Maternal microchimerism in pediatric inflammatory bowel disease

Abstract

Inflammatory bowel disease (IBD) shares many immunologic and clinical characteristics with graft versus host disease caused by allogeneic T lymphocytes after hematopoietic cell transplantation. Since maternal cells are known to enter the fetal circulation in a high proportion of pregnancies, we hypothesized that maternal engraftment in the fetus results in immune sequelae that can lead to IBD.

Method

The presence and extent of maternal microchimerism in tissues and blood samples from patients with Crohn's, Ulcerative colitis (UC), and control groups were determined using kinetic Polymerase Chain Reaction (kPCR) to detect maternal- and patient-specific HLA types. In addition, fluorescent in situ hybridization (FISH) was employed to detect maternal cells in biopsies from patients with IBD.

Results

Using kPCR, maternal microchimerism was observed in 9 of the 16 (56%) patients with IBD and 6 out of 15 of the control group (40%) (P=NS). Five of 10 Crohn's patients had evidence of maternal microchimerism (50%) (P=NS). Four of six UC patients had evidence of maternal microchimerism in gut tissues (67%) (P=NS). There was no correlation between maternal michrochimerism and disease activity, disease location or granulomas in patients with IBD. Using FISH, five male Crohn's and five male UC patient's intestinal biopsies were analyzed for maternal microchimerism. No maternal cells were identified.

Conclusion

There is nothing in the data to suggest that patients with IBD differ from disease controls in their frequency of maternal microchimerism in either blood or gut mucosal tissues. These data suggest that maternal microchimerism in blood and biopsies is a relatively common phenomenon that has neither positive nor negative impact on IBD.

Introduction

Inflammatory Bowel Disease (IBD) is a chronic disease characterized by acute and chronic inflammation of the bowel, IBD includes two primary diagnoses: Crohn disease (CD) and ulcerative colitis (UC). The etiology is unknown. Many hypotheses exist regarding the primary cause or trigger of the disease including possible autoimmune etiology.¹

Another possibility is that IBD has an alloimmune rather than an autoimmune etiology. Maternal microchimerism occurs when maternal cells reside in the body of offspring. Maternal-fetal lymphocyte transfer is known to occur during pregnancy starting as early as the tenth week of gestation and continuing up to delivery.^2,3 Disease has been shown to occur in mothers who have had fetal microchimerism, and children/adults who have had maternal microchimerism. Children who have immunodeficiency are at risk for graft versus host disease owing to engraftment of maternal lymphocytes. In both SCID (severe combined immunodeficiency) and DiGeorge syndrome, maternal microchimerism causing graft versus host disease (GVHD) has been described.^4,5

Although significant differences exist between the pathophysiology of IBD and acute GVHD, they do share some histologic and clinical characteristics. In both IBD and GVHD, primary disease can occur within the gastrointestinal tract, liver and skin. Although GVHD is primarily characterized by apoptotic cells within the gastrointestinal tract, both GVHD as well as IBD can have an inflammatory infiltrate composed of neutrophils and mononuclear cells within the lamina propria and epithelium. In each disease, cystic dilatation of glands or crypts lined by regenerative epithelium, crypt abscesses, and frank epithelial destruction have been described.⁶ The similarities can be seen within the pathophysiology of both disease processes. The cellular effectors of GVHD and of IBD include CD4⁺ T cells, CD8⁺ T cells, natural killer (NK) cells, and probably monocytes and granulocytes. In both, CD4⁺ T-lymphocytes are important in the development of the enteropathy. Although neither disease process is completely elucidated, it appears that CD4⁺ T-cells cause intestinal cell damage via proinflammatory secretions versus a direct cytotoxic effect (i.e. CD8⁺ T-cells). CD4⁺ T-cells recognize MHC class II. MHC class II expression is increased in both IBD and GVHD.^6–12

Animal models for IBD take many forms, from chemically induced to gene knockout models. One model that has shown much promise is the adoptive transfer model. In this model there is a selective transfer of alloimmune cells to immunocompromised host animals which causes a disease similar to GVHD.^2,13–18 Another such model uses proinflammatory CD4⁺CD45RB^High T-cells from wild-type donor mice transferred to immunodeficient immunodeficient mice. In these mice wasting syndrome with transmural intestinal inflammation is seen starting 5 to 8 weeks after cell transfer.^19–21

Clinically, IBD and GVHD also share many similarities. Both cause gastrointestinal distress and nutritional compromises. In both, treatment regimes center on immunomodulating medications. Systemic glucocorticoid medications are the mainstay of treatment for acute severe disease for GVHD and IBD. Because of these similarities, we hypothesized that patients with IBD have maternal microchimerism and that this microchimerism leads to alloimmune type disease similar to GVHD.

Results

A total of 44 mother/patient pairs took part in the study with 22 mother/patient pairs in the IBD and the control group. The group participants were comparable in age and sex with the mean age of the patients with IBD being 13 ± 3.1 years and the controls 12 ± 2.6 years (age range 8–18 years for both groups). Using the informative loci of the patients and their mothers, evidence of maternal microchimerism was explored within blood specimens and a minimum of seven tissue specimens (stomach, duodenum, rectosigmoid, descending colon, transverse colon, ascending colon, cecum and/or ileum). Thirteen patients' results were excluded from final analysis secondary to high background on negative controls which decreased assay sensitivity.

Maternal microchimerism was observed using kPCR in nine of the sixteen (56%) patients with IBD and six out of fifteen of the control group (40%) (P=0.4). Two patients with IBD (12.5%) and three controls (20%) (P=0.5) had maternal microchimerism in the blood, while eight patients with IBD (50%) and four control patients had chimerism in their tissue specimens (26%) (P=0.3). Five of ten Crohn patients had evidence of maternal microchimerism (50%) (P=0.7). Microchimerism was found in the blood of two of the Crohn patient (0.2 to 2 copies of maternal DNA per 2.5 × 10⁴ copies of patient's DNA) and in the tissue samples of five of the Crohn patients (0.4 to 2.5 copies of maternal DNA per 2.5 × 10⁴ copies of patient DNA). Four of six ulcerative colitis patients had evidence of maternal microchimerism in gut tissues (67%) (P=0.8). Microchimerism was not found in the blood of patients with ulcerative colitis. In the tissue samples of the ulcerative colitis patients, maternal microchimerism was found in a range of 0.2 to 13.9 copies of maternal DNA per 2.5 × 10⁴ copies of patient DNA. In controls, maternal microchimerism in the blood ranged from 0.5 to 1.2 copies of maternal DNA per 2.5 × 10⁴ copies of patient DNA and in the tissue ranged from 0.2 to 5.2 copies of maternal DNA per 2.5 × 10⁴ copies of patient DNA. Disease activity was defined as no disease, mild disease, and moderate/severe disease. Using Spearman's rank correlation there was no correlation between maternal michrochimerism and disease activity, disease location or granulomas in patients with IBD (Table 1).

Table 1.

Maternal microchimerism in inflammatory bowel disease

Study #	Age at biopsy	Sex	Primary Dx	Secondary Dx	Blood Chimerism	Tissue Disease Activity
						Rec/sig	DC	TC	AC	Cecum	Ileum	Sto	Duod
1	8	M	IBD	UC	A	S	S	S	S	S		S	M
2	16	F	Control	FAPS	A		O	O	O	O		O	O
3	12	M	Control	FAP and esophagitis	A		O	O	O	O		O	O
4	9	M	IBD	Crohn	P	S	S	S	S	S	S	S	O
5	12	M	Control	Juvenile polyposis syndrome	A		O	O	O	O		O	O
6	11	M	Control	FAPS	P	O	O	O	O	O		O	O
7	13	M	IBD	Crohn	A	M	M	M	M	M	S	M	O
8	9	F	IBD	Crohn	P	S	S	S	S	S	S	S	S
9	12	F	Control	FAPS	A	O	O	O	O	O	O	M	O
10	17	M	IBD	Crohn	A	S	S	S		O		O	O
11	11	M	IBD	Crohn	A	S	S	S	S	S		S	O
12	15	M	Control	Eosinophilic gastroenteritis	A	M	M	M		M	M	O	O
13	18	M	Control	Eosinophilic gastroenteritis	P	O	O	O		O	O	M	S
14	11	M	Control	Esophagitis amd FAP	A	O	O	O		O	O	O	O
15	8	F	Control	Fructose intolerance	A	O	O	O		O	O	O	O
16	12	M	Control	FAPS	P	O	O	O		O	O	O	O
17	16	F	IBD	Crohn	A	O	O	O	O	O	O	M	M
18	10	M	IBD	UC	A	O	O	O		O	O	O	O
19	11	F	Control	Cecal diverticulum/bacterial overgrowth	A	O	M	M		S	O	O	O
20	11	F	Control	Eosinophilic colitis	A	O	O	O	O	O		O	O
21	15	M	IBD	Crohn	A	O	O	O		S	S	M	M
22	12	M	Control	Peutz-Jaeger	A	M	M	M	M	M	M	M
23	18	M	IBD	UC	A	S	S	S		S	M	M	M
24	8	M	Control	Acute self-limited colitis	A	O	O	O		O	M	O	O
25	12	M	Control	Acute self-limited colitis	A	M	M	M		M	M	O	M
26	15	F	IBD	Crohn	A	O	O	O		M	S	M	M
27	15	F	IBD	UC	A	S	S	S	S	S		O	O
28	15	M	IBD	Crohn	A	G,S	M	M		G,S	G,S	M	O
29	11	M	IBD	UC	A	S	S	S	S	S		O	M
30	12	M	IBD	Crohn	A	G,S	O	O		M		G,S	O
31	14	F	IBD	UC	A	S	S	S	S	S	O	M	O

Key for tissue samples: O, No disease activity; M, Mild disease activity; S, Moderate/severe disease activity; G, Granuloma; P, Microchimerism Present in blood; A, Microchimerism absent in blood; Bold areas, Microchimerism present in tissue specimen; FAPS, Functional abdominal pain syndrome; DC, Descending colon; TC, Transverse colon; AC, Ascending colon

Ten male patients with IBD (five Crohn disease and five ulcerative colitis) were further analyzed for maternal microchimerism using FISH. Individuals with the highest degree of maternal microchimerism on kPCR where chosen for this part of the study. Strict criteria was used to determine maternal microchimerism as depicted by two X chromosomes, i.e. two red signals, with no Y chromosome (green signal) within the blue nuclear materials. Distance required between X chromosomes was greater than the individual signals themselves. Although multiple potential chimeric cells were identified, none fit all the criteria.

Discussion

Maternal microchimerism is common in healthy individuals,^3,22 an observation further confirmed by this study. The high occurrence of maternal microchimerism in healthy individuals has been suggested to have a tolerogenic effect that may contribute to long-term microchimerism. In contrast, in utero transplantation of haploidentical cells does not always lead to immune tolerance²³ and may lead to immune sensitization. Thus, the immunological consequences of the migration of maternal cells to the fetus appear to be variable. In most cases the maternal cells are likely to be cleared by the host's immune system or they may escape destruction by the immune system leading to engraftment. Mold, et al have shown that suppressive T_reg cells are generated against noninherited maternal antigens in utero and persists into adulthood thus causing tolerance for maternal chimerism.²⁴ Conversely, it is felt in some autoimmune disorders, engrafted maternal cells in the fetus may subsequently be rejected by the immune system of the offspring, leading to an inflammatory process.²⁵ Thus, among the possible outcomes of haplogeneic maternal chimerism, ranging from tolerance to sensitization, the hypothesis that maternal chimerism disrupts the normal development of self-tolerance leading to autoimmune disease is worthy of consideration and testing.

From this study maternal microchimerism does not appear to be increased in patients with either Crohn disease or ulcerative colitis as compared with controls. In patients with chimerism, the chimerism was not associated with disease location, disease activity or granuloma formation. To our knowledge this is the first study demonstrating maternal chimerism in the gut. Fetal chimerism has been detected in maternal intestinal tissue.^22,26 These data further confirm that microchimerism itself is a relatively common phenomenon that had no obvious positive or negative impact on IBD.

There are some fundamental limitations with this study. The study sample size is small. Although our initial aim was to have a sample size of 75 patients to properly power this study, we where unable to reach this number given economic and timing constraints. In this study, there is a trend towards increase chimerism in IBD. This trend is not statistically significant but could become so with increase number of patients studied. Despite this possibility, it seems unlikely that even if there was a statistical significant difference, that this would translate into either an etiologic or clinical significance given the relative small amount of chimerism and the lack of association between disease activity and chimerism. This too may be an oversimplication of the immune regulation involving microchimerism. Nonetheless, this study is the first to show that maternal microchimerism does not play a local effect on an inflammation within the intestines in inflammatory bowel disease.

Methods

A prospective case control study was conducted involving assessment of HLA markers in persons with IBD, controls and their mothers. The HLA type of patients and their mothers was initially determined. Then using kinetic polymerase chain reaction (kPCR), the presence and extent of maternal microchimerism in the tissues and blood samples from patients was determined.

Study Subjects.

The protocol was approved by the Seattle Children's Hospital Institutional Review Board. Patients were recruited from the outpatient and inpatient services of the Division of Pediatric Gastroenterology at University of Washington Seattle Children's Hospital. Informed consent was obtained from all subjects. Patients with suspected/possible IBD based on history, physical and laboratory evidence undergoing endoscopy and colonoscopy for diagnostic purposes were invited to participate in this study. Exclusion criteria for this study included use of immunosuppressive medications including azathiopurine, mercaptopurine, methotrexate and any biologic medications. The study patients are summarized in Table 1. Once consent was obtained, blood samples were obtained from patient and mother. Seven biopsy samples from each patient were obtained from both inflamed and non-inflamed appearing tissue during diagnostic endoscopy/colonoscopy. Tissue samples were taken from stomach, duodenum, terminal ileum, cecum, ascending colon, transverse colon, descending colon and rectosigmoid areas. Final diagnosis of Crohn disease and ulcerative colitis were based upon history, physical, radiographic studies, laboratory studies and histology. Control patients diagnosis included both inflammatory and noninflammatory conditions including eosinophilic gastroenteritis, acute/self limiting colitis, polyposis syndrome and irritable bowel disease/functional abdominal pain.

HLA typing and Kinetic Polymerase Chain Reaction (kPCR).

HLA-A, -B, and DRB1 alleles of children and mothers were determined by sequence specific oligonucleotide probe hybridization with microbead arrays (SSO) (One Lambda, Canoge Park, CA). Using the informative HLA types of the patients and their mothers (i.e., types that are different), evidence of maternal DNA was explored within the patient's blood and tissue specimens, using a previously reported method of kPCR.²⁷ DNA was isolated from whole blood and tissue specimens using silica-gel-membrane columns (QIAGEN Inc., Valencia, CA). The minor modifications to the kinetic PCR protocol included use of iCyler from BioRad (Hercules, CA USA) for amplification. Assays have been previously developed to quantify all of the common low resolution HLA-DR types as well as many of the most common HLA-A and B types. For mother-child pairs with no informative HLA alleles for which assays were available, informative polymorphisms in non-HLA loci (e.g., HA-1, KIR0) were informative. For the remaining mother-child pairs, extended HLA typing (i.e., high resolution HLA typing and/or typing of additional loci HLA-C, DQ, and DP) was performed to identify informative polymorphisms. Each sample was tested with three PCR reactions. Each reaction contained 25,000 genomic equivalences (gEq). Our sensitivity was 1/25,000 gEq or 0.004%. Samples with non-shared HLA types were used as negative controls.

Fluorescent in situ hybridization (FISH) of male intestinal biopsies. We used a previously reported method of FISH²⁸ using X and Y-chromosomes probes. Y-chromosomes were stained with a green fluorescent dye, fluorescein isothiocyanate (FITC). X chromosomes were stained with a red fluorescent dye, cyanine 3 (Cy-3). Nuclear material was stained with a blue fluorescent dye 4″, 6″-diamidino-2-phenylindole (DAPI). After initial digestion and staining of the biopsy specimens, we examined a 5 mm intestinal biopsy sections per patient for evidence of female cells, depicted by two red signals, i.e. X chromosomes, with no green signal within the blue nuclear material.