ISSN 0006-2979, Biochemistry (Moscow), 2026, Vol. 91, No. 6, pp. 923-938 © Pleiades Publishing, Ltd., 2026.
923
Dose-Dependent Effects of Soy Lecithin Intake
on Synaptic Ultrastructure in Brain Neurons
and Behavioral Patterns of C57BL/6 Laboratory Mice
Lyubov A. Suldina
1,2#
, Ksenia N. Morozova
1,3#
, Konstantin S. Pavlov
2
,
Elena V. Kiseleva
1
, and Lidiya V. Boldyreva
1,a
*
1
Institute of Cytology and Genetics, Siberian Branch of the Russian Academy of Sciences,
630090 Novosibirsk, Russia
2
Scientific Research Institute of Neurosciences and Medicine, 630117 Novosibirsk, Russia
3
Novosibirsk State University, 630090 Novosibirsk, Russia
a
e-mail: boldyrevalv@neuronm.ru
Received January 16, 2026
Revised June 2, 2026
Accepted June 14, 2026
Abstract— Phospholipid preparations, including lecithin, are widely used as hepatoprotective and neuropro-
tective agents, while soy lecithin is intensively used in food industry, with its total dose in the modern human
diet potentially reaching high levels. Soy lecithin contains up to 70% of biologically active phospholipids: phos-
phatidylcholine, phosphatidylethanolamine, phosphatidylinositol, and phosphatidic acid. These compounds
perform a spectrum of key cellular functions, including neuromediation processes, and ensure formation of
the cellular membrane structures and vesicles. Previously, in the mouse model of chronic intestinal inflam-
mation, behavioral changes were observed together with the significant increase in the relative content of
several phospholipid classes in the intestinal epithelial cells. The animals fed with soy lecithin, which con-
tains a mixture of these phospholipids, showed similar behavioral changes in the absence of inflammation:
impaired social recognition and behavior, reduced signs of compulsivity and anxiety, and increased aggres-
sion in males. In this study, we found that reducing the dose of soy lecithin in short-term administration
restores normal social recognition and behavior in the healthy C57BL/6 animals, while reduction in anxiety
is maintained. Comparative electron microscopy analysis of the neurons and synapses in the amygdala,
hypothalamus, and frontal motor cortex of the C57BL/6 animals treated with soy lecithin was conducted.
Dose-dependent and region-specific changes were observed. High-dose lecithin administration, both long-term
and short-term, reduced synapse density in the neuropil, and irregular synaptic vesicles were detected in
the amygdala and hypothalamus. Threefold reduction in the lecithin dosage resulted in the increase in the
number of vesicles per synapse in the hypothalamus and motor cortex. The obtained results demonstrate
dose-dependent effect of soy lecithin on the synaptic ultrastructure in the hypothalamus, amygdala, and motor
cortex of the frontal lobe, as well as on the behavioral patterns of the healthy C57BL/6 laboratory mice.
DOI: 10.1134/S0006297926600109
Keywords: phospholipids, soy lecithin, social behavior, anxiety, C57BL/6 mice, electron microscopy, ultrastruc-
tural analysis, synapse, synaptic vesicles, neuropil
* To whom correspondence should be addressed.
# These authors contributed equally to this study.
INTRODUCTION
Various classes of phospholipids (PLs) perform
key molecular and cellular functions in the nervous
system, and changes in their metabolism correlate
with the diseases and chronic inflammatory processes
[1-3]. The major phospholipids in the human brain are
phosphatidylcholine, phosphatidylethanolamine, and
phosphatidylserine, with their enrichment observed
in all cortical regions of the brain [1]. The most com-
mon fatty acids in phosphatidylcholine are palmitic
SULDINA et al.924
BIOCHEMISTRY (MOSCOW) Vol. 91 No. 6 2026
and stearic acids at the sn-1 position, while in the
brain, phosphatidylcholine contains polyunsaturated
fatty acids at the sn-2 position, with docosahexaenoic
acid and arachidonic acid being the most common
in the human brain [1, 2]. It has been suggested that
this structural composition of phosphatidylcholine
in the brain in combination with the properties of
cholesterol and other PLs ensures conformation and
fluidity necessary for functional activity of the neuro-
nal membranes. Phospholipids are not only the main
component of cell membranes and transport mole-
cules but also serve as substrates and precursors for
a wide spectrum of biologically active molecules that
regulate metabolic pathways in the gut–brain axis
[4, 5]. In the central nervous system (CNS), PLs are
closely linked to the processes critical for neurome-
diation mechanisms, such as adenosine triphosphate
(ATP) production, mitochondrial transport, and pro-
duction and secretion of neurotransmitters. For ex-
ample, it has been estimated that oxidation of PLs
provides, on average, 20% of the total energy produc-
tion in the brain [6]. Phosphatidylcholine also serves
as a choline depot for the synthesis of the neurotrans-
mitter acetylcholine and, at the same time, acts as
a substrate for the synthesis of phosphatidic acid.
Sphingolipids, glycerophospholipids, and cholesterol
are involved not only in transmission of a spectrum
of cellular signals and formation of the lipid rafts,
maintaining energy and oxidative balance, but also
in axon myelination, blood-brain barrier formation,
and neuroinflammatory homeostasis regulation [6-8].
Phospholipids play an important role in synapse for-
mation, neurotransmitter release, and signal trans-
mission, and phospholipid metabolism deregulation
is closely associated with various neurodegenerative
diseases. Enzymes and co-factors of phospholipid me-
tabolism are considered as therapeutic targets [8, 9].
The membranes of synaptic vesicles consist of
phospholipids, with the largest proportions being
phosphatidylcholine, phosphatidylethanolamine, and
cholesterol. These phospholipids ensure physicochem-
ical properties of the neuronal membranes, which
are critical for neuromediation mechanisms [10, 11].
Phosphatidylserine and phosphatidic acid act as key
signaling phospholipids in regulation of the synaptic
vesicle cycle [11]. For example, phosphatidylserine in
the membrane of synaptic vesicles ensures interac-
tion with Rab and SNARE family proteins, providing
anchoring the synaptic vesicles at the presynaptic
membrane [10-12]. Additionally, phosphatidylserine
mediates initiation of synaptic exocytosis by the Ca
2+
sensor synaptotagmin-1, via its functional domains
(C2A and C2B) activation by phosphatidylserine only
in the presence of Ca
2+
[12]. Physicochemical proper-
ties of phosphatidic acid facilitate membrane bending
and instantaneous fusion upon neural impulse arrival.
Furthermore, phosphatidic acid acts as a critical co-
factor in stabilization of the synaptophysin complex,
as well as in positioning and activity of dynamin in
the mechanism of synaptic vesicles recycling [6, 8].
Deficiency of both phosphatidic acid and phosphati-
dylinositol, which are precursors for PI(4,5)P
2
, leads
to deregulation of the synaptic vesicles docking on
the presynaptic membrane, inhibiting their recycling
cycle, and thus impairing the neuromediation [6].
Connection between the regulation of the body’s
metabolic pathways and psychoemotional state has
become a highly relevant topic in the last decade.
A number of studies on the patients and animal
models indicate a much broader and more signifi-
cant influence than previously thought [13, 14]. Mod-
ern scientific paradigm of the gut–brain axis has un-
dergone significant evolution: from a model focusing
exclusively on microbiome, the emphasis has shifted
to studying metabolic pathways that mediate modu-
lation of the CNS functions by active components en-
tering with food: through interaction with the enteric
nervous system; through changes in the composition
of the gut microbiota; and through various mecha-
nisms of delivery of active metabolites and their sub-
sequent effect on molecular and cellular processes in
the brain [15, 16].
Previously, in the genetic model of chronic colitis
in mice with mutation in the Mucin-2 (Muc2) gene
it was shown that, on the one hand, there is a sig-
nificant change in the behavioral characteristics of
animals and, on the other hand, there is a change in
the metabolomic profile of intestinal and brain cells
[17, 18]. In Muc2 animals, social behavior disorders
were accompanied by the significant increase in the
levels of several forms of phospholipids in the intesti-
nal epithelial cells, most notably phosphatidylcholine,
phosphatidylserine, and phosphatidic acid [17]. Since
metabolism of phospholipids is critically important
for the brain function, and phospholipids can cross
the intestinal and blood-brain barriers in various
forms of packaging, we have then performed studies
that revealed a significant effect of the dietary intake
of both a phospholipid mixture (phosphatidylcholine,
phosphatidylserine, and phosphatidic acid) and soy
lecithin (70% total content of a phospholipid mixture:
phosphatidylcholine, phosphatidylethanolamine, phos-
phatidylinositol, and phosphatidic acid) on the social
and sexual behavior of mice [19, 20]. As a result, im-
pairments in social recognition and behavior, reduced
signs of compulsivity and anxiety, and increased ag-
gression in the males were observed with both long-
term and short-term intake of doses of soy lecithin by
the healthy animals, reproducing the findings of both
the same behavioral traits and the phospholipid en-
richment in the intestinal epithelium during chronic
inflammation in the Muc2 mice [17, 18].
EFFECTS OF SOY LECITHIN ON SYNAPTIC STRUCTURE AND BEHAVIOR 925
BIOCHEMISTRY (MOSCOW) Vol. 91 No. 6 2026
This study is focused on the comparative analy-
sis of the effects of perinatal and short-term dietary
soy lecithin in high and reduced doses on the neu-
rons and synapses in the hypothalamus, amygdala,
and motor zone of the frontal cortex of the brain, as
well as on the behavioral patterns mediated by these
brain regions in the healthy C57BL/6 mice. The study
comprehensively investigates the effects of dietary
soy lecithin intake on neuromediation processes at
both the cellular level (ultrastructural analysis of syn-
apses) and the CNS level (behavioral phenotyping).
MATERIALS AND METHODS
Experimental animals. The study used C57BL/
6JNskrc (local subcolony of C57BL/6J, hereafter
C57BL/6) and BALB/cNskrc (local subcolony of BALB/c,
hereafter BALB/c) mouse strains. Animals were kept
in same-sex groups of 3-5 individuals in cages mea-
suring 37×21×15 cm (length×width×height) with
wood sawdust bedding in a conventional vivarium at
the Research Institute of Neurosciences and Medicine,
Novosibirsk. The 12D:12N (reversed) light regime was
maintained with temperature 20-22°C, free access to
standard full-ration dry granulated feed and puri-
fied water was provided to the laboratory rodents.
Exclusion criteria for animals from the experiment
included: weight loss of more than 20% of normal;
organic disorders of CNS functions, alopecia, infertil-
ity, abscesses, or injuries.
Diet. Experimental groups of C57BL/6 mice re-
ceived standard feed uniformly mixed with soy
lecithin (Solgar, USA; declared composition: 70%
phospholipid mixture: phosphatidylcholine, phospha-
tidylethanolamine, phosphatidylinositol, phosphatidic
acid) at a calculated daily dose per mouse, multi-
plied by the number of mice in the cage. Doses of
180 mg/mouse/day (high) and 60 mg/mouse/day (re-
duced) were chosen based on the previous metabolo-
mic data [17, 18] and extrapolation of the potentially
achievable total levels of phospholipid consumption
within the modern human diet [3]. The control
group received standard feed without additives. Feed
was placed in the feeder daily. Feed leftovers were
weighed daily for 2 weeks. Control of the daily intake
of the lecithin dose was carried out by calculating av-
erage amount of feed consumed per day per animal.
Perinatal feeding was carried out as follows:
pregnant C57BL/6 females (20 animals) from the sec-
ond week of pregnancy received feed with soy lec-
ithin at a dose of 180 mg/day per animal. The off-
spring continued to receive feed with the same dose
of substances until sample collection. For newborn
pups, lecithin intake occurred through the mother’s
milk, which continued to receive the lecithin diet.
Experimental groups were formed from adult
males of the same age (10-12 weeks), from the litters
of 20 females of the same age. At the age of 3 weeks,
males were separated into combined randomized
groups of 3-6 animals for subsequent perfusion and
sample preparation for transmission electron micros-
copy (TEM) (number of animals is indicated taking
into account subsequent exclusions during experi-
ments).
Short-term feeding: mature C57BL/6 males (age
10-12 weeks) received feed with soy lecithin at a dose
of 180 mg/day per animal for 2 weeks.
Short-term feeding with reduced dose: mature
C57BL/6 males (age 10-12 weeks) received feed with
soy lecithin at a dose of 60 mg/day per animal for
2 weeks.
Behavioral testing. The following experimental
groups of animals were formed:
1. Control group (n = 10): sexually mature C57BL/6
males (age 10-12 weeks) receiving standard feed;
2. Reduced-dose lecithin group (n = 10): sexually
mature C57BL/6 males (age 10-12 weeks) receiv-
ing soy lecithin at a dose of 60 mg/day per ani-
mal for 2 weeks.
Testing was conducted during the dark period
(for animals) under red lighting (unless otherwise
specified). Between tests, the arena of the setups was
cleaned with 70% ethanol solution to remove odors.
Open field test. A square plastic setup (40×40 cm)
with transparent walls and an opaque bottom was
used. Center of the field was defined as a 20×20 cm
square. Testing was conducted for 6 min. A mouse
was placed in the center of the field, and the fol-
lowing parameters were measured: distance traveled,
number of rearing, time spent in the center of the
field. All parameters were recorded and processed
using Ethovision XT10 software (Noldus Internation-
al Technology). Number of defecations and grooming
time were recorded visually.
Light–dark box test. A rectangular setup (42×
21×25 cm) consisting of two compartments separat-
ed by a partition with a 3×4 cm hole was used. The
dark compartment constitutes 1/3 of the setup. Test
was conducted with bright white lighting in the light
compartment and no lighting in the dark compart-
ment. A mouse was placed in the dark compartment,
facing away from the hole connecting the compart-
ments. Over 5 min, the following parameters were re-
corded using a camera positioned above: time of first
exit from the dark compartment, duration in the light
compartment, distance traveled in the light compart-
ment. All parameters were recorded and processed
using Ethovision XT10 software (Noldus International
Technology).
Marble burying test. The test was conducted
in clean plastic animal cages (37×21×15 cm). Wood
SULDINA et al.926
BIOCHEMISTRY (MOSCOW) Vol. 91 No. 6 2026
sawdust was poured onto the bottom of the cages
(4-cm depth), on which 20 glass marbles (diame-
ter = 1.0 cm) were evenly distributed. Each mouse
was placed in the separate cage for 30 min. After-
ward, the mouse was removed from the cage, and
number of marbles covered with sawdust by more
than 70% was counted.
Social preference test (two-intruder test: male
and female). Four days before the test, males that
had received sexual experience using a previously
described method [21] were placed in individual cag-
es. 10 BALB/c females and 10 BALB/c males [22] were
used as intruders. One day before the test, sexually
mature female intruders were marked with a safe
dye in the scapular region. To perform the test, a fe-
male and male intruder were simultaneously placed
in the home cage with the test male, and the animals
were allowed to interact freely for 15 min. During
this time, a video recording was made, and duration
of chasing and sniffing (social contact), mounting
(sexual preference), and attacks (aggression) of the
test male toward the female and male intruders was
counted.
Social odor test. To assess olfactory preference, a
test male, which had previously received sexual expe-
rience [21], was presented with soiled bedding sam-
ples from BALB/c females and males placed in two
mesh metal containers (diameter 6 cm) in the home
cage, since it is known that presence of sexual expe-
rience promotes manifestation of sexual motivation
in males [23, 24]. A male explored the samples for
5 min, during which the time spent sniffing the soiled
bedding from females and males was counted. The
result was expressed as (time sniffing bedding from
female or male)/(total sniffing time) in percentages.
Transmission electron microscopy (TEM) ul-
trastructural analysis. For electron microscopy, an-
imals were kept and fed under the same conditions
as those used for behavioral phenotyping [19, 20] to
exclude the influence of stress encountered during
testing on the ultrastructure of neurons. From each
animal in the comparison groups (control (n = 3 for
each experiment), perinatal lecithin intake (n = 3),
short-term lecithin intake (n = 3), reduced-dose leci-
thin intake (n = 3)), 3 brain regions (amygdala, hypo-
thalamus, frontal cortex) were obtained. The control
group of animals was formed from the animals of
the same age that were kept under the same condi-
tions (the only difference was the diet) as each of the
experimental groups. Prior to euthanasia mice were
anesthetized with Domitor intraperitoneal injection
of 75 μl per 10 g of mouse weight (Orion Pharma,
Espoo, Finland) and Zoletil intraperitoneal injection
of 60 μL per 10 g of mouse weight (Virbac Sante An-
imale, France). After 20 min intracardiac perfusion
was performed with 15 ml of physiological saline
followed by 15 ml 6% glutaraldehyde (AppliChem
GmbH, Germany) in physiological saline per animal.
For the experimental and control groups, perfusion of
animals and sample preparation were carried out si-
multaneously and under identical conditions. In total,
TEM analysis was performed for 6 paired comparison
groups (18 animals, 3 per group) in 54 brain region
samples (3 regions from each animal).
Before ultrastructural analysis of the obtained
samples, an analysis of the quality of perfusion, tis-
sue fixation, and preservation of intracellular mem-
brane and non-membrane structures was performed.
After perfusion with 6% glutaraldehyde in physi-
ological saline, brain regions were fixed in a 2.5%
glutaraldehyde solution in 0.1 M sodium cacodylate
buffer (pH 7.4) for 1 h at room temperature. Next,
they were washed three times with a 0.1 M sodium
cacodylate buffer and post-fixed in a 1% osmium
tetroxide solution with 0.8% potassium ferricyanide
for 1 h. Fixed samples were contrasted with a 1% ura-
nyl acetate solution in water. Samples were dehy-
drated and embedded in an epoxy resin (Epon 812).
Semi-thin cross-sections (250 nm) were prepared,
stained with methylene blue, and preliminarily an-
alyzed using an Axioscope-4 microscope (Zeiss).
Ultra- thin sections (60 nm) for TEM were prepared
with a diamond knife using a Leica EM UC6 ultra-
microtome (Leica), and after that examined using a
JEM1400 transmission electron microscope (JEOL).
For each block, serial ultra-thin sections were pre-
pared, and at least 10 fields of view (neuropil) were
analyzed for the presence of synapses. TEM was per-
formed using equipment of the Shared Access Center
for Microscopic Analysis of Biological Objects of the
SB RAS (FWNR-2026-0024).
Morphometric analysis. For morphometric
analysis, 3 animals per each of the 4 groups (control,
perinatal lecithin intake, short-term lecithin intake,
reduced-dose lecithin intake) were used – a total
of 12 animals. From each animal, 3 brain regions
(amygdala, hypothalamus, frontal cortex) were tak-
en. For each region, at least 10 fields of view were
analyzed. Relative density of synapses was calculat-
ed on electron micrographs of the neuropil at 8000×
magnification. For this, total area of the neuropil was
measured, then the number of synapses was counted
and related to the area of the neuropil as units/μm
2
.
Similarly, on micrographs of synapses at 25,000× mag-
nification, the synapse area was measured, the num-
ber of vesicles was counted, and related to the area
of the presynaptic terminal. Synaptic vesicles were
considered heterogeneous in case of significant de-
viation from a round shape and noticeable variation
in size parameters. Mitochondria were considered de-
fective if presence of pronounced matrix clarification,
fragmentation, or absence of cristae, or membrane
EFFECTS OF SOY LECITHIN ON SYNAPTIC STRUCTURE AND BEHAVIOR 927
BIOCHEMISTRY (MOSCOW) Vol. 91 No. 6 2026
integrity disorders were observed. The analysis was
performed in a blind manner.
Statistical analysis. Data are presented graph-
ically as a mean and standard error of the mean
(M ± SEM). Statistical analysis of the data was per-
formed using STATISTICA 12.0 (StatSoft TIBCO Soft-
ware) with the Student’s t-test for normally distrib-
uted data. Normality of the data distribution was
evaluated using the Shapiro–Wilk test. For comparing
samples with distribution different from normal, the
Mann–Whitney U test was used. The significance level
was set at p < 0.05.
RESULTS
To identify cellular effects of the soy lecithin in-
take on neuromediation processes that may underly
the previously identified behavioral pattern disorders
in animals [18-20], in this study, we performed com-
parative analysis of the ultrastructure of neurons and
synapses in the mice brain using transmission elec-
tron microscopy (TEM). Comparative ultrastructural
analysis of the samples from three brain regions was
conducted: amygdala and hypothalamus, which are
most involved in the regulation of social and sexual
behavior, anxiety, and aggression in mice [25], as well
as motor zone of the frontal cortex for the following
groups of experimental animals:
1. Adult C57BL/6 males receiving soy lecithin long-
term perinatally;
2. Adult C57BL/6 males receiving soy lecithin short-
term (2 weeks);
3. Adult C57BL/6 males receiving a reduced dose of
soy lecithin short-term (2 weeks);
4. Control adult C57BL/6 males receiving standard
feed (n = 3 per brain region and group, total
36 samples).
In all three studied brain regions of the exper-
imental groups, neurons exhibited typical ultrastruc-
tural morphology, with no deviations in the frequen-
cy of cell death (apoptosis and autophagy) traits,
such as nuclear and cellular fragmentation, plasma
membrane integrity disorders, vacuole formation,
and apoptotic bodies, observed in the brain samples
of animals from the lecithin-receiving groups and
control animals. Neurons in all three studied brain
regions exhibited morphological characteristics typi-
cal for each region.
Neurons of the amygdala (Fig. 1, a, d, g, j) have
a complex ultrastructure characteristic of highly ac-
tive neurons involved in emotion processing, memo-
ry, and decision-making. Neurons in this brain region
have a large nucleus, numerous mitochondria, rough
endoplasmic reticulum, and Golgi apparatus in the cy-
toplasm. This brain region is notable for its high syn-
apse density and axon myelination. Neurons of the
hypothalamus (Fig. 1, b, e, h, k) with characteristic
electron-light cytoplasm and a large nucleus showed
nuclear envelope invagination in all experimental
groups. The cytoplasm contained numerous mito-
chondria, rough endoplasmic reticulum, a developed
Golgi apparatus, and electron-dense neurosecretory
granules. Neurons of the motor zone of the frontal
cortex (Fig. 1, c, f, i, l) are relatively large, with large
light nucleus. The cytoplasm contains mitochondria,
endoplasmic reticulum cisternae, and free ribosomes,
and numerous synaptic contacts with other neurons
are observed. Representative electron micrographs in
Fig. 1 also show synaptic contacts with other neurons,
which are marked with a blue background. Thus, the
ultrastructural analysis of neurons from three brain
regions – amygdala, hypothalamus and motor zone
of the frontal cortex – did not reveal morphological
signs of neuronal death or other cellular disorders.
Next, we performed a detailed ultrastructural
analysis of synapses, since synaptic terminals are
directly involved in neuromediation processes.
Ultrastructural changes in the synapses could indicate
changes in mediatory processes in the studied brain
regions, which may underlie the observed behavioral
deviations in the animals receiving soy lecithin.
Effects of perinatal soy lecithin intake on
synaptic ultrastructure in the brain of C57BL/6
mice. TEM micrographs show representative imag-
es of the synaptic terminals from the studied brain
regions, containing synaptic vesicles, synaptic cleft,
and postsynaptic membrane (Fig. 2). For accurate as-
sessment of the extent of the synaptic ultrastructure
disruptions, morphometric and subsequent statistical
analysis was performed (Fig. 3). Active zone of the
synapse – the area of the presynaptic membrane
where vesicle fusion occurs for neurotransmitter re-
lease– is marked in Fig. 2 with white arrows. In this
zone, accumulations of synaptic vesicles containing
neurotransmitters are observed, with size of approx-
imately 20-25 nm, as well as mitochondria necessary
for supplying energy to the synaptic transmission
process. Opposite the active zone of the presynaptic
membrane, the postsynaptic density is visible, char-
acterized by the dense accumulation of protein mol-
ecules – receptors, ion channels, and signaling mole-
cules (marked in the images with black arrowheads).
Comparing ultrastructure of the synapses in the
animals that received lecithin perinatally, it should be
noted that in the neuropil of the amygdala and hy-
pothalamus, synaptic terminals are less frequent than
in the control (Fig. 2, Fig. 3). Moreover, synapses in
the brain samples of the perinatally lecithin-receiving
group are filled with the small vesicles of irregular
size and shape (Fig. 2, b and d), unlike the regular
vesicles in the control (Fig. 2, a and c). Number and
SULDINA et al.928
BIOCHEMISTRY (MOSCOW) Vol. 91 No. 6 2026
a
d
g
j
b
e
h
k
c
f
i
l
Fig. 1. Representative electron micrographs of neurons in the studied brain regions: amygdala (a, d, g, j), hypothalamus (b, e,
h, k), and motor zone of the frontal cortex (c, f, i, l) of healthy adult C57BL/6 males (a, b, c); adult C57BL/6 males long-term
perinatally receiving soy lecithin (d, e, f); adult C57BL/6 males short-term (2 weeks) receiving soy lecithin (g, h, i); adult
C57BL/6 males receiving a reduced dose of soy lecithin (j, k, l). Designations: N – nucleus, mt – mitochondria, aG – Golgi
apparatus, gr – secretory granules, ms – myelin sheath, s – synapses.