The body weights of mice in the control group, low-Trp group and Trp-free group were comparable at the beginning of feeding ( Figure 1A ). The body weights of mice in the control group were significantly higher than that those of mice in the low-Trp and Trp-free groups after 2 months of feeding ( Figure 1A ).

To investigate the effects of Trp deficiency on early pregnancy, we examined mouse uteri in each group on day 8 of pregnancy. There was no obvious difference in the implantation sites between the low-Trp group and the control group on day 8 of pregnancy. In the Trp-free group, no implantation sites were observed in the uterus ( Figure 1B ). To explore the cause of pregnancy abnormalities on day 8 of pregnancy in the Trp-free group, we analyzed mouse uteri on day 5 of pregnancy. However, implantation sites were identified in each group, and the numbers of implantation sites were similar among the three groups ( Figures 1C, D ). We further examined the levels of decidualization-related markers.

Decidualization is critical for establishing pregnancy in mice (2). The prolactin family 8 subfamily A member 2 (Prl8a2) is a reliable marker for in vitro mouse decidualization (31). Compared with that in the control group, the expression of Prl8a2 in the low-Trp group was significantly higher, while Prl8a2 expression in the Trp-free group was significantly lower ( Figure 1E ). An increase in ALP activity is a marker for mouse decidualization (32). The density of alkaline phosphatase staining in the control group was higher than that in the low-Trp group, while alkaline phosphatase staining was not observed in the Trp-free group ( Figure 1F ). These results suggest that there was a decidualization abnormality in the Trp-free group.

A unique crypt (implantation chamber) is established when luminal epithelial evaginations move towards the antimesometrial pole and are usually “spear shaped” (33, 34). On day 5 of pregnancy, the morphology of the implantation chamber was similar between the control group and the low-Trp group. Compared to that in the control group, the morphology of the implantation chamber in the Trp-free group was abnormal. There were significantly more abnormal implantation chambers in the Trp-free group than in both the control and low-Trp groups ( Figures 2A–C ).

Planar cell polarity (PCP) signaling is necessary for crypt development (35). Van Gogh-Like Protein 2 (VANGL2) and Rbpsuh (RBPJ) are key PCP components. Epithelial Vangl2 null mice exhibit shallower luminal epithelial eversions and defects in crypt shape and size (36). In Rbpj-deficient mice, a seriously deviated uterus embryo axis at the implantation site, abnormal decidual morphology, and impaired embryonic development were observed (37). Therefore, we examined the expression of PCP signaling molecules in the uterus on day 5 of pregnancy in each group. The immunofluorescence intensity of VANGL2 and RBPJ in the uterus of the Trp-free group was lower than that in the control group, while the immunofluorescence signal in the low-Trp group was slightly greater than that in the control group ( Figure 2D ). These results suggest that PCP signaling is impaired in the Trp-free group.

A successful pregnancy requires proper interaction between the activated embryo and the receptive uterus. On day 4 of pregnancy, uterine epithelial cells stop proliferating and begin to differentiate into the receptive phase (38, 39). Compared to that in the control group, immunostaining for Ki67, a marker of cell proliferation, was increased in epithelial cells on day 4 of pregnancy in the Trp-free group, while Ki67 immunostaining in the low-Trp group was similar to that in the control group ( Figure 3A ). Compared to that in the control and low-Trp groups, the number of subluminal stromal cells immunostained for Hand2 was obviously lower in the Trp-free group ( Figure 3A ). Lactoferrin (LTF) and complement C3 (C3) are target genes of oestrogen signaling (40, 41). Compared to that in the control group, Ltf and C3 expression in the Trp-free group was significantly higher. In the low-Trp group, the expression of Ltf was similar to that in the control group, but C3 expression was significantly higher than that in the control group ( Figures 3B, C ). Suppression of oestrogen-mediated uterine epithelial proliferation by progesterone is a prerequisite for successful implantation. HAND2, a progesterone-stimulated gene, is expressed in the uterine stroma during the peri-implantation period and is essential for mouse embryo implantation (39). These results suggest that a Trp-free diet for two months might have adverse effects on uterine receptivity.

Although Trp is metabolized to Kyn mainly via IDO1 and Tdo2, Trp is also used to synthesize serotonin via the TPH1 enzyme (13). To further explore the reasons for the adverse effects of Trp deficiency on early pregnancy, we examined the key enzymes and their products involved in Trp metabolism. Compared to those in the control and low-Trp groups, there was an increase in IDO1 immunofluorescence and a decrease in TPH1 immunofluorescence in the Trp-free group ( Figure 4A ). The uterine Kyn concentrations in the Trp-free and low-Trp groups were significantly higher than those in the control group ( Figure 4B ). However, the uterine serotonin concentration in the Trp-free group was significantly lower than that in the control and low-Trp groups ( Figure 4C ). These data suggest that Trp is metabolized mainly into Kyn. Under in vitro decidualization, the expression of Prl8a2, a marker of mouse in vitro decidualization (42), was significantly increased, which was significantly inhibited by either 0.5 or 1 mM Kyn ( Figure 4D ). In addition, treatment of mouse stromal cells with 5-HT under in vitro decidualization conditions revealed that low concentrations of 5-HT had minimal effects on Prl8a2 expression, while higher concentrations of 5-HT suppressed Prl8a2 expression ( Figure 4E ). Therefore, the high uterine Kyn concentration resulting from strong IDO1 expression impaired mouse decidualization in the Trp-free group.

Because Trp and Kyn can activate AHR (43) and the AHR target genes CYP1A1 and CYP1B1 (44, 45), the expression of AHR and its target genes was examined. Compared with that in the low-Trp and control groups, AHR immunofluorescence was strongly detected in the subluminal stromal cells at the implantation site in the Trp-free group ( Figure 5A ). Additionally, the expression of CYP1A1 and CYP1B1 was significantly higher in the Trp-free and low-Trp groups than in the control group ( Figures 5B, C ). Cytochrome CYP1A1 and CYP1B1 catalyze the oxidative metabolism of oestradiol to produce the catechol oestrogens 2-hydroxy oestradiol (2OE-E2) and 4-hydroxy oestradiol (4OE-E2), respectively (46, 47). Under in vitro decidualization, Prl8a2 expression was significantly increased, which was suppressed by 2OE-E2 (20 μM) and 4OE-E2 (2 and 20 μM) ( Figures 5D, E ). Treatment of mouse stromal cells with Kyn under in vitro decidualization significantly increased Cyp1a1 and Cyp1b1 expression. CH223191, an AHR inhibitor, abrogated Kyn-induced Cyp1a1 and Cyp1b1 upregulation ( Figures 5F, G ). Moreover, treatment of mouse stromal cells with Kyn under in vitro decidualization significantly downregulated Prl8a2 expression, while CH223191 could improve the inhibition of decidualization by Kyn ( Figure 5H ). Taken together, these results suggest that activated AHR and downstream genes contribute to abnormal decidualization in the Trp-free group.

Female CD-1 mice (6 weeks old) were bought from Hunan Slaike Jingda Laboratory Animal Co., Ltd. (Chaina, Hunan) and kept in a temperature controlled environment with a 12 h light photoperiod. The animals were given free food and water. All animal protocols were approved by the Animal Care and Use Committee of South China Agricultural University (No. 2021f085). Female mice were co-caged with fertile or vasectomized males to induce pregnancy or pseudopregnancy. The day when vaginal plug was observed was defined as day 1 of pregnancy. From days 1 to 4, pregnancy was confirmed by collecting embryos from oviducts or uterus. Implantation sites on days 5 and 6 of pregnancy were visualized by intravenous injection of 0.1 ml of 1% Chicago blue dye (Sigma-Aldrich). Artificial decidualization was induced by intraluminal injection of 10 μl of sesame oil (Sigma-Aldrich). The uninjected contralateral uterine horn served as control.

According to the nutrient requirements of laboratory animals, the Trp content in the maintenance feed for mice is 0.1%, and the Trp content in the breeding feed is 0.2% (70). In a previous study, Trp amount in a Trp-limited diet is 0.062% (52). Because we mainly studied the effects of Trp content on early pregnancy, female mice were fed different diets containing 0.2% Trp (for control group, N=35 mice), 0.062% Trp (for low-Trp group, N=35 mice) and 0% Trp (for Trp-free group, N=35 mice), respectively. Mouse diets containing different concentrations of Trp were purchased from Jiangsu Xietong Pharmaceutical Bio-engineering Co., Ltd. (Jiangsu, China),. All proteins in the mouse diets were replaced with amino acid premixes in the process of making feed to ensure that the Trp content in the food was 0.2%, 0.062% and 0 of the total feed (see Table 1 for specific feed formulations). In vivo experiments with dietary for Trp addition or deficiency are generally a long-term process for 12 weeks, 8 weeks, or 70 days in previous studies (71–74). Therefore, we first fed female mice with different amounts of Trp for 2 months and then performed different treatments. Pregnant or pseudopregnant mice were also fed with the diets containing different concentrations of Trp until they were sacrificed by cervical dislocation. Female body weights were comparable in three groups at the beginning of feeding. The uteri and serum were collected on days 4, 5 and 8 of pregnancy for further analysis.

Alkaline phosphatase (ALP) staining was performed as previously described (75). Frozen sections (10 μm) of mouse uteri were quickly fixed for 10 min in 4% paraformaldehyde (158127, Sigma Aldrich, St. Louis, MO). After three washes in PBS, sections were incubated with nitro blue tetrazolium (0885, Life Science) and 5-bromo-4-chloro-3-indolyl phosphate (0329, Life Science, St. Louis, MO) in PBS solution for 10 min to show alkaline phosphatase activity. The staining density was examined under a microscope. Each experiment was repeated at least three times.

The immunofluorescence method was performed as previously described (76, 77). Frozen sections (10 μm) were fixed with 4% paraformaldehyde (158127, Sigma Aldrich) for 10 min. Paraffin sections (5 μm) were dewaxed, hydrated and antigen-retrieved by boiling in 10 mM citric acid buffer for 10 min. After permeabilized with 0.1% Triton 100 and blocked with 10% horse serum at 37° C for 1 h, sections were incubated with appropriate dilutions of primary antibodies overnight at 4° C. Primary antibodies used in this study included anti-KI67 (1:200, GB111141, Exilon, Guangzhou), anti-RBPJ (1:200, 14613, Proteintech, Wuhan, China), anti-VANGL2 (1:200, 21492, Proteintech), anti-HAND2 (1:200, sc-9409, Santa Cruz Biotechnogoly, Santa Cruz, USA), anti-IDO1 (1:200, 66528, Proteintech), anti-TPH1 (1:100, CSB-PA024100LA01HU, Cusabio, Wuhan, China). After washing with PBS, sections were incubated with 488-conjugated secondary antibodies (2.5 μg/ml, G21234, Invitrogen, Carlsbad, CA) for 40 min, counterstained with propidium iodide (5 μg/ml, PI, P4170, Sigma-Aldrich) and mounted with Extended™ Diamond Antifade Mountant (Thermo Fisher, Waltham, MA). Images were taken with a laser scanning confocal microscope (Leica, Germany). Each experiment was repeated at least three times.

Kyn concentration was determined as previously described (78). In short, after the mouse uterus was collected at implantation sites on day 5 of pregnancy, the endometrium was isolated from the myometrium and homogenized in sample diluent. After centrifugated at 5000×g for 10 min, a total of 360 μl supernatant was mixed with 180 μl of 30% trichloroacetic acid (TCA) (T6399, Sigma) and incubated at 50° C for 30 min. After centrifugation at 3000×g for 10 min, the supernatant was mixed with equal volume of current Ehrlich reagent (2% p-dimethylaminobenzaldehyde, D109644, Aladdin, Shanghai, China) dissolved in glacial acetic acid) thoroughly, and incubated at room temperature for 12-30 min. Absorption was measured at 492 nm and compared with the standard curve of L-Kyn (HY-104026, MedChemExpress, NJ, USA). Each experiment was repeated at least three times.

After removing myometrium, the endometrium at implantation sites on day 5 of pregnancy was homogenized for serotonin measurement according to the manufacturer’s instructions (Elabscience, E-EL-0033c, Wuhan, China). This kit’s sensitivity is greater than 9.38 ng/ml. The standard working solution or samples of 50 μL were mixed with 50 μL biotinylated antibody in working solution and incubated at 37° C for 45 min. Following washing 3 times, 100 μL HRP enzyme conjugate working solution was added to each well and incubated at 37° C for 30 min. After washing 5 times, 90 μL of substrate solution was added to each well and incubated at 37° C for 15 min. Absorbance at 450 nm wavelength was measured following adding 50 μL of termination solution to each well. Each experiment was repeated at least three times.

Mouse endometrial stromal cells were isolated as previously mentioned (77). Mouse uteri on day 4 of pseudopregnancy were longitudinally cut, washed in HBSS, and incubated with 1% (W/V) trypsin and 6 mg/ml dispase in 3.5 mL HBSS for 1 h at 4°C, 1 h at room temperature, and 10 min at 37°C. The uterine tissues were washed with Hanks’ balanced salt solution and incubated with 6 ml of HBSS with 0.15 mg/ml Collagenase I (Invitrogen, 17100-017) at 37°C for 35 min. Primary endometrial stromal cells were cultured in DMEM/F12 with 10% FBS.

Mouse endometrial stromal cells were induced for in vitro decidualization as previously described (79). Endometrial stromal cells were treated with 10 nM of estradiol-17 β and 1 μM of P4 in DMEM/F12 containing 2% charcoal-treated FBS (cFBS, Biological Industries) to induce decidualization in vitro for 72 h. Stromal cells were treated with L- Kyn (0.25, 0.5 and 1 mM, HY-104026, MedChemExpress), 4-Hydroxyestradiol (0.2, 2 or 20 mM, GC19552, Glpbio, Shanghai, China), 2-Hydroxyestradiol (0.2, 2 or 20 mM, GC13460, Glpbio) in DMEM/F12 containing 2% cFBS, respectively. Each experiment was repeated at least three times.

RT-qPCR was performed as previously described (80). Total RNAs were digested with RQ1 deoxyribonuclease I (Promega, Fitchburg, WI) to remove DNA, and reverse-transcribed into cDNA with Prime Script Reverse Transcriptase Reagent Kit (Takara, Japan). Real-time PCR was performed using a SYBR Premix Ex Taq Kit (TaKaRa) on the CFX96 TouchTM Real-Time System (Bio-Rad) The 2-△△Ct method was used to calculate the data, which were standardized to the RPL7 (mouse) or RPL7 (human) level. The primer sequences for each gene were provided in Table 2 . Each experiment was repeated at least three times.

Data were processed using GraphPad Prism 8.0. The student’s t test was employed to compare the difference between the two groups. The ANOVA test was utilized to compare the difference of more than two groups. At least three independent replications of each experiment were conducted. Each group in the mouse study contained a minimum of three mice. Standard deviation (SD) and mean were used to present the data. A result with a p-value of 0.05 was deemed significant.