Betaines and related ammonium compounds in chestnut (Castanea sativa Mill.)

Article abstract of DOI:10.1016/j.foodchem.2015.10.070

Chestnut fruits, being poor of simple sugars and consisting mainly of fibers and starch, are among the constituents of Mediterranean diet. While numerous studies report on content of proteins and amino acids in chestnut, no one has appeared so far on betaines, an important class of nitrogen compounds ubiquitous in plants for their protective action in response to abiotic stress. In this study, we analyzed by HPLC-ESI-tandem mass spectrometry, in fruits and flours of varieties of chestnut cultivated in Italy, the composition of betaines and ammonium compounds intermediates of their biosynthesis. Besides the parent amino acids, the compounds quantified were choline, glycerophosphocholine, phosphocholine, glycine betaine, N-methylproline, proline betaine (stachydrine), β-alanine betaine, 4-guanidinobutyric acid, trigonelline, N,N,N-trimethyllysine. Interestingly, some uncommon derivatives of pipecolic acid, such as N-methylpipecolic acid, 4-hydroxypipecolic acid and 4-hydroxy-N-methylpipecolic acid were identified for the first time in chestnut samples and characterized by MSⁿ tandem mass spectrometry.

Full text of DOI:10.1016/j.foodchem.2015.10.070

Food Chemistry 196 (2016) 1301–1309
Contents lists available at ScienceDirect
Food Chemistry
journal homepage: www.elsevier.com/locate/foodchem
Betaines and related ammonium compounds in chestnut (Castanea sativa
Mill.)
a
a
a
b
Luigi Servillo a, , Alfonso Giovane , Rosario Casale , Maria Luisa Balestrieri , Domenico Cautela ,
⇑
Marina Paolucci c,d, Francesco Siano d, Maria Grazia Volpe d, Domenico Castaldo b,e,f
a Dipartimento di Biochimica, Biofisica e Patologia Generale, Seconda Università degli Studi di Napoli, Via L. De Crecchio 7, 80138 Napoli, Italy
b Stazione Sperimentale per le Industrie delle Essenze e dei derivati dagli Agrumi, Azienda Speciale della Camera di Commercio di Reggio Calabria, Via Generale Tommasini 2,
89127 Reggio Calabria, Italy
c Dipartimento di Scienze e Tecnologie, Università degli Studi del Sannio, Via Port’Arsa 11, 82100 Benevento, Italy
d Consiglio Nazionale delle Ricerche, Istituto di Scienze dell’Alimentazione, Via Roma 64, 83100 Avellino, Italy
e Ministero dello Sviluppo Economico, Via Molise 2, Roma, Italy
f Dipartimento di Ingegneria Industriale e ProdAL scarl, Università degli Studi di Salerno, Via Ponte Don Melillo 1, 84084 Fisciano, Salerno, Italy
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 18 May 2015
Received in revised form 9 September 2015
Accepted 16 October 2015
Available online 17 October 2015
Chestnut fruits, being poor of simple sugars and consisting mainly of fibers and starch, are among the con-
stituents of Mediterranean diet. While numerous studies report on content of proteins and amino acids in
chestnut, no one has appeared so far on betaines, an important class of nitrogen compounds ubiquitous in
plants for their protective action in response to abiotic stress. In this study, we analyzed by HPLC–ESI-
tandem mass spectrometry, in fruits and flours of varieties of chestnut cultivated in Italy, the composition
of betaines and ammonium compounds intermediates of their biosynthesis. Besides the parent amino
acids, the compounds quantified were choline, glycerophosphocholine, phosphocholine, glycine betaine,
N-methylproline, proline betaine (stachydrine), b-alanine betaine, 4-guanidinobutyric acid, trigonelline,
N,N,N-trimethyllysine. Interestingly, some uncommon derivatives of pipecolic acid, such as
N-methylpipecolic acid, 4-hydroxypipecolic acid and 4-hydroxy-N-methylpipecolic acid were identified
for the first time in chestnut samples and characterized by MSn tandem mass spectrometry.
Ó 2015 Elsevier Ltd. All rights reserved.
Keywords:
Chestnut
Mediterranean diet
Amino acids
Quaternary ammonium compounds
Betaines
b-Alanine betaine
Stachydrine
Glycine betaine
Pipecolic acid
4-Guanidinobutyric acid
N-Methylpipecolic acid
4-Hydroxypipecolic acid
4-Hydroxy-N-methylpipecolic acid
1. Introduction
and milled fruits, which is utilized for bread and cake preparations
or as a sauce thickener.
The fruits of chestnut are among the constituents of Mediter-
ranean diet, a nutritional model which utilizes foods rich of fibers
and carbohydrates with low glycemic index, able to reduce the risk
of cardiovascular diseases and maintain health status (Jenkins
et al., 2002). Chestnut belongs to the family Fagaceae, the same
as oak and beech, and genus Castanea which includes several spe-
cies. In Europe only one species, the Castanea sativa Mill., is
exploited, particularly in the Mediterranean basin, for fruit and
wood production. Fruits of chestnut are rarely eaten raw but, more
frequently, roasted or boiled. A flour can be prepared from dried
From composition point of view, chestnut fruit is described as a
product poor of simple sugars, mainly consisting of fibers and
starch (De Vasconcelos, Bennett, Rosa, & Cardoso, 2007, 2010), with
low glycemic index and free from gluten, a protein found in wheat
which can induce severe gut damages in people with celiac disease.
As for nitrogen constituents, literature data are mainly focused on
protein and free amino acid content (De Vasconcelos, Bennett,
Rosa, & Cardoso, 2009, 2010) but no study has appeared so far on
the occurrence and quantification of betaines in chestnut. Betaines
are quaternary ammonium compounds, ubiquitous in the vegetal
world, produced by specific biosynthetic pathways involving the
exhaustive nitrogen methylation of amino and imino acids. These
substances tend to accumulate in the cytoplasm and intercellular
⇑
Corresponding author.
E-mail address: luigi.servillo@unina2.it (L. Servillo).
http://dx.doi.org/10.1016/j.foodchem.2015.10.070
0308-8146/Ó 2015 Elsevier Ltd. All rights reserved.

1302
L. Servillo et al. / Food Chemistry 196 (2016) 1301–1309
fluids where they exert protective functions on the structures of
proteins, nucleic acids, and cell membranes (Kavi Kishor et al.,
2005; Street, Bolen, & Rose, 2006) in response to plant abiotic
stresses, such as reduced availability of water, high soil salinity,
hypoxia, cold, freezing. Plants express characteristic patterns and
levels of betaines according to their species. As an example, proline
betaine and 4-hydroxyprolinebetaine occur at high levels in citrus
plants (Servillo, Giovane, Balestrieri, Cautela, et al., 2011; Servillo,
Giovane, Balestrieri, Bata-Csere, et al., 2011), proline betaine and
pipecolic acid betaine in alfalfa (Wood et al., 1991) and glycine
betaine occurs at high levels in wheat, spinach and beet (Zeisel,
Mar, Howe, & Holden, 2003). Some betaines are of particular inter-
est in human diet for the maintenance of health status. Glycine
betaine has the biological role in humans of methyl donor for the
formation of methionine from homocysteine (Craig, 2004;
McKeever, Weir, Molloy, & Scott, 1991), thus counteracting accu-
mulation of homocysteine in plasma, which constitutes a risk fac-
tor for cardiovascular diseases (Shai et al., 2004). Also, choline, a
quaternary ammonium compound abundant in plants, from which,
by oxidation, glycine betaine derives, is an important nutrient
(Zeisel & Blusztajn, 1994) acting as biosynthetic precursor of phos-
pholipids and the neurotransmitter acetylcholine. More recently,
proline betaine (stachydrine), occurring at high levels in citrus fruit
juice, has attracted noticeable interest as a nutraceutical substance
and a biomarker for citrus juice consumption (Heinzmann et al.,
2010). Stachydrine ameliorates high-glucose induced endothelial
cell senescence and SIRT1 downregulation (Servillo et al., 2013),
attenuates norepinephrine-induced cardiomyocyte hypertrophy
(Zhang, Shan, et al., 2014), and interferes with the endoplasmic
reticulum stress mediated apoptosis, ultimately leading to sup-
pression of renal tubular epithelial cell apoptosis (Zhang, Lu,
et al., 2014). On the basis of these results, it seemed of interest
to investigate the content of betaines and some of their precursors
in various cultivars of chestnut by employing a specific and sensi-
tive mass spectrometric technique, which we successfully used to
study betaines in citrus plant genus (Servillo et al., 2012; Servillo,
Giovane, Balestrieri, Bata-Csere, et al., 2011). Chemical structures
of betaines and related biosynthetic intermediates examined in
this study, grouped on the basis of precursor amino acids, are
reported in Fig. 1.
2. Materials and methods
2.1. Reagents
Glycine betaine, N-methylproline, trans-4-hydroxyproline, cho-
line, phosphocholine chloride calcium salt tetrahydrate, L-
ophosphorylcholine, (3-carboxypropyl)trimethylammonium
chloride ( -aminobutyric acid betaine), N-methylnicotinic acid
(trigonelline), 4-guanidinobutyric acid, -N,N,N-trimethyllysine
a-glycer
c
e
hydrochloride, pipecolic acid, 1-methylpiperidine-2-carboxylic
acid hydrochloride, 1-methylpiperidine-3-carboxylic acid, and 1-
methylpiperidine-4-carboxylic acid hydrochloride were from
Sigma–Aldrich (Milan, Italy). N,N-dimethyl-L-proline (stachydrine)
and 4-hydroxy-L-prolinebetaine (betonicine) were purchased from
extrasynthese (Genay, France). Pipecolic acid betaine, b-alanine
betaine (also known as homobetaine or propiobetaine), N-methyl-
4-hydroxypipecolic acid and N,N-dimethyl-4-hydroxypipecolic
acid (4-hydroxypipecolic acid betaine) were synthesized as
described by Servillo et al. (2012), Boc-4-hydroxypiperidine-2-car
boxylic was purchased from Sigma–Aldrich and used to prepare
4-hydroxypipecolic acid by reaction with trifluoroacetic acid. The
Fig. 1. Chemical structures of substances analyzed in this study deriving from (1)
serine, (6) b-alanine, (8) proline, (11) hydroxyproline, (14) pipecolic acid, (21)
c
-
aminobutyric acid, (24) lysine and (26) nicotinic acid. (2) choline, (3) phosphocholine,
(4) glycerophosphorylcholine, (5) glycine betaine, (7) b-alanine betaine, (9)
N-methylproline, (10) N,N-dimethylproline, (12) N-methylhydroxyproline, (13) 4-
hydroxyproline betaine, (15) N-methylpipecolic acid, (16) pipecolic acid betaine, (17)
4-hydroxypipecolic acid, (18) 5-hydroxypipecolic acid, (19) 4-hydroxy-N-methylpi-
standard mixture of L-amino acids, containing Ala, Arg, Asp, Glu,
His, Iso, Leu, Lys, Met, Phe, Pro, Ser, Thr, Val and Cys at 2.5 mM con-
centration in 0.01 M HCl, was from Pierce company (Milan, Italy).
pecolic acid, (20) 4-hydroxy-N,N-dimethylpipecolic acid, (22)
c-aminobutyric acid
betaine, (23) 4-guanidinobutyric acid, (25) -N,N,N-trimethyllysine, (27) trigonelline.
e

L. Servillo et al. / Food Chemistry 196 (2016) 1301–1309
1303
AccQ Fluor reagent for amino acid analysis was from Waters (Milan,
Italy). Milli-Q water was used for all the preparations of solutions
and standards. The solution of formic acid 0.1% in water used for
the HPLC–ESI-MS/MS analyses was from Sigma–Aldrich (Milan,
Italy).
the respective calibration curve built with standard solutions.
HPLC–ESI-MS/MS analyses were performed with an Agilent LC-
MSD SL quadrupole ion trap, in positive multiple reaction monitor-
ing (MRM) mode using the following MS2 transitions: 104.1 ? 60
for choline, 118.1 ? 59 for glycine betaine, 132.1 ? 73 for b-
alanine betaine, 146.1 ? 87 for
c-aminobutyric acid, 144.1 ? 84
2.2. Fresh fruit and chestnut flour samples
for N,N-dimethylproline (stachydrine), 130.1 ? 84 for N-
methylproline, 132.1 ? 86 for 4-hydroxyproline, 146.1 ? 100 for
4-hydroxy-N-methylproline, 130.1 ? 84 for pipecolic acid,
146.1 ? 100 for 4-hydroxypipecolic acid, 160.1 ? 114 for N-
methy-4-hydroxypipecolic acid, 174.1 ? 156 for N,N-dimethyl-4-
hydroxypipecolic acid, 144.1 ? 98 for 1,2-N-methylpipecolic acid,
158.1 ? 72 for pipecolic acid betaine (homostachydrine),
Fourteen different batches of about 6 kg of fresh chestnut fruits,
all of Italian origin, were sampled in three different Italian regions
(Calabria, Campania and Piedmont). All analytical determinations
(total proteins, percent of dry matter, content of free amino acids,
betaines and related compounds) were conducted on samples from
the three regions obtained using 40 ground fruits, deprived of peri-
carp and endocarp, taken from individual batches. As for flours, 4
lots made with chestnuts of Italian origin, provided by three differ-
ent Italian companies, were analyzed. All analyzes were conducted
in triplicate.
146.1 ? 87 for 4-guanidinobutyric acid, 189.1 ? 130 for e-N,N,N-
trimethyllysine, and 138.1 ? 110 for trigonelline. The mass
spectrometer was operated utilizing nitrogen as the nebulizing
and drying gas. The instrumental conditions were as follows: neb-
ulizer pressure, 30 psi; drying temperature, 350 °C; drying gas
7 l/min. The ion charge control (ICC) was applied with target set
at 30,000 and maximum accumulation time at 20 ms. Each extract
was analyzed in triplicate and the mean concentration of each com-
pound was calculated and expressed in mg/kg of dry product. The
concentrations of each compound were determined by comparison
with the relative calibration curve. Standard stock solutions of each
analyte were prepared at 2 mg/L. Additional calibration levels (0.2;
0.1; 0.05, 0.02, 0.002 and 0.001 mg/L) were prepared by serial dilu-
tion with water containing 0.1% formic acid. The calibration curves
were built using these standard solutions. The linear regression
analysis was carried out by plotting the peak areas of the monitored
fragment ions versus the concentrations of the analyte standard
solutions. The linearity of the instrumental response was assessed
by correlation coefficients (r2) >0.99 for all analytes.
2.3. Total proteins and dry matter
Total proteins were determined according to the Kjeldahl pro-
cedure using a selenium catalyst. The determination of dry matter
(total solids) was performed by oven drying the samples under
vacuum at 105 °C until constant weight was obtained.
2.4. Amino acid sample preparation and HPLC analysis
The amino acids were extracted from fresh chestnut fruits or
chestnut flours using 0.1% formic acid in water in the ratio (w/w)
1:10, under stirring for 1 h. The extract clarified at 18,000g for
60 min was filtered first through 5 lm and then 0.45 lm Millipore
filters, and finally 1 mL was applied on a cation-exchange column
(100 ꢀ 6 mm AG50WX8-H+ Bio Rad). After column washing with
50 mL of Milli-Q water, amino acids were eluted with 3.0 M NH3
(about 10 mL) and collected. The eluate was evaporated to dryness,
2.6. Statistical analysis
The mean values of the concentrations of analytes and their
standard deviations were calculated from the results of analyses
carried out in triplicate.
recovered with 0.1 M HCl (10 mL), and filtered through 0.45 lm fil-
ter. Quantification of free amino acids was performed by reverse
phase (RP) HPLC. A Waters chromatograph model 2690 equipped
with fluorescence detector model 474 was employed. Samples
3. Results and discussion
(10 lL) were derivatized with the Waters ACCQFLuor reagent
according to the method of Van Wandelen and Cohen (1997).
Quantification was made using the peak area of the fluorescence
emission intensity by excitation at 350 nm and recording fluores-
cence emission at 395 nm. Amino acids were identified on the
basis of their retention times and quantified by comparison of
the corresponding peak area with the respective calibration curve.
3.1. Quantification of betaines by LC–ESI-MS/MS
HPLC analysis performed on aqueous extracts of chestnut fruits
and flours allowed determination of betaines and related com-
pounds in about 30 min by using a 25 cm C8 column in isocratic
condition with 0.1% formic acid in water as eluent, at flow rate of
100 lL/min. Most of the betaines and other metabolites investi-
2.5. Analysis of betaine by HPLC–ESI-MS/MS
gated in this study eluted well separated in such chromatographic
conditions. For the sake of clarity, results of our study are reported
by grouping the quaternary ammonium compounds and their
biosynthetic intermediates, when present, on the basis of the pre-
cursor amino acid. The levels of the compounds examined in this
study, expressed as mean values and standard deviations of results
on samples of the same type, that is fruits or flours, are reported in
Table 1.
Betaines and related compounds were extracted from the fresh
chestnut fruits or flours using 0.1% formic acid in water in the ratio
(w/w) 1:5 or 1:10 according to Servillo, Giovane, Balestrieri,
Cautela, et al. (2011). The suspension was kept under stirring for
about 3 h and finally centrifuged at 18,000g for 60 min. The clarified
extract was filtered first through 5 lm and then 0.45 lm Millipore
filters and stored frozen until used. The analyses were performed by
HPLC–ESI-MS/MS according to Servillo, Giovane, Balestrieri, Bata-
Csere, et al., 2011 with an Agilent 1100 series liquid chromatograph
using a Supelco Discovery C8 column, 250 ꢀ 3.0 mm, particle size
3.2. Serine derivatives
5
l
m. The chromatography was conducted isocratically with 0.1%
Serine is the amino acid precursor of ethanolamine from which
choline is formed by methylation. Glycine betaine is biosynthesized
from choline through two oxidative steps. In the first, choline is oxi-
dized into betaine aldehyde from which, by further oxidation, gly-
cine betaine is finally generated. Glycine betaine is almost
ubiquitous in vegetal species. In some foods such as spinach, beet,
formic acid in water at flow rate of 100 lL/min. Volumes of 20 lL
of standard solution or sample were injected. Betaines were identi-
fied on the basis of their retention times and MS2 fragmentation
patters. Quantification of each betaine was generally obtained by
comparison of the peak area of its most intense MS2 fragment with

1304
L. Servillo et al. / Food Chemistry 196 (2016) 1301–1309
Table 1
Concentrations of the free amino acids precursors of betaines, total proteins, betaines and related ammonium compounds in Italian chestnut products. Concentrations are
expressed as mg/kg of dry matter. NQ = not quantified. Total protein levels are expressed as g/100 g of dry matter. Numbers in brackets cross-reference to Fig. 1.
Compounds
Fresh chestnuts
Min
Chestnut flours
Min
Aminoacids precursors of betaines
Max
Mean Std dev
Max
Mean Std dev
Serine (1)
Proline (8)
Hydroxyproline (11)
Pipecolic acid (14)
270
210
1.8
780
730
8.9
458 147
470 120
6.7 1.6
250
370
2.0
2
350
520
10.1
6
300 46
420 65
7.7 1.9
0.2
5.0
1.6 1.2
4
1.1
c
-Aminobutyric acid (21)
810
110
2.8
1550
810
5.8
1100 180
280 180
4.5 0.5
820
170
4.5
1200
330
7.0
940 140
220 70
5.5 0.9
Lysine (24)
Total Proteins
Betaines and ammonium compounds
Choline (2)
Glycine betaine (5)
b-Alanine betaine (7)
Phosphocholine (3)
Glycerophosphocholine (4)
N,N-Dimethylproline (10)
N-Methylproline (9)
4-Hydroxypipecolic acid (17)
N-Methylpipecolic acid (15)
4-Hydroxy-N-methylpipecolic acid (19)
17.0
0.8
0.5
5
33.0
2.1
2.7
12
28
1.5 0.4
1.6 1.0
8
7
0.5 0.3
0.4 0.2
NQ
0.3 0.1
NQ
0.4 0.1
0.7 0.1
0.5 0.1
6
42.0
2.0
1.1
3
55.0
5.5
5.0
15
46 3.5
3.3 1.4
3.0 1.1
1.0
2.0
8
6
1.0
2.0
4
10
2
14
0.1
0.2
NQ
0.1
NQ
0.2
0.4
0.3
0.7
0.9
NQ
0.6
NQ
0.6
1.0
0.8
0.8
0.8
NQ
0.4
NQ
0.7
0.6
1.3
1.4
1.3
NQ
1.0
NQ
1.7
1.1
1.8
1.1 0.3
1.2 0.1
NQ
0.5 0.1
NQ
1.3 0.4
0.8 0.1
1.7 0.1
c
-Guanidinobutyric acid (23)
-N,N,N-Trimethyllysine (25)
Trigonelline (27)
e
Fig. 2. Representative MS2 extracted ion chromatograms conducted in MRM mode. HPLC analysis was performed by using a 25 cm C8 column in isocratic condition with 0.1%
formic acid in water as eluent, at flow rate of 100
L/min. (a) Chestnut fruit extract. (b) Standard mixture of the analyzed compounds. The MS2 transitions followed were
104.1 ? 60 for choline (2), 189.1 ? 130 for -N,N,N-trimethyllysine (25), 132.1 ? 73 for b-alaninebetaine (7), 184.1 ? 86 for phosphocholine (3), 118.1 ? 59 for glycine
betaine (5) and 258.1 ? 104 for glycerophosphorylcholine (4). Retention times in minutes were 13.0 (2), 13.1 (25), 13.8 (7), 15.2 (3), 15.4 (5) and 15.8 for (4).
l
e
and wheat, the substance is present at high levels (De Zwart et al.,
2003; Slow et al., 2005; Zeisel et al., 2003). In our chromatographic
conditions, glycine betaine eluted at the retention time of 15.4 min
and was quantified by using the MS2 strongest precursor to product
ion transition 118.1 ? 59, while 118.1 ? 58 was used as the confir-
matory transition. A representative HPLC–ESI-MS2 chromatogram
of a chestnut fruit extract, where mass spectrometric data were
collected in multiple reaction monitoring (MRM) mode, is shown
in Fig. 2a. Besides, glycine betaine (5), choline (2), -N,N,N-
trimethyllysine (25), b-Alanine betaine (7), phosphocholine (3)
e
and glycerophosphorylcholine (4) are well represented in the
chestnut extracts (Fig. 2a).
e-N,N,N-Trimethyllysine and choline
coeluted in the chromatographic conditions used, instead
other betaines appeared well resolved in the chromatogram.

L. Servillo et al. / Food Chemistry 196 (2016) 1301–1309
1305
Fig. 3. Chromatographic separation of the three positional isomers of N-methylpiperidinecarboxylic acid. The mixture of the three isomers was subjected to HPLC analysis
performed as described in the legend of Fig. 2. (Panel a) MS2 (left) and MS3 (right) fragmentation patterns of 1,4-N-methylpiperidinecarboxylic acid, isolating at m/z 144.1
(MS2) and at m/z 98 (MS3). (Panel b) MS2 (up) and MS3 (down) fragmentation patterns of 1,3-N-methylpiperidinecarboxylic acid, isolating at m/z 144.1 (MS2) and at m/z 98
(MS3). (Panel c) MS2 (up) and MS3 (down) fragmentation patterns of 1,2-N-methylpiperidinecarboxylic acid, alias 1,2-N-methylpipecolic acid, isolating at m/z 144.1 (MS2) and
at m/z 98 (MS3).
HPLC–ESI-MS/MS analysis of a standard solution of the same
metabolites is shown in Fig. 2b. The average glycine betaine levels
were 1.5 mg/kg of dry matter for fresh chestnut fruits and 3.3 mg/
kg of dry matter for chestnut flours (Table 1). In contrast, choline
resulted to be the most abundant quaternary ammonium com-
pound in the metabolic profile of chestnut (Table 1). Choline eluted
at the retention time of 13.0 min and was quantified through the
MS2 transition 104.1 ? 60. In fresh chestnut fruits, the choline con-
tent showed a rather wide variability, ranging from 17 to 33 mg/kg
of dry matter (Table 1), while, in chestnut flours, this interval was
narrower, in the range 42-55 mg/kg of dry matter (Table 1). Phos-
phocholine, the first intermediate in the biosynthetic route leading
from choline to lecithin, eluted at retention time of 15.2 min. Its
quantification was achieved by using the MS2 strongest transition
184.1 ? 86, while 184.1 ? 125 was used as the confirmatory tran-
sition. Phosphocholine levels were in the range of 5–12 mg/kg of
dry matter for fresh chestnut fruits and 3–15 mg/kg of dry matter
for flours. Glycerophosphorylcholine (4), like phosphocholine (3),
was present in all samples of chestnut although at levels slightly
lower than phosphocholine. Its quantification was achieved by
using the MS2 transition 258.1 ? 104. Since in our instrumental
conditions only the MS2 fragment at m/z 104 (corresponding to cho-
line) was produced, the presence of glycerophosphorylcholine (4)
was confirmed in MS3 by isolating the MS2 fragment at m/z 104,
from which only the MS3 fragment at m/z 60 was formed, as
expected for choline (data not shown). Glycerophosphorylcholine

1306
L. Servillo et al. / Food Chemistry 196 (2016) 1301–1309
only through methylation steps of b-alanine (Rathinasabapathi
et al., 2000).
3.4. Proline derivatives
Low but significant levels of stachydrine (proline betaine or N,
N-dimethylproline) were found in all chestnut samples examined
(Table 1). Stachydrine (10) occurs at high levels in alfalfa and citrus
genus plants where also its hydroxylated derivative, betonicine
(4-hydroxy-N,N-dimethylproline), occurs. Stachydrine is believed
to derive from proline methylation through the intermediate N-
methylproline, while it is unclear if betonicine (13) is formed
through stachydrine hydroxylation or by successive methylation
of hydroxyproline. We examined the presence in the chestnut sam-
ples of all the possible precursors of stachydrine and betonicine.
Proline was found at rather high levels, with mean values of
420 mg/kg of dry matter for flour and 470 mg/kg of dry matter
for fresh fruits (Table 1). Also 4-hydroxyproline was detected in
all samples but at much lower levels than proline, in the range of
2–10 mg/kg of dry matter (Table 1). Interestingly, N-
methylproline (9), the immediate precursor of stachydrine, was
found in all samples (Table 1). Instead, betonicine (4-hydroxy-N,
N-dimethylproline) was always absent and, although a compound
eluting at r.t. 20 min with m/z 146 was detected in all samples,
mass spectrometric analyses ruled out that it was 4-hydroxy-N-
methylproline (12) (see below).
3.5. Pipecolic acid derivatives
In plants in which stachydrine is present, frequently also
homostachydrine (pipecolic acid betaine) occurs (Servillo et al.,
2012; Wood, 1991). Homostachydrine probably derives from N-
methylation of pipecolic acid. We examined the presence of the
components of the entire pathway leading from pipecolic acid to
its betaine. Homostachydrine was absent in all chestnut samples.
However, both pipecolic acid and a substance showing the same
m/z 144 as stachydrine but eluting at higher retention time
(20 min) were detected. This unknown compound might be N-
methylpipecolic acid (alias 1-methylpiperidine-2-carboxylic acid),
supposed the homostachydrine precursor (Essery, McCaldin, &
Marion, 1962; Wood, 1991). Little information is available in the
literature on the occurrence of N-methylpipecolic acid in plants.
As far as is known, Erythrochiton brasiliensis is the unique vegetal
species in which it was identified (Sargenti et al., 1993). Consider-
ing the absence of homostachydrine in all chestnut samples, in
order to avoid false assignment for that unknown substance, we
compared the three possible isomers of N-methylpipecolic acid
by their chromatographic and mass spectrometric features to
assess if one of them could be the unknown compound. The mix-
ture of the three isomers was completely resolved in our chro-
matographic conditions (Fig. 3). The 1,2 isomer was that eluting
later, with a retention time of 19.8 min, virtually coincident to
the retention time of the unknown compound. As for MS2 fragmen-
tation patterns, both the positional isomers 1,3 and 1,4 showed the
most intense fragment at m/z 126, likely due to the neutral loss of
H2O from the carboxyl group of the parent ion, and a less intense
fragment at m/z 98, which correspond to the loss of CO + H2O from
the parent ion (Fig. 3a and b). Instead, the fragmentation pattern of
1,2 isomer showed only one intense MS2 fragment at m/z 98
(Fig. 3c). This is not surprising as the easier formation of the highly
stable immonium ion as a consequence of the carboxyl loss favors
mostly its loss from position 2, closer to piperidine nitrogen atom,
than from the other two positions. The unknown compound also
showed the MS2 fragmentation pattern with only one intense frag-
ment at m/z 98. Furthermore, the MS3 fragmentation patterns of
the unknown compound and 1,2 isomer, obtained by isolating
Fig. 4. Comparison of MS2 fragmentation patterns of 4-hydroxypipecolic acid, 5-
hydroxypipecolic acid, and 4-hydroxy-N-methylproline. (a) MS2 fragmentation
pattern of 4-hydroxy-N-methylproline isolating at m/z 146.1. (b) MS2 fragmenta-
tion pattern of 4-hydroxypipecolic acid isolating at m/z 146.1. (c) MS2 fragmenta-
tion pattern of 5-hydroxypipecolic acid isolating at m/z 146.1. Note that in the
fragmentation pattern of 5-hydroxypipecolic acid the most intense fragment is at
m/z 128, while in that of 4-hydroxypipecolic acid it is at m/z 100. Instead, fragments
below m/z 82 are not present in the MS2 fragmentation pattern of 4-hydroxy-N-
methylproline.
levels resulted in the range of 4–10 mg/kg of dry matter in chestnut
fruits and in the range of 2–14 mg/kg of dry matter for flours.
3.3. b-Alanine derivatives
b-Alanine is straightforwardly produced in prokaryotes by
decarboxylation of aspartic acid. Instead, in plants, it is biosynthe-
sized by rather complex routes starting from the polyamines sper-
midine and spermine (Terano & Suzuki, 1978) or from propanoate
(Rathinasabapathi, 2002). This non-proteinogenic amino acid is the
direct precursor of b-alanine betaine by successive methylation
steps. We found that b-alanine betaine, rather uncommon in
plants, was present in all chestnut samples examined. Results
showed that the content of b-alanine betaine was in the range of
0.5–2.7 mg/kg of dry matter for both fruits and flours (Table 1).
b-Alanine betaine eluted at the retention time of 13.8 min
(Fig. 2b). Its quantification was achieved by using the strongest
MS2 transition 132.1 ? 73, while 132.1 ? 60 was used as the con-
firmatory transition. The occurrence in chestnut of b-alanine
betaine is rather surprising as it was reported so far in stress toler-
ant herbaceous plants such those of the Plumbaginaceae family
(Hanson et al., 1994). b-Alanine betaine is believed to be a more
suitable osmoprotectant than glycine betaine under hypoxic condi-
tions (Hanson et al., 1994; Rathinasabapathi, Sigua, Ho, & Gage,
2000) because the biosynthesis of b-alanine betaine, differently
from glycine betaine, does not require oxidative steps but occurs

L. Servillo et al. / Food Chemistry 196 (2016) 1301–1309
1307
Fig. 5. Chromatographic separation of 4-hydroxypipecolic acid, 4-hydroxy-N-methylpipecolic acid, and 4-hydroxy-N,N-dimethylpipecolic acid. The mixture of the three
compounds was subjected to HPLC analysis performed as described in the legend of Fig. 2. (A) MS2 extracted ion chromatogram performed in MRM mode. The MS2 transitions
followed were 146.1 ? 100 (peak 17), 160.1 ? 114 (peak 19) and 174.1 ? 156 (peak 20). (B) MS2 fragmentation patterns of 4-hydroxy-N-methylpipecolic acid. (C) MS2
fragmentation patterns of 4-hydroxy-N,N-dimethylpipecolic acid.
the MS2 fragment at m/z 98, resulted identical (Fig. 3c). All
together, these data leave no doubt that the unknown compound
is the 1,2 isomer, that is, N-methylpipecolic acid. Quantitative anal-
ysis on the chestnut samples revealed that its levels were in the
range of 0.1–1 mg/kg (Table 1).
In the course of the chromatographic analyses of chestnut
extracts, also two somewhat intense peaks at m/z 146 were
observed. The one eluting at r.t. 15.9 min was identified as 4-
guanidinobutyric acid (see below), the other, eluting at r.t.
14.3 min, showed a MS2 fragmentation pattern that at first sight
dates are a rich source (Ali, Alhaj, Al-Khalifa, & Bruckner, 2014).
In order to have reference standards, we prepared
4-hydroxypipecolic acid from the hydrolysis of its t-BOC deriva-
tive, which was commercially available, and aqueous extract from
ripe dates to obtain 5-hydroxypipecolic. In our chromatographic
conditions, the two substances were badly resolved, being 4-
hydroxypipecolic eluted at r.t. 13.6 min and 5-hydroxypipecolic
at r.t. 14.0 min. However, their MS2 fragmentation spectra were
different (Fig. 4b and c) and, most importantly, the chestnut
unknown substance perfectly coeluted with 4-hydroxypipecolic
acid and showed identical MS2 fragmentation pattern (data not
shown).
Interestingly, one more derivative of pipecolic acid was found in
all chestnut samples. Indeed, a substance eluting at r.t. 14.4 min
showed m/z 160 and a MS2 fragmentation pattern which suggested
that it was 4-hydroxy-N-methylpipecolic acid. Actually, after hav-
ing synthesized the substance by methylation of 4-
hydroxypipecolic acid (Fig. 5), we found that it coeluted with the
unknown peak and showed identical MS2 fragmentation pattern
(data not shown). It is worth to note that 4-hydroxy-N-
methylpipecolic acid is the likely precursor of 4-hydroxy-N,N-
dimethylpipecolic acid, an uncommon betaine reported to occur
in Lamium species (Yuan, Patel, Blunden, & Turner, 1992). We
sought to see if this betaine might occur also in chestnut, but our
suggested it could be 4-hydroxy-N-methylproline,
a proline
derivative occurring in various plant species (Blunden, Patel,
Adrian-Romero, & Melendez, 2004). Indeed, this possibility was
discarded when, after having synthesized the substance, we found
that it had a retention time slightly higher than the unknown com-
pound and different MS2 fragmentation pattern. In particular, the
unknown compound showed two MS2 fragments at m/z 55
and 74, which are absent in the fragmentation spectrum of 4-
hydroxy-N-methylproline (Fig. 4a).
Therefore, we sought to see if the unknown compound was a
hydroxylated derivative of pipecolic acid. In plants, only two of
such derivatives are reported to occur, they are 4-
hydroxypipecolic acid, particularly abundant in Acacia species
(Clark-Lewis & Mortimer, 1959) and 5-hydroxypipecolic of which

1308
L. Servillo et al. / Food Chemistry 196 (2016) 1301–1309
effort was unsuccessful. Indeed, we did not find any compound in
the chestnut extracts at m/z 174 coeluting with the synthesized
trans-4-hydroxypipecolic acid betaine. In summary, we have
observed that of the three pathways leading to the formations
of the three betaines N,N-dimethylproline (stachydrine),
N,N-dimethylpipecolic acid betaine (homostachydrine), and 4-
hydroxy-N,N-dimethylpipecolic acid betaine, only the first path-
way is complete in chestnut, whereas the other two end at level
of the N-methylpipecolic acid and 4-hydroxy-N-methylpipecolic
acid, respectively.
3.8. Nicotinic acid derivatives
Trigonelline (27), the betaine of nicotinic acid (26), is commonly
found in cotyledons of seeds and has an important role as hormone
which controls the cell cycle in plants by promoting preferential
cell arrest in G2 phase (Tramontano, Hartnett, Lynn, & Evans,
1982). Recently, it has been reported that trigonelline possesses
hypoglycemic properties and could assist in maintaining glycemic
control in diabetes mellitus (Adams et al., 2014). In our chromato-
graphic conditions, trigonelline eluted at retention time of
16.4 min and its quantification was achieved by using 138.1 ?
110, which was the MS2 strongest precursor to product ion transi-
tion, while 138.1 ? 94 was used as the confirmatory transition.
The average trigonelline levels resulted both for fruits and flours
lower than 2 mg/kg of dry matter (Table 1).
3.6.
c
-Aminobutyric acid derivatives
c
-Aminobutyric acid (21) is present at high levels in chestnut
(Table 1), for this reason we looked for the possible occurrence also
of its betaine that might be produced by direct methylation as it
happens for the formation of b-alanine betaine from b-alanine.
Actually, it appeared reasonable to suppose that the same enzyme
which methylates b-alanine could do the same also with
Conflict of interest statement
The authors declare no conflicts of interest.
Acknowledgements
c-aminobutyric acid, which is the next higher homolog of
b-alanine. Indeed, we did not detect the substance in any chestnut
sample examined. However, it is interesting to note that another
This research was supported by PSR Campania Region
2007/2013 (Regulation (CE) n. 1698/2005), Tool 124, ‘‘VAL.I.CO.”
Project, to Dr. Maria Grazia Volpe.
This research is also part of the PhD thesis of F. Siano at the
University of Sannio.
compound, related to
z 146 as -aminobutyric acid betaine, was detected in all chestnut
samples. This substance turned out to be -guanidinobutyric acid
c-aminobutyric acid and with the same m/
c
c
(23) and, for all we know, its occurrence has never been reported
before in chestnut fruits and flours. The biosynthesis of
c
-guanidinobutyric acid (23) involves transamidination of the ami-
dino group from arginine to the amino group of -aminobutyric
acid (Irreverre, Evans, Hayden, & Silber, 1957; Tachikawa &
Hosoya, 2011). Indeed, also arginine, like -aminobutyric acid,
References
c
Adams, G. G., Imran, S., Wang, S., Mohammad, A., Kok, M. S., Gray, D. A., ... Harding, S.
E. (2014). The hypoglycemic effect of pumpkin seeds, Trigonelline (TRG),
Nicotinic acid (NA), and D-Chiro-inositol (DCI) in controlling glycemic levels in
diabetes mellitus. Critical Reviews in Food Science and Nutrition, 54(10),
1322–1329.
Ali, H. S., Alhaj, O. A., Al-Khalifa, A. S., & Bruckner, H. (2014). Determination and
stereochemistry of proteinogenic and non-proteinogenic amino acids in Saudi
Arabian date fruits. Amino Acids, 46, 2241–2257.
c
occurs at high levels in chestnut, with mean values of 900 mg/kg
of dry matter for fruits and 1300 mg/kg of dry matter for flours.
The reaction, catalyzed by an amidinotransferase enzyme, is:
Arginine þ
c
ꢁ Aminobutyric acid
Blunden, G., Patel, A. V., Adrian-Romero, M.,
& Melendez, P. (2004). The
!
c
ꢁ Guanidinobutyric acid þ Ornithine
accumulation of trans-4-hydroxyl-N-methylproline and N-methylproline by
some plant species. Biochemical Systematics and Ecology, 32, 1153–1158.
Clark-Lewis, J. W., & Mortimer, P. I. (1959). Occurrence of 4-hydroxypipecolic acid in
Acacia species. Nature, 184, 1234–1235.
c
-guanidinobutyric acid has a MS2 fragmentation pattern with
numerous fragment ions which make reliable its identification also
in a complex matrix. Its retention time was 15.9 min. Quantification
was achieved by the MS2 transition 146.1 ? 87 while 146.1 ? 104
was used as the confirmatory transition. The levels in chestnut sam-
ples were in the range of 0.2–1.7 mg/kg of dry matter (Table 1). It is
worth noting that beneficial effects in healing of gastric lesions have
been recently attributed to this compound (Hwang & Jeong, 2012).
Craig, S. A. S. (2004). Betaine in human nutrition. American Journal of Clinical
Nutrition, 80(3), 539–549.
De Vasconcelos, M. D. C. B. M., Bennett, R. N., Rosa, E. A. S., & Cardoso, J. V. F. (2007).
Primary and secondary metabolite composition of kernels from three cultivars
of Portuguese chestnut (Castanea sativa Mill.) at different stages of industrial
transformation. Journal of Agricultural and Food Chemistry, 55(9), 3508–3516.
http://dx.doi.org/10.1021/jf0629080.
De Vasconcelos, M. D. C. B. M., Bennett, R. N., Rosa, E. A. S., & Cardoso, J. V. F. (2009).
Industrial processing effects on chestnut fruits (Castanea sativa Mill.). 2. Crude
protein, free amino acids and phenolic phytochemicals. International Journal of
Food Science & Technology, 44, 2613–2619.
De Vasconcelos, M. C. D. B. M., Bennett, R. N., Rosa, E. A. S., & Cardoso, J. V. F. (2010).
Composition of European chestnut (Castanea sativa Mill.) and association with
health effects: Fresh and processed products. Journal of the Science of Food and
Agriculture, 90, 1578–1589. http://dx.doi.org/10.1002/jsfa.4016.
De Zwart, F. J., Slow, S., Payne, R. J., Lever, M., George, P. M., Gerrard, J. A., &
Chambers, S. T. (2003). Glycine betaine and glycine betaine analogues in
common foods. Food Chemistry, 83(2), 197–204. http://dx.doi.org/10.1016/
S0308-8146(03)00063-3.
3.7. Lysine derivatives
e
-N,N,N-Trimethyllysine is a particular betaine as it can be con-
sidered also an amino acid, being the
a-amino group still present in
its structure (Table 1). In mammals,
e-N,N,N-trimethyllysine is the
precursor of
forms carnitine, a metabolite essential for the fatty acid transport
and utilization in mitochondria. Recently, we showed that -N,N,
c-aminobutyric acid betaine, which, by hydroxylation,
Essery, J. M., McCaldin, D. J., & Marion, L. (1962). The biogenesis of stachydrine.
Phytochemistry, 1, 209–213.
Hanson, A. D., Rathinasabapathi, B., Rivoal, J., Burnet, M., Dillon, M. O., & Gage, D. A.
e
N-trimethyllysine is not rare in vegetables, as it was supposed,
but, on the contrary, it is ubiquitous in vegetal world (Servillo,
Giovane, Cautela, Castaldo, & Balestrieri, 2014). Actually, we found
(1994). Osmoprotective compounds in the Plumbaginaceae:
a natural
experiment in metabolic engineering of stress tolerance. Proceedings of the
National Academy of Sciences of the United States of America, 91(1), 306–310.
http://dx.doi.org/10.1073/pnas.91.1.306.
the occurrence of
e-N,N,N-trimethyllysine also in all chestnut sam-
Heinzmann, S. S., Brown, I. J., Chan, Q., Bictash, M., Dumas, M. E., Kochhar, S., ...
Nicholson, J. K. (2010). Metabolic profiling strategy for discovery of nutritional
biomarkers: Proline betaine as a marker of citrus consumption. The American
Journal of Clinical Nutrition, 92(2), 436–443.
Hwang, I. Y., & Jeong, C. S. (2012). Inhibitory effects of 4-guanidinobutyric acid
against gastric lesions. Biomolecules and Therapeutics, 20, 239–244.
ple examined. -N,N,N-Trimethyllysine eluted at retention time of
e
13.1 min and its quantification was achieved by using 189.1 ?
130, which was the MS2 strongest precursor to product ion transi-
tion, while 189.1 ? 84 was used as the confirmatory transition.
The average
e-N,N,N-trimethyllysine levels resulted in the range
Irreverre, F., Evans, R. L., Hayden, A. R., & Silber, R. (1957). Occurrence of
c-
0.4–1.1 ppm (Table 1).
guanidinobutyric acid. Nature, 180, 704–705.

L. Servillo et al. / Food Chemistry 196 (2016) 1301–1309
1309
Jenkins, D. J. A., Kendall, C. W. C., Augustin, L. S. A., Franceschi, S., Hamidi, M.,
Atherosclerosis,
177(2),
375–381.
http://dx.doi.org/10.1016/j.
Marchie, A., ... Axelsen, M. (2002). Glycemic index: Overview of implications in
health and disease. The American Journal of Clinical Nutrition, 76(1), 266S–273S.
Kavi Kishor, P. B., Sangam, S., Amrutha, R. N., Sri Laxmi, P., Naidu, K. R., Rao, K. R. S. S.,
... Sreenivasulu, N. (2005). Regulation of proline biosynthesis, degradation,
uptake and transport in higher plants: Its implications in plant growth and
abiotic stress tolerance. Current Science, 88, 424–438.
McKeever, M. P., Weir, D. G., Molloy, A., & Scott, J. M. (1991). Betaine-homocysteine
methyltransferase: organ distribution in man, pig and rat and subcellular
distribution in the rat. Clinical Science (London, England: 1979), 81(4), 551–556.
Rathinasabapathi, B. (2002). Propionate, a source of b-alanine, is an inhibitor of b-
alanine methylation in Limonium latifolium, Plumbaginaceae. Journal of Plant
Physiology, 159, 671–674.
Rathinasabapathi, B., Sigua, C., Ho, J., & Gage, D. A. (2000). Osmoprotectant b-alanine
betaine synthesis in the Plumbaginaceae: S-adenosyl-L-methionine dependent
N-methylation of b-alanine to its betaine is via N-methyl and N, N-dimethyl b-
alanines. Physiologia Plantarum, 109, 225–231. http://dx.doi.org/10.1034/
j.1399-3054.2000.100302.x.
Sargenti, S. R., Fernandes, J. B., Dasilva, M., Vieira, P. C., Salatino, A., & Pirani, J. R.
(1993). Betaines, amino-acids and waxes from Erythrochiton-brasiliensis.
Biochemical Systematics and Ecology, 21(6–7), 723–727. http://dx.doi.org/
10.1016/0305-1978(93)90078-6.
Servillo, L., Giovane, A., Balestrieri, M. L., Cautela, D., & Castaldo, D. (2011a). A).
Proline derivatives in fruits of bergamot (Citrus bergamia Risso et Poit):
Presence of n-methyl-l-proline and 4-hydroxy-l-prolinebetaine. Journal of
Agricultural and Food Chemistry, 59(1), 274–281. http://dx.doi.org/10.1021/
jf102833v.
Servillo, L., Giovane, A., Balestrieri, M. L., Bata-Csere, A., Cautela, D., & Castaldo, D.
(2011b). B). Betaines in fruits of citrus genus plants. Journal of Agricultural and
Food Chemistry, 59(17), 9410–9416. http://dx.doi.org/10.1021/jf2014815.
Servillo, L., Giovane, A., Balestrieri, M. L., Ferrari, G., Cautela, D., & Castaldo, D.
(2012). Occurrence of pipecolic acid and pipecolic acid betaine
(homostachydrine) in Citrus genus plants. Journal of Agricultural and Food
Chemistry, 60(1), 315–321. http://dx.doi.org/10.1021/jf204286r.
Servillo, L., D’Onofrio, N., Longobardi, L., Sirangelo, I., Giovane, A., Cautela, D., ...
Balestrieri, M. L. (2013). Stachydrine ameliorates high-glucose induced
endothelial cell senescence and SIRT1 downregulation. Journal of Cellular
Biochemistry, 114(11), 2522–2530. http://dx.doi.org/10.1002/jcb.24598.
Servillo, L., Giovane, A., Cautela, D., Castaldo, D., & Balestrieri, M. L. (2014). Where
atherosclerosis.2004.07.020.
Slow, S., Donaggio, M., Cressey, P. J., Lever, M., George, P. M., & Chambers, S. T.
(2005). The betaine content of New Zealand foods and estimated intake in the
New Zealand diet. Journal of Food Composition and Analysis, 18(6), 473–485.
http://dx.doi.org/10.1016/j.jfca.2004.05.004.
Street, T. O., Bolen, D. W., & Rose, G. D. (2006). A molecular mechanism for
osmolyte-induced protein stability. Proceedings of the National Academy of
Sciences of the United States of America, 103(38), 13997–14002. http://dx.doi.org/
10.1073/pnas.0606236103.
Tachikawa, M.,
& Hosoya, K.-I. (2011). Transport characteristics of guanidino
compounds at the blood-brain barrier and blood-cerebrospinal fluid barrier:
relevance to neural disorders. Fluids and Barriers of the CNS, 8(1), 13. http://dx.
doi.org/10.1186/2045-8118-8-13.
Terano, S.,
& Suzuki, Y. (1978). Formation of b-alanine from spermine and
spermidine in maize shoots. Phytochemistry, 17, 148–149. http://dx.doi.org/
10.1016/S0031-9422(00)89700-9.
Tramontano, W. A., Hartnett, C. M., Lynn, D. G., & Evans, L. S. (1982). Relationship
between trigonelline concentration and promotion of cell arrest in G2 in
cultured roots of Pisum sativum. Phytochemistry, 21(6), 1201–1206. http://dx.
doi.org/10.1016/0031-9422(82)80111-8.
Van Wandelen, C., & Cohen, S. A. (1997). Using quaternary high-performance liquid
chromatography eluent systems for separating 6-aminoquinolyl-N-
hydroxysuccinimidyl carbamate-derivatized amino acid mixtures. Journal of
Chromatography A, 763, 11–22. http://dx.doi.org/10.1016/S0021-9673(96)
00891-6.
Wood, K. V., Stringham, K. J., Smith, D. L., Volenec, J. J., Hendershot, K. L., Jackson, K.
A., ... Rhodes, D. (1991). Betaines of alfalfa: characterization by fast atom
bombardment and desorption chemical ionization mass spectrometry. Plant
Physiology, 96(3), 892–897. http://dx.doi.org/10.1104/pp.96.3.892.
Yuan, Z. X., Patel, A. V., Blunden, G., & Turner, C. H. (1992). Trans-4-hydroxypipecolic
acid betaine from Lamium maculatum. Phytochemistry, 31, 4351–4352.
Zeisel, S. H., & Blusztajn, J. K. (1994). Choline and human nutrition. Annual Review of
Nutrition, 14, 269–296. http://dx.doi.org/10.1146/annurev.nutr.14.1.269.
Zeisel, S. H., Mar, M.-H., Howe, J. C., & Holden, J. M. (2003). Concentrations of
choline-containing compounds and betaine in common foods. The Journal of
Nutrition, 133(5), 1302–1307.
Zhang, C., Lu, Y., Zhou, Y. J., Tong, Q. Q., Qu, C., & Kang, T. J. (2014a). The effect of
stachydrine on the expression of caspase-12 in rats with unilateral ureteral
obstruction. Journal of Urology, 192(5), 1549–1554.
does N(e)-trimethyllysine for the carnitine biosynthesis in mammals come
from? PloS One, 9(1), e84589. http://dx.doi.org/10.1371/journal.pone.0084589.
Shai, I., Stampfer, M. J., Ma, J., Manson, J. E., Hankinson, S. E., Cannuscio, C., ... Rimm,
E. B. (2004). Homocysteine as a risk factor for coronary heart diseases and its
association with inflammatory biomarkers, lipids and dietary factors.
Zhang, C., Shan, X. L., Liao, Y. L., Zhao, P., Guo, W., Wie, H. C., & Lu, R. (2014b). Effects
of stachydrine on norepinephrine-induced neonatal rat cardiac myocytes
hypertrophy and intracellular calcium transients. BMC Complementary and
Alternative Medicine, 14(1), 474.