This study was approved by the Medical Ethics Committee of Beijing Children’s Hospital and conducted according to Declaration of Helsinki principles. Clinical information (including the demographic information, details of surgery, complications, clinical course, imaging examination results) and blood or tissue samples were obtained from the patient, parents, and controls. Informed consent has been signed by all the participants.

The PBMCs were freshly isolated from peripheral blood by density gradient centrifugation using Ficoll-Hypaque (GE Healthcare). Briefly, 4 ml freshly isolated EDTA anticoagulated blood was diluted 1:1 with RPMI-1640 medium (Thermo Fisher Scientific), and layered onto 4 ml Ficoll-Hypaque. After centrifuged at 800 g for 20 minutes at room temperature, the buffy layer containing PBMCs was collected, washed and re-suspended in RPMI-1640 medium. Cell viability was determined with trypan-blue (Sigma-Aldrich) staining and counted on an EVE™ automated cell counter (NanoEnTek).

Tissues were fixed in 10% buffered formalin for at least 24 hours, embedded into paraffin. Tissues were then sectioned into 5 μm slice and stained with hematoxylin and eosin (H&E). For Immunohistochemistry, tissue sections were deparaffinized, rehydrated and treated by epitope retrieval using citrate buffer (pH 6.0). The slides were incubated overnight at 4°C with a rabbit anti-human TTC7A antibody (Proteintech, #21600-1-AP) diluted 1:60 in PBS containing 1% BSA. After rinsing three times with PBST, the sections were incubated with biotinylated goat anti-rabbit secondary antibody (Vector Laboratories) and subsequently with streptavidin-labeled HRP (Vector Laboratories). The sections were finally visualized with 3, 3’-diaminobenzidine (DAB) substrate and counterstained with hematoxylin. Digital images were recorded by an Olympus BX53 microscope.

Genomic DNA was extracted with a QIAamp DNA blood midi kit (Qiagen) using the peripheral blood samples from the patient and parents. SureSelect Human All Exon Kit v6.0 (Agilent Technologies) was used to build an exome library. We performed the targeted regions sequencing with 150 bp pair-end runs by a NovaSeq 6000 system (Illumina). The mean depth of sequencing was > 100×. Raw data were mapped to the GRCh37/hg19 human reference sequence. The SAMtools was used to call single nucleotide variants and indels. All the variants were annotated by ANNOVAR. We filtered the common variants with minor allele frequency > 0.05 according to Single Nucleotide Polymorphism databases (gnomAD, dbSNP 147 and 1000 Genomes Project Database) and disease databases (ClinVar and Human Gene Mutation Database, HGMD). Following programs were used to predict the potential impact of the candidate variants on protein function: SIFT (sift.jcvi.org/www/SIFT_enst_submit.html), Polyphen-2 (genetics.bwh.harvard.edu/pph2/), Mutation Taster (www.mutationtaster.org/), and CADD (cadd.gs.washington.edu/). All candidate variants were interpreted according to guidelines of the American College of Medical Genetics and Genomics (ACMG). Sanger sequencing was performed to verify the mutation of TTC7A gene. Primers were designed online using Primer 3.0 software. Primers sequence was listed in Table 1 .

Amino acid conservation was evaluated using the UGENE database with the sequences of Homo sapiens, Pan troglodytes, Macaca mulatta, Bos taurus, Mus musculus, Rattus norvegicus, Gallus gallus, Danio rerio and Xenopus (Silurana) tropicalis.

The sequences of TTC7A were obtained from the UniProtKB (UniProtKB: Q9ULT0). The crystal structure of TTC7A (SMTL ID: 5dse.2.A) was used as the template, which was determined by X-ray diffraction at a resolution of 2.9 Å. The structural impact of mutant Leu69Pro was analyzed by PyMOL software. Residue 69 and certain nearby residues within 3 Å were illustrated. For clear demonstration of the inter-residue relationship, some residues were highlighted in colors with the computed hydrogen bonds being labeled.

DNA extraction was performed from 300 μl blood according to the manufacturer’s instructions (Qiagen). Multiple fluorescence qPCR was used to determine the copy numbers of KREC and TREC, and a real-time PCR assay was performed on ABI 7500 platform (Life Technologies). RPP30 gene was used as a housekeeping gene. For each PCR reaction, 50 ng of genomic DNA was used in a final volume of 20 μl. The PCR setup was as follows: 1× (30s 95°C), and 45× (5s 95°C, 34s 60°C). All DNA samples were run in triplicates alongside no-template controls. KREC and TREC copy numbers were determined by extrapolating the values from a unique standard curve, which were generated by PCR amplification from serial dilution a known concentration of TREC/KREC plasmids. Primers and probes sequence are available on request.

Red blood cells-depleted peripheral blood cells were incubated with the fluorochrome-conjugated monoclonal antibodies for 20 minutes at room temperature in the presence of Fc blocking (BD Biosciences) reagents. All the antibodies used were from BD Pharmingen and listed below: CD3 (clone OKT3), CD19 (clone 4G7), CD45RA (clone HI100), CD4 (clone OKT4), CD8 (clone RPA-T8), CD24 (clone ML5), CD38 (clone HIT2), CD16 (clone 3G8), and CD56 (clone B159). The samples were acquired on a BD LSRFortessa flow cytometer (BD Biosciences), and analyzed using FlowJo software (Tree Star).

Cell lysates from PBMCs were prepared in cold RIPA lysis buffer supplemented with protease and phosphatase inhibitor cocktails (Roche Life Science). Equivalent amounts of total proteins (20-50 μg) were separated by 10% SDS-PAGE gel electrophoresis and subsequently transferred onto a PVDF membrane (Millipore). After blocking in 5% BSA solution, the membranes were incubated with appropriate primary and secondary antibodies, visualized using ECL substrate. The blots were scanned on Fusion Solo Imaging system (VILBER) and analyzed with ImageJ software. The following primary antibodies were used: TTC7A (Proteintech, 21600-1-AP, 1:300 dilution), β-actin (Cell Signaling, 4970, 1:2000 dilution).

Total RNA was extracted from a portion of the colectomy specimen using a RNeasy Kit (Qiagen). qPCR was performed using SYBR Green on an ABI 7500 system (Life Technologies). Primer sequences were listed in Table 1 . The narrow colon tissues of the patient were obtained from surgical resection at the age of 81 days. Control samples were obtained from the noninflammatory location of colon tissues proximal to the dilated segment (relatively normal colon tissue) as controls during the surgery of classic Hirschsprung’s disease (HD). We excluded the possibility of intestinal atresia or CID in both HD patients before the surgery. The selected colon tissues of HD were consistent with the location of the patient in this study.

Total RNAs were extracted from PBMCs. The RNA libraries were constructed by poly(A) protocol. Fragmentation was performed using Fragmentation Buffer. The two strands of cDNA were synthesized using a random hexamer primer and M-MuLV Reverse Transcriptase (RNase H Minus) and DNA polymerase I and RNase H, respectively. The Illumina NovaSeq 6000 system was employed to sequence the cDNA libraries with paired-end 150 bp reads.

The paired-end reads were aligned to human reference genome by Hisat v2.2.1 (5) with Ensembl gene annotation (GRCh37/hg19 assembly). The mapped reads were then processed by Stringtie v2.1.4 to quantitate the gene expression against the Ensembl transcript annotation (6). The RNA-seq data of healthy controls were downloaded from National Center for Biotechnology Information (NCBI) Gene Expression Omnibus (GEO) database (GSE111459) (7). The count-based gene expression data was normalized and processed by R DESeq2 package (8). The genes were considered as differentially expressed genes (DEGs) if their fold changes between the patient and healthy controls were greater than 2. The gene set enrichment analysis (GSEA) was conducted to identify the Gene Ontology (GO) terms or pathways overrepresented by the DEGs. The pathways used for GSEA were curated by Kyoto Encyclopedia of Genes and Genomes (KEGG) (9), Reactome (10), and WikiPathway (11), and collected from MSigDB database. The GSEA was implemented by R enricher function in clusterProfiler package (12).

Differences between groups were calculated using unpaired Student t-test with GraphPad Prism 9 software. All experiments were repeated at least three times, and data were presented as the mean ± SD. The statistical significances were indicated as *, p < 0. 05; **, p < 0.01; ***, p < 0.001; ****, p < 0.0001.

The proband was a boy born to consanguineous parents of Chinese origin, and he was the fifth child of the couple ( Figure 1A ). The first child was diagnosed with MIA and died in the first month of life due to postoperative complications including severe septicemia and multiple organ failure. The second and fourth child manifested with constant incidence of vomiting due to digestive obstruction after birth and died in the neonatal period. The third child was an 11-year-old boy without any obvious healthy issues. The mother had a history of induced abortion (the fifth pregnancy, Figure 1A ). The patient’s uncle (father’s younger brother) is otherwise healthy with congenital absence of the left forearm, and other known relatives did not have congenital malformations or diseases.

The patient was born with a vaginal delivery at an estimated gestational age of 34 weeks. After birth, neonatal intestinal perforation and meconium peritonitis caused by intestinal obstruction were suspected. The patient was diagnosed with MIA involving the pyloric diaphragm, ileum atresia, and colon stenosis after three times of surgery. Furthermore, pathological examination of the ileum adjacent to the atresia areas revealed mucosal erosion, epithelial depolarization, small vascular hyperplasia and dense immune-cell infiltration within the lamina propria. Glands located in the lamina propria were dilated and twisted, with diffused submucosal lymphoproliferative changes ( Figure 1B ).

During hospitalization, the patient experienced seven episodes of sepsis characterized by hyperpyrexia and elevated C-reactive protein (CRP), which in many occasions were accompanied by increased numbers of white blood cell (WBC) and neutrophils ( Supplementary Table 1 ). Coinfection of Klebsiella oxytoca and Methicillin-Resistant Staphylococcus epidermidis were detected by two sets of blood culture drawn from different sites at 21 days after birth. Recurrent and refractory sepsis was diagnosed with poor response to antibiotics including Aztreonam, Amikacin, Meropenem, and Fluconazole. No serious pulmonary infection, vaccine-associated pathogen dissemination, and no signs of fungal and other opportunistic infections had been found. The oftenest episodes of antibiotic-resistant infections could not be explained by the common postoperative complications, which would otherwise be effectively contained by above treatment. Unfortunately, the patient eventually died of respiratory and circulatory failure, electrolyte disturbance, and severe infection at 3 months of age.

To dissect the genetic element that could possibly associate with the disease, we carried out the WES using genomic DNA from the patient and the parents. We identified a novel homozygous missense mutation c. 206T>C in exon 2 of TTC7A gene in the patient that had not been documented in the Clinvar and the HGMD databases. Both unaffected father and mother were heterozygous for TTC7A missense mutation c. 206T>C. Sanger sequencing confirmed one allele of the TTC7A gene was inherited from his father, and the other from the mother ( Figure 1C ). This homozygous missense mutation was predicted as pathogenic by MutationTaster software.

Phylogenetic analysis indicated that the amino acid sequences surrounding the proline 69 residue were highly conserved across different species by UGENE software ( Figure 2A ). As shown in Figure 2B , this site-mutation would cause a conversion from leucine 69 into proline in exon 2 (p. Leu69Pro). In silico analysis indicated that the hydrogen bond present between Gly65 and Leu69 in the wild-type TTC7A was disrupted by the Leu69Pro mutation ( Figure 2C ).

A group of MIA patients with TTC7A mutation has been shown to be associated with CID (13). In light of multiple episodes of antibiotic-resistant infections, we strongly suspected the immunodeficiency of this patient. The immunologic profiling was delineated by blood cell flow cytometry at the age of 80 days. While the absolute lymphocyte count was kept within the normal range (3020 cells/μl, reference 2400-9500 cells/μl, Figure 3A ), the frequency of T cells was markedly reduced (6.01% CD3⁺ cells in the patient versus 68.9% and 60.4% in two healthy controls, Supplementary Figure 1A ). There was a significantly reduced absolute number of T (both CD4 and CD8) cells (CD3, 182 cells/μl, reference 2179-4424 cells/μl; CD4, 47 cells/μl, reference 1461-3018 cells/μl; CD8, 123 cells/μl, reference 556-1687 cells/μl, Table 2 ). In addition, a marked reduction in the proportion and absolute cell number of naive CD4 T-cell (CD4⁺CD45RA⁺) and naive CD8 T-cell (CD8⁺CD45RA⁺) were observed (CD4⁺CD45RA⁺, 6.87%, reference 69.15-88.10%, 13 cells/μl, reference 1170-2595 cells/μl; CD8⁺CD45RA⁺, 16.61%, reference 68.90-94.60%, 31 cells/μl, reference 503-1276 cells/μl, Table 2 ). Moreover, the dropped naive T cells were consistent with a subdued thymic output as the recent thymic emigrants evaluated by our TRECs determination indicated an obvious reduction (29.4 copies/μl in the patient versus 454.8, 153.9, 123.4 copies/μl in three age- and sex-matched healthy controls, Figure 3B ). Chest X-ray examination also showed that the thymus of the patient was smaller than aged-matched healthy controls. Besides, the patient presented with a markedly elevated double negative T (DNT, CD3⁺CD4^-CD8^-, 6.04% versus 15.9%, Figure 3C ) cells, which suggested that an impeded or skewed T-cell differentiation does exist.

Likewise, the proportion of CD19⁺ B cells in lymphocytes was markedly reduced (0.82% CD19⁺ cells in the patient versus 18.0% and 22.9% in two healthy controls, Supplementary Figure 1B ). The patient exhibited a significantly decreased number of peripheral B cells (25 cells/μl, reference 734-2265 cells/μl, Table 2 ). The BCR recombination activity assessed by KREC levels was shown to decrease significantly (54.3 copies/μl in the patient versus 423.8, 145.6, 181.7 copies/μl in three age- and sex-matched healthy controls, Figure 3D ). The residue B cells in circulation were predominantly transitional B cells (CD24^brightCD38^bright) and virtually no CD24⁺CD38^dim naive B cells or CD24⁺CD38^- memory B cells were presented (naive B cells, 11.4% versus 60.5%; memory B cells, 1.49% versus 14.2%; transitional B cells, 77.2% versus 13.6% Figure 3E ).

Our study found that the frequency of NK cells in this patient was comparable with that of age- and sex-matched healthy controls (5.10% CD3^-CD56⁺ cells in the patient versus 5.41% and 5.67% in two healthy controls, Supplementary Figure 1C ). Whereas, the absolute number of NK cells was significantly reduced (154 cells/μl, reference 290-780 cells/μl, Table 2 ).

Collectively, these results demonstrated that the homozygous TTC7A missense mutation led to a compromised lymphocyte development.

In the present study, we speculated this novel homozygous missense mutation might affect the TTC7A expression. Indeed, we found that patient PBMCs expressed very low level of TTC7A compared with age- and sex-matched healthy controls by western blot ( Figures 4A, B ). We then tested the TTC7A expression in mRNA level using colon tissue of the patient, qPCR analysis indicated that the expression level of TTC7A was lower in the patient compared with age- and sex-matched HD patients as control ( Figure 4C ). Immunohistochemical staining of paraffin slice confirmed a reduced level of TTC7A expression in patient ileum tissue compared with that from meconium ileus patient as control ( Figure 4D ). These results confirmed that this homozygous missense mutation in TTC7A gene led to a reduced expression in lymphocytes and intestinal tissues.

To shed light on the functional impact of TTC7A variants on the gene expression, we compared the gene expression profile in patient PBMCs with those of healthy controls, and identified a total of 7,396 differentially expressed genes (2,583 upregulated and 4,813 downregulated; adjusted p-value < 0.05, fold change > 2). The upregulated and downregulated genes were then subjected to GSEA analysis. The downregulated genes significantly enriched in GO terms and KEGG pathways were clustered into the function of T- and B-cell proliferation, differentiation and activation, as well as NK-cell mediated cytotoxicity ( Figures 5A, B ). The downregulation of genes implicated in lymphocyte functions might explain the T^lowB^-NK^low immune-phenotype we observed in this patient.

It has been confirmed that TTC7A mutations directly hinder the transport of PI4KIIIα, a major TTC7A-interacting protein, into plasma membrane for membranous stabilization (15). We next examined the transcription level of PI4KA and PI4KB in colon tissue and found that PI4KA and PI4KB expression were significantly reduced in patient colon tissue compared with that of the same tissue location in HD controls ( Figures 6A, B ). It has been shown that the TTC7A-PI4K complex is stabilized by the adaptor protein FAM126 and tethered by EFR3 at the plasma membrane (1, 16). It was our great curiosity to know how these two important factors expressed in patient colon tissue. Through qPCR analysis, we found that patient expressed much lower levels of FAM126A and FAM126B mRNA compared with that of HD controls ( Figures 6C, D ). TTC7A has been previously described to interact with EFR3 protein, which might serve as a membrane anchor for PI4K (1). Interestingly, we observed a reduced expression of EFR3A mRNA in patient colon tissue but not EFR3B ( Figures 6E, F ).

Previous studies reported that TTC7A protein could act as a repressor of RhoA signaling pathway, and crucial to normal ROCK function (2, 17). Reduced levels of ROCK1 and ROCK2 expression were found in the patient colon tissue as shown in Figures 6G, H .

Taken together, these results indicated that the homozygous TTC7A missense mutation of the current patient resulted in the downregulation of genes implicated in the PI4K-FAM126A-EFR3A pathway of colon tissues.