https://orcid.org/0000-0001-6317-4305
Abstract
Very-early-onset ulcerative colitis (VEO-UC) is a severe form of inflammatory bowel disease that manifests before the age of 6 years. Compared to typical pediatric UC, it is characterized by distinct genetic and immunological factors. This study aimed to investigate the roles of specific human leukocyte antigen (HLA) alleles and maternal microchimerism (MMc) in the pathogenesis of VEO-UC in a Japanese population.
This study included 27 VEO-UC patients, including 4 patients treated with colorectal resection. HLA typing was performed by polymerase chain reaction–sequence-specific oligonucleotide probing (PCR-SSOP) and compared with the Japanese general population. Immunohistochemistry and fluorescence in situ hybridization were used to evaluate MMc in intestinal tissues. Statistical comparisons of HLA were performed against data from the general Japanese population, with Bonferroni corrections applied to handle multiple comparisons.
HLA-B52 and HLA-DR15 were more prevalent in cases of VEO-UC than in the general Japanese population, although the statistical significance decreased after the Bonferroni correction. MMc was found in the intestinal tissues of three VEO-UC cases, whereas it was absent in the control UC cases. Maternal HLA concordance with specific alleles associated with VEO-UC was noted in several cases, suggesting maternal immune involvement in the pathogenesis of the disease.
VEO-UC seems to share genetic traits with adult UC, such as an association with HLA-B52 and HLA-DR15, and is also affected by maternal immune contributions, as shown by the presence of MMc in the affected tissues. These findings highlight the complex interaction between genetic and immunological factors in the pathogenesis of VEO-UC and underscore the need for further research to develop targeted therapeutic strategies that address these mechanisms.
Lay Summary
In cases of Japanese very-early-onset ulcerative colitis, human leukocyte antigens B52 and DR15 were identified, and the presence of maternal microchimerism in intestinal tissue suggests that maternal immunity may also be involved in disease onset, showing the complex interactions between genetic predisposition and maternal immune factors.
Very-early-onset ulcerative colitis (VEO-UC) involves genetic and immunological factors in its etiology, unlike typical pediatric UC.
Japanese VEO-UC cases have high frequencies of specific human leukocyte antigen alleles and maternal microchimerism, suggesting that VEO-UC is different from typical UC and that genetic and maternal immunity are involved in its development.
The pathology of VEO-UC may identify a therapeutic approach differing from that used for standard UC.
Introduction
Ulcerative colitis (UC) is a chronic inflammatory bowel disease (IBD) characterized by continuous mucosal inflammation starting from the rectum and potentially spreading proximally throughout the colon.1 Along with Crohn’s disease, UC is one of the primary categories of IBD and manifests through symptoms such as abdominal pain, diarrhea, and bloody stool. Although the precise causes of UC are still unknown, both genetic and environmental factors are thought to play key roles.2,3 Increasing rates of IBD, including pediatric cases, underscore the need for further research to understand disease mechanisms and improve treatment approaches.4
Ulcerative colitis often shows more severe clinical features in pediatric populations and progresses faster than in adults. An incredibly aggressive subset, termed very-early-onset UC (VEO-UC), develops in children under 6 years of age and frequently involves extensive disease and rapid progression.5 VEO-UC is thought to arise from a specific combination of genetic and immunological factors that differ from those found in adult-onset UC. Current research links pediatric UC to genetic abnormalities, often involving single-gene mutations, and suggests potential immune deficiencies that may contribute to this early and severe disease phenotypes.6
Recent studies highlight the role of the human leukocyte antigen (HLA) complex, found on chromosome 6, in various immune responses, including the presentation of antigens to T cells. Genome-wide association studies have implicated specific HLA alleles in UC susceptibility, suggesting that variations in HLA alleles may predispose individuals to autoimmune processes, thus contributing to the pathogenesis of UC.3 In adults, HLA alleles such as HLA-B52 and HLA-DR15 have shown associations with UC risk, but there is limited research on HLA’s roles in pediatric or VEO-UC cases.7 Moreover, specific HLA alleles have been linked to UC therapy outcomes, indicating a nuanced role for HLA in both disease onset and progression.8,9 Another factor potentially relevant to VEO-UC development is maternal microchimerism (MMc), in which fetal or maternal cells remain in the offspring’s tissues. MMc has been linked to autoimmune diseases such as systemic lupus erythematosus, juvenile dermatomyositis, and biliary atresia, suggesting that maternal cells in the child’s tissues may trigger immune responses.10–12 Given these parallels, MMc could similarly contribute to the immune dysregulation observed in VEO-UC.
This study investigated the possible associations between specific HLA alleles and MMc in Japanese children with VEO-UC. The aim was to clarify the genetic and immunological underpinnings of VEO-UC by analyzing HLA allele frequencies in VEO-UC patients compared with the general Japanese population and assessing the presence and localization of maternal cells within affected tissues. This dual focus on genetic predisposition and maternal influence may offer new insights into the different mechanisms behind this severe type of pediatric UC.
Materials and Methods
Study Population and Examinations
Japanese VEO-UC cases and UC cases with a surgical history for immunohistochemical comparison diagnosed by findings of colonoscopy (CS) and histological findings due to gastrointestinal symptoms such as bloody stool and abdominal pain at institutions participating between April 2010 and August 2024 were included in this study. Patients were confirmed by medical questionnaires to have no relatives of other ethnicities within the third degree of consanguinity. Exclusion criteria were failure of total CS, insufficient medical history, and genetic abnormalities identified by targeted gene panel sequencing for already known primary immunodeficiency syndrome and monogenic IBD.13
All patients who met the inclusion criteria and none of the exclusion criteria were evaluated for the biomarkers and clinical, endoscopic, and pathological findings. The clinical activity of UC was assessed using the pediatric ulcerative colitis activity index (PUCAI).14 The PUCAI and blood tests were conducted within 2 weeks of the preceding endoscopic examination.
The ethics committee of each institution approved the study according to the provisions of the World Medical Association’s Declaration of Helsinki, and the parents of all patients gave written informed consent.
Endoscopic Examinations
Initial VEO-UC and UC cases underwent CS and were diagnosed based on colonoscopic and histological findings. Pediatric gastroenterologists with over 8 years of experience performed CS using a standard endoscope (Olympus). The extent of disease at onset in initial cases of VEO-UC was defined according to the Paris classification: E1 = ulcerative proctitis; E2 = left-sided UC (distal to the splenic flexure); E3 = extensive (hepatic flexure distally); and E4 = pan-colitis (proximal to the hepatic flexure).15
Typing of Human Leukocyte Antigen
Human leukocyte antigen typing for female and male VEO-UC cases was conducted using the polymerase chain reaction–sequence-specific oligonucleotide probing (PCR-SSOP) method combined with the Luminex multiplex assay system (SRL, Inc.).
The detected HLA allele frequencies were compared with those of the previously evaluated Japanese control populations.16 HLA haplotypes were inferred from a database of Japanese families.17
The mother’s HLA in the VEO-UC cases was also assessed individually, and a written informed consent for this was obtained separately.
Immunohistochemistry
Very-early-onset ulcerative colitis cases and UC cases with a history of total colectomy were examined. Immunohistochemical staining of colon sections from VEO-UC and UC cases was performed to identify the existence and specificity of MMc in VEO-UC cases. Colonic specimens from patients were prepared as serial, 3-µm-thick paraffin sections from formalin-fixed tissues. Immunohistochemical staining was performed using the ultraView universal diaminobenzidine (DAB) Detection Kit and Hematoxylin Counterstain II (Ventana Medical Systems) in an automated immunostainer (Benchmark GXTM; Ventana) according to the manufacturer’s instructions. Mouse anti-human CD3 (1:5000; Proteintech) was used as the primary antibody.
After immunohistochemistry (IHC), cells were washed with 0.2-N hydrogen chloride (HCl) for 20 minutes and then with purified water for 2 minutes, repeated twice. They were subsequently placed in a constant-temperature chamber at 88 °C for 30 minutes using 2× saline-sodium citrate buffer (SSC)/0.1% nonidet P-40 (NP-40), then washed in purified water for 1 minute and in 2×SSC wash buffer for 5 minutes, repeated twice. Cells were treated with 0.1% pepsin in 0.01 N HCl at 37 °C for 10 minutes, followed by a wash in 2×SSC buffer for 5 minutes, repeated twice. Next, 1-2 μL of X/Y chromosome-specific probe was applied dropwise and hybridized at 80 °C for 8 minutes, and then at 37 °C for 72 hours. Slides were immersed in 2×SSC/0.3% NP-40, heated at 60 °C ± 2 °C for 2 minutes, then transferred to 2×SSC. Fluorescent in situ hybridization (FISH) was performed using 2-4 μL of an X-chromosome-specific probe (Spectrum Orange) and a Y-chromosome-specific probe (Spectrum Green). The Vysis CEP X SpectrumGreen/Y SpectrumRed DNA Probe Kit (Abbott) was used, and slides were counterstained with 4’,6-diamidino-2-phenylindole (DAPI).
Following multiplex staining for IHC and FISH, images were captured using a ZEN pro microscope (Carl Zeiss) equipped with a triple bandpass filter at 100× magnification. All images were acquired from the same unit. Cells were only counted if they displayed 2 signals within a clearly defined nucleus; cells exhibiting 2 closely proximal X signals were excluded to avoid counting mitotic signals. The same researcher performed all cell counts manually, blinded to the patient information to ensure objectivity.
Statistical Analysis
The characteristics of the samples and the statistics of the examined data are presented as medians (interquartile range [IQR]) or as percentages, as appropriate. For qualitative variables, group comparisons were performed using the chi-squared test. The comparison of HLA allele aggregation with previously reported HLA data was performed using Fisher’s exact test. In each analysis, multiple comparisons were conducted, with additional Bonferroni correction to reduce the risk of false-positive results due to the small sample size. This analysis reduced the possibility of random variance and increased the reliability of the results. Statistical analyses were performed with GraphPad Prism (version 9.4.1: GraphPad Software), and P < .05 was considered significant.
Results
Clinical Characteristics
Of the 27 examined cases, 21 with VEO-UC and 4 with UC for comparison of histological evaluation (see Table S1, which shows patient characteristics according to UC for historical comparison), irrespective of disease activity and current therapy, were included in this study. Excluded were 2 cases of vasculitis syndrome. The baseline characteristics and clinical data of the VEO-UC group are presented in Table 1; 17 cases (80.9%) were male, and 6 cases (28.5%) had a family history. All VEO-UC cases were diagnosed as pan-colitis type at the onset of endoscopic examination, and extensive lesions were observed. With respect to treatment, mesalazine intolerance was observed in 2 cases (9.5%), but biological agents and total colectomy were required in 4 cases (19.0%) and 4 cases (19.0%), respectively.
Characteristics of the very-early-onset ulcerative colitis cases included in this study
| Characteristic | Median [interquartile range or percentage] | |
| [n = 21] | ||
| Male, number of cases [%] | 17 | [80.9] |
| Disease onset, months | 30 | [23-43] |
| Age at diagnosis, months | 36 | [27-58] |
| Family history of IBD, number of cases [%] | 6 | [28.5] |
| PUCAI | 40 | [15-50] |
| Blood test | ||
| White blood cells, /μL | 10 000 | [7700-13 200] |
| Hemoglobin, g/dL | 11.4 | [9.7-12.2] |
| ESR, mm/hour | 22 | [12-43] |
| Albumin, g/dL | 3.7 | [3.5-4.2] |
| C-reactive protein, mg/dL | 0.23 | [0.08-0.63] |
| Disease extent, Paris classification, number of cases [%] | ||
| E1-3 | 0 | |
| E4 | 21 | [100] |
| Treatment at study inclusion, number of cases [%] | ||
| Mesalazine | 19 | [90.5] |
| Immunomodulators | 8 | [38.0] |
| Biologics | 4 | [19.0] |
| Total colectomy | 4 | [19.0] |
| Characteristic | Median [interquartile range or percentage] | |
| [n = 21] | ||
| Male, number of cases [%] | 17 | [80.9] |
| Disease onset, months | 30 | [23-43] |
| Age at diagnosis, months | 36 | [27-58] |
| Family history of IBD, number of cases [%] | 6 | [28.5] |
| PUCAI | 40 | [15-50] |
| Blood test | ||
| White blood cells, /μL | 10 000 | [7700-13 200] |
| Hemoglobin, g/dL | 11.4 | [9.7-12.2] |
| ESR, mm/hour | 22 | [12-43] |
| Albumin, g/dL | 3.7 | [3.5-4.2] |
| C-reactive protein, mg/dL | 0.23 | [0.08-0.63] |
| Disease extent, Paris classification, number of cases [%] | ||
| E1-3 | 0 | |
| E4 | 21 | [100] |
| Treatment at study inclusion, number of cases [%] | ||
| Mesalazine | 19 | [90.5] |
| Immunomodulators | 8 | [38.0] |
| Biologics | 4 | [19.0] |
| Total colectomy | 4 | [19.0] |
Abbreviations: ESR, erythrocyte sedimentation rate; IBD, inflammatory bowel disease; PUCAI, pediatric ulcerative colitis activity index.
Characteristics of the very-early-onset ulcerative colitis cases included in this study
| Characteristic | Median [interquartile range or percentage] | |
| [n = 21] | ||
| Male, number of cases [%] | 17 | [80.9] |
| Disease onset, months | 30 | [23-43] |
| Age at diagnosis, months | 36 | [27-58] |
| Family history of IBD, number of cases [%] | 6 | [28.5] |
| PUCAI | 40 | [15-50] |
| Blood test | ||
| White blood cells, /μL | 10 000 | [7700-13 200] |
| Hemoglobin, g/dL | 11.4 | [9.7-12.2] |
| ESR, mm/hour | 22 | [12-43] |
| Albumin, g/dL | 3.7 | [3.5-4.2] |
| C-reactive protein, mg/dL | 0.23 | [0.08-0.63] |
| Disease extent, Paris classification, number of cases [%] | ||
| E1-3 | 0 | |
| E4 | 21 | [100] |
| Treatment at study inclusion, number of cases [%] | ||
| Mesalazine | 19 | [90.5] |
| Immunomodulators | 8 | [38.0] |
| Biologics | 4 | [19.0] |
| Total colectomy | 4 | [19.0] |
| Characteristic | Median [interquartile range or percentage] | |
| [n = 21] | ||
| Male, number of cases [%] | 17 | [80.9] |
| Disease onset, months | 30 | [23-43] |
| Age at diagnosis, months | 36 | [27-58] |
| Family history of IBD, number of cases [%] | 6 | [28.5] |
| PUCAI | 40 | [15-50] |
| Blood test | ||
| White blood cells, /μL | 10 000 | [7700-13 200] |
| Hemoglobin, g/dL | 11.4 | [9.7-12.2] |
| ESR, mm/hour | 22 | [12-43] |
| Albumin, g/dL | 3.7 | [3.5-4.2] |
| C-reactive protein, mg/dL | 0.23 | [0.08-0.63] |
| Disease extent, Paris classification, number of cases [%] | ||
| E1-3 | 0 | |
| E4 | 21 | [100] |
| Treatment at study inclusion, number of cases [%] | ||
| Mesalazine | 19 | [90.5] |
| Immunomodulators | 8 | [38.0] |
| Biologics | 4 | [19.0] |
| Total colectomy | 4 | [19.0] |
Abbreviations: ESR, erythrocyte sedimentation rate; IBD, inflammatory bowel disease; PUCAI, pediatric ulcerative colitis activity index.
Human Leukocyte Antigen Allele
A total of 25 types of HLA were detected in the VEO-UC cases, including 6 A types, 10 B types, and 9 DR types, as shown in Table 2. The HLA types only observed in mothers of VEO-UC cases were B56 and DR16, as shown in Figure 1. HLA types B52, DR13, and DR15 were found in 9 cases (26.5%, P = .010), 6 cases (17.6%, P = .021), and 13 cases (38.2%, P = .006), respectively, and the HLA genotype frequencies tended to be higher than those reported in the Japanese population, but no statistically significant differences were observed with Bonferroni correction (Table 2).
Detection rate of human leukocyte antigen alleles in very-early-onset ulcerative colitis cases.
| HLA allele | VEO-UC cases [%] | Control data [%]a | P value | Significance Bonferroni | |
| HLA-Class I | [n = 34] | [n = 18 604] | |||
| A2 | 6 | [17.6] | 24.7 | .428 | 0.0083 |
| A11 | 2 | [5.9] | 9.3 | .766 | 0.0083 |
| A24 | 13 | [44.1] | 37.2 | .478 | 0.0083 |
| A26 | 3 | [8.8] | 11.5 | .792 | 0.0083 |
| A31 | 3 | [8.8] | 8.4 | .761 | 0.0083 |
| A33 | 5 | [14.7] | 7.5 | .108 | 0.0083 |
| B7 | 2 | [5.9] | 5.6 | .718 | 0.005 |
| B35 | 4 | [11.8] | 8.2 | .524 | 0.005 |
| B44 | 5 | [14.7] | 7.1 | .091 | 0.005 |
| B46 | 1 | [2.9] | 4.7 | 1 | 0.005 |
| B51 | 3 | [8.8] | 9.1 | 1 | 0.005 |
| B52 | 9 | [26.5] | 11.0 | .010 | 0.005 |
| B58 | 1 | [2.9] | 0.5 | .180 | 0.005 |
| B61 | 3 | [8.8] | 13.1 | .614 | 0.005 |
| B62 | 3 | [11.8] | 8.3 | .527 | 0.005 |
| B71 | 1 | [5.9] | 1.5 | .095 | 0.005 |
| HLA allele | VEO-UC cases [%] | Control data [%]a | P value | Significance Bonferroni | |
| HLA-Class I | [n = 34] | [n = 18 604] | |||
| A2 | 6 | [17.6] | 24.7 | .428 | 0.0083 |
| A11 | 2 | [5.9] | 9.3 | .766 | 0.0083 |
| A24 | 13 | [44.1] | 37.2 | .478 | 0.0083 |
| A26 | 3 | [8.8] | 11.5 | .792 | 0.0083 |
| A31 | 3 | [8.8] | 8.4 | .761 | 0.0083 |
| A33 | 5 | [14.7] | 7.5 | .108 | 0.0083 |
| B7 | 2 | [5.9] | 5.6 | .718 | 0.005 |
| B35 | 4 | [11.8] | 8.2 | .524 | 0.005 |
| B44 | 5 | [14.7] | 7.1 | .091 | 0.005 |
| B46 | 1 | [2.9] | 4.7 | 1 | 0.005 |
| B51 | 3 | [8.8] | 9.1 | 1 | 0.005 |
| B52 | 9 | [26.5] | 11.0 | .010 | 0.005 |
| B58 | 1 | [2.9] | 0.5 | .180 | 0.005 |
| B61 | 3 | [8.8] | 13.1 | .614 | 0.005 |
| B62 | 3 | [11.8] | 8.3 | .527 | 0.005 |
| B71 | 1 | [5.9] | 1.5 | .095 | 0.005 |
| HLA-Class II | [n = 34] | [n = 18 604] | |||
| DR1 | 1 | [2.9] | 5.5 | .720 | 0.0056 |
| DR4 | 4 | [11.8] | 22.8 | .109 | 0.0056 |
| DR8 | 1 | [2.9] | 13.3 | .116 | 0.0056 |
| DR9 | 1 | [2.9] | 13 | .079 | 0.0056 |
| DR11 | 1 | [2.9] | 2.6 | .583 | 0.0056 |
| DR12 | 2 | [5.9] | 7 | .715 | 0.0056 |
| DR13 | 6 | [17.6] | 7.8 | .021 | 0.0056 |
| DR14 | 5 | [14.7] | 5.5 | .081 | 0.0056 |
| DR15 | 11 | [38.2] | 17.4 | .006 | 0.0056 |
| HLA-Class II | [n = 34] | [n = 18 604] | |||
| DR1 | 1 | [2.9] | 5.5 | .720 | 0.0056 |
| DR4 | 4 | [11.8] | 22.8 | .109 | 0.0056 |
| DR8 | 1 | [2.9] | 13.3 | .116 | 0.0056 |
| DR9 | 1 | [2.9] | 13 | .079 | 0.0056 |
| DR11 | 1 | [2.9] | 2.6 | .583 | 0.0056 |
| DR12 | 2 | [5.9] | 7 | .715 | 0.0056 |
| DR13 | 6 | [17.6] | 7.8 | .021 | 0.0056 |
| DR14 | 5 | [14.7] | 5.5 | .081 | 0.0056 |
| DR15 | 11 | [38.2] | 17.4 | .006 | 0.0056 |
Abbreviations: HLA, human leukocyte antigen; VEO-UC, very-early-onset ulcerative colitis.
aThe data cited in reference 15 were used.
Detection rate of human leukocyte antigen alleles in very-early-onset ulcerative colitis cases.
| HLA allele | VEO-UC cases [%] | Control data [%]a | P value | Significance Bonferroni | |
| HLA-Class I | [n = 34] | [n = 18 604] | |||
| A2 | 6 | [17.6] | 24.7 | .428 | 0.0083 |
| A11 | 2 | [5.9] | 9.3 | .766 | 0.0083 |
| A24 | 13 | [44.1] | 37.2 | .478 | 0.0083 |
| A26 | 3 | [8.8] | 11.5 | .792 | 0.0083 |
| A31 | 3 | [8.8] | 8.4 | .761 | 0.0083 |
| A33 | 5 | [14.7] | 7.5 | .108 | 0.0083 |
| B7 | 2 | [5.9] | 5.6 | .718 | 0.005 |
| B35 | 4 | [11.8] | 8.2 | .524 | 0.005 |
| B44 | 5 | [14.7] | 7.1 | .091 | 0.005 |
| B46 | 1 | [2.9] | 4.7 | 1 | 0.005 |
| B51 | 3 | [8.8] | 9.1 | 1 | 0.005 |
| B52 | 9 | [26.5] | 11.0 | .010 | 0.005 |
| B58 | 1 | [2.9] | 0.5 | .180 | 0.005 |
| B61 | 3 | [8.8] | 13.1 | .614 | 0.005 |
| B62 | 3 | [11.8] | 8.3 | .527 | 0.005 |
| B71 | 1 | [5.9] | 1.5 | .095 | 0.005 |
| HLA allele | VEO-UC cases [%] | Control data [%]a | P value | Significance Bonferroni | |
| HLA-Class I | [n = 34] | [n = 18 604] | |||
| A2 | 6 | [17.6] | 24.7 | .428 | 0.0083 |
| A11 | 2 | [5.9] | 9.3 | .766 | 0.0083 |
| A24 | 13 | [44.1] | 37.2 | .478 | 0.0083 |
| A26 | 3 | [8.8] | 11.5 | .792 | 0.0083 |
| A31 | 3 | [8.8] | 8.4 | .761 | 0.0083 |
| A33 | 5 | [14.7] | 7.5 | .108 | 0.0083 |
| B7 | 2 | [5.9] | 5.6 | .718 | 0.005 |
| B35 | 4 | [11.8] | 8.2 | .524 | 0.005 |
| B44 | 5 | [14.7] | 7.1 | .091 | 0.005 |
| B46 | 1 | [2.9] | 4.7 | 1 | 0.005 |
| B51 | 3 | [8.8] | 9.1 | 1 | 0.005 |
| B52 | 9 | [26.5] | 11.0 | .010 | 0.005 |
| B58 | 1 | [2.9] | 0.5 | .180 | 0.005 |
| B61 | 3 | [8.8] | 13.1 | .614 | 0.005 |
| B62 | 3 | [11.8] | 8.3 | .527 | 0.005 |
| B71 | 1 | [5.9] | 1.5 | .095 | 0.005 |
| HLA-Class II | [n = 34] | [n = 18 604] | |||
| DR1 | 1 | [2.9] | 5.5 | .720 | 0.0056 |
| DR4 | 4 | [11.8] | 22.8 | .109 | 0.0056 |
| DR8 | 1 | [2.9] | 13.3 | .116 | 0.0056 |
| DR9 | 1 | [2.9] | 13 | .079 | 0.0056 |
| DR11 | 1 | [2.9] | 2.6 | .583 | 0.0056 |
| DR12 | 2 | [5.9] | 7 | .715 | 0.0056 |
| DR13 | 6 | [17.6] | 7.8 | .021 | 0.0056 |
| DR14 | 5 | [14.7] | 5.5 | .081 | 0.0056 |
| DR15 | 11 | [38.2] | 17.4 | .006 | 0.0056 |
| HLA-Class II | [n = 34] | [n = 18 604] | |||
| DR1 | 1 | [2.9] | 5.5 | .720 | 0.0056 |
| DR4 | 4 | [11.8] | 22.8 | .109 | 0.0056 |
| DR8 | 1 | [2.9] | 13.3 | .116 | 0.0056 |
| DR9 | 1 | [2.9] | 13 | .079 | 0.0056 |
| DR11 | 1 | [2.9] | 2.6 | .583 | 0.0056 |
| DR12 | 2 | [5.9] | 7 | .715 | 0.0056 |
| DR13 | 6 | [17.6] | 7.8 | .021 | 0.0056 |
| DR14 | 5 | [14.7] | 5.5 | .081 | 0.0056 |
| DR15 | 11 | [38.2] | 17.4 | .006 | 0.0056 |
Abbreviations: HLA, human leukocyte antigen; VEO-UC, very-early-onset ulcerative colitis.
aThe data cited in reference 15 were used.
The human leukocyte antigen (HLA) types included in very-early-onset ulcerative colitis cases. The HLA-A, -B, and -DR for 17 very-early-onset ulcerative colitis cases (Pt) and their mothers (Mo). The cells colored gray and blue indicate the detected HLA types with or without concordance with maternal HLA, respectively. Dots represent homozygous alleles.
Maternal HLA characteristics were then investigated, and the concordance rate between the 3 HLA types showing potential predominance in VEO-UC cases and the maternal HLA types was examined. Notable concordance was observed in B52 (30.8%), DR13 (14.3%), and DR15 (50.0%) (Table 3). In 6 cases (Nos. 1, 2, 3, 5, 6, and 7) with a family history of UC, A24 and B52 were observed in 5 cases (83.3%) and 3 cases (50.0%), respectively. Duplicated alleles and fully matched maternal HLAs were found in 10 (62.5%) and 6 (37.5%) cases, respectively, showing that 12 (75.0%) cases were observed with either of them (Figure 1 and Table 3).
Characteristics of maternal human leukocyte antigens in this study.
| Maternal HLA matched with child, number of cases [%] | ||
| [n = 13] | ||
| A24 | 7 | [53.8] |
| A33 | 3 | [23.1] |
| B44 | 2 | [15.4] |
| B52 | 4 | [30.8] |
| [n = 14] | ||
| DR13 | 2 | [14.3] |
| DR14 | 2 | [14.3] |
| DR15 | 7 | [50.0] |
| Maternal HLA matched with child, number of cases [%] | ||
| [n = 13] | ||
| A24 | 7 | [53.8] |
| A33 | 3 | [23.1] |
| B44 | 2 | [15.4] |
| B52 | 4 | [30.8] |
| [n = 14] | ||
| DR13 | 2 | [14.3] |
| DR14 | 2 | [14.3] |
| DR15 | 7 | [50.0] |
| Homozygous maternal alleles, number of cases [%] | [n = 16] | |
| Total | 10 | [62.5] |
| A24 | 6 | [37.5] |
| A33 | 1 | [6.2] |
| B35 | 1 | [6.2] |
| B62 | 1 | [6.2] |
| DR15 | 3 | [18.8] |
| Homozygous maternal alleles, number of cases [%] | [n = 16] | |
| Total | 10 | [62.5] |
| A24 | 6 | [37.5] |
| A33 | 1 | [6.2] |
| B35 | 1 | [6.2] |
| B62 | 1 | [6.2] |
| DR15 | 3 | [18.8] |
| Full matched maternal HLAs, number of cases [%] | [n = 16] | |
| Total | 6 | [37.5] |
| HLA-A | 4 | [25.0] |
| HLA-B | 3 | [18.8] |
| HLA-DR | 2 | [12.5] |
| Full matched maternal HLAs, number of cases [%] | [n = 16] | |
| Total | 6 | [37.5] |
| HLA-A | 4 | [25.0] |
| HLA-B | 3 | [18.8] |
| HLA-DR | 2 | [12.5] |
HLA, human leukocyte antigen.
Characteristics of maternal human leukocyte antigens in this study.
| Maternal HLA matched with child, number of cases [%] | ||
| [n = 13] | ||
| A24 | 7 | [53.8] |
| A33 | 3 | [23.1] |
| B44 | 2 | [15.4] |
| B52 | 4 | [30.8] |
| [n = 14] | ||
| DR13 | 2 | [14.3] |
| DR14 | 2 | [14.3] |
| DR15 | 7 | [50.0] |
| Maternal HLA matched with child, number of cases [%] | ||
| [n = 13] | ||
| A24 | 7 | [53.8] |
| A33 | 3 | [23.1] |
| B44 | 2 | [15.4] |
| B52 | 4 | [30.8] |
| [n = 14] | ||
| DR13 | 2 | [14.3] |
| DR14 | 2 | [14.3] |
| DR15 | 7 | [50.0] |
| Homozygous maternal alleles, number of cases [%] | [n = 16] | |
| Total | 10 | [62.5] |
| A24 | 6 | [37.5] |
| A33 | 1 | [6.2] |
| B35 | 1 | [6.2] |
| B62 | 1 | [6.2] |
| DR15 | 3 | [18.8] |
| Homozygous maternal alleles, number of cases [%] | [n = 16] | |
| Total | 10 | [62.5] |
| A24 | 6 | [37.5] |
| A33 | 1 | [6.2] |
| B35 | 1 | [6.2] |
| B62 | 1 | [6.2] |
| DR15 | 3 | [18.8] |
| Full matched maternal HLAs, number of cases [%] | [n = 16] | |
| Total | 6 | [37.5] |
| HLA-A | 4 | [25.0] |
| HLA-B | 3 | [18.8] |
| HLA-DR | 2 | [12.5] |
| Full matched maternal HLAs, number of cases [%] | [n = 16] | |
| Total | 6 | [37.5] |
| HLA-A | 4 | [25.0] |
| HLA-B | 3 | [18.8] |
| HLA-DR | 2 | [12.5] |
HLA, human leukocyte antigen.
In addition, based on the HLA of VEO-UC patients and their mothers, HLA haplotypes for 3 loci (A-B-DR) were determined in 13 cases, and for 2 loci (B-DR) in 2 cases. HLA haplotypes were examined for the combinations of 2 and 3 loci, respectively, showing potential differences compared with previous reports15,16 in the combinations of 2 loci (A24-B52 [P = .010], A24-DR15 [P = .054], B44-DR13 [P = .081], and B52-DR15 [P = .065]) and in the combinations of 3 loci (A24-B52-DR15 [P = .02], A33-B44-DR13 [P = .11]) (Table 4). Bonferroni correction demonstrated significant differences for A24-B52 (P = .010, significance level: 0.017) and A24-B52-DR15 (P = .023, significance level: 0.025).
Two- or 3-locus haplotypes of human leukocyte antigen retention in this study.
| Haplotypes | VEO-UC cases [%] | Control data [%] | P value | Significance Bonferroni | |
| [n = 26 or 30a] | [n = 18 604]b | ||||
| A24-B7 | 2 | [7.7] | 4.3 | .307 | 0.017 |
| A24-B52 | 7 | [26.9] | 9.7 | .010 | 0.017 |
| A33-B44 | 3 | [11.5] | 5.9 | .195 | 0.017 |
| A33-DR13 | 3 | [11.5] | 4.8 | .128 | 0.025 |
| A24-DR15 | 6 | [20.6] | 10.7 | .054 | 0.025 |
| B44-DR13 | 4a | [13.3] | 5.5 | .081 | 0.025 |
| B52-DR15 | 6a | [20.0] | 9.7 | .065 | 0.025 |
| A24-B52-DR15 | 6 | [23.1] | 8.5 | .02 | 0.025 |
| A33-B44-DR13 | 3 | [11.5] | 4.5 | .11 | 0.025 |
| Haplotypes | VEO-UC cases [%] | Control data [%] | P value | Significance Bonferroni | |
| [n = 26 or 30a] | [n = 18 604]b | ||||
| A24-B7 | 2 | [7.7] | 4.3 | .307 | 0.017 |
| A24-B52 | 7 | [26.9] | 9.7 | .010 | 0.017 |
| A33-B44 | 3 | [11.5] | 5.9 | .195 | 0.017 |
| A33-DR13 | 3 | [11.5] | 4.8 | .128 | 0.025 |
| A24-DR15 | 6 | [20.6] | 10.7 | .054 | 0.025 |
| B44-DR13 | 4a | [13.3] | 5.5 | .081 | 0.025 |
| B52-DR15 | 6a | [20.0] | 9.7 | .065 | 0.025 |
| A24-B52-DR15 | 6 | [23.1] | 8.5 | .02 | 0.025 |
| A33-B44-DR13 | 3 | [11.5] | 4.5 | .11 | 0.025 |
Abbreviations: HLA, human leukocyte antigen; VEO-UC, very-early-onset ulcerative colitis.
aIncluding four alleles for which only the A locus could not be determined.
bThe data cited in reference 15 were used.
Two- or 3-locus haplotypes of human leukocyte antigen retention in this study.
| Haplotypes | VEO-UC cases [%] | Control data [%] | P value | Significance Bonferroni | |
| [n = 26 or 30a] | [n = 18 604]b | ||||
| A24-B7 | 2 | [7.7] | 4.3 | .307 | 0.017 |
| A24-B52 | 7 | [26.9] | 9.7 | .010 | 0.017 |
| A33-B44 | 3 | [11.5] | 5.9 | .195 | 0.017 |
| A33-DR13 | 3 | [11.5] | 4.8 | .128 | 0.025 |
| A24-DR15 | 6 | [20.6] | 10.7 | .054 | 0.025 |
| B44-DR13 | 4a | [13.3] | 5.5 | .081 | 0.025 |
| B52-DR15 | 6a | [20.0] | 9.7 | .065 | 0.025 |
| A24-B52-DR15 | 6 | [23.1] | 8.5 | .02 | 0.025 |
| A33-B44-DR13 | 3 | [11.5] | 4.5 | .11 | 0.025 |
| Haplotypes | VEO-UC cases [%] | Control data [%] | P value | Significance Bonferroni | |
| [n = 26 or 30a] | [n = 18 604]b | ||||
| A24-B7 | 2 | [7.7] | 4.3 | .307 | 0.017 |
| A24-B52 | 7 | [26.9] | 9.7 | .010 | 0.017 |
| A33-B44 | 3 | [11.5] | 5.9 | .195 | 0.017 |
| A33-DR13 | 3 | [11.5] | 4.8 | .128 | 0.025 |
| A24-DR15 | 6 | [20.6] | 10.7 | .054 | 0.025 |
| B44-DR13 | 4a | [13.3] | 5.5 | .081 | 0.025 |
| B52-DR15 | 6a | [20.0] | 9.7 | .065 | 0.025 |
| A24-B52-DR15 | 6 | [23.1] | 8.5 | .02 | 0.025 |
| A33-B44-DR13 | 3 | [11.5] | 4.5 | .11 | 0.025 |
Abbreviations: HLA, human leukocyte antigen; VEO-UC, very-early-onset ulcerative colitis.
aIncluding four alleles for which only the A locus could not be determined.
bThe data cited in reference 15 were used.
Evaluation of Maternal Microchimerism
Each of the 4 cases of VEO-UC and UC with a history of total colectomy was included in this study. Immunohistochemical staining demonstrated that MMc lymphocytes were identified in 3 of the 4 VEO-UC cases, but not in the control UC cases (Figure 2). MMc lymphocytes were observed in multiple regions of the epithelium or submucosa of the rectum. One of these cases had CD3-positive MMc (Figure 2). A CD3-positive cell was found in a case showing MMc lymphocytes.
The findings of immunohistochemistry (IHC) and fluorescent in situ hybridization (FISH). A positive XX and CD3 signal cell is identified in the submucosal layer of very-early-onset ulcerative colitis. A arrowhead and arrow indicates X and CD3, respectively (A). IHC and FISH findings of a control ulcerative colitis case indicating no evidence of XX. A arrowhead and arrow indicates X and Y, respectively (B). Standard control. A arrowhead and arrow indicates X and Y, respectively (C).
Discussion
This study provides new insights into the genetic and immunological factors that may contribute to the development of VEO-UC, a severe form of IBD that occurs before the age of 6 years. The present results suggest a strong association between specific HLA alleles, such as HLA-B52 and HLA-DR15, and the onset of VEO-UC, aligning with results observed in adult UC studies in Japan. Thus, specific HLA alleles are strongly associated with the onset of VEO-UC, highlighting the potential influence of genetic predisposition in pediatric UC. In addition, MMc was found in the intestinal tissues of several patients with VEO-UC, providing further evidence of the complex immune mechanisms involved in this disease.
Features of Human Leukocyte Antigen Alleles and Clinical Significance
This study showed a certain tendency for specific HLA types (B52, DR13, and DR15) in VEO-UC cases compared to the general Japanese population data. Consistent with previous Japanese HLA and adult-onset UC studies, it suggested that VEO-UC may be a spectrum of typical UC rather than a distinctly different disease. In particular, the fact that both HLA-B52 and -DR15, which have been reported to be associated with adult-onset UC in Japanese,7,18 were detected in VEO-UC, may support this hypothesis. In addition, the VEO-UC patients included in the present study had no identified mutations in the primary immunodeficiency or monogenic IBD gene panels, and no significant novel mutations have been identified in the simultaneous whole genome sequencing to date. Thus, despite the early onset and severity of VEO-UC, its genetic background may be homologous to that of adult UC.
Maternal Microchimerism in Very-Early-Onset Ulcerative Colitis
The present study explored the possible role of MMc in developing VEO-UC. MMc occurs when fetal cells persist in the mother’s body or maternal cells remain in the child after birth. This phenomenon is linked to several autoimmune diseases, and its presence in VEO-UC patients may indicate an immunological mechanism contributing to disease development.10–12 The present study identified MMc lymphocytes in the intestinal tissues of 3 of 4 VEO-UC patients, a finding not observed in control UC cases. This suggests that MMc could play a role in the pathogenesis of VEO-UC, particularly given the similarities in histopathological features between VEO-UC and graft-versus-host disease (GVHD), which may also involve immune-mediated damage to the gastrointestinal tract.19 In addition, a particularly valuable finding of the present study was the identification of MMc in VEO-UC patients, which had not been identified in a previous study conducted in children over 7 years of age,20 suggesting that VEO-UC may have a pathology that differentiates it from other pediatric and adult UC cases. This observation aligns with diseases such as dermatomyositis and biliary atresia, where maternal immune contributions have been implicated. The presence of MMc in intestinal tissues, particularly within the rectal epithelium and submucosa, supports the hypothesis that maternal cells may participate in the immune dysregulation seen in VEO-UC. This mechanism may be unique to VEO-UC, in which maternal immune factors interact more prominently during the early developmental stages of the immune system.
MMc in patients with VEO-UC presents intriguing possibilities. However, it is essential to note that the present study could not establish a direct causal relationship between MMc and the onset or progression of the disease. The present results support the hypothesis that MMc may contribute to immune dysregulation in VEO-UC. Nonetheless, further research is needed to determine whether MMc is a primary driver of the disease or merely a bystander phenomenon. The histopathological similarities between VEO-UC and GVHD suggest that MMc cells might contribute to the immune-mediated tissue damage seen in VEO-UC.21 Still, additional studies are required to clarify the exact mechanisms by which maternal cells affect the disease process.
However, there were also some reports regarding the frequency of MMc in healthy children. In a Scandinavian report evaluating MMc in the tonsil tissue of 20 healthy children aged 2-15 years, MMc was identified in 4 cases (20%).22 In addition, a report evaluating circulating MMc cells in Japanese persons with type 1 diabetes mellitus was able to identify circulating MMc cells in blood in 16% of the normal control group.23 Therefore, MMc cells can be identified in healthy individuals, but at a lower rate than the 75% in the present study. However, a larger sample size may be required to demonstrate whether this is a significantly higher rate in VEO-UC.
Clinical Implications
The findings of this study highlight the importance of considering both genetic and immunological factors in the diagnosis and treatment of VEO-UC. Whereas HLA-B52 and HLA-DR15 may serve as genetic markers for susceptibility, the detection of MMc introduces a new dimension to understanding disease mechanisms. Therapies targeting immune regulation, particularly those addressing maternal immune contributions, may hold promise for managing VEO-UC. However, additional studies are needed to validate these findings and explore their therapeutic potential.
The results of the present study suggest that genetic predisposition involving HLA loci, particularly B52 and DR15, may contribute to disease onset, but they are insufficient to fully explain the severe and early-onset nature of VEO-UC. Including or proposing further investigations, such as a mixed lymphocyte reaction test comparing patients with their siblings, could indeed provide more clarity on the immunological mechanisms involved. The observed MMc raises intriguing possibilities about the interplay between maternal immune cells and the host’s immune system in this disease context. Future research with larger sample sizes and detailed functional analyses is required to clarify these relationships. In addition, investigating the similarities between VEO-UC and other immune-mediated diseases, such as GVHD, may offer insights into shared mechanisms and potential therapeutic approaches. Expanding research to include diverse populations will also be crucial for understanding the global relevance of these findings.
Study Limitations
This study has several limitations. First, the sample size was small, particularly for the VEO-UC cohort, which may limit the generalizability of the findings. Furthermore, the study focused on patients from a limited geographical area, which could introduce population bias. Second, VEO-UC is a rare disease, and the accurate incidence rate in Japan is unclear. Because similar conditions are considered to apply in other countries, it is difficult to discuss ethnic variation. Third, although HLA typing suggested a possible accumulation of certain alleles, the functional mechanisms underlying these potential associations remain unclear. Fourth, Bonferroni correction was conducted for the statistical analysis, but due to its potential conservatism, clinically valuable associations may have been underestimated. Fifth, the functional mechanisms related to HLA and MMc to VEO-UC remain unclear. It is expected that these issues will be clarified by conducting large-scale, multicenter studies including large numbers of various populations and advanced molecular analysis in future research.
In conclusion, the present study demonstrated that VEO-UC has genetic and immunological characteristics similar to adult UC and may also have a background of maternal-derived immunological abnormalities in Japanese children. This provides evidence of a genetic predisposition to this severe form of UC. Detecting MMc in intestinal tissues suggests a possible immunological mechanism contributing to disease development. Though the exact roles of these factors remain to be fully elucidated, the present findings suggest that genetic and immunological factors likely contribute to the complex etiology of VEO-UC. Future research with larger cohorts and functional studies will be crucial to better understand the interplay among HLA alleles, MMc, and other factors in the development of pediatric UC.
Acknowledgments
The authors would like to thank the patients and families and paramedical staff who contributed to this study. The authors would also like to express their sincere gratitude to Dr Toshihiro Muraji of the Kirishima City Medical Centre, Kagoshima Prefecture, Japan, and Dr Hidetaka Eguchi of the Biomedical Research Center, Juntendo University Graduate School of Medicine, Tokyo, Japan, for their invaluable contributions to the methodology, HLA analysis, and fluorescence immunostaining of this study.
Funding
This work was supported by a Japan Society for the Promotion of Science KAKENHI Grant-in-Aid for Young Scientists (grant number 23K14991).
Conflicts of Interest
The authors have no conflicts of interest to declare.
Ethical Considerations
All participants or their guardians provided a written informed consent. The study was conducted according to Good Clinical Practice guidelines. Our institute’s institutional review board or ethics committee approved the study protocols (No. E22-0258). An opt-out consent model was implemented, allowing participants to refuse participation via the hospital’s website.
