Subfertility of Offspring and Transgenerational Transmission of Splenic Transcriptome Induced by Lipopolysaccharide in Hens

Background: Transgenerational phenotypes are linked to genetic background, environmental factors, and diseases. Maternal stimulation is one of the suspected contributing factors to transgenerational phenotypes, and recent evidence from human studies suggests that maternal immune stimulation affects the phenotype and gene expression of the offspring, but few similar studies were reported in poultry. Here, we used laying hens as a model organism to investigate the effects of maternal stimulation on hens' immunity and reproductive performance, as well as the regulatory role of the splenic transcriptome in hens induced by lipopolysaccharide (LPS). Methods: To investigate the effect of maternal stimulation on egg-laying rate of hens and their offspring, laying hens were rst intravenously injected with LPS. RNA-seq from the spleens of F0, F1, and F2 generations of hens was used to identify differentially expressed genes (DEGs) between the LPS group and controls. We investigated the effect of LPS maternal stimulation on the transcriptome of laying hens and its multigenerational transmission effect by analyzing shared genes, enrichment pathways, and protein-protein interaction networks across generations. We concentrated on immune and reproduction-related pathways and genes that existed across generations. Subsequently, we performed a correlation analysis between the DEGs associated with the multi-generational transmission effect and egg-laying rate of hens. We The study

research into the multi-generational transmission of maternal stimulation, and the selected marker genes are of great signi cance to the breeding of laying hens. Background Prenatal exposure to infectious or in ammatory environmental stimulating factors can cause mental illnesses in offspring, such as schizophrenia, autism and bipolar psychosis [1], as well as retarded growth and low fertility in offspring [2][3][4][5]. These environmental exogenous stimulus are being referred to as maternal stimulants. Lipopolysaccharide (LPS) is the main component of the outer membrane in gramnegative bacteria and can be used to simulate bacteria. It is a common immune-related maternal stimulant, which can be used to study the impact of maternal stimulation on animals and their offspring [6,7]. LPS was reported to cause a strong immune response in animals, thereby causing immune stress.
It can bind to the TLR4, CD14 and MD2 receptors of many types of cells, such as macrophages, dendritic cells, monocytes and B cells, and further promote the secretion of in ammatory cytokines, NO and eicosanoids [8]. Therefore, LPS stimulation could affect the immune system of laying hens and their offspring.
Many studies have shown that maternal immunostimulation could affect the reproductive function of animals [9,10], and these effects can be transmitted across generations [11]. Yet, the impact of maternal stimulation on the reproductive performance of livestock and poultry offspring is rarely reported. The current environmental impact on livestock and poultry is large and there are many pathogenic microorganisms such as bacteria and viruses, which easily affect the immune systems of livestock and poultry, resulting in the decline of the animals' feed intake, production e ciency and reproductive performance. These adverse effects may also be passed on to the offspring, bringing huge economic losses to animal husbandry. Since the bacterial mimic LPS may affect the immune system, we presumed that the LPS could affect the immune system and reproductive performance of laying hens and their offspring.
Spleen is a vital immune organ in laying hens, and transcriptomes allow inferring gene expression patterns and phenotypic transmission. In this study, we investigated the effect of maternal stimulation on layers immunity and reproduction and the underlying mechanism regulating the changes in the splenic transcriptome across the F0, F1 and F2 generations. It was found that the maternal stimulation by LPS caused a decrease in the immunity and egg-laying rate of laying hens and their unchallenged offspring.
This effect may be regulated by the intergenerational transmission effect of the transcriptome. Our results will provide important reference materials for animal breeding, future multigenerational transmission of maternal immune stimulation and immune stress-related research.

Ethics Statement
All the following animal works were performed in strict accordance with guidelines and regulations established by the Animal Welfare Committee of China Agricultural University in Beijing, China (permit number: DK996). All the procedures involving animals have been previously approved by the committee.
We made all efforts to minimize the hen's suffering during the entire experiment.
Animal and LPS challenge A total of 58 Rhode Island White hens were randomly selected from the nuclear brood of a poultry breeding farm. Hens were then divided into two groups, which were separately challenged with a singledose intravenous injection of lipopolysaccharide (LPS, 0.2 mg/kg; Escherichia coli; MilliporeSigma, Burlington, MA, USA) (group L, n = 29) or equivalent sterile saline (group C, n = 29) under the wing of the hens at 53 weeks of age. Arti cial insemination was performed and generated 60 LPS-exposed (L) and 75 control (saline, C) F1 hens. The hens were reared in wire cages and had free access to water and feed.
There was no LPS challenge in F1 and F2 hens. The overall framework of this study was shown in Fig. 1.

Egg-laying rate testing
Eggs were collected and recorded for two weeks before and three weeks post the LPS challenge (52-56 weeks of age) in the F0 generation. In the F1 generation, the eggs collection was performed from 21-66 weeks of age. First of all, we calculated the total egg-laying rate before (two weeks) and after (three weeks) the maternal stimulation and studied its changes in the LPS group. The difference in egg-laying rate between LPS group and the controls was examined. Furthermore, we investigated the effect of maternal stimulation on the daily egg-laying rate of the chicken ock in different conditions in both F0 and F1 generations. Finally, the daily egg-laying rate of the F1 generation was divided into early, middle and late laying periods for further study.

Sampling
For tissue collection, six hens (group L, n = 3; group C, n = 3) were randomly sampled at 24 h postinjection at 53 weeks of age in the F0 generation. Spleens were immediately snap-frozen in liquid nitrogen and then stored at -80 °C for RNA extraction. To investigate the transgenerational transmission effect of transcriptome resulted from the maternal stimuli, we also sampled six spleens (group L, n = 3; group C, n = 3) with the same operation to perform RNA-seq at 53 weeks of age in the F1 generation and 1 day of age in the F2 generation, respectively.
The total RNA of chicken spleens were isolated using Trizol (Thermo Fisher Scienti c, Waltham, MA, USA) following the manufacturer's protocol after grinding the frozen spleen sample into powder in the liquid nitrogen environment. The quality of the RNA sample was assessed by 1% agarose gel electrophoresis and quanti ed using Qubit RNA Assay Kit (Thermo Fisher Scienti c, Waltham, MA, USA) and NanoPhotometer spectrophotometer (Implen, Germany). Bioanalyzer 2100 System (Agilent Technologies, Santa Clara, CA, USA) was used to monitor the RNA integrity number (RIN, a score from 0 to 10). All RNA samples were of high quality (RIN > 7, OD260/280 > 1.96, Table S1), and 2 μg of total RNA for each sample were used to construct cDNA libraries using NEBNext ® Ultra ™ RNA library Prep Kit for Illumina ® (NEB, Ipswich, MA, USA). At last, the cDNA libraries were clustered with TruSeq Rapid PE Cluster Kit (Illumina, USA) and sequenced on an Illumina HiSeq 2500 platform (Illumina, USA) and 125 paired-end reads were generated.
Sequencing data quality control and differentially expressed genes (DEGs) identi cation FastQC v.0.11.5 was used to assess the quality of the raw data. To get the high-quality data, Trimmomatic-0-2.38 [12] was used to trim the raw data with the default parameters. In detail, the clean reads were obtained by removed adapters, leading and trailing low quality, N bases, reads that below the 36 bases long and reads with the average quality below 15. Before the downstream analyses, FastQC v.0.11.5 was used again to examine the quality of clean data. The GRCg6a reference genome (FASTA format) and gene annotation les (GTF format) were downloaded from the Ensembl database (ftp://ftp.ensembl.org/pub/release-99/). The clean data was aligned to the reference genome using HISAT2 [13] with basic options. The output les of HISAT2 were sam format. Then Samtools was used to convert the sam format to bam format. Cuffdiff v2.2.1 [14] was used to compute the DEGs with the bam les between group L vs. group C (L vs. C) in F0, F1 and F2 generations. The DEGs were ltered with criteria of P < 0.05 and fold change (FC) > 1.5.

Functional enrichment and annotation
The DEGs in F0 and F1 generations were used to annotate biological functions, respectively. Gene ontology (GO) term of Biological Process analysis predicted with DEGs was conducted using the WEBbased Gene SeT AnaLysis Toolkit (WebGestalt, http://www.webgestalt.org) [15].

Statistical analyses
Signi cance chi-square test was used to analyze the difference in egg-laying rate. In RT-qPCR, 2 -ΔΔCT was used as the relative quantitative value, and the signi cant difference between different conditions was tested by the student's t-test. In order to evaluate the consistency of gene expression level between RNAseq data and RT-qPCR results, log 2 FC was calculated with 2 -ΔΔCT in different conditions. Then regression analysis was carried out with linear model tting function in R software (v4.0.2). In addition, Pearson correlation analysis between DEGs and the egg-laying rate was performed in R. P < 0.05 was considered to be statistically signi cant.

LPS maternal stimulation reduces the reproductive performance of laying hens and their offsprings
To explore the possible intergenerational transmission effects caused by the LPS maternal stimulation, we rst measured the egg-laying rate of F0 hens. The results showed that the total egg-laying rate of the three weeks post the LPS treatment was 17.17% lower than that of the two weeks before the treatment.
Also, the total egg-laying rate of two weeks before the LPS treatment was 3.20% higher than that of the control group with non-signi cant differences ( Fig. 2A, P > 0.05). However, the egg-laying rate of LPS group became 8.31% signi cantly lower than that of the controls in the three weeks after the treatment ( Fig. 2A, P < 0.05). The difference between the two groups widened to 11.51% after the stimulation, indicating that LPS stimulation could signi cantly reduce the reproductive performance of laying hens.
Furthermore, the daily egg-laying rate of the LPS group was signi cantly lower than that of the control group ( Fig. 2B, P < 0.05), which proved that LPS treatment could signi cantly reduce the reproductive performance of the laying hens. We further examined the two groups of hens as a whole union and calculated the differences in egg-laying rate between the two groups for 22 consecutive days and found that the daily egg-laying rate of the LPS group that was lower than the control group accounted for 72.73% of the total statistics ( Fig. 2C), which was extremely signi cantly higher than the probability of the LPS group that was higher than that of the controls (18.18%, P < 0.001). These ndings support the notion that LPS stimulation could reduce the egg-laying rate of hens.
For the aim of studying the effect of LPS maternal stimulation on the egg-laying rate of the corresponding untreated offspring, we calculated the daily egg-laying rate of F1 hens for 304 consecutive days (21-66W, except 39-40W), and found that the daily egg-laying rate of the F1 hens was signi cantly lower than that of the control group ( Fig. 2D, P < 0.001). Furthermore, we compared the daily egg-laying rate between the two groups during the 304 days, and found that the egg-laying rate of the LPS group was lower than that of the control group almost during the entire laying period in the F1 generation ( Fig.  2E), especially in the early and late laying stages. We also found that 90.79% (276/304) of the daily egglaying rate in the LPS group was lower than that of the control, of which 30.07% were signi cantly different ( Fig. 2E and 2F, P < 0.05). The daily egg-laying rate of the LPS group was signi cantly lower than that of the control group accounted for 28%, 18.25% and 54.76% in early, middle and late laying stages, respectively (Fig. 2F). These results showed that the LPS maternal stimulation has a greater negative impact on the early and late laying stages of F1 hens.

Intergenerational transmission of the transcriptome caused by LPS maternal stimuli
To reveal the molecular mechanism of LPS maternal stimulation that regulates the decreased egg-laying rate of the hens and their offspring from the perspective of immunity, we examined the transcriptome changes in the spleen using 53-week-old laying hens in F0 and F1 generations. It was found that LPS maternal stimulation caused changes in the F0 and F1 transcriptomes (Fig. 3A, B). Using the screening criteria of P <0.05 and FC > 1.5, 634 and 577 differentially expressed genes (DEGs) were obtained in F0 and F1, respectively, of which up-regulated genes in the LPS group accounted for 48.58% and 66.90%, respectively. With the criteria of P <0.05, FC > 2, 282 and 272 DEGs were separately detected in F0 and F1, of which up-regulated genes accounted for 50.35% and 59.56% (Fig. 3C). It can be seen that LPS activates the expression of some genes. The follow-up studies are based on P < 0.05, FC > 1.5.
To investigate the transcriptome's intergenerational transmission of maternal effect on laying hens, we examined the changes in the expression of common DEGs in the two generations of laying hens. The results showed that a total of 1077 DEGs were involved in the F0 and F1 generations, including 134 common DEGs (coDEGs), of which 63 coDEGs (47.01%, 42 up-regulated, 21 down-regulated) showed the same expression trend between the two generations ( Fig. 3D). Furthermore, we used the FPKM values of the above 1077 DEGs to perform the correlation analysis between the two generations in each group. The results showed that the gene expression between the two generations of the same group had a very high positive correlation, and the correlation coe cients reached 0.9 ( Fig. 3E, P < 2.2E-16), indicating that the effect of maternal stimulation on the transcriptome of laying hens has a transmission effect. We lists the 134 coDEGs between the two generations, including 63 coDEGs with the same expression trend, which can be used as marker genes for further studies, such as APLNR, BFSP1, C3, CCL1, CATHL2, CEP162, EXFABP, F13A1, and FOS (Fig. 3F).
The RNA-seq data used in this study was veri ed by RT-qPCR by randomly selecting three genes in the F0 and F1 generations. We compared the log 2 FC predicted by RNA-seq with the actual log 2 FC obtained by RT-qPCR. These results showed that all genes have the same expression trend (Fig. 4A). In addition, the regression analysis results showed that log 2 FC obtained by RNA-Seq is highly correlated with log 2 FC obtained by RT-qPCR (r = 0.92, P < 0.01), con rming the reliability of the current RNA-seq data (Fig. 4B).

LPS could in uence the immune response of hens in F0 and F1 generations
To analyze the biological functions separately involved in the DEGs of F0 and F1 generations, we used Ingenuity Pathway Analysis (IPA) to predict Ingenuity Canonical Pathways using the DEGs of F0 and F1 generations. A total of 103 and 44 signi cantly enriched pathways were identi ed in F0 and F1 generations, respectively (Table S3 and S4, P < 0.05). Then 33 and 18 Ingenuity Canonical Pathways were separately retained with the screening criteria of P <0.01, of which 8 pathways were shared by both F0 and F1 generations. All of the eight pathways are directly or indirectly related to immune response. The eight shared pathways of F0 and F1 contained 34 and 23 genes, respectively, of which 15 genes were shared between the two generations, such as ALB, AMBP, C3, EGR1, F13A1, F2, FGA, FGB, FOS, GC, HPX, MST1, RBP4, SERPINC1, VTN (Table 1). These results indicated that the treatment of LPS could in uence the immune system of the F0 generation, and the effect could be transmitted to the next generation.
We also used IPA to study upstream regulators of the DEGs in F0 and F1 generations and found that there were 6 (30%) shared upstream regulators in the top 20 upstream regulators of F0 and F1 generations, of which, all were related to immune-in ammatory responses, such as lipopolysaccharide, TNF, IL1B, PPARA, dexamethasone and IL6 (Table 2), further indicated that the immune system of F0 and F1 was affected. The discovery of a common regulator such as lipopolysaccharide suggests that the maternal response to LPS affects the immune system and has a long-lasting intergenerational transmission effect.
In order to further understand the intergenerational transmission effect of the LPS maternal stimulus, the 63 coDEGs with the same expression trend in F0 and F1 generations were used to perform GO term enrichment analysis. We only focused on GO terms related to biological processes, and top30 signi cantly enriched terms are shown in Fig. 5A. The majority of the top terms are related to immune system (from the bottom up), such as defense response to bacterium, in ammatory response, response to external stimulus, innate immune response, response to external biotic stimulus, immune response, humoral immune response, defense response to Gram-negative bacterium (Fig. 5A), suggesting that the LPS maternal stimulation could cause the immune-in ammatory response of the F0 generation, and the effect had been transmitted to the F1 generation. Further analysis of the seven terms (terms in red font) with FDR < 0.05 showed that these seven terms were related to immune-in ammatory response, including defense response to bacterium, response to external stimulus, antimicrobial humoral response, innate immune response and in ammatory response. In addition, we found that these seven terms were mainly associated with 10 DEGs such as AVD, HPS5 CATHL2, S100A12, EXFABP, RSFR, LY86, PKD4, XCL1 and FOS (Fig. 5B). Among these 10 DEGs, only XCL1 and FOS were signi cantly down-regulated in the LPS group (P < 0.05), while the remaining ones were signi cantly up-regulated (P < 0.05). These results indecate that the F1 hens produces a certain immune memory to the F0 LPS maternal stimulation, and the innate immunity and humoral immunity of the F1 hens are activated, which further autonomously stimulates the regulatory pathways to external stimuli. These genes could be used as genetic markers for the transmission of maternal immune stimulation between generations.
The decrease of egg-laying rate might result from the immune system deterioration of hens It was reported that the diversity of LPS structures and the differential recognition of these structures by TLR4 had been associated with several bacterial diseases [18]. LPS could induce innate immunity with Alzheimer's disease [19] and airway disease [20]. It also increases the invasive ability of pancreatic cancer cells through the TLR4/MyD88 signaling pathway [21]. So the IPA was used to analyze the Diseases or Functions Annotation to study the effect of LPS maternal stimulation on the immunity of the challenged hens and their offspring. We extracted the top20 terms of both F0 and F1 generations and found that 90% (18) of the terms are shared in the two generations, and these terms were mostly involved in diseases such as cancer and tumors (Table 3, all terms are shown in Table S5 and S6), revealing that LPS could affect the health of F0 and the unchallenged F1 hens. We also found several terms related to reproduction in the F0 such as development of female reproductive tract, development of ovary, abnormal morphology of ovary and in the F1 such as development of reproductive system. Two common terms morphology of ovary and tumorigenesis of reproductive tract were also found in the two generations (Table S5 and S6). These terms indicated that the LPS maternal stimulation might negatively impact on the reproductive system of laying hens, and this in uence has a succession effect, which might also explain the decline in egg-laying rate of F0 and F1 hens.
Moreover, the IPA was used to do interaction network analysis using the DEGs of F0 and F1 (Fig. 6A, B and Fig. S1), and found that the second network of F0 was related to Organismal Injury and Abnormalities (Fig. S1B) and the third network was related to Antimicrobial Response and In ammatory Response (Fig. S1C), which indicates that maternal response to LPS caused an immune in ammatory response. The top3 networks in the F1 generation are mainly enriched in metabolism and development-related pathways, such as drug metabolism, carbohydrate metabolism, lipid metabolism and embryonic development (Fig. 6A,B and Fig. S1D). It is worth noting that we found TREM2 and MST1 in the rst and third networks of the F1 generation respectively (Fig. 6A,B). Our previous studies found these two genes to be signi cantly negatively correlated with the egg-laying rate [22]. At the same time, we also found TREM2 and MST1 were in the PPI network predicted with the shared DEGs of F0 and F1 (Fig. 6C), indicating the importance of these two genes in the intergenerational transmission effect of maternal stimulation. Moreover, MST1 gene is highly negatively correlated with the egg-laying rate of F1 53-weekold hens (r = -0.97, P = 0.001, Fig. 6D). The signi cantly up-regulation of MST1 in the LPS group of F0 and F1 generations (Fig. 6E) may be closely related to the decrease in the egg-laying rate of the two generations hens. These ndings suggested that the LPS could induce subfertility and in uence the immunity of the unchallenged offspring via the transmissible transcriptome. Additionally, the decline in the immunity may compromise the laying rate.

Transgenerational transmission of transcriptome induced by the LPS maternal stimulation
Although the F2 generation RNA-seq data comes from the spleen of 1-day-old chicken, the common DEGs obtained from the spleen tissues of F0, F1, and F2 chickens are still of reference value for the study of multi-generational transmission effects. The splenic transcriptome of the F2 generation was further

Regulation of transgenerational transmission of transcriptome on egg-laying rate of hens
In addition, we extracted the FPKM of the F1 hens of the above 10 DEGs with transgenerational transmission effect and conducted a Pearson correlation analysis with the egg-laying rate of the F1 hens at 53 weeks of age. Five genes (50%) were found to have a correlation of more than 0.5 with egg-laying rate, including MMR1L3, C3, F13A1, LY86 and GDPD2 (Fig. 8). Among them, MMR1L3, C3, F13A1 and LY86 genes were negatively correlated with egg-laying rate, depicting that these genes inhibit the reproductive performance of laying hens. The GDPD2 gene was positively correlated with the egg-laying rate, indicating that this gene can moderately promote the reproductive performance of hens (Fig. 8). The signi cantly up-regulation of MMR1L3, C3, F13A1 and LY86 genes and the signi cantly down-regulation of GDPD2 in the LPS group of the three generations of hens indicate that the LPS maternal stimulation may promote the expression of MMR1L3, C3, F13A1, LY86 genes while inhibiting the expression of GDPD2 gene to reduce the egg-laying rate of hens, and this effect may has a transgenerational transmission effect.

Discussion
Stimulation from the maternal source can cause the transmission of phenotypes across generations [1], and this phenomenon is often controlled by epigenetics [23]. Research on this phenomenon mainly focuses on the epigenetic multi-generational transmission and reprogramming events during gamete formation, differentiation, and development [24,25]. As the interface between the environment and gene expression, epigenetics can regulate gene expression after being in uenced by environmental factors, which in turn in uences the phenotype [26]. The transcriptome of an organism can be regarded as the intermediate phenotype of the nal phenotype of the organism, and it often plays an important role in the transmission of maternal stimulation across generations. However, in the study of maternal stimulation of multi-generational transmission, the multi-generational transmission of the transcriptome directly related to the phenotype is often ignored. In this study, laying hens were used as model organisms to study the multi-generational transmission effect of the bacterial mimic LPS maternal stimulation. It was observed that the maternal stimulation had an impact on the egg-laying rate, immunity and transcriptome of the challenged hens and their offspring. The results of this experiment will be discussed in three points: The rst point is that the LPS maternal stimulation could signi cantly decrease the egg-laying rate of F0 and F1 generation hens. More and more studies have shown that prenatal exposure to immune stress can inhibit the development and reproductive performance of mice [27,28]. As a result, this negative effect can be transferred from parents to the next generations [29][30][31][32]. The current study found that LPS treatment signi cantly reduced the egg-laying rate of hens. The overall egg-laying rate in three weeks after treatment decreased by 17.17% compared with the previous two weeks ( Fig. 2A). This negative impact was passed over to the F1 generation. In the F1 generation (Fig. 2C), the overall egg-laying rate of 304 days in LPS group was signi cantly lower than that of the controls, and the daily egg-laying rate of 276 (90.79%) days was lower than that of the control group (Fig. 2E). In addition, we discovered that the negative effect of the LPS maternal stimulation on the egg-laying rate of F1 generation hens was severe in the early and late laying stages ( Fig. 2E and 2F). Studies have reported that toxic substances such as disinfectants and pesticides in the environment can lead to ovarian diseases in female animals and affect the DNA methylation and transcriptome levels of offspring across generations [33]. Our results suggest that LPS maternal stimulation has a negative effect on the hens' ovaries and thus their offspring; it inhibits follicle maturation in the early stage and promotes follicle senescence in the late stage, resulting in a decrease in the egg-laying rate of F1 generations.
The second point is that LPS maternal stimulation leads to the transgenerational transmission of splenic transcriptome and affects the immunity of hens and their offspring. The successive generational transmission includes intergenerational and transgenerational transmission. Intergenerational transmission means that the maternal stimulus has the same or similar effects on the treated parents and the offspring that are directly stimulated, and transgenerational transmission means that the offspring are not directly affected by the stimulus factor, but have the same or similar phenotype or gene expression to the treated parents [34]. Using the poultry maternal stimulation model, intergenerational and transgenerational transmission effects can be observed in F1 and F2 generation, respectively [22]. In this study, we found that LPS maternal stimulation induced 134 common DEGs in F0 and F1 generations, of which 47.01% of the DEGs shared the same expression trend (Fig. 3D and 3F). The correlation of the splenic transcriptome levels in the same condition between F0 and F1 hens reached 0.9 (Fig. 3E, P < 2.2E-16). Furthermore, 37 shared genes were found in F0, F1, and F2 generations, in which 10 DEGs had the same expression trend among the three generations ( Fig. 7B and 7C), indicating that the LPS maternal stimulation could result in transcriptome changes, and this alteration has a transgenerational transmission effect.
External immune stimulation given to the mother prior to childbirth will result in an immune response in the offspring [35]. We also found that LPS could affect the immunity of laying hens and their offspring. Using the DEGs shared by F0 and F1 generations, we have discovered many canonical pathways and biological processes related to immune response, such as IL-12 signaling and production in macrophages, defense response to bacterium, humoral immune response, innate immune response, in ammatory response (Table 1 and Fig. 5). Meanwhile, we discovered that lipopolysaccarige (LPS) is a common upstream regulator of F0 and F1 generation DEGs ( Table 2), indicating that the F1 generation generates immune memory to the F0 generation's LPS maternal stimulation, and the innate immunity and humoral immunity are activated, which further autonomously stimulates the regulatory pathways to external stimuli. These ndings suggest that maternal LPS stimulation induced an intergenerational transmission effect of the immune response in laying hens, and this effect may be regulated by the transcriptome and the related pathways. The immune-related DEGs inherited between generations can be used as genetic markers for the intergenerational transmission effects of maternal immune stimulation, such as AVD, HPS5, CATHL2, S100A12, EXFABP, RSFR, LY86, PKD4, XCL1 and FOS.
Thirdly, the intergenerational transmission of reduced egg-laying rate may be affected by immunity and transcriptome. Maternal exposure to infectious agents or maternal stimuli can increase disease risk in the offspring through epigenetic regulation [27-30, 36, 37]. In the present study, we found that DEGs in F0 and F1 generations were enriched in tumorigenesis of reproductive tract, development of female reproductive tract, morphology of ovary, development of ovary, abnormal morphology of ovary, development of reproductive system and other reproduction-related disease pathways (Table S5 and S6), verifying that toxic substances in the environment can cause ovarian diseases in female animals [33,38]. It indicates that the LPS maternal stimulation may have a negative impact on the reproductive system of laying hens, and this in uence has the effect of intergenerational transmission, which may explain the decline of the egg-laying rate of F0 and F1 hens. Our previous study found that all of the genes that were signi cantly negatively related to egg-laying rate (P <0.05) were up-regulated in the LPS group [22]. The expression of the MST1 gene in the F1 generation was highly negatively correlated with the egg-laying rate of 53-weekold hens, and the correlation coe cient reached -0.97 (Fig. 6D, P = 0.001). MST1 is an immune-related gene that is signi cantly enriched in IL-12 signaling and production in macrophages in this study, indicating that the intergenerational transmission of reduced egg-laying rate may be caused by immunity decline and transcriptional regulation.
In the current study, among the 10 shared DEGs transmitted across the three generations, the correlation coe cients between ve genes and the egg-laying rate of F1 hens at 53 weeks of age were above 0.5, of which MMR1L3, C3, F13A1, LY86 were up-regulated in the LPS group and negatively correlated with the egg-laying rate (Fig. 8). GDPD2 was down-regulated in the LPS group and positively correlated with egglaying rate (Fig. 8). These genes may be the key regulatory genes for the intergenerational transmission of the drop in egg-laying rate of hens. Among these genes, LY86 (lymphocyte antigen 86) and MMR1L3 (macrophage mannose receptor 1-like 3, macrophage mannose receptor 1-like 3) are two important genes. LY86 and MMR1L3 expressions in the LPS group of the three generations were consistently signi cantly up-regulated, and both of them are linked to TLR4 ( Fig. 7D and 7E), which is the receptor of LPS [39], suggesting that these two genes may play an important role in the transgenerational transmission effect of LPS maternal stimulation. LY86 gene is closely related to the innate immune system and in ammatory response [40], suggesting that maternal stimulation may continue to affect the innate immunity of the unchallenged offspring. In addition, the PPI shows that the proteins MMR1L3 and TREM2 are also directly related (Fig. 7D). Our previous study has shown that the TREM2 gene is signi cantly negatively correlated with the egg-laying rate [22], indicating that the decline of the egglaying rate in F1 hens may also be related to the signi cant up-regulation of the MMR1L3 gene. And the signi cant up-regulation of MMR1L3 in the F2 generation suggests that the egg-laying rate of F2 hens may also be reduced.

Conclusions
In summary, the current study is the rst to explore    Table 3. Top20 diseases or functions annotation predicted by IPA in F0 and F1 generations Figure 1 The technical route of the study caused by the LPS maternal stimulation.    were mainly associated with these seven terms; red and blue denote the up-and down-regulated genes in the LPS group, respectively. The right side of the circle represents the terms of the biological process.   Correlation analysis between transgenerational heritable DEGs and egg-laying rate.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.