Abstract

   It has been suggested that dietary intake of lipids and fatty acids may
   influence cognitive function, however, the effect of lard intake during
   pregnancy and postpartum periods on cognitive function of mother
   remains to be elucidated. We investigated the effect and mechanism of
   consuming soybean oil (SO), the mixed oil of lard and soybean oil at
   the ratio of 1:1 (LS) and lard oil (LO) during the pregnancy and
   postpartum periods on cognitive function of the maternal mice. All
   pregnant C57BL/6JNifdc mice were fed with soybean oil diet during day
   0–10 (the day when vaginal plugs appeared in female mice was recorded
   as day 0), and then randomly assigned to SO, LS and LO groups (n = 10)
   from day 11 to day 44. The time in center zone and the number of times
   to enter in center zone were significantly higher in the SO group than
   in the LO group detected by the open-field test. The levels of
   neuroglial cells, NOD-like receptor family pyrin domain containing 3
   (NLRP3) inflammasome complex and pyroptosis related proteins in brain
   of the LO group were significantly higher than those in the SO group.
   RNA-sequencing results showed that the calcium signaling pathway
   related genes in brain, including Adcy8, Ntsr1, Trhr, Oxtr, Htr5b and
   Camk2d levels significantly higher in the LO group than in the SO
   group. Lipidomic analysis indicated that PG 18:2_18:2, PG 20:5_22:6,
   and CL 12:0_16:0_22:3_22:5 of glycerophospholipid metabolism in brain
   significantly connected with Htr5b of calcium signaling pathway. In
   conclusion, the intake of lard during the pregnancy and postpartum
   periods is detrimental to the cognitive function of maternal mice,
   which probably due to changes in the composition of fatty acid in the
   brain, thereby activating neuroinflammation via calcium signaling
   pathway in brain.

   Keywords: Lard, Postpartum cognitive function, Omega-3 fatty acid,
   Neuroinflammation

Graphical abstract

   [37]Image 1
   [38]Open in a new tab

Highlights

     * •
       Lard altered the brain fatty acid profile and glycerophospholipid
       metabolism.
     * •
       Lard could active neuroinflammation via calcium signaling pathway.
     * •
       Lard has negative effects on the postpartum cognitive function than
       soybean oil.

1. Introduction

   Women may experience significant physiological and behavioral
   fluctuations during the pregnancy and postpartum period ([39]Xavier et
   al., 2021), moreover, it is noteworthy that no less than one in every
   five women are at risk of developing anxiety and/or depression either
   during pregnancy, postpartum or both periods ([40]Falah-Hassani et al.,
   2017; [41]Woolhouse et al., 2015). The epidemiological study also
   demonstrated a decline in free recall, working memory, and executive
   function among pregnant women ([42]Anderson and Rutherford, 2012;
   [43]Davies et al., 2018). The involvement of fatty acids as fundamental
   constituents in membrane structure is associated with the prevalence of
   brain disorders. Previous studies have proved that dietary lipids
   (soybean oil or flaxseed oil) rich in n-3 polyunsaturated fatty acid
   (PUFA) were associated with preventing cognitive decline compared with
   a dietary lipids (safflower oil) rich in n-6 PUFA in postpartum mice
   ([44]Harauma et al., 2022; [45]Tang et al., 2018b).

   Lard, a traditional culinary oil in China, offers numerous health
   benefits including diuretic properties, blood purification, promotion
   of blood circulation, and detoxification for the human body ([46]Wang
   et al., 2017). But most studies believed that lard has negative effects
   on cognition function in animal obese model. For instance, high-fat
   lard diet enriched in saturated fatty acid (SFA) could lead to poorer
   performance of male rats in an open-field test ([47]Alghamdi, 2021),
   similarly, male rats were treated with high-fat lard and high sucrose
   diet for more than 8 weeks and therefore exhibited the anxiety-like
   behavior ([48]Nakajima et al., 2020). Regrettably, the widely used
   AIN-93M (10% energy from fat) and AIN-93G (16% energy from fat) are
   promulgated by the American Institute of Nutrition and applied to the
   maintenance and breeding periods, respectively ([49]Reeves et al.,
   1993), so the energy from high-fat diet far exceeds what is needed for
   normal growth and reproduction in previous studies on lard and these
   findings fail to demonstrate typical physiological responses in the
   body. Inflammation is proposed to be an important pathophysiological
   mechanism underlying cognitive impairment ([50]Miller and Spencer,
   2014) and constantly changes in inflammatory responses occur throughout
   the pregnancy and postpartum period ([51]Payne and Maguire, 2019).
   Numerous research of fatty acids biological function suggested that
   dietary oil rich in monounsaturated fatty acid (MUFA) and n-3 PUFA
   could diminish neuroinflammation, and dietary oil rich in n-6 PUFA and
   SFA increase neuroinflammation ([52]de Andrade et al., 2017; [53]Fan et
   al., 2022). Compared with widely used soybean oil in the control group
   of the animal experiment, lard is rich in SFA and MUFA, but the fatty
   acid proportion of PUFA is lower than soybean oil. To date, whether
   lard can affect cognitive function of postpartum mice by altering
   neuroinflammation is currently unknown.

   Our previous studies have shown that the mixed oil of lard and soybean
   oil could be more beneficial for body weight and liver function than
   soybean oil or lard ([54]Liu et al., 2023), the aim of the present
   study was to investigate the effect of consuming soybean oil (SO), the
   mixed oil of lard and soybean oil at the ratio of 1:1 (LS) and lard oil
   (LO) during the pregnancy and postpartum periods on cognitive function
   of the maternal mice. Previous studies suggested that insufficient
   level of n-3 PUFA in placenta were potential risk factors for early
   spontaneous pregnancy loss ([55]Li et al., 2018), and supplementation
   with fish oil rich in n-3 PUFA could prevent recurrent miscarriage in
   persistent antiphospholipid syndrome ([56]Rossi and Costa, 1993).
   Therefore, all pregnant mice were administered with soybean oil diet
   for day 0–10 of pregnancy to prevent abortion. In the present study, we
   evaluated the maternal neuroimmune status, explored the transcriptional
   profiles by RNA-sequencing and analyzed the brain fatty acid
   composition by GC and UPLC-MS/MS.

2. Experimental section

2.1. Animals and treatments

   All experiments were conducted in accordance with the Guidelines for
   Care and Use of Laboratory Animals of Qingdao University, with approval
   from the Ethics Committee of Medical College of Qingdao University
   (No.QDU-AEC-2023408). Female C57BL/6JNifdc mice (7-week-old) were
   purchased from Beijing Charles River Laboratory Animal Technology Co.,
   Ltd (Beijing, China, license number: SCXK (Jing) 2021-0006). All mice
   were housed in a room with a temperature at 22 ± 1 °C, humidity at
   50 ± 5%，a 12 h/12 h light/dark cycle (07:00–19:00), and allowed to eat
   diet and drink water freely. After completing 1 week of
   acclimatization, female and male mice (C57BL/6JNifdc mice of
   9-week-old) were caged together. The female vaginal suppository was
   checked the next morning, if available, the day was recorded as day 0.
   Firstly, all pregnant mice were fed with SO diet for consecutive 10
   days, and then consumed the standard diet with the same fat content but
   only different types of fat (15.8% energy from fat), which were
   produced by Jiangsu Xietong Pharmaceutical Bio-engineering Co., Ltd
   (Jiangsu, China). They were randomly assigned to the following three
   groups (n = 10), viz. SO, LS and LO on the day 11 until day 44
   ([57]Fig. S1A). The ingredients of the above 3 standard diets were
   showed in [58]Table 1 and the fatty acid composition of each dietary
   oil were described in [59]Figs. S1B–C. On day 36–42, the maternal mice
   were tested by an open-field test, Y-maze test and Morris water maze
   test to evaluate their cognitive function. Maternal mice were fasted
   overnight and sacrificed after the behavioral test. The brain samples
   were randomly selected and fixed in 4% paraformaldehyde for hematoxylin
   and eosin (H&E), Nissl staining and immunofluorescence assays, and the
   other brains were frozen in liquid nitrogen and stored at −80 °C until
   analysis.

Table 1.

   Nutrition ingredients of standard diet (g/kg).
            Ingredient           SO    LS    LO
   Casein                       200   200   200
   L-Cystine                    3     3     3
   Corn starch                  397   397   397
   Maltodextrin 10              132   132   132
   Sucrose                      100   100   100
   Cellulose                    50    50    50
   Soybean oil                  70    35    70
   Lard oil                     0     35    0
   T-butylhydroquinone          0.014 0.014 0.014
   Mineral mix SA0022M          35    35    35
   Vitamin mix V1OO37           10    10    10
   Choline bitartrate           2.5   2.5   2.5
   Total energy (1000g)         3850  3850  3850
   Energy from protein (%)      20.3  20.3  20.3
   Energy from carbohydrate (%) 63.9  63.9  63.9
   Energy from fat (%)          15.8  15.8  15.8
   [60]Open in a new tab

   SO, soybean oil; LS, mixed oil of lard oil and soybean oil at the ratio
   of 1:1; LO, lard oil.

2.2. Behavioral testing

   Spontaneous activity and anxious state of the maternal mice were
   evaluated by the open-field test. The mice were placed in the bottom
   center of the open box (40 × 40 × 40 cm), allowing them to move freely
   for 5 min, and total distance traveled, the time in center zone and the
   number of times to enter in center zone were analyzed. And then, the
   working memory was evaluated by Y-maze test, which was composed of
   three equally long arms (35 cm × 5 cm × 15 cm). Similar to the open
   field test, the mice were allowed to explore maze freely for 5 min, and
   the total arm entries, alternation triplet (enter three different arms
   in sequence), and alternation triplet (%) were analyzed. Y-maze
   alternation triplet (%) was the evaluation criterion for working memory
   and calculated as follows: Alternation triplet (%) = alternation
   triplet/(the total arm entries-2) × 100%. Finally, the spatial learning
   and memory function were evaluated by Morris water maze test. The
   circular pool used in Morris water maze test was 150 cm in diameter
   filled with water mixed with TiO[2], and the next operational
   procedures were performed based on previous research ([61]Zhang et al.,
   2022). The time of escape latency, entries in platform zone and time in
   quadrant platform zone (%) of visible platform trial, hidden platform
   trial and probe test were analyzed. In the above behavioral tests, the
   movement traces of all maternal mice were recorded by an automated
   tracking system (PACKWIN 2.0.5 software, Panlab, USA).

2.3. H&E and nissl staining

   Brain tissues were placed in 4% paraformaldehyde for more than 24 h,
   embedded in paraffin, and sectioned at 3–5 μm thickness. The slices
   were dewaxed and sequentially stained with hematoxylin and eosin
   (Servicebio, Wuhan, China) to complete the H&E staining or stained with
   toluidine blue (Servicebio, Wuhan, China) to complete Nissl staining.
   The all slices were captured with a NIKON DS-U3 (Tokyo, Japan), the
   number of intact neurons were counted manually counted and the total
   number of neurons were counted automatically by Image J (National
   Institutes of Health, Bethesda, USA) software.

2.4. Enzyme-linked immunosorbent assay

   The brain samples were mixed and ground with precooled saline, and then
   centrifuged at 3000 rpm for 10 min at 4 °C, respectively. According to
   the protocol of commercial enzyme-linked immunosorbent assay (ELISA)
   kit manufacturer, we measured the levels of TNF-α, IL-6, IL-1β and
   IL-18. These ELISA kits were purchased from Jiangsu Jingmei
   Biotechnology Co., Ltd (Jiangsu, China).

2.5. Immunofluorescence staining

   The total brains were collected for immunofluorescent detection of
   glial fibrillary acidic protein (GFAP) and ionized calcium binding
   adapter molecule 1 (IBA1). In brief, deparaffined and rehydrated slices
   were placed in a box filled with EDTA (pH = 8.0, Servicebio, Wuhan,
   China) antigen retrieval solution for antigen retrieval. Next, the
   slices were blocked for 30 min with 3% bovine serum albumin (BSA,
   Servicebio, Wuhan, China) and incubated with the primary antibody
   solutions of GFAP (1:1000, Servicebio, Wuhan, China) and IBA1 (1:500,
   Servicebio, Wuhan, China) at 4 °C overnight. PBS (pH = 7.4, Servicebio,
   Wuhan, China) washing was implemented, followed by incubation with
   secondary antibodies solution for 50 min at room temperature. The
   neurons nucleuses were stained with 4′6-diamidino-2-phenylindole
   (DAPI).

2.6. Western blotting

   In brief, the brain samples were thoroughly homogenized through 60 Hz
   for 1 min by mixing with RIPA lysis buffer (Beyotime, Beijing, China),
   PMSF (Beyotime, Beijing, China), protease inhibitor cocktail (Beyotime,
   Beijing, China) and phosphatase inhibitor cocktail (Beyotime, Beijing,
   China), and then the mixtures were centrifuged at 14000×g for 15 min at
   4 °C. The supernatants were collected and the BCA protein assay kits
   (Beyotime, Beijing, China) were used to quantify brain tissue protein
   concentrations. Equal amount of protein per sample was separated by
   8–12% SDS-PAGE and then transferred onto polyvinylidene fluoride
   membranes (Merck Millipore, Billerica, USA) activated with methanol.
   After being blocking with 5% skim milk for 2 h at room temperature, the
   membranes were washed 3 times with TBST and incubated with the
   following primary antibodies, viz. NOD-like receptor family pyrin
   domain containing 3 (NLRP3, 1:1000, Affinity, USA), Cleaved-Caspase 1
   (1:1000, Affinity, USA), apoptotic speck protein containing a caspase
   recruitment domain, (ASC, 1:1000, Affinity, USA), GFAP (1:1000,
   Affinity, USA), IBA1 (1:1000, Abcam, USA) and Gasdermin D (GSDMD,
   1:1000, Abcam, USA) at 4 °C overnight. The membranes were washed 3
   times with TBST and incubated with specific secondary antibodies for
   2 h at room temperature. The enhanced chemiluminescence (EpiZyme,
   Shanghai, China) was used for protein signal detection and the protein
   expression levels were analyzed Image J software. GAPDH (1:10000,
   Abcam, USA) was used as the protein loading control.

2.7. RNA-sequencing

   According to the instruction manual of the TRlzol Reagent (Life
   technologies, California, USA), the total brain was extracted, measured
   RNA concentration, purified by NanoDrop 2000 (Thermo Fisher Scientific,
   Wilmington, DE), and assessed RNA integrity by RNA Nano 6000 Assay Kit
   of the Agilent Bioanalyzer 2100 system (Agilent Technologies, CA, USA).
   Next, the library for transcriptome sequencing was prepared and
   sequenced on an Illumina NovaSeq platform to generate 150 bp paired-end
   reads. The raw reads generated for the present study were further
   processed and analyzed with the BMK Cloud ([62]www.biocloud.net) online
   platform. Genes between the two groups with an P < 0.05 and fold change
   ≥1.5 found by DESeq2 were assigned as differentially expressed genes.
   Finally, the gene functional annotation of identified differentially
   expressed genes were analyzed based on the Gene Ontology (GO) and Kyoto
   Encyclopedia of Genes and Genomes (KEGG) database. P < 0.05 was the
   decision criteria for GO terms and KEGG pathways enrichment.

2.8. Quantitative real time PCR (QRT-PCR)

   Total RNA of the brain was extracted by Trizol reagent (TaKaRa, Dalian,
   China), and the reverse transcription reaction was performed based on
   the protocol of PrimeScript RT Reagent Kit (Yeasen Biotechnology,
   Shanghai, China). QRT-PCR was completed amplification reactions by
   using the Hieff® qPCR SYBR Green Master Mix kit (Yeasen Biotechnology,
   Shanghai, China) on Quantstudio 1 (Thermo Fisher Scientific, US). The
   sequences of housekeeping gene (β-actin) and target genes, including
   adenylate cyclase 8 (Adcy8), neurotensin receptor 1(Ntsr1), thyrotropin
   releasing hormone receptor (Trhr), oxytocin receptor (Oxtr),
   5-hydroxytryptamine (serotonin) receptor 5B (Htr5b) and
   calcium/calmodulin dependent protein kinase II delta (Camk2d), were
   synthesized by Shanghai Sangon Biotech Co., Ltd and listed in [63]Table
   S1. The conditions of PCR were as follows: 95 °C for 5 min and 40
   cycles of 95 °C for 10 s, followed by 60 °C for 20 s and finally 72 °C
   for 34 s. The relative expression levels of target genes were
   calculated as 2^ΔΔCt = 2^Δ (ΔCt Target gene−ΔCt housing keeping gene).

2.9. Analysis of fatty acid profile in the maternal mice brain

   The extraction and detection of fatty acids in the brain were described
   previously ([64]Zhang et al., 2022). Subsequently, gas chromatography
   (Agilent 7890A, USA) was employed to detect the fatty acids, followed
   by analysis of their relative contents including SFA, MUFA, n-6 PUFA,
   n-3 PUFA, and individual fatty acid monomers. Lipidomic analysis was
   performed by Shanghai Bioprofile Technology Co., Ltd (Shanghai, China).
   Shortly, the entire analysis of the brain samples was performed at 4 °C
   in the autosampler. The samples were separated on an ACQUITY UPLC ® HSS
   C18 (2.1 × 100 mm, 1.9 μm) (Waters, Milford, MA, USA) column using a
   SHIMADZU-LC30 ultra-high performance liquid chromatography (UHPLC)
   system (Shimadzu, Kyoto, Japan). Subsequently, the samples were
   analyzed by Exactive plus mass spectrometer (Thermo Fisher Scientific,
   USA). Lipid identification and quantification as well as data
   processing were performed by MSDAIL (Version 4.0.9). Finally, the
   multidimensional statistical analysis of the data was carried out by
   using R software.

2.10. Statistical analysis

   All data were presented as mean ± standard deviation (SD) or median
   (interquartile range), and analyzed by GraphPad Prism version 9.1.1
   (GraphPad Software, CA, USA). Significant differences between groups
   were analyzed using non-parametric tests or one-way analysis of
   variance (one-way ANOVA) followed by post hoc Tukey test. The figure of
   the potential mechanism was depicted by Figdraw ([65]www.figdraw.com).

3. Results

3.1. Effect of dietary oil on cognitive function in the maternal mice

   The effect of SO, LS and LO on the spontaneous activity and anxious
   state of the maternal mice were described in [66]Fig. 1A–F. The results
   revealed no significant difference in the total distance traveled among
   the group of SO, LS and LO. However, the LO group exhibited
   significantly lower time spent in the center zone and a lesser number
   of entries into the center zone compared to the SO group. As shown in
   [67]Fig. 1G–I, the results of working memory evaluated by Y-maze test
   suggested that the Y-maze alternation triplet and total arm entries in
   the LO and LS groups were significantly lower than in the SO group,
   respectively, however, there was no statistical difference in
   alternation triplet (%) among these groups. The effect of SO, LS and LO
   on the spatial learning and memory function evaluated by Morris water
   maze test ([68]Fig. 1J-L) indicated that no significant differences
   were observed in the time of escape latency during total Morris water
   maze test period, including visible platform trial, hidden platform
   trial and probe test. And also, there were no significant effect on the
   entries in platform zone and time in quadrant platform zone (%) in
   probe test.

Fig. 1.

   [69]Fig. 1
   [70]Open in a new tab

   Effect of dietary oils on cognitive function in the maternal mice. (A)
   The distance traveled (cm), (B) time in zone (seconds)-center, (C)
   entries in zone-center, the representative traveled paths of the (D) SO
   group, (E) LS group and (F) LO group mice in the open-field test. (G)
   The Y-maze alternation triplet, (H) total arm entries, and (I)
   alternation triplet (%) in Y-maze test. (J) The escape latency (s) of
   visible platform trial, hidden platform trial and probe test, (K)
   entries in zone-platform, and (L) time in zone (%)-quadrant platform of
   probe test in Morris water maze test. Data are presented as mean ± SD
   (n = 8). * (P < 0.05) and ** (P < 0.01) indicate the significant
   difference between groups. SO, soybean oil; LS, mixed oil of lard oil
   and soybean oil at the ratio of 1:1; LO, lard oil.

3.2. Effect of dietary oil on histopathology and the levels of inflammatory
cytokines in the maternal mice brain

   H&E and Nissl staining were determined for neuronal damage in the
   cortex and hippocampus of maternal mice. H&E staining indicated that
   neurons presented nuclei pyknosis in the cortex of LO group ([71]Fig.
   2A–B), indeed, the percent of normal neurons in the cortex of LO group
   was significantly lower compared with the SO group (white arrow), but
   no significant differences were observed in the percent of normal
   neurons in hippocampal circuit regions (CA1, CA3) and dentate gyrus
   (DG) between these groups. As shown in [72]Fig. 2C–D, neuron Nissl
   bodies in the cortex of LO group was disintegrated and even disappeared
   (red arrow), and the percent of normal neurons in the cortex of the LO
   and LS groups was significantly lower than in the SO group detected by
   Nissl staining, respectively. The hippocampus (CA1, CA3, and DG) of
   maternal mice between these groups were consistent with the results of
   H&E staining. We also detected inflammatory cytokines in the brain
   ([73]Fig. 2E–H), and the results suggested that the levels of IL-6 and
   IL-18 in the LO group were significantly higher than in the SO group.
   In addition, the level of TNF-α in the LO group was significantly
   higher than in the SO and LS groups, respectively; the level of IL-1β
   in the LO and LS groups was significantly higher than in the SO group,
   respectively. And it presented a certain dose-response relationship.

Fig. 2.

   [74]Fig. 2
   [75]Open in a new tab

   Effect of dietary oils on histopathology and the levels of inflammatory
   cytokines in the maternal mice brain. (A) The H&E staining diagrams of
   cortex and hippocampus in maternal mice (scale bar = 20 μm). (B) The
   percent of normal neurons (%) per field in maternal mice determined by
   the H&E staining. (C) The Nissl staining diagrams of cortex and
   hippocampus in maternal mice (scale bar = 20 μm). (D) The percent of
   normal neurons (%) per field in maternal mice determined by the Nissl
   staining. (E) The levels of TNF-α, (F) IL-6, (G) IL-1β and (H) IL-18 in
   maternal mice brain (pg/g). Data are presented as mean ± SD (n = 3). *
   (P < 0.05) and ** (P < 0.01) indicate the significant difference
   between groups. SO, soybean oil; LS, mixed oil of lard oil and soybean
   oil at the ratio of 1:1; LO, lard oil.

3.3. Effect of dietary oil on the activation of neuroglial cells in the
maternal mice brain

   Based on the histopathology and the levels of inflammatory cytokines,
   activation of astrocyte and microglia in maternal mice brain were
   evaluated through GFAP-positive cells and IBA1-positive cells, which
   were determined by immunofluorescence ([76]Fig. 3A and C). The
   representative immunofluorescent images indicated that the astrocyte
   and microglia in the cortex of LO group were more active than in the
   groups of SO and LS, and no obvious differences were observed in the
   protein expression of GFAP in CA1, CA3 and DG. Indeed, the western
   blotting results demonstrated that the protein expression of GFAP in
   the brain of LO group was significantly higher than in the groups of SO
   and LS ([77]Fig. 3B), respectively. Similarly, the western blotting
   results shown that the protein expression of IBA1 in the brain of LO
   group was significantly higher than in SO group ([78]Fig. 3D).

Fig. 3.

   [79]Fig. 3
   [80]Open in a new tab

   Effect of dietary oils on the activation of neuroglial cells in the
   maternal mice brain. (A) The GFAP expression of cortex and hippocampus
   determined by immunofluorescence (scale bar = 20 μm). (B) The GFAP
   expression of brain determined by western blotting as SO, LO, and LS
   (n = 3). (C) The IBA1 expression of cortex and hippocampus determined
   by immunofluorescence (scale bar = 20 μm). (D) The IBA1 expression of
   brain determined by western blotting as SO, LO, and LS (n = 3). Data
   are presented as mean ± SD. * (P < 0.05) and ** (P < 0.01) indicate the
   significant difference between groups. SO, soybean oil; LS, mixed oil
   of lard oil and soybean oil at the ratio of 1:1; LO, lard oil.

3.4. Effect of dietary oil on the activation of NLRP3 inflammasome complex
and pyroptosis in the maternal mice brain

   The relative expression of the NLRP3 inflammasome complex related
   proteins consisting of the NLRP3, ASC, and Cleaved-Caspase 1 in the
   brain were detected by western blotting ([81]Fig. 4B–D) and the results
   indicated that the relative proteins expression of NLRP3 and ASC in the
   LO group were significantly higher compared with the groups of SO and
   LS, and Cleaved-Caspase 1 in the LO group was only significantly higher
   than in the SO group. The interaction of NLRP3 inflammasome complex
   related proteins were closely associated with the activation of
   pyroptosis, and we observed that the relative proteins expression of
   GSDMD and GSDMD-N in the LO group was significantly higher than in the
   SO and LS groups, respectively ([82]Fig. 4E–F).

Fig. 4.

   [83]Fig. 4
   [84]Open in a new tab

   Effect of dietary oils on the expression of NLRP3 inflammasome complex
   and pyroptosis related proteins in the maternal mice brain. (A) The
   representative bands of NLRP3 inflammasome complex and pyroptosis
   related proteins detected by western blotting as SO, LO, and LS. The
   relative proteins levels of (B) NLRP3, (C) ASC, (D) Cleaved-Caspase 1,
   (E) GSDMD, and (F) GSDMD-N. Data are presented as mean ± SD (n = 3). *
   (P < 0.05) and ** (P < 0.01) indicate the significant difference
   between groups. SO, soybean oil; LS, mixed oil of lard oil and soybean
   oil at the ratio of 1:1; LO, lard oil.

3.5. Effect of dietary oil on the expression of differentially expressed
genes and validation of genes related to calcium signaling pathway in the
maternal mice brain

   Summary of the RNA-sequencing raw date results of the maternal mice
   brain were described in [85]Table S2. Compared with the LO group, there
   were 46 up-regulated genes and 200 down-regulated genes in the LS
   group, in addition, there were 108 up-regulated genes and 124
   down-regulated genes in the SO group ([86]Figs. S2A–B). Interestingly,
   we found that the shared differential genes were 34 in both LO vs LS
   and LO vs SO ([87]Fig. S2C). The biological function of differentially
   expressed genes was explored through KEGG pathway enrichment analysis
   and GO functional analysis. The results of KEGG pathway enrichment
   analysis suggested that the most significantly enriched pathways of
   differentially expressed genes in LO vs LS included neuroactive
   ligand-receptor interaction, cAMP signaling pathway, cocaine addiction,
   calcium signaling pathway and so on, and the most significantly
   enriched pathways of differentially expressed genes in LO vs SO
   included neuroactive ligand-receptor interaction, calcium signaling
   pathway, pathway in cancer and proteoglycans in cancer and so on
   ([88]Figs. S3A–B). GO functional analysis also showed that the
   differentially expressed genes in LO vs LS and LO vs SO both involved
   the molecular function of calcium ion binding ([89]Figs. S3C–D).
   Therefore, a total of 6 shared differentially expressed genes (Adcy8,
   Ntsr1, Trhr, Oxtr, Htr5b and Camk2d) were screened out from the calcium
   signaling pathway. Our transcriptome sequencing results ([90]Fig. 5A)
   were consistent with those observed in the RNA sequencing results
   ([91]Fig. 5B). Compared with the SO group, the relative mRNA expression
   levels of Adcy8, Ntsr1, Trhr, Htr5b and Camk2d were significantly
   higher in the LO group, in addition, only the relative mRNA expression
   level of Oxtr was significantly higher in the LO group than in the
   groups of SO and LS, respectively.

Fig. 5.

   [92]Fig. 5
   [93]Open in a new tab

   The genes expression related to calcium signaling pathway in the
   maternal mice brain. (A) The relative FPKM level of genes detected by
   RNA-sequencing. (B) The relative mRNA expression levels of genes
   detected by quantitative real time PCR. Data are presented as mean ± SD
   (n = 3). * (P < 0.05) and ** (P < 0.01) indicate the significant
   difference compared with LO group. SO, soybean oil; LS, mixed oil of
   lard oil and soybean oil at the ratio of 1:1; LO, lard oil.

3.6. Effect of dietary oil on fatty acid profile in the maternal mice brain

   The results presented in [94]Fig. 6A–C and [95]Table S3 demonstrated
   that the proportion of SFA in brain was significantly higher in the LO
   group compared with the SO group, and the proportion of n-3 PUFA in
   brain of the LO group was significantly lower than in the SO and LS
   groups, respectively. On the contrary, the proportion of n-6 PUFA in
   the LO group was significantly higher than in the LS group. There was
   no significant effect of different kinds of dietary oil on MUFA
   proportion of maternal mice. No significant differences were also
   observed in the proportion of C18:0, C18:1, C18:3, C20:4 and EPA among
   these groups. However, the proportion of DHA and the ratio of n-3/n-6
   PUFA were significantly lower in the LO group compared with the SO and
   LS groups, respectively, and the proportion of C16:0 and C18:2 was
   significantly higher in the LO group than in the LS group. What's more,
   we observed a significant positive correlation between the proportion
   of n-3 PUFA, DHA, and n-3/n-6 PUFA with the duration spent in the
   center and number of entries into the center during the open-field test
   ([96]Fig. 6D–K). Notably, only the proportion of SFA exhibited a
   significantly negative correlation with both time spent in the center
   and entries into the center during the open-field test. It is worth
   mentioning that there was no significant correlation found between the
   proportion of C16:0, C18:2, n-6 PUFA and time spent in the center or
   entries into the center ([97]Fig. S4).

Fig. 6.

   [98]Fig. 6
   [99]Open in a new tab

   Effect of dietary oils on fatty acid profile in the brain and its
   correlation with open-field test data. (A) The proportion of fatty acid
   with different saturations (%), (B) the proportion of fatty acid
   monomers (%) and (C) the ratio of n-3/n-6 PUFA in the maternal mice.
   Person's correlation of (D) SFA (% of total fatty acids), (E) DHA (% of
   total fatty acids), (F) n-3 PUFA (% of total fatty acids), and (G) the
   ratio of n-3/n-6 PUFA with time in center (seconds) of the open-field
   test. Person's correlation of (H) SFA (% of total fatty acids), (I) DHA
   (% of total fatty acids), (J) n-3 PUFA (% of total fatty acids) and (K)
   the ratio of n-3/n-6 PUFA with entries in center (times) of the
   open-field test. Data are presented as mean ± SD (n = 6). * (P < 0.05)
   and ** (P < 0.01) indicate the significant difference between groups.
   SO, soybean oil; LS, mixed oil of lard oil and soybean oil at the ratio
   of 1:1; LO, lard oil.

   Based on the effect of dietary oils on brain fatty acid in the maternal
   mice and its correlation with open-field test data, we conducted
   lipidomics to explore more nuanced alterations in the lipid metabolism
   of the maternal mice brain. The principal component analysis (PCA) of
   lipid molecules in the maternal mice brain revealed that the quality
   control samples exhibited strong aggregation, indicating the
   instrument's stability throughout sample detection and analysis
   ([100]Fig. S5A). According to the classification of the International
   Lipid Classification and Nomenclature Committee, lipid molecules were
   divided into various lipid categories and lipid subclasses. In our
   study, we identified a total of 67 lipid subclasses and 1380 lipid
   molecules, with Ether-linked phosphatidylcholine (EtherPC),
   Ether-linked phosphatidylethanolamine (EtherPE), and several other
   lipid subclasses exhibiting relatively high abundance ([101]Fig. S5B).
   Notably, Phosphatidylcholine (PC) exhibited the highest abundance
   across all three groups ([102]Fig. S5C). The identified lipid molecules
   for each comparative analysis group were visualized using volcano plot
   ([103]Figs. S6A–B) and lipid category bubble chart ([104]Figs. S6C–D),
   and the results indicated that the differences in some lipid molecules
   of glycerol phospholipids (GP) between groups were found to be the most
   pronounced. The samples from different groups were successfully
   classified into distinct clusters using orthogonal partial least
   squares-discriminant analysis (OPLS-DA) for LO vs SO and LO vs LS
   comparisons ([105]Figs. S7A and D), indicating a pronounced
   differentiation of lipid molecules in the brain of maternal mice. The
   following screening parameters were employed to identify differentially
   expressed lipid molecules in our study: 1) variable importance for
   projection (VIP) score >1 in the OPLS-DA model; 2) fold change >1.5
   or < 0.667; and 3) p-value <0.05. Compared with the LO group, there
   were 29 up-regulated lipid molecules and 32 down-regulated lipid
   molecules in the LS group, in addition, there were 37 up-regulated
   lipid molecules and 37 down-regulated lipid molecules in the SO group,
   and the hierarchical clustering analysis of all differentially
   expressed lipid molecules were presented in [106]Fig. S8.
   Interestingly, the lipid category exhibiting the highest abundance of
   differentially expressed lipid molecules corresponded to GP ([107]Fig.
   S7B-C and E-F). The biological function of differentially expressed
   lipid molecules was investigated using KEGG pathway enrichment
   analysis, revealing that glycerophospholipid metabolism exhibited the
   most significant enrichment in both LO vs SO and LO vs LS comparisons
   ([108]Fig. S9).

3.7. Correlations between lipid molecules of glycerophospholipid metabolism
and genes of calcium signaling pathway in the maternal mice brain

   Based on the aforementioned findings, it was evident that there were
   discernible alterations in both transcriptional and lipid profiles
   between the brain of different groups. Therefore, a network analysis
   was conducted to assess the potential association between
   differentially expressed lipid molecules and differentially expressed
   genes associated with the calcium signaling pathway in LO vs SO and LO
   vs LS comparisons ([109]Fig. 7A–B), and the results suggested that a
   correlation existed between the calcium signaling pathway and
   glycerophospholipid metabolism. Further, a Pearson's correlation
   analysis was employed to evaluate the potential correlation between
   differentially expressed lipid molecules involved in
   glycerophospholipid metabolism and differentially expressed genes
   related to calcium signaling pathway in LO vs SO and LO vs LS
   comparisons ([110]Fig. 8). Mainly the genes of Htr5b and Oxtr were
   significantly correlated with differentially expressed lipid molecules
   in both LO vs SO and LO vs LS comparisons. Interestingly, we observed
   that the shared differentially expressed lipid molecules were 19 in LO
   vs LS and LO vs SO comparisons, mainly including the identified
   subclasses of Phosphatidylglycerol (PG) and Cardiolipin (CL), followed
   by Lysophosphatidylethanolamine (LPE) and Phosphatidylethanolamine
   (PE). Meanwhile, the number of lipid molecules exhibiting significant
   correlations with the genes of Htr5b and Oxtr was found to be higher in
   LO vs SO than in LO vs LS.

Fig. 7.

   [111]Fig. 7
   [112]Open in a new tab

   Network analysis between significantly differentially expressed lipid
   molecules and differentially expressed genes of calcium signaling
   pathway in (A) LO vs SO and (B) LO vs LS. SO, soybean oil; LS, mixed
   oil of lard oil and soybean oil at the ratio of 1:1; LO, lard oil.

Fig. 8.

   [113]Fig. 8
   [114]Open in a new tab

   Pearson correlation analysis between significantly differentially
   expressed lipid molecules of glycerophospholipid metabolism and
   differentially expressed genes of calcium signaling pathway in (A) LO
   vs SO and (B) LO vs LS. * (P < 0.05), ** (P < 0.01) and *** (P < 0.001)
   indicate the significant difference between groups.

4. Discussion

   In this study, we have demonstrated for the first time that the effect
   of lard intake during pregnancy and postpartum periods on cognitive
   function of the maternal mice and elaborated on its potential
   mechanism. We observed that the administration of lard resulted in
   cognitive impairment as detected by the open-field test, through
   altering the brain fatty acid profile and activating neuroinflammation
   via calcium signaling pathway in brain and when compared with soybean
   oil. The present study had one limitation, we did not record baseline
   data, including food intake and litter size, because we observed that
   repetitive procedures such as weighing induce stress in the pregnant
   mice during our previous study, leading to miscarriages or postpartum
   stress, and even causing them to cannibalize their pups.

   Neuroinflammation plays a role in the postpartum cognitive impairment,
   which characterized by loss of normal neurons, increased expression of
   inflammatory cytokines, and activated glial cells and inflammasomes in
   brain ([115]Wang et al., 2022; [116]Zhang et al., 2016). The present
   study revealed that the percentage of normal neurons in the LO group
   were significantly lower than in the SO group, and the inflammatory
   cytokines levels were opposite. The persistent neuroinflammation
   responses involve the activation of microglia and astrocytes, which
   play crucial roles in regulating inflammatory responses within the
   central nervous system ([117]Wang et al., 2021). Consistent with our
   results, a previous study indicated that substantial morphological and
   functional changes have occurred in glia cells during the pregnancy and
   postpartum periods, which were associated with postpartum anxiety and
   highly sensitive to dietary oil changes ([118]Leyrolle et al., 2021).
   In order to further explore the mechanisms of neuroglial cells
   activation in the maternal mice, we found that the expression levels of
   NLRP3 inflammasomes related proteins in brain of the LO group were
   significantly higher than in the SO and LS groups, respectively. The
   NLRP3 inflammasomes is known to respond to microbial infection,
   environmental stimuli and so on, and its administration has been
   proposed as a promising therapeutic target for preventing cognitive
   impairment ([119]Huang et al., 2021). It has been reported that fish
   oil attenuated postpartum depression via inhibition of NLRP3
   inflammasomes driven inflammatory pathway in a rat model ([120]Aziz et
   al., 2021). Similarly, another study suggested that the supplementation
   of n-3 fatty acids in pregnancy period of rats alleviated
   neuroinflammation with a decrease in protein levels of NLRP3 in certain
   brain regions ([121]Tang et al., 2018a). When the NLRP3 inflammasomes
   are activated, caspase 1 will also be activated through self-cleavage,
   which induce the maturation of the inflammatory cytokines including
   IL-1β and IL-18 ([122]Lei et al., 2018). Besides, Cleaved-Caspase 1 can
   cleave GSDMD and release its N-terminal domain, and then GSDMD-N forms
   pores on the cell membrane, which mediates the release of IL-1β, IL-18
   and other cellular contents, and induces the inflammatory cell death
   known as pyroptosis ([123]Shi et al., 2015). In the present study, we
   observed a more severe pyroptosis phenotype in brain of the LO group
   compared to SO and LS groups. It has been previously reported that
   dietary fatty acids can modulate pyroptosis through regulation of NLRP3
   inflammasomes via the TLR4/NF-κB signaling pathway, which is consistent
   with our findings ([124]Jin et al., 2022).

   We conducted RNA-sequencing to further elucidate the underlying
   molecular mechanisms responsible for the upregulation of NLRP3
   inflammasome expression in the maternal mice, and we proposed that the
   calcium signaling pathway played a pivotal role in its activation in
   this study. Ca^2+ serves as a crucial secondary messenger, and its
   dysregulation has been implicated in the pathophysiology of cognitive
   dysfunction ([125]Clapham, 2007), and a previous review has elaborated
   that the activation of NLRP3 inflammasomes can be regulated by enhanced
   Ca^2+ signaling ([126]Kong et al., 2021). Compared to the SO group, the
   maternal mice in the LO group exhibited significantly elevated relative
   mRNA expression levels of Ntsr1, Trhr, Htr5b and Oxtr genes in the
   present study. The actions of neurotensin are mediated through its
   binding to two G-protein-coupled receptors, namely Ntsr1 and Ntsr2,
   which exhibit widespread expression in the brain ([127]Schroeder and
   Leinninger, 2018), and the activation of Ntsr1 could induce
   intracellular Ca^2+ release in neurons ([128]Tabarean, 2020). The
   thyrotropin-releasing hormone binds to the cell-surface TRH receptor
   (Trhr) and triggers the activation of G proteins Gq/G11, thereby
   initiating phospholipase C activation and elevation in intracellular
   free Ca^2+ level ([129]Mulla et al., 2009). The Oxtr is also
   characterized as a G protein-coupled receptor, thereby being coupled to
   a trimeric complex of G proteins comprising one Gα and one β/γ unit
   ([130]Jurek and Neumann, 2018), and previous studies have demonstrated
   the induction of Ca^2+ release from intracellular stores by Oxtr
   ([131]Prole and Taylor, 2016; [132]Saleem et al., 2013). The rodent
   5-HT5 receptor family comprises two receptors, namely 5-HT 5A and 5B
   (Htr5a and Htr5b), while human Htr5b is believed to be non-functional
   due to the presence of stop codons in its first exon. Consequently,
   only a limited number of studies have been conducted thus far to
   investigate the role of Htr5b in the central nervous system
   ([133]Maekawa et al., 2010). A previous study demonstrated that
   overexpression of Htr5b in brain neurons of mice with cognitive
   impairment was sufficient to ameliorate anxiety-like behaviors and
   spatial memory deficits by rescuing the decreased intracellular Ca^2+
   level ([134]Tang et al., 2020). According to the above research
   results, the genes of Htr5b, Ntsr1, Trhr and Oxtr can cause the
   increase of Ca^2+ concentration, which may account for the upregulated
   expression levels of these genes in activating the NLRP3 inflammasomes.
   Moreover, we found that the relative mRNA expression levels of Adcy8
   and Camk2d genes in brain were significantly higher in the LO group
   than in the SO group. Cyclic AMP acts as an additional second messenger
   to control Ca^2+ homeostasis, with adenylyl cyclases (Adcys) serving as
   its synthetic sources. The activity of Adcys has the potential to be
   activated by Ca^2+, with Adcy8 content increasing in response to
   elevated level of Ca^2+ ([135]Halls and Cooper, 2011).
   Calcium/calmodulin dependent protein kinase II is also involved in the
   regulation of Ca^2+ homeostasis, with Camk2d identified as its
   predominant isoform that can be activated upon binding of calmodulin to
   local influx of Ca^2+ ([136]Dalal et al., 2021). In other words, there
   exists a positive correlation between the expression levels of Camk2d
   and intracellular Ca^2+ content.

   Extensive research has indicated the consumption of different lipids
   and fatty acids could alter the composition of brain fatty acids, which
   play a crucial role in both brain development and cognitive behavior
   ([137]Fan et al., 2022; [138]Mizunoya et al., 2013). In our study, we
   demonstrated that the relative contents of SFA and n-6 PUFA in brain of
   the LO group were significantly higher than in the SO and LS groups,
   respectively. These findings were consistent with a previous research
   that the SFA and n-6 PUFA concentrations of brain were significantly
   higher in the n-3 PUFA deficiency group than in the n-3 PUFA balanced
   group and hence induced worse spatial memory in the adult offspring
   ([139]Labrousse et al., 2018). But no significant correlation was found
   between n-6 PUFA in brain and the behavioral outcomes in our study.
   Indeed, the findings regarding the association between n-6 PUFA and
   cognitive function have generated conflicting results. A previous study
   reported no significant differences in the association between n-6 PUFA
   and major depression disorder ([140]Thesing et al., 2018). Therefore,
   the effect of n-6 PUFA on cognitive function need to be further
   investigated. Additionally, we discovered a significant positive
   correlation between the levels of DHA, n-3 PUFA, and the ratio of
   n-3/n-6 PUFA in the brain and the outcomes of the open-field test. This
   is in line with a previous research showing that the mice brain fatty
   acid proportions of n-3 PUFA, DHA and the ratio of n-3/n-6 PUFA in
   peony seed oil and fish oil groups rich in n-3 PUFA was significantly
   higher than the control and model groups deficient in n-3 PUFA
   ([141]Zhang et al., 2022). Regrettably, we showed that no significant
   difference on time in center and entries in center detected by the
   open-field test between the LO group and the LS group. Based on the
   function related to inflammation of fatty acids, we speculate that it
   may be related to accounting for a vast proportion of the
   pro-inflammatory fatty acids, namely SFA in brain have no significant
   difference between the LO group and the LS group.

   Lipidomics analysis enables efficient investigation of alterations and
   functionalities within lipid category and molecule across diverse
   biological processes, thereby elucidating pertinent mechanisms
   underlying biological activities ([142]Züllig and Köfeler, 2021). In
   our results, we found that the GPs in brain exhibited the highest
   abundance of differentially expressed lipid molecules and the
   glycerophospholipid metabolism exhibited the most significant
   enrichment of KEGG pathway in both LO vs SO and LO vs LS comparisons.
   These findings suggest a significant involvement of glycerol
   phospholipids in cognitive function within the scope of our study. In
   addition to structural integrity role to neural membranes, GPs are
   integral components of the signal transduction network that mediates
   extracellular signals from the cell surface to the nucleus, thereby
   eliciting a biological response at the genetic level ([143]Frisardi et
   al., 2011). Some degradation products of GPs could exhibit
   pro-inflammatory properties, triggering the activation of astrocytes
   and microglia in the brain and subsequent release of inflammatory
   cytokines. And these cytokines could perpetuate and intensify oxidative
   stress and neuroinflammation ([144]Farooqui et al., 2000;
   [145]Stephenson et al., 1999). In our study, the levels of PG 18:2_18:2
   and PG 20:5_22:6 in brain were significantly elevated, exhibiting a
   negative correlation with Htr5b in LO vs SO. Evidence from the present
   study suggested the idea that PG belonging to GPs exerted
   anti-inflammatory effects. PGs were discovered in Scenedesmus,
   comprising a glycerol backbone-linked glycerol headgroup and two fatty
   acyl chains ([146]Benson and Maruo, 1958). Consistent with our
   findings, previous studies have reported that the presence of PG
   competes with lipopolysaccharides in disrupting the activation of the
   TLR pathway, thereby inhibiting the formation of downstream
   inflammatory molecules ([147]Kandasamy et al., 2011; [148]Kuronuma et
   al., 2009). Additionally, the lipid molecule of CL 12:0_16:0_22:3_22:5
   levels in brain was significantly decreased, exhibiting a positive
   correlation with Htr5b in LO vs SO. The findings of this study
   indicated that the pro-inflammatory properties of CLs may exert a
   significant influence on cognitive impairment in maternal mice. CLs are
   tetra-acylated diphosphatidylglycerol lipids, which exhibit specific
   localization on the mitochondria and constitute approximately 15–20% of
   the total mitochondrial lipid content ([149]Chen et al., 2018).
   Cardiolipin can enhance microglial phagocytosis, regulate the secretion
   of inflammatory mediators, and promote neuroprotective effects mediated
   by microglia ([150]Pointer et al., 2019). Furthermore, it has been
   reported that CLs could bind to NLRP3 located in the mitochondria and
   trigger the activation of the NLRP3 inflammasome, leading to a
   neuroinflammatory response ([151]Liu et al., 2018). Although changes in
   the lipid profiles of the maternal mice brain were observed through
   lipidomics, the determination of the specific fatty acid that
   predominantly contributes to cognitive function in the maternal mice
   proved challenging in our study and further investigations are needed.

5. Conclusion

   The present results demonstrated for the first time that the
   consumption of lard during the pregnancy and postpartum periods has a
   negative impact on the cognitive function of the maternal mice compared
   with soybean oil. The mechanism may be related to altering fatty acid
   composition of the brain, affecting the calcium signaling pathway, and
   then activating neuroinflammation in brain.

6. Date availability statement

   The datasets presented in this study can be found in online
   repositories. Transcriptome data at the National Center for
   Biotechnical Information (NCBI) (SRA ID: PRJNA1030946), Lipidomic data
   at the MetaboLights (MTBL ID: MTBLS9017).

CRediT authorship contribution statement

   Runjia Shi: Writing – original draft, Methodology, Investigation,
   Funding acquisition, Data curation, Conceptualization. Xiaoying Tian:
   Writing – original draft, Methodology. Tianyu Zhang: Investigation,
   Data curation. Andong Ji: Resources. Huina Xu: Resources, Methodology.
   Zhongshi Qi: Resources, Investigation. Chunhui Zhao: Visualization,
   Validation. Duo Li: Project administration.

Declaration of competing interest

   The authors declare that they have no known competing financial
   interests or personal relationships that could have appeared to
   influence the work reported in this paper.

Acknowledgements