, 1,1-diphenyl-2-picrylhydrazyl;
FISH, fluorescence in situ hybridization; FRAP, ferric reducing activity power; (H)MWM, (high) molecular weight material; HOSC, hydroxyl
radical scavenging capacity; IC
50
, 50% inhibitory concentration; MIC, minimum inhibitory concentration; MMP, matrix metalloprotease; ORAC,
oxygen radical absorbance capacity.
910 | Food Funct., 2012, 3, 903–915 This journal is ªThe Royal Society of Chemistry 2012
(glutathione peroxidase and glutathione reductase), whose
activities are increased when cells are submitted to oxidative
stress.
93
Also, the high molecular weight fractions of coffee brews
isolated by stepwise ultrafiltration and subjected to in vitro
fermentation for 24h with human fecal bacteria showed antioxi-
dant activity,
94
suggesting a possibly role of coffee melanoidins in
the protection against radical stress in the colon, which is associ-
ated with the development of colon cancer.
95
The HMWM from
instant coffee can also inhibit the formation of hydroperoxide free
radicals, and secondary lipoxidation products during simulated
gastric digestion of turkey meat.
96
The ability of melanoidins to
inhibit the lipoxidation can contribute to their health benefits,
since lipoxidation products are involved in the development of
atherosclerosis and other diseases.
97
The inhibition of lip-
oxidation by coffee melanoidins was also reported to occur in a rat
liver microsome system,
98
and model glucose–glycine melanoidins
also exerted a protective effect against lipoxidation in isolated rat
hepatocytes submitted to oxidative stress.
99
4.2. Inhibition of matrix metalloproteases
Matrix metalloproteases (MMPs) are a family of endo-peptidases
that are thought to play a central role in tumor growth and
metastasis.
100
The potential activity of coffee melanoidins as
MMP inhibitors was reported based on the ability of the
HMWM isolated from coffee brews to inhibit the in vitro activity
of selected human MMPs (MMP-1, MMP-2, and MMP-9). For
all MMPs (and roasting times >10 min), IC
50
values ranged
between 0.2 and 1.1 mg mL
1
. The HMWM from green coffee
brew showed no significant inhibitory activity against any of the
MMPs at concentrations up to 2.5 mg mL
1
. Also, the inhibitory
potential increased with the degree of roast.
36
Concerning possible effects on human health, the inhibitory
activity of coffee melanoidins may offer protective effects against
colon cancer, since MMP-1, MMP-2, and MMP-9 are thought to
be involved in the pathogenesis of colon cancer.
101
As estimated
by Fogliano and Morales,
11
the daily intake of coffee melanoi-
dins range between 0.5 to 2.0 g for moderate and heavy
consumers, respectively. This allows to estimate, based on the
assumption that the colon accumulates its content over at least
24 h in a maximum volume of 2 L,
36
a concentration of coffee
melanoidins in the order of 0.25 to 1 mg mL
1
in colon. These
values are comparable to the IC
50
values obtained with the
HMWM of (roasted) coffee brews, suggesting that the regular
intake of melanoidins from coffee brews could have a MMP
inhibitory activity and, consequently, be involved in the protec-
tion against colon cancer.
4.3. Antimicrobial activity
The antimicrobial activity of coffee melanoidin fractions (refer-
ring to the HMWM isolated by ultrafiltration from coffee brews)
was evaluated against bacterial strains of Gram-positive
(Staphylococcus aureus and Bacillus cereus) and Gram-negative
(Escherichia coli,Proteus mirabilis,Pseudomonas aeruginosa, and
Salmonella typhymurium), including those that produce high-
affinity iron chelating compounds (siderophores).
32,33
The anti-
microbial activity was tested as the minimum inhibitory
concentration (MIC), defined as the lowest concentration of
melanoidin fractions that did not produce any detected cell
growth.
32
For all strains studied, the MIC of coffee melanoidin
fractions ranged between 2 and 10 mg mL
1
.
32,33
Gram-positive
bacteria were more sensitive to the antimicrobial activity of
coffee melanoidin fractions, showing lower MIC values (2–3
mg mL
1
) than Gram-negative bacteria ($4mgmL
1
). As coffee
brews, depending on the method of preparation, have an esti-
mated melanoidin content ranging between 2 to 4 mg mL
1
,
11
it
can be inferred that all coffee brews present antibacterial activity
in the mouth against these Gram-negative bacteria and possibly
also against some Gram-positive. On the other hand, based on
the estimated melanoidin concentration of 0.25 to 1 mg mL
1
in
colon of moderate and heavy consumers (discussed in point 4.2),
it can be expected that the regular intake of coffee melanoidins
has an inhibitory/regulatory effect on colon microflora, but does
not prevent microbial growth.
Comparing instant coffees with different degrees of roast, it
was observed that the HMWM from the higher degree of roast
exerts a higher inhibitory bacterial growing activity, measured
against Geobacillus stearothermophylus. It was also observed, for
all instant coffees, that the non-covalently linked compounds
released from the HMWM after incubation in 2 M NaCl exert
a lower antibacterial activity than that of the remaining
HMWM.
102
In another study, it was showed that the antibacte-
rial activity against Escherichia coli of low and intermediate
molecular weight compounds isolated from a coffee brew by
sequential ultrafiltration steps exert a lower antibacterial activity
than that of the HMWM isolated in the first ultrafiltration
(referred as melanoidins).
32
The antibacterial activity reported
for the coffee melanoidins is in line with other studies demon-
strating the antibacterial activity of whole coffee brews, though
other compounds, such as caffeine and a-dicarbonyl compounds
formed during the roasting process, have been suggested to
contribute greatly for their antimicrobial activity against certain
bacterial strains, such as Salmonella enterica and Staphylococcus
aureus, respectively.
103–106
Maillard reaction products, obtained
using sugar-amino acid model systems in solution, were also
pointed out to possess antibacterial activity.
107–109
The high
molecular weight Maillard reaction products were more inhibi-
tory than the low molecular weight products when tested with
Bacillus subtilis,Escherichia coli, and Staphylococcus aureus.
107
The antimicrobial capacity of coffee melanoidins has been
ascribed to their metal chelating properties.
32,33
In particular,
three different mechanisms for the antibacterial activity of coffee
melanoidins were suggested by Rufi
an-Henares and de la
Cueva,
33
as follows: (1) at low concentrations, melanoidins may
exert a bacteriostatic activity mediated by iron chelation from the
culture medium; (2) in bacterial strains that are able to produce
siderophores for iron acquisition, melanoidins may chelate the
siderophore-Fe
3+
complex, which could decrease the virulence of
such pathogenic bacteria; and (3) coffee melanoidins may also
exert a bactericide activity at high concentrations by removing
Mg
2+
cations from the outer membrane, promoting the disrup-
tion of the cell membrane and allowing the release of intracellular
molecules. The mechanism behind the distinct sensitiveness of
Gram-negative and Gram-positive bacteria is not yet known
beyond the assumption that the absence in Gram-positive
microorganisms of the outer membrane makes them more
susceptible to antimicrobial substances.
33
This journal is ªThe Royal Society of Chemistry 2012 Food Funct., 2012, 3, 903–915 | 911
No reported studies on the antimicrobial activity of melanoi-
dins in yeasts or fungi are available.
4.4. Modulation of the bacterial colon population
In vitro fermentation of coffee brew high molecular weight
fractions (>100, 50–100, 10–50, and 3–10 kDa) with human fecal
bacteria showed that melanoidins are degraded or modified by
the human gut bacteria, as indicated by the decrease of the
absorbance at 405 nm of the solutions. For all coffee fractions, it
was observed an increase of Bacteroides-Prevotella,
94
as had been
observed in a previous study with an ethanol soluble high
molecular weight fraction characterized by 2D NMR.
28
Because
in these studies the melanoidins were probably present in
mixtures with polysaccharides, it cannot be disclosed if this effect
can be attributed only to the melanoidins. However, model
melanoidins, prepared from a mixture of glucose and amino acid
or protein, were shown to affect the growth of human gut
bacteria, their cell numbers were increased or decreased
depending on the bacteria and the time of thermal treatment or
incubation.
110,111
The capacity of coffee melanoidin fractions to modulate the
bacterial colon population is in accordance with a human
volunteer study demonstrating that the moderate consumption
of an instant coffee produces an increase in the number of Bac-
teroides-Prevotella bacteria detected in faeces.
112
However, in this
in vivo study, the main bacteria increase was observed for Bifi-
dobacterium spp., known for their probiotic effects.
113,114
The
selective modulation of the bacterial colon population by coffee
consumers observed by this in vivo study is in accordance with the
estimated amount of daily coffee intake discussed in point 4.3,
providing melanoidins in the colon in concentrations not far but
lower than their MIC, promoting the growth of the least
sensitive.
4.5. Anticariogenic activity
Streptococcus mutans is a bacterial species known to be involved
in the development of human dental caries. Its cariogenic
potential is in part related to its ability to adhere to the tooth
surface and form a biofilm.
115,116
The potential anticariogenic
activity of coffee melanoidins was reported based on the ability
of HMWM isolated from a coffee brew, and brown-colored
melanoidin fractions derived from the HMWM, to affect S.
mutans sucrose-dependent adhesion to and detachment from
saliva-coated hydroxyapatite beads and to inhibit biofilm
formation on microliter plates. The presence of HMWM at
a concentration of 6 mg mL
1
inhibited the S. mutans adhesion to
the beads in 91% and abolished the biofilm production. The
bacterial detachment from the beads after 2 h incubation was
three times higher with HMWM (23%) than with the controls
(7%). Based on these results and those obtained from the mela-
noidin fractions, it was suggested that the coffee melanoidins
may exert an anticariogenic activity.
117
This hypothesis was also
supported by a previous study demonstrating that the anti-
adhesive effect of coffee brews on the adhesive properties of
S. mutans are due to both naturally occurring and roasting-
induced molecules, the latter including melanoidins.
118
Also,
based on the estimated melanoidin concentration in the coffee
brews (2–4 mg mL
1
),
11
it is expected that coffee preparations
promote dental caries protection.
4.6. Anti-inflammatory activity
Several inflammatory markers were quantified in liver samples
from rats subjected to a high-fat diet for 3 months and to the
ingestion of different beverages from the beginning of the second
month. Rats drinking decaffeinated coffee or melanoidins (the
HMWM isolated from the decaffeinated coffee), compared with
control rats drinking water, showed reduced concentrations of
proinflammatory cytokines such as tumor necrosis factor
a(TNF-a) and interferon-g(IFN-g) and increase of anti-
inflammatory ones such as interleukin-4 (IL-4). These and other
results obtained in this study suggested that coffee melanoidins
may exert an anti-inflammatory activity, particularly in liver.
Concerning possible implications on human health, it was sug-
gested that the anti-inflammatory activity of coffee melanoidins
may play a role in counteracting the progression of liver diseases,
namely nonalcoholic steatohepatitis, a chronic inflammation
state in which radical oxygen species and several immunomod-
ulatory factor contribute to liver injury.
119
In agreement with this
study, other studies conducted in mice suggested the anti-
inflammatory action of coffee brews.
120,121
4.7. Antihypertensive activity
The potential antihypertensive activity of the HMWM isolated
by ultrafiltration from instant coffees was evaluated in vitro by
monitoring the angiotensin-I converting enzyme (ACE)-inhibi-
tory activity.
34
ACE (EC 3.4.15.1) is a circulating enzyme that
catalyzes the cleavage of a dipeptide from the C-terminal of the
decapeptide angiotensin I to form the potent vasopressor
angiotensin II. It also inactivates the vasodilator bradykinin by
sequential removal of two C-terminal dipeptides. ACE is a key
element of the renin-angiotensin system that regulates blood
pressure, and ACE inhibitors are important for the treatment of
hypertension.
122
The HMWM isolated from instant coffees
showed ACE-inhibitory activity (37–45% of ACE inhibition at
a concentration of 2 mg mL
1
), being this activity attributed to
the melanoidins. The high molecular weight fraction recovered
after overnight incubation of the HMWM in 2 M NaCl showed
much higher ACE-inhibitory activity (53–59%) than the low
molecular weight fraction (12–20%). Also, it was observed that
the ACE-inhibitory activity of HMWM fractions (containing the
melanoidins) increase with the degree of roast.
34
In line with this
study, another study showed that concentrations higher than 1.5
mg mL
1
of the HMWM isolated from coffee brews are neces-
sary to inhibit in vitro ACE activity by 50%.
36
The antihyper-
tensive activity of melanoidins was also suggested based on the
study of the antihypertensive activity of Maillard reaction
products obtained from several glucose-amino acid model
systems in solution.
108
Based on the estimated intake of coffee
melanoidins of 0.5 to 2.0 g per day,
11
their absorption and resi-
dence and dilution in the blood, it can be estimated that the
amount of melanoidins from coffee intake are far from reaching
the required concentration to have an effect on ACE inhibition.
The mechanism of action for potential ACE-inhibitory activity
of melanoidins is not known, but different mechanisms have
912 | Food Funct., 2012, 3, 903–915 This journal is ªThe Royal Society of Chemistry 2012
been suggested, as previously described by Rufi
an-Henares and
Morales.
34,108
Because ACE is a zinc-dependent enzyme,
122
the
inhibitory activity of melanoidins can come from their metal
chelating properties. On other hand, melanoidins can act as an
ACE non-competitive inhibitor, bind to the enzyme in an area
other than the active center, deform the enzyme, and hinder
binding to the substrate.
123
4.8. Antiglycative activity
Advanced glycation end products (AGEs) are considered
important mediators of diabetes complications. Thus, inhibitors
of glycation reactions are of great interest because of their
preventive or therapeutic potential.
124
The potential antiglycative
activity of coffee melanoidins was reported based on the ability
of the HMWM isolated from a coffee brew to inhibit in vitro
glycation of bovine serum albumin with glucose. The concen-
tration of the HMWM which is able to inhibit 50% of the gly-
cation (IC
50
) was 274 mgmL
1
. However, the accumulation of
Amadori products during glycation in the presence of the
HMWM was observed, suggesting that its antiglycative action is
paramount in the post-Amadori phase of the reaction. The
HMWM showed a lower antiglycative activity compared to the
low molecular weight fraction (IC
50
of 60 mgmL
1
).
35
Based on
the estimated intake of coffee melanoidins of 0.5 to 2.0 g per
day,
11
although dependent on the proportion of melanoidins
absorbed and present in the blood, it can be estimated that the
concentration of melanoidins in blood are in the range of the
effective dose for a possible antiglycative activity of coffee brew.
4.9. Other biological activities
The reaction of methanol-insoluble fractions isolated from
instant coffee (ascribed to melanoidin fractions) with nitrous acid
and thiocyanate was studied under acid conditions simulating
the mixture of coffee melanoidins, saliva, and gastric juice. Based
on the results obtained, it was suggested that coffee melanoidins
may react with salivary nitrite and thiocyanate in the gastric
lumen, producing nitric oxide (NO). It was also suggested that
the mechanism of reaction involves o-diphenol groups in mela-
noidins.
125
Concerning possible physiological effects, the
formation of NO in the stomach may contribute to the inhibition
of the microbial growth as well as the regulation of mucosal flow,
mucosal formation and gastric mobility.
125
Model melanoidins (>12 kDa) prepared from a glucose–
glycine mixture (dry-heated for 2 h at 125 C) exhibited modest
but significant genotoxic effects in human lymphocytes
cultures.
126
However, to the best of our knowledge, there is no
study on coffee melanoidins reporting their genotoxic effects in
human lymphocytes. This is an aspect that deserves future
attention.
5. Conclusions
Future studies are needed to better understand the structures of
coffee melanoidins, mechanisms of their formation, and their
potential health impacts, since there are still several unanswered
questions. With respect to their potential health impacts, most of
the studies undertaken to date on biological activities of coffee
melanoidins were developed using the high molecular weight
material isolated from coffee brews without subsequent purifi-
cation. However, it is known that different melanoidin pop-
ulations (considering their structural features) are present in
coffee brew. Thus, further studies are required to understand the
relationship between the chemical structure of the different
melanoidin populations and their biological activities. Also, it is
important to confirm whether the biological activities attributed
to coffee melanoidins based on in vitro studies are also observed
in in vivo studies.
Beyond the antioxidant activity provided by coffee brews due
to the melanoidins and other components, the studies published
to date suggest that the amount of melanoidins ingested during
regular coffee intake should provide protection against colon
cancer by inhibition of matrix metalloproteases, which prevent
bacterial growth in the mouth and the appearance of dental
caries, promote selective bacterial growth in the colon, and exert
anti-inflammatory and antiglycative effects.
Acknowledgements
The authors thank the financial support provided to Ana S. P.
Moreira (PhD grant, SFRH/BD/80553/2011), project PTDC/
QUI-QUI/100044/2008, QOPNA (project PEst-C/QUI/UI0062/
2011), and RNEM by the Portuguese Foundation for Science
and Technology (FCT).
References
1 A. Illy, E. Illy, R. Macrae, M. Petracco, M. R. Sondahl, S. Valussi
and R. Viani, Espresso coffee: the chemistry of quality, Academic
Press, London, 1995.
2 K. Kumazawa and H. Masuda, J. Agric. Food Chem., 2003, 51,
2674–2678.
3 S. Schenker, C. Heinemann, M. Huber, R. Pompizzi, R. Perren and
R. Escher, J. Food Sci., 2002, 67, 60–66.
4 S. Mussatto, E. Machado, S. Martins and J. Teixeira, Food
Bioprocess Technol., 2011, 4, 661–672.
5 J.-K. Moon and T. Shibamoto, J. Agric. Food Chem., 2009, 57,
5823–5831.
6 F. Nunes and M. A. Coimbra, Phytochem. Rev., 2010, 9, 171–185.
7 J. E. Hodge, J. Agric. Food Chem., 1953, 1, 928–943.
8 S. I. F. S. Martins, W. M. F. Jongen and M. A. J. S. van Boekel,
Trends Food Sci. Technol., 2000, 11, 364–373.
9 G. P. Rizzi, Food Rev. Int., 1997, 13, 1–28.
10 H.-Y. Wang, H. Qian and W.-R. Yao, Food Chem., 2011, 128, 573–
584.
11 V. Fogliano and F. J. Morales, Food Funct., 2011, 2, 117–123.
12 V. Faist, M. Lindenmeier, C. Geisler, H. F. Erbersdobler and
T. Hofmann, J. Agric. Food Chem., 2002, 50, 602–606.
13 T. D. Obretenov, S. D. Ivanova, M. J. Kuntcheva and G. T. Somov,
J. Agric. Food Chem., 1993, 41, 653–656.
14 A. Adams, R. C. Borrelli, V. Fogliano and N. De Kimpe, J. Agric.
Food Chem., 2005, 53, 4136–4142.
15 H.-D. Belitz, W. Grosch and P. Schieberle, in Food Chemistry, ed.
H.-D. Belitz, W. Grosch and P. Schieberle, Springer, Berlin, 4th
edn, 2009, ch. 21, pp. 938–951.
16 F. M. Nunes and M. A. Coimbra, J. Agric. Food Chem., 2007, 55,
3967–3977.
17 D. Gniechwitz, N. Reichardt, J. Ralph, M. Blaut, H. Steinhart and
M. Bunzel, J. Sci. Food Agric., 2008, 88, 2153–2160.
18 F. J. Morales and S. Jim
enez-P
erez, Eur. Food Res. Technol., 2004,
218, 515–520.
19 F. J. Morales, Food Chem., 2002, 76, 363–369.
20 E. K. Bekedam, M. J. Loots, H. A. Schols, M. A. J. S. Van Boekel
and G. Smit, J. Agric. Food Chem., 2008, 56, 7138–7145.
21 R. C. Borrelli, A. Visconti, C. Mennella, M. Anese and V. Fogliano,
J. Agric. Food Chem., 2002, 50, 6527–6533.
This journal is ªThe Royal Society of Chemistry 2012 Food Funct., 2012, 3, 903–915 | 913
22 E. K. Bekedam, H. A. Schols, M. A. J. S. van Boekel and G. Smit, J.
Agric. Food Chem., 2006, 54, 7658–7666.
23 E. K. Bekedam, M. P. F. C. De Laat, H. A. Schols, M. A. J. S. Van
Boekel and G. Smit, J. Agric. Food Chem., 2007, 55, 761–768.
24 E. K. Bekedam, E. Roos, H. A. Schols, M. A. J. S. Van Boekel and
G. Smit, J. Agric. Food Chem., 2008, 56, 4060–4067.
25 E. K. Bekedam, H. A. Schols, M. A. J. S. Van Boekel and G. Smit, J.
Agric. Food Chem., 2008, 56, 2055–2063.
26 T. Hofmann, M. Czerny, S. Calligaris and P. Schieberle, J. Agric.
Food Chem., 2001, 49, 2382–2386.
27 H. G. Maier, W. Diemair and J. Ganssmann, Z. Lebensm.-Unters.
Forsch., 1968, 137, 287–292.
28 D. Gniechwitz, N. Reichardt, E. Meiss, J. Ralph, H. Steinhart,
M. Blaut and M. Bunzel, J. Agric. Food Chem., 2008, 56, 5960–
5969.
29 C. Delgado-Andrade and F. J. Morales, J. Agric. Food Chem., 2005,
53, 1403–1407.
30 C. Delgado-Andrade, J. A. Rufi
an-Henares and F. J. Morales, J.
Agric. Food Chem., 2005, 53, 7832–7836.
31 J. A. Rufi
an-Henares and F. J. Morales, J. Agric. Food Chem., 2007,
55, 10016–10021.
32 J. A. Rufi
an-Henares and F. J. Morales, J. Agric. Food Chem., 2008,
56, 2357–2362.
33 J. A. Rufi
an-Henares and S. P. de la Cueva, J. Agric. Food Chem.,
2009, 57, 432–438.
34 J. A. Rufi
an-Henares and F. J. Morales, J. Agric. Food Chem., 2007,
55, 1480–1485.
35 E. Verzelloni, D. Tagliazucchi, D. Del Rio, L. Calani and A. Conte,
Food Chem., 2011, 124, 1430–1435.
36 L. M. de Marco, S. Fischer and T. Henle, J. Agric. Food Chem.,
2011, 59, 11417–11423.
37 F. M. Nunes and M. A. Coimbra, J. Agric. Food Chem., 2001, 49,
1773–1782.
38 F. M. Nunes and M. A. Coimbra, J. Agric. Food Chem., 2002, 50,
7046–7052.
39 F. M. Nunes and M. A. Coimbra, J. Agric. Food Chem., 2002, 50,
1429–1434.
40 F. J. Morales, C. Fern
andez-Fraguas and S. Jim
enez-P
erez, Food
Chem., 2005, 90, 821–827.
41 M. Takenaka, N. Sato, H. Asakawa, X. Wen, M. Murata and
S. Homma, Biosci., Biotechnol., Biochem., 2005, 69, 26–30.
42 R. J. Redgwell, D. Curti, M. Fischer, P. Nicolas and L. B. Fay,
Carbohydr. Res., 2002, 337, 239–253.
43 F. M. Nunes, A. Reis, A. M. S. Silva, M. R. M. Domingues and
M. A. Coimbra, Phytochemistry, 2008, 69, 1573–1585.
44 P. Capek, M. Matulov
a, L. Navarini and F. Suggi-Liverani,
Carbohydr. Polym., 2010, 80, 180–185.
45 M. Matulov
a, P. Capek, S. Kaneko, L. Navarini and F. S. Liverani,
Carbohydr. Res., 2011, 346, 1029–1036.
46 R. J. Redgwell, V. Trovato, D. Curti and M. Fischer, Carbohydr.
Res., 2002, 337, 421–431.
47 H. G. E. Maier and H. Buttle, Z. Lebensm.-Unters. Forsch., 1973,
150, 331–334.
48 R. Kl€
ocking, R. Hofmann and D. M€
ucke, Z. Lebensm.-Unters.
Forsch., 1971, 146, 79–83.
49 L. Heinrich and W. Baltes, Z. Lebensm.-Unters. Forsch., 1987, 185,
366–370.
50 A. G. W. Bradbury and D. J. Halliday, J. Agric. Food Chem., 1990,
38, 389–392.
51 F. M. Nunes, M. R. Domingues and M. A. Coimbra, Carbohydr.
Res., 2005, 340, 1689–1698.
52 A. Oosterveld, J. S. Harmsen, A. G. J. Voragen and H. A. Schols,
Carbohydr. Polym., 2003, 52, 285–296.
53 A. Oosterveld, A. G. J. Voragen and H. A. Schols, Carbohydr.
Polym., 2003, 54, 183–192.
54 A. S. P. Moreira, M. A. Coimbra, F. M. Nunes, J. Sim~
oes and
M. R. M. Domingues, J. Agric. Food Chem., 2011, 59, 10078–
10087.
55 F. M. Nunes, A. Reis, M. R. M. Domingues and M. A. Coimbra, J.
Agric. Food Chem., 2006, 54, 3428–3439.
56 R. J. Redgwell, C. Schmitt, M. Beaulieu and D. Curti, Food
Hydrocolloids, 2005, 19, 1005–1015.
57 R. Acu~
na, R. Bass€
uner, V. Beilinson, H. Cortina, G. Cadena-
G
omez, V. Montes and N. C. Nielsen, Physiol. Plant., 1999, 105,
122–131.
58 M. B. Coelho, M. L. g. R. Macedo, S. r. Marangoni, D. S. d. Silva,
I. Cesarino and P. Mazzafera, J. Agric. Food Chem., 2010, 58, 3050–
3055.
59 W. J. Rogers, G. B
ezard, A. Deshayes, I. Meyer, V. P
etiard and
P. Marraccini, Plant Physiol. Biochem., 1999, 37, 261–272.
60 S. Casal, E. Mendes, M. B. P. P. Oliveira and M. A. Ferreira, Food
Chem., 2005, 89, 333–340.
61 P. Montavon, A.-F. Mauron and E. Duruz, J. Agric. Food Chem.,
2003, 51, 2335–2343.
62 A. Farah and C. M. Donangelo, Braz. J. Plant Physiol., 2006, 18, 23–
36.
63 M. N. Clifford, J. Sci. Food Agric., 2000, 80, 1033–1043.
64 R. Jaiswal, M. A. Patras, P. J. Eravuchira and N. Kuhnert, J. Agric.
Food Chem., 2010, 58, 8722–8737.
65 J.-K. Moon, H. S. Yoo and T. Shibamoto, J. Agric. Food Chem.,
2009, 57, 5365–5369.
66 D. Perrone, A. Farah, C. M. Donangelo, T. de Paulis and
P. R. Martin, Food Chem., 2008, 106, 859–867.
67 A. Farah, T. de Paulis, L. C. Trugo and P. R. Martin, J. Agric. Food
Chem., 2005, 53, 1505–1513.
68 V. Leloup, A. Louvrier and R. Liardon, in Proceedings of the 16th
ASIC Colloquium (Kyoto), ASIC, Paris, 1995, pp. 192–198.
69 W. Fiddler, W. E. Parker, A. E. Wasserman and R. C. Doerr, J.
Agric. Food Chem., 1967, 15, 757–761.
70 R. K. Sharma, T. S. Fisher and M. R. Hajaligol, J. Anal. Appl.
Pyrolysis, 2002, 62, 281–296.
71 R. H. Stadler, D. H. Welti, A. A. St€
ampfli and L. B. Fay, J. Agric.
Food Chem., 1996, 44, 898–905.
72 O. Frank, S. Blumberg, C. Kunert, G. Zehentbauer and
T. Hofmann, J. Agric. Food Chem., 2007, 55, 1945–1954.
73 A. Farah, T. de Paulis, D. P. Moreira, L. C. Trugo and P. R. Martin,
J. Agric. Food Chem., 2006, 54, 374–381.
74 A. L. Waterhouse, in Current Protocols in Food Analytical
Chemistry, ed. R. E. Wrolstad, John Wiley & Sons, Inc., New
York, 2002, Supplement 6, pp. I1.1.1–I1.1.8.
75 H. M. Rawel, S. Rohn and J. Kroll, Deut. Lebensm.-Rundsch., 2005,
101, 148–160.
76 V. M. Totlani and D. G. Peterson, J. Agric. Food Chem., 2007, 55,
414–420.
77 F. Hayase, T. Usui and H. Watanabe, Mol. Nutr. Food Res., 2006,
50, 1171–1179.
78 R. Tressl, G. T. Wondrak, L.-A. Garbe, R.-P. Kr€
uger and
D. Rewicki, J. Agric. Food Chem., 1998, 46, 1765–1776.
79 R. Tressl, G. T. Wondrak, R.-P. Kr€
uger and D. Rewicki, J. Agric.
Food Chem., 1998, 46, 104–110.
80 T. Hofmann, Z. Lebensm.-Unters. -Forsch. A, 1998, 206, 251–258.
81 T. Hofmann, J. Agric. Food Chem., 1998, 46, 3891–3895.
82 T. Hofmann, J. Agric. Food Chem., 1998, 46, 3896–3901.
83 L. W. Kroh, T. Fiedler and J. Wagner, Ann. N. Y. Acad. Sci., 2008,
1126, 210–215.
84 B. C€
ammerer and L. W. Kroh, Food Chem., 1995, 53, 55–59.
85 B. C€
ammerer, W. Jalyschko and L. W. Kroh, J. Agric. Food Chem.,
2002, 50, 2083–2087.
86 T. Hofmann, W. Bors and K. Stettmaier, in Free Radicals in Food,
ed. F. S. Michael J. Morello, Chi-Tang Ho, American Chemical
Society, 2002, vol. 807, ch. 4, pp. 49–68.
87 T. Hofmann, W. Bors and K. Stettmaier, J. Agric. Food Chem., 1999,
47, 391–396.
88 V. Faist and H. F. Erbersdobler, Ann. Nutr. Metab., 2001, 45, 1–12.
89 V. Somoza, Mol. Nutr. Food Res., 2005, 49, 663–672.
90 M. Richelle, I. Tavazzi and E. Offord, J. Agric. Food Chem., 2001,
49, 3438–3442.
91 M. C. Nicoli, M. Anese, L. Manzocco and C. R. Lerici, LWT–Food
Sci. Technol., 1997, 30, 292–297.
92 M. D. del Castillo, J. M. Ames and M. H. Gordon, J. Agric. Food
Chem., 2002, 50, 3698–3703.
93 L. Goya, C. Delgado-Andrade, J. A. Rufi
an-Henares, L. Bravo and
F. J. Morales, Mol. Nutr. Food Res., 2007, 51, 536–545.
94 N. Reichardt, D. Gniechwitz, H. Steinhart, M. Bunzel and M. Blaut,
Mol. Nutr. Food Res., 2009, 53, 287–299.
95 T. A. Ullman and S. H. Itzkowitz, Gastroenterology, 2011, 140,
1807–1816.
96 D. Tagliazucchi, E. Verzelloni and A. Conte, J. Agric. Food Chem.,
2010, 58, 2513–2519.
97 U. Koji, Free Radical Biol. Med., 2000, 28, 1685–1696.
914 | Food Funct., 2012, 3, 903–915 This journal is ªThe Royal Society of Chemistry 2012
98 M. Daglia, A. Papetti, C. Aceti, B. Sordelli, C. Gregotti and
G. Gazzani, J. Agric. Food Chem., 2008, 56, 11653–11660.
99 V. Valls-Bell
es, M. C. Torres, P. Mu~
niz, L. Boix, M. L. Gonz
alez-
Sanjose and P. Codo~
ner-Franch, J. Sci. Food Agric., 2004, 84,
1701–1707.
100 M. A. Forget, R. R. Desrosiers and R. B
eliveau, Can. J. Physiol.
Pharmacol., 1999, 77, 465–480.
101 A. G. Mysliwiec and D. L. Ornstein, Clin. Colorectal Cancer, 2002, 1,
208–219.
102 J. A. Rufi
an-Henaresand F. J. Morales, FoodRes. Int.,2006,39, 33–39.
103 M. Daglia, M. T. Cuzzoni and C. Dacarro, J. Agric. Food Chem.,
1994, 42, 2270–2272.
104 A. A. P. Almeida, A. Farah, D. A. M. Silva, E. A. Nunan and
M. B. A. Gl
oria, J. Agric. Food Chem., 2006, 54, 8738–8743.
105 M. Daglia, A. Papetti, P. Grisoli, C. Aceti, V. Spini, C. Dacarro and
G. Gazzani, J. Agric. Food Chem., 2007, 55, 10208–10213.
106 M. Mart
ınez-Tom
e, A. M. Jim
enez-Monreal, L. Garc
ıa-Jim
enez,
L. Almela, L. Garc
ıa-Diz, M. Mariscal-Arcas and M. A. Murcia,
Eur. Food Res. Technol., 2011, 233, 497–505.
107 H. Einarsson, B. G. Snygg and C. Eriksson, J. Agric. Food Chem.,
1983, 31, 1043–1047.
108 J. A. Rufi
an-Henares and F. J. Morales, Food Res. Int., 2007, 40,
995–1002.
109 M. L. Stecchini, P. Giavedoni, I. Sarais and C. R. Lerici, Lett. Appl.
Microbiol., 1991, 13, 93–96.
110 C. Dell’Aquila, J. M. Ames, G. R. Gibson and A. G. Wynne, Eur.
Food Res. Technol., 2003, 217, 382–386.
111 J. M. Ames, A. Wynne, A. Hofmann, S. Plos and G. R. Gibson, Br.
J. Nutr., 1999, 82, 489–495.
112 M. Jaquet, I. Rochat, J. Moulin, C. Cavin and R. Bibiloni, Int. J.
Food Microbiol., 2009, 130, 117–121.
113 C. Picard, J. Fioramonti, A. Francois, T. Robinson, F. Neant and
C. Matuchansky, Aliment. Pharmacol. Ther., 2005, 22, 495–
512.
114 G. R. Gibson and X. Wang, J. Appl. Microbiol., 1994, 77, 412–
420.
115 W. F. Liljemark and C. Bloomquist, Crit. Rev. Oral Biol. Med., 1996,
7, 180–198.
116 D. M. Krol, Curr. Probl. Pediatr. Adolesc. Health Care, 2003, 33,
253–270.
117 M. Stauder, A. Papetti, D. Mascherpa, A. M. Schito, G. Gazzani,
C. Pruzzo and M. Daglia, J. Agric. Food Chem., 2010, 58, 11662–
11666.
118 M. Daglia, R. Tarsi, A. Papetti, P. Grisoli, C. Dacarro, C. Pruzzo
and G. Gazzani, J. Agric. Food Chem., 2002, 50, 1225–
1229.
119 P. Vitaglione, F. Morisco, G. Mazzone, D. C. Amoruso,
M. T. Ribecco, A. Romano, V. Fogliano, N. Caporaso and
G. D’Argenio, Hepatology, 2010, 52, 1652–1661.
120 I. Paur, T. R. Balstad and R. Blomhoff, Free Radical Biol. Med.,
2010, 48, 1218–1227.
121 Y. Fukushima, M. Kasuga, K. Nakao, I. Shimomura and
Y. Matsuzawa, J. Agric. Food Chem., 2009, 57, 11100–11105.
122 J. Riordan, GenomeBiology, 2003, 4, 225.1–225.5.
123 B. Hern
andez-Ledesma, P. J. Mart
ın-
Alvarez and E. Pueyo, J. Agric.
Food Chem., 2003, 51, 4175–4179.
124 C. Guillet, Nutr. Clin. Metab., 2010, 24, 109–114.
125 U. Takahama and S. Hirota, J. Agric. Food Chem., 2008, 56, 4736–
4744.
126 S. Gl€
osl, K.-H. Wagner, A. Draxler, M. Kaniak, S. Lichtenecker,
A. Sonnleitner, V. Somoza, H. Erbersdobler and I. Elmadfa, Food
Chem. Toxicol., 2004, 42, 1487–1495.
This journal is ªThe Royal Society of Chemistry 2012 Food Funct., 2012, 3, 903–915 | 915
