近日，中国药科大学药学院Kai Han(第一作者)、美国密西根大学Fang Xie(共同第一作者)、美国密西根大学James J. Moon(通讯作者)在国际顶级期刊《Nature Materials》(IF：37.2)发表了题为“Inulin-gel-based oral immunotherapy remodels the small intestinal microbiome and suppresses food allergy”的研究性论文。

食物过敏已经成为全球主要的公共健康问题，特别是在工业化国家。对过敏原的意外接触可能引起威胁生命的低血容量性休克。干预长期局限在严格的过敏原避免、紧急处理和试验治疗。美国FDA最近批准首个口服免疫治疗(OIT)药物-Palforzia，其通过抑制病理学二型免疫反应来减少对花生过敏反应的概率和严重程度。然而由于负面事件，胃肠道事件是最常见原因，10-20%患者中止Palforzia。而且，由于在OIT或OIT中断只有2-24周期间间歇给药、不足的每天维持剂量，大多数患者没有发展长期持久的免疫无应答性。

越来越多证据支持肠道微生态失衡、微生物代谢物和食物过敏之间存在一定的关系。细菌疗法包括粪便微生物移植和益生菌已经被报道能够修复健康肠道微生物和促进过敏原脱敏，其中微生物代谢物是一个关键连接。然而，安全关注、缺乏对有益微生物属核心组的共识和较差的移植率大大限制了其临床应用。特别是，对肠道微生物如何促进过敏原脱敏的以前研究主要分析粪便样品中的肠道共生菌和特定微生物代谢物包括短链脂肪酸和胆汁酸。尽管粪便取样是一个无创且方便的方法，由于胃肠道中部位不同，粪便样品只是整个肠道微生物的一个近似值。重要的是，抗原启动和对食物抗原的忍耐在小肠中开始。新证据表明小肠中的微生物和代谢物在食物过敏中发挥关键作用。然而，仍然不知道如何调节肠道微生物及其代谢物来对食物过敏进行治疗干预。

将膳食抗原递送到小肠抗原采样树突细胞(DCs)对于肠道稳态和口服耐受至关重要。然而，小肠的平均转运时间较短，小鼠和人在60-70分钟，从而在OIT期间限制了抗原采样过程和耐受诱导。以前研究报道了一个益生元基凝胶，其能够延长胃排空和胃肠道滞留时间。在该研究中，研究人员基于菊粉凝胶来制备一个新OIT平台用于口服抗原递送到肠DCs和原位调控微生物-代谢物-免疫轴。研究人员证实菊粉凝胶与食物抗原配制可以延长菊粉和抗原肠道停留，改善小肠中耐受性树突状细胞对抗原的摄取。在食物过敏小鼠模型中，菊粉凝胶基OIT可以抑制TH2细胞亚群和诱导IFNγ+和IL-10+调节T(Treg)细胞，正常化小肠中失调的微生物和代谢物。菊粉凝胶基OIT建立免疫耐受，持久对抗反复发生的食物过敏挑战。由于菊粉是一种被广泛消费的膳食纤维，被FDA普遍认为是安全的，该工作显示了菊粉凝胶基OIT的治疗和转化潜力。

Fig. 1: Inulin gel/OVA protects mice against repeated allergen challenges with chicken egg white.

a, Images of inulin gel and inulin gel/AF647-labelled OVA. b, Schematic of the intestinal anaphylaxis and therapeutic regimen. As a healthy control group, naive BALB/c mice were not sensitized with alum/OVA. Mice in all other treatment groups were sensitized with alum/OVA on days 0 and 14. From day 29, alum/OVA-sensitized mice were orally gavaged with PBS, OVA (1 mg per dose) or inulin gel (55 mg per dose)/OVA (1 mg per dose) for three times every four days. From day 49, mice from all groups (including naive mice) were challenged intragastrically (i.g.) with OVA (50 mg per dose) on six alternating days (days 49, 51, 53, 56, 58 and 60). c, After the sixth intragastric challenge, changes in the average and individual core body temperature were measured. d–h, Over the six consecutive intragastric challenges, mice were analysed for anaphylactic scores (d), diarrhoea occurrence rate (e) and diarrhoea severity score (f); images of diarrhoea occurrence during the third intragastric challenge (red boundary marks mice with diarrhoea, and the X indicates mortality) (g); body-weight drop (h). i–k, Mice were analysed for serum-OVA-specific IgE levels on day 48 (i), serum MMCP-1 concentrations at 60 min after the sixth intragastric challenge on day 60 (j), mast cell counts (each dot represents one mouse) and toluidine blue staining of jejunal mast cells (arrows) on day 62 (k). l–n, Splenocytes were restimulated ex vivo with 250 µg ml–1 OVA on day 62. After 72 h, the supernatants were analysed for cytokines. The concentrations of IFNγ (l) and IL-4 and IL-13 (m) and the ratios of IFNγ/IL-4, IFNγ/IL-5, IL-2/IL-4 and TNF/IL-5 (n) are shown. Data represent mean ± s.e.m. from a representative experiment of two independent experiments (n = 5 for naive, 7 for OVA or 8 for PBS and inulin gel/OVA (c,j); n = 10 for naive or 17 for other groups (d,h); n = 10 for naive or 16 for other groups (e,f); n = 10 for naive or 13 for other groups (i); n = 5 for naive, 6 for OVA, 7 for PBS or 8 for inulin gel/OVA (k); n = 5 for naive and OVA, 7 for PBS or 8 for inulin gel/OVA (l); n = 5 for naive and 6 for other groups (m); n = 5 for naive in IFNγ/IL-4 and IL-2/IL-4 datasets and 4 for naive in IFNγ/IL-5 and TNF/IL-5 datasets; n = 6 for all other groups (n)). Data were analysed by two-way ANOVA (c,d,f,h) or one-way ANOVA (i–n) with Bonferroni’s multiple comparisons test.

Alum/OVA-sensitized BALB/c mice were treated as that shown in Fig. 1b. a, After the sixth intragastric challenge, the changes in the core body temperature and body weight were measured. b,c, Over the six consecutive intragastric challenges, anaphylactic scores (b) and diarrhoea occurrence rate and severity (c) were recorded. d,e, Mice were analysed with OVA-specific IgE levels in serum on days 48 and 55 (d) and the serum MMCP-1 concentration at 60 min post-intragastric challenge on day 60 (e). f, Dynamic rheological and viscosity measurements of inulin gel/OVA, inulin/OVA; G′ (elastic modulus) and G″ (viscous modulus). g,h, Alum/OVA-sensitized mice were orally gavaged with FITC-labelled inulin gel/Texas Red-labelled OVA, or FITC-labelled inulin/Texas Red-labelled OVA on day 29. The visualization of Texas Red-labelled OVA under Texas Red channel in the small intestine over time (g) and measurement of the total fluorescence intensity of Texas Red-labelled OVA in the small intestine, duodenum, jejunum and ileum (h) are shown. i, Mice were treated as in Fig. 1b and analysed for FITC-dextran in serum after oral gavage on day 57. j,k, Alum/OVA-sensitized BALB/c mice received OIT as in Fig. 1b. Mice were orally gavaged with inulin gel/AF647-labelled OVA or inulin/AF647-labelled OVA on day 49. After 3 h, the mice were euthanized for the analyses of OVA uptake by CX3CR1+ DCs (j) and frequency of CD103+ DCs (k) in SI-LP via flow cytometry. Data represent the mean ± s.e.m. from one of two independent experiments (n = 8 (a–d), 3 (f), 4 (h) and 6 (j,k); n = 7 for inulin/OVA or 8 for inulin gel/OVA (e); n = 6 for OVA; or n = 7 for other groups (i) biologically independent samples). Data were analysed by two-way ANOVA (a–c,f, h) or one-way ANOVA (i–k) with Bonferroni’s multiple comparisons test or two-tailed Mann–Whitney test (e) or unpaired, two-sided Student’s t-test (a,d).

Fig. 3: Inulin gel/OVA induces immune regulatory phenotype in the small intestine.

BALB/c mice were treated as shown in Fig. 1b. The SI-LP were isolated on day 61, pooled from three mice per group and subjected to scRNA-seq. a,b, Uniform manifold approximation and projection (UMAP) in all samples, pie chart of the relative proportion of various cell types (a) and percentages of mast cells and CD69+ TH2 cells (b) were analysed. c–g, Small intestine and MLN were analysed before and after intragastric challenges. The frequencies of GATA3+CD4+ TH2 cells on day 57 (c), T-bet+CD4+ TH1 cells on day 61 (d) in MLN; FOXP3+CD4+ Treg cells in SI-LP on day 49 (before the intragastric challenge; e), and in MLN on day 61 (f) and representative flow cytometry plots; and frequency of GATA3+CD4+ Treg cells on day 57 (g) are shown. h–k, MLN cells were activated ex vivo with PMA and ionomycin on day 61. After 4 h, mice were analysed for the ratio of IFNγ+/IL-4+ in CD4+ T cells (h), the frequencies of IL-10+ among IL-4+CD4+ T cells (i), IFNγ+ Treg cells (j) and IL-10+ Treg cells (k). l–n, Alum/OVA-sensitized BALB/c mice were treated as in l, 100 µg of antibody against IL-10+ cells, IFNγ+ cells or isotype IgG1 control was administered intraperitoneally, as indicated. The changes in core body temperature (m) and body-weight drop at the sixth intragastric challenge (n) are shown. Data represent the mean ± s.e.m. (n = 3 (b); n = 4 for naive and OVA, 5 for inulin/OVA, 8 for PBS or 7 for inulin gel/OVA (c); n = 5 for naive, 8 for inulin gel/OVA or 6 for other groups (d,f,i–k); n = 5 for naive, 6 for OVA, 7 for inulin gel/OVA or 8 for other groups (e); n = 4 for naive and OVA, 6 for inulin/OVA, 8 for PBS or 7 for inulin gel/OVA (g); n = 8 for inulin gel/OVA or 6 for other groups (h); n = 5 for naive, 7 for PBS and inulin gel/OVA plus isotype IgG1, or 8 for other groups (m); n = 5 for naive, 7 for inulin gel/OVA plus isotype IgG1 or 8 for other groups (n)). Data were analysed by one-way ANOVA (b–k,n) or two-way ANOVA (m) with Bonferroni’s multiple comparisons test.

Fig. 4: Inulin gel/OVA establishes durable protection with sustained unresponsiveness.

a, Schematic of the intestinal anaphylaxis and therapeutic regimen. BALB/c mice were sensitized with alum/OVA. From day 29, mice were orally gavaged with inulin gel/OVA or free OVA. The dose of inulin gel was 55 mg, and the OVA (heated at 70 °C for 2 min) doses were 0.25, 0.5, 1, 2, 4, 8, 12 and 16 mg for the dose escalation phase and 20 mg for the maintenance phase. From day 49, mice were intragastric. challenged with OVA (50 mg per dose) on six alternating days (b–g) or challenged for four times and discontinued for 25 days, followed by intragastric re-challenge of OVA (h–m). b, After the third intragastric challenge, the body-weight change was recorded. c,d, During the three consecutive intragastric challenges, mice were analysed for anaphylactic scores (c) as well as diarrhoea occurrence rate and diarrhoea severity (d). e,f, Mice were analysed for MMCP-1 in serum on day 60 (e) and mast cell counts and toluidine blue staining of jejunal mast cells (arrows) on day 62 (f). g, Splenocytes were restimulated ex vivo with 250 µg ml–1 OVA on day 62. After 72 h, the supernatants were analysed for cytokines. h–m, Mice were re-challenged with OVA 25 days after the fourth intragastric challenge. Body-weight changes, anaphylactic scores and diarrhoea-free mice ratio on day 78 were measured (h); on day 79, mice were further analysed for the frequencies of FOXP3+CD4+ Treg cells in SI-LP (i); CD103+ DCs in MLN (j); the ratios of ROR-γt+/GATA3+ among CD4+ T cells and FOXP3+CD4+ Treg cells in SI-LP and MLN (k); the ratio of T-bet+/GATA3+ among CD4+ T cells in SI-LP (l) and frequency of IL-4+FOXP3+CD4+ T cells and ratio of IFNγ+/IL-4+ among FOXP3+CD4+ Treg cells in MLN (m). Data represent the mean ± s.e.m. from a representative experiment of two independent experiments (n = 15 for OVA or 14 for inulin gel/OVA (b); n = 16 for OVA or 15 for inulin gel/OVA (c,d); n = 8 for OVA or 4 for inulin gel/OVA (e); n = 6 for OVA or 4 for inulin gel/OVA (f,g); n = 6 for OVA or 7 for inulin gel/OVA (h,j); n = 5 for OVA or 7 for inulin gel/OVA (i and l); n = 5 for OVA in SI-LP dataset and 6 for OVA in MLN dataset or 7 for inulin gel/OVA (k); n = 6 (m) biologically independent samples). Data were analysed by two-way ANOVA with Bonferroni’s multiple comparisons test (c,d) or unpaired, one-sided Student’s t-test (f) or two-tailed Mann–Whitney test (e) or unpaired, two-sided Student’s t-test (b,g–m).

Fig. 5: Inulin gel/OVA normalizes the dysbiotic ileal microbiota in food allergy.

a,b, Alum/OVA-sensitized mice were fed with antibiotic cocktails during OIT. From day 49, the mice received repeated intragastric challenges of OVA. a, Over the four consecutive intragastric challenges, changes in the anaphylactic scores, diarrhoea occurrence rate and diarrhoea severity were measured. b, After the fourth challenge, the body-weight changes were recorded. c–l, Alum/OVA-sensitized BALB/c mice were treated as that shown in Fig. 1b. The microbial communities in the ileal contents and faeces were analysed via 16S rRNA gene sequencing. The analyses of the observed OTU, inverse Simpson diversity (c); the NMDS score plot (based on Bray–Curtis) in the ileal contents (d) and faeces (e); the relative abundances of the gut commensal microorganisms at the phylum and family levels in the ileal contents (f) and faeces (h); the heat maps showing the normalized Z-score values of the relative abundances of differentially abundant microorganisms at genus level in the ileal contents (g) and faeces (i) are shown. j, Relative abundances of Eggerthellaceae and Enterorhabdus in the ileal contents and faeces were compared. k, Spearman’s correlation coefficient analyses between the level of OVA-specific IgE in serum and the relative abundances of various differentially abundant microorganisms in the ileal contents or faeces. l, Spearman’s correlation coefficient analyses of Eggerthellaceae and Enterorhabdus with OVA-specific IgE in the ileal contents and faeces were compared. p, phylum; f, family; g, genus. Data represent the mean ± s.e.m. (n = 5 (a,b); n = 5 for naive, 7 for OVA in ileal contents dataset or inulin/OVA in the faeces dataset or 8 for all other groups (c–e,j) biologically independent samples). Data were analysed by an analysis of molecular variance (AMOVA) (d,e); one-way (c,j) or two-way (a) ANOVA with Bonferroni’s multiple comparisons test or unpaired, two-sided Student’s t-test (b); or two-tailed Spearman’s rank correlation test (k,l). NS, not significant.

Fig. 6: Therapeutic efficacy of inulin gel/allergen OIT in various food allergy models.

a, Therapeutic regimen of cow’s milk allergy in BALB/c mice (a). b,c, Mice were analysed for anaphylactic scores on day 76 (b) and casein-specific IgE in serum on days 70 and 77 (c). d, Therapeutic regimen of peanut allergy in alum/peanut extract (PE)-sensitized C3H/HeJ mice with intragastric challenges. e,f, Mice were analysed for anaphylactic scores on day 72 (e) and changes in the average core body temperature (f) after repeated intragastric challenges of peanut powder. g, Therapeutic regimen of peanut allergy in cholera toxin/peanut extract (CT/PE)-sensitized C3H/HeJ mice. h–j, Parenteral antigen-induced anaphylaxis (g), changes in the average (h) and individual core body temperature (i), and anaphylactic scores (j) after intraperitoneal challenge of peanut extract were measured. Data represent the mean ± s.e.m. from a representative experiment of two independent experiments (n = 5 for naive or 7 for all other groups (b,c); n = 5 (e,f); n = 7 for naive or 5 for all other groups (h–j) biologically independent samples). Data were analysed by two-way (f,h) or one-way (b,c,e,j) ANOVA with Bonferroni’s multiple comparisons test.

https://doi.org/10.1038/s41563-024-01909-w