Complete human CD1a deficiency on Langerhans cells due to a rare point mutation in the coding sequence

To the Editor: The family of CD1 molecules is structurally similar to MHC class I molecules, but the 2 protein families mediate fundamentally different immune functions. MHC class I molecules present peptides to T cells, whereas CD1 molecules present lipids to natural killer T cells and other CD1-restricted T cells. CD1a is highly expressed on human Langerhans cells (LCs), a specialized mononuclear phagocyte that is prevalent in the epithelial cell layer of the skin and mucosal surfaces. Epidermal LCs can function as classical antigen-presenting cells (APCs) to induce naive T-cell responses in draining lymph nodes, but also have a regulatory function in the skin via local induction of regulatory T cells and maintenance of epithelial barrier integrity. Human dermal dendritic cells (DCs) also express CD1a, but in much lower amounts compared with LCs. CD1a dermal DCs, which coexpress CD1c, have been shown to efficiently stimulate CD4 and CD8 T cells in vitro. However, immune deficiencies due to selective CD1a defects have not been previously described, and it has proved difficult to dissect the specific role of CD1a in immune regulation. During the course of a clinical study that involvedanalysis ofAPC subsets in human skin biopsies by flow cytometry, we identified a healthy Vietnamese individual, donor 007, who showed complete absence of CD1a expression on skinAPCs (Fig 1,A). This case presented an opportunity to study the biological significance of CD1a expression. To check whether LCs were absent altogether in donor 007, we obtained a second skin biopsy, separated the epidermis from the underlying structures, and stained the epidermal tissues with antibodies binding to CD1a and to HLA-DR. Donor 007 LCs displayed intense HLA-DR staining with typical dendritic morphology, but CD1a staining was minimal (Fig 1, B). We next addressed whether the CD1a deficiency represented a generic expression defect, using monocyte-derived dendritic cells (moDCs) as a model. In keeping with our earlier observations, moDCs from donor 007 showed no surface CD1a expression by flow cytometry or immunohistochemistry (see Fig E1, A, in this article’s Online Repository at www.jacionline.org), in contrast to moDCs derived from a normal healthy control donor. Staining with other anti-human CD1a clones, OKT6 and NA1/34-HLK, showed the same result as staining with clone HI149 (see Figs E2 and E3 in this article’s Online Repository at www. jacionline.org). In addition, no costain with early endosome antigen-1 and CD1awas observed, excluding CD1a accumulation in early endosomes in donor 007 (Fig E1, B). To address whether the CD1a defect was caused by a mutation in the CD1a gene, we invited the parents and all 4 siblings of

Complete human CD1a deficiency on Langerhans cells due to a rare point mutation in the coding sequence To the Editor: The family of CD1 molecules is structurally similar to MHC class I molecules, but the 2 protein families mediate fundamentally different immune functions. MHC class I molecules present peptides to T cells, whereas CD1 molecules present lipids to natural killer T cells and other CD1-restricted T cells. 1 CD1a is highly expressed on human Langerhans cells (LCs), a specialized mononuclear phagocyte that is prevalent in the epithelial cell layer of the skin and mucosal surfaces. Epidermal LCs can function as classical antigen-presenting cells (APCs) to induce naive T-cell responses in draining lymph nodes, but also have a regulatory function in the skin via local induction of regulatory T cells and maintenance of epithelial barrier integrity. 2,3 Human dermal dendritic cells (DCs) also express CD1a, but in much lower amounts compared with LCs. CD1a 1 dermal DCs, which coexpress CD1c, have been shown to efficiently stimulate CD4 1 and CD8 1 T cells in vitro. 4,5 However, immune deficiencies due to selective CD1a defects have not been previously described, and it has proved difficult to dissect the specific role of CD1a in immune regulation.
During the course of a clinical study that involved analysis of APC subsets in human skin biopsies by flow cytometry, we identified a healthy Vietnamese individual, donor 007, who showed complete absence of CD1a expression on skin APCs (Fig 1, A). This case presented an opportunity to study the biological significance of CD1a expression. To check whether LCs were absent altogether in donor 007, we obtained a second skin biopsy, separated the epidermis from the underlying structures, and stained the epidermal tissues with antibodies binding to CD1a and to HLA-DR. Donor 007 LCs displayed intense HLA-DR staining with typical dendritic morphology, but CD1a staining was minimal (Fig 1, B).
We next addressed whether the CD1a deficiency represented a generic expression defect, using monocyte-derived dendritic cells (moDCs) as a model. In keeping with our earlier observations, moDCs from donor 007 showed no surface CD1a expression by flow cytometry or immunohistochemistry (see Fig E1, A, in this article's Online Repository at www.jacionline.org), in contrast to moDCs derived from a normal healthy control donor. Staining with other anti-human CD1a clones, OKT6 and NA1/34-HLK, showed the same result as staining with clone HI149 (see Figs E2 and E3 in this article's Online Repository at www. jacionline.org). In addition, no costain with early endosome antigen-1 and CD1a was observed, excluding CD1a accumulation in early endosomes in donor 007 (Fig E1, B).
To address whether the CD1a defect was caused by a mutation in the CD1a gene, we invited the parents and all 4 siblings of donor 007 for a clinical assessment and CD1a expression analysis. Summary clinical information for the family members is presented in Table E1 in this article's Online Repository at www.jacionline.org. Apart from donor 007's father, who had severe Parkinson's disease, the family members were generally healthy and displayed apparently normal skin barrier function and wound healing.
Both parents (001 and 002) and siblings 003, 004, and 006 showed normal CD1a surface expression on skin DCs and/or moDCs by immunohistochemistry and flow cytometry ( Fig E4, A, in this article's Online Repository at www.jacionline.org). However, skin DCs of sibling 005 showed complete absence of surface CD1a expression, similar to donor 007 (Fig E4, B). Blood DC subsets from family members, and from Singaporean healthy controls, were also analyzed by flow cytometry; the absence of CD1a had no impact on the development of blood DC subsets, and did not affect the expression of CD1c and CD1d molecules, excluding an intracellular CD1 protein trafficking defect ( Fig E4, C-F).
To establish the genetic cause of the CD1a deficiency, we isolated RNA from moDCs for CD1a mRNA length and sequence analysis (see Fig E5, A, in this article's Online Repository at www. jacionline.org). The lengths of the CD1a open reading frame from donor 007, from the parents, and from 1 sibling were identical, ruling out a shorter splice variant as the cause of the CD1a expression defect in donor 007. However, sequencing of the mRNA identified a single nucleotide polymorphism (SNP) (rs761269454) (Fig E5, B) that differed between donor 007 and nonaffected family members. The rs761269454 T to C conversion results in an amino acid change from Leucine to Proline at position 285 of the CD1a protein, located in the a3 domain of CD1a (Fig 2, A). Interestingly, parent 001 exhibited a double peak at this nucleotide position, suggesting that both the normal and mutant allele were expressed at the mRNA level, resulting in a normal CD1a phenotype at the protein level ( Fig E5, B, and Fig E4, A).
We next isolated whole blood genomic DNA from all family members and sequenced the CD1a gene and 5000bases upstream and downstream using Illumina MiSeq (see Fig E5, C [Sanger sequencing] and Table E2 [MiSeq] in this article's Online Repository at www.jacionline.org). Donors 007 and 005 were heterozygous for rs761269454 (Fig E5,C,and Table E2), but expressed only the variant form of CD1a ( Fig E5, B), in contrast to parent 001 and sibling 006, who were also heterozygous but expressed both alleles or at least the normal allele, respectively (Fig E4 and Fig E5, B). Intriguingly, we identified a second SNP rs538916791 that introduces a stop codon at amino acid 94 of the CD1a protein. The hereditary distribution of this SNP could explain the CD1a expression pattern: in the presence of the L285P SNP on one allele, the other allele was expressed normally. However, if one allele contained the L285P SNP and the other allele contained the stop codon SNP, as for 005 and 007, only the mutant L285P form could be expressed.
To test whether the L285P mutation was sufficient to abrogate surface CD1a expression, we recombinantly expressed both the reference/wild-type and the mutant forms of CD1a in human embryonic kidney cells (a fibroblast cell line) and K562 cells (a granulocytic/monocytic cell line) (Fig 2, B). We chose 2 cell lines to address potential cell-type-specific differences in expression. Flow cytometry analysis showed that only the reference but not the mutant form of CD1a was expressed on the cell surface ( Fig  2, B), whereas both forms were transcribed equally ( Fig E3). Immunohistochemistry of transfected HEK cells confirmed this finding (Fig 2, C). Different transfection ratios of normal to mutant CD1a resulted in the expected expression level of normal CD1a and excluded competition at the translational level (Fig 2, B).
In summary, we describe complete CD1a deficiency in 2 apparently healthy Vietnamese adults, and have identified a novel mutation responsible for the expression defect. This did not result in any apparent CD1a-related skin abnormalities, or in systemic immune impairment in either individual.
CD1a-restricted T cells specific for the mycobacterial lipopeptide didehydroxymycobactin can be detected in the blood of tuberculin-positive individuals ex vivo. Besides a potential role of CD1a-restricted T cells in antibacterial responses, presentation of natural skin lipids to CD1a-autoreactive T cells has been suggested to be essential for maintenance of the skin immune barrier. According to this hypothesis, a skin injury causes CD1aexpressing epidermal LCs to activate dermal CD1a-restricted T cells, resulting in IL-22 secretion, which, in turn, helps to repair any epithelial damage. 6,7 Moreover, the inflammation caused by bee and wasp venom is mediated via CD1a-restricted self-reactive T cells in the skin. These venoms contain phospholipase A2, FIG 1. CD1a deficiency on skin DCs of a healthy adult. A, Single cells were isolated from skin biopsies from a healthy individual and donor 007 and skin APCs were analyzed by flow cytometry. B, Epidermal sheets from a control donor and from donor 007 were stained for HLA-DR (red) and CD1a expression (green). Hoechst (blue) was used to stain cell nuclei, and samples were analyzed by confocal microscopy. Control data are representative of more than 20 healthy donors. AF, Autofluorescence; DAP I, 49-6-diamidino-2phenylindole, dihydrochloride; FSC-W, forward scatter-width; SSC, side scatter.
which processes skin lipids that are then presented as neoantigens on CD1a, resulting in the activation of CD1a-restricted T cells. 8 None of the family members described here had a history of tuberculosis, although all are likely to have been exposed because tuberculosis is endemic in the region. Similarly, there was no apparent difference in the occurrence of common skin infections, or in wound healing, between family members displaying different CD1a expression patterns, and no family members recalled unusual reactions to bee or wasp stings.
These findings suggest that it is unlikely that CD1a surface expression is an essential element in the proposed pathway by which LCs are thought to function to maintain the integrity of the skin immune barrier. Pre-existing anti-PEG antibodies are associated with severe immediate allergic reactions to pegnivacogin, a PEGylated aptamer

To the Editor:
PEGylation is commonly used to extend half-life and limit volume of distribution of an increasing number of nucleic acid, peptide, and small molecule therapeutics. Pegnivacogin is a modified 31-nucleotide RNA aptamer that binds to and inhibits factor IXa conjugated to an inert 40-kD branched methoxypolyethylene glycol polymer. Although early clinical testing did not identify any safety concerns, the phase IIb Randomized, Partially Blinded, Multicenter, Active-Controlled, Dose-Ranging Study Assessing the Safety, Efficacy, and Pharmacodynamics of the REG1 Anticoagulation System in Patients with Acute Coronary Syndromes (RADAR) trial was stopped after 3 allergic reactions. 1 An extensive investigation demonstrated elevated levels of IgG anti-PEG antibodies in the 3 patients with allergic events, suggesting that the PEG moiety, and not the oligonucleotide, was the causative allergic agent. 2 On the basis of previous safety record of PEGylated products, investigators and regulatory authorities agreed that pegnivacogin should undergo additional definitive testing incorporating a risk mitigation and action plan in a phase III trial (Randomized, Open-label, Multi-Center, Active-Controlled, Parallel Group Study to Determine the Efficacy and Safety of the REG1 Anticoagulation System Compared to Bivalirudin in Patients Undergoing Percutaneous Coronary Intervention [REGULATE-PCI]) in which subjects undergoing percutaneous coronary intervention were randomized to pegnivacogin or bivalirudin. 3 Methodology of the trial, planned biochemical analyses, statistical analyses, and allergy definitions are available in the first and second sections in this article's Online Repository at www.jacionline.org. REGULATE-PCI was ultimately terminated after enrollment of 3,232 of a planned 13,200 patients after an excess of allergic reactions in pegnivacogin-treated patients. 3 The incidence and timing of allergic reactions are summarized in Table I. Descriptions of allergies meeting reporting criteria, as judged by the investigators, are provided in the third section in this article's Online Repository at www.jacionline. org. Assignment to pegnivacogin was associated with a statistically significant increase in allergic reactions. Of the clinical variables assessed, female sex, allergic reactions in the past year, current smoking, and previous percutaneous coronary intervention were associated with severe allergic reactions (see Table E1 in this article's Online Repository at www.jacionline. org). There was no evidence of altered risk of allergic reactions in patients premedicated with H1 or H2 blockers, corticosteroids, beta blockers, or angiotensin-converting enzyme inhibitors (see Table E2 in this article's Online Repository at www.jacionline.org).
As stipulated in the risk mitigation and action plan, measurements of complement activation, tryptase release, and anti-PEG IgG antibodies were performed in all patients experiencing allergic reactions within 24 hours of pegnivacogin or bivalirudin

Clinical methods
To study the index case and his family members, ethical approval was obtained from the Ethical Committee of the Hospital for Tropical Diseases of Ho Chi Minh City and the Oxford Tropical Research Ethics Committee. Following written informed consent, a detailed clinical assessment was performed by a single physician, and a blood sample was obtained for a full hematology/biochemistry panel together with a sample for the immunological studies described below. Four of 7 participants also consented to a shave skin biopsy performed by the same physician.
Control samples consisted of anonymized blood specimens provided by healthy donors to the National University Hospital of Singapore Blood Bank. All donors gave written informed consent.

Skin biopsies
Shave biopsies were taken under local lignocaine anesthesia using DermaBlades (Personna Medical, AccuTec Blades Inc, Verona, Va). Biopsies were collected in RPMI medium and cut into 2 parts. One part was fixed overnight at 48C in PBS containing 30% sucrose and 2% paraformaldehyde, then washed in 30% sucrose for 2 hours, and kept in PBS at 48C until use. The other section of the biopsy was cut into small pieces and digested overnight at 378C in RPMI medium containing 0.8 mg/mL collagenase (Type IV, Worthington-Biochemical, Lakewood, NJ) and 100 U/mL DNase (Roche, Ho Chi Minh City, Vietnam). Digested tissue was disrupted by manual pipetting, and connective tissue and debris were removed by filtering the cells through a 70-mm filter. Cells were labeled with the following antibodies: HLA-DR-PE-Cy7 (clone L243), CD45-V500 (clone HI30), CD1a-APC (clone HI149), and CD14-PE (clone M5E2) (all from Becton Dickinson, BD Biosciences, San Jose, Calif) before flow cytometry analysis. The gating strategy has been described elsewhere. E1

Differentiation of monocyte-derived DCs, histology, and flow cytometry
PBMCs were isolated by Ficoll density gradient (GE Healthcare Life Science, Singapore) and frozen in liquid nitrogen for later analysis. Monocytes were isolated using CD14-microbead positive selection (STEMCELL Technologies Canada Inc, Singapore). CD141 monocytes were cultured in RPMI medium supplemented with 10% FCS with 50 ng/mL recombinant human GM-CSF and 10 ng/mL IL-4 (both from Immuno Tools, Friesoythe, Germany) for 6 to 7 days to generate moDCs. Anti-CD1a clone HI149 (Biomarkers and Immunoassays, Biolegend, San Diego, Calif) or clone OKT6 E2 was used for flow cytometry and histology.
For microscopy analysis, cells were seeded on Poly-L-Lysine-coated chamber slides (Ibidi, Planegg/Martinsried, Germany) and fixed with 4% paraformaldehyde for 20 minutes. Cells were then permeabilized with 0.1% Triton X-100 and blocked with 3% BSA for 2 hours before the antibody was added at room temperature and incubated for 2 hours. Hoechst (200 ng/mL; Invitrogen, Thermo Fisher Scientific, Carlsbad, Calif) was added for 5 minutes and slides were washed 3 times before addition of Prolong Gold (Life Technologies Corporation, Thermo Fisher Scientific, Carlsbad, Calif). Images were taken on an Olympus IX81 confocal microscope.

PBMC flow cytometry analysis
PBMCs were isolated using Ficoll (GE Healthcare) or cell preparation tubes (Becton Dickinson) and frozen for later analysis. Thawed cells were stained and analyzed using a FACS-LSRII (Becton Dickinson). For the DC subset analysis, the negative fractions of the CD14-positive sort (see moDC generation) were used. DC subsets were determined following a previously described gating strategy E3

Sequencing of messenger RNA and genomic DNA
For mRNA sequencing, moDCs were collected into catch buffer (10 mM Tris, pH 8, RNase inhibitor RNasin [Promega Corporation, Madison, Wis]) and frozen at 2808C until use. cDNA was generated and amplified using the One-Step PCR Kit (Qiagen, Hilden, Germany) with the following primers: 59-CTA CTT CCA TTG TTA GCT GTT CTC CC and 59-TGT CTT AAC AGA AAC AGC GTT TCC T. The PCR product was loaded on an agarose gel and the product was isolated with a gel extraction kit (Qiagen).
Genomic DNA was isolated from whole blood using the DNeasy Blood & Tissue Kit (Qiagen). Sequencing of the CD1a gene was first performed using overlapping primers and sequencing the PCR products with Sanger sequencing (Fig E5, C).
For a deeper coverage, we used Illumina MiSeq to sequence the CD1a gene and approximately 5-kb region upstream and downstream of the coding regions, inclusive of the 59 and 39 UTRs. The entire 13.8-kb genomic sequences (->hg38_refGene_NM_001763 range5chr1:158249137-158262944) were subdivided into 5 shorter regions and were amplified for library preparation. Primer sequences and their positions can be found in the table below (Table E3).
Amplification was performed using 10 ng of total genomic DNA prepared from blood as template using LongAmp Taq DNA polymerase (New England Biolabs, Singapore) according to manufacturer's instructions. PCR cycling conditions for amplicon regions 1, 3, 4, and 5 were 948C for 30 seconds (initial denaturation), followed by 25 cycles of 948C for 15 seconds (denaturation), 588C for 30 seconds (annealing), and 658C for 30 seconds (extension), with a final elongation of 658C for 10 minutes. For amplifying the amplicon 2 region, the cycling profile was similar to above except that the annealing was carried out at a higher temperature (608C for 30 seconds). Amplified products were purified using Agencourt Ampure XP beads (Beckman Coulter). Equimolar amounts of PCR products were pooled together for each sample separately for library preparation.
Libraries of pooled amplicons for each sample were prepared using the Nextera XT kit (Illumina, San Diego, Calif) according to manufacturer's instructions. Libraries were constructed using 0.8 ng of pooled amplicons as starting material. Briefly, fragmentation of template DNA (5 mL) was carried out in 10 mL of Tagment DNA buffer using 5 mL of Amplicon Tagment Mix. Indexes were added to Tagmented DNA using 12 PCR cycles. The amplified and indexed libraries were purified and size selected using Agencourt Ampure XP beads (Beckman Coulter). The length distribution of the libraries was monitored using DNA 1000 kits on the Agilent 2100 Bioanalyzer (Agilent Technologies, Singapore). Equimolar amounts of purified libraries were pooled and sequenced using indexed PE sequencing runs of 2 3 250bp on an Illumina MiSeq Personal Sequencer (MiSeq Control Software Version 2.4.1.3). Each individual sample was sequenced at an average depth of 2.3 million reads to detect SNPs in the genomic region of interest.

Data analysis
MiSeq reads were mapped to the HG38 reference genome with bowtie 2. E4 SNP calls were made with Samtools Mpileup and BCFtools. E5 Annotation of rs numbers was performed with an in-house custom script and data from dbSNP for the targeted region. SNP function was determined with snpEff and the Gencode V24 annotation. E6 The data from this study are available in the NCBI BioProject database (ID: 315777).

Database screen for L285P prevalence
A comprehensive genetic database comprising more than 2000 Vietnamese individuals genotyped with the Illumina Human Exome array E7,E8 was screened but did not reveal the CD1a 285L location to be polymorphic. Further examination of exome-sequenced East Asian samples, E9 as well as the 1000 Genomes project Phase 3 cosmopolitan database, E10 also did not reveal any polymorphisms in the CD1 285L genomic location.

Generation of L285P mutant and expression in HEK and K562 cells
Construction of a BCMGSneo CD1a-expressing plasmid has been described elsewhere. E11 To introduce the L285P mutant, an overlap PCR was performed using primer pair CD1a_XhoI_FOR 59-CTTCTCGAGATGCTGTTTTTGCTACTTCC-39 and CD1a_mut_REV 59-ATGTCCTGGCCCTCTGGACTGCTGTGCTTCAC-39 and primer pair CD1a_NotI_REV 59-CCACAGCGGCCGCTTAACAGAAACAGCGTTTC-39 and CD1a_mut_FOR 59-TGAAGCACAGCAGTCCAGAGGGCCAGGA CATC-39. Products were purified, mixed, and used as a template for an overlap PCR using primer pair CD1a_XhoI_FOR and CD1a_NotI_REV. The final product was digested with XhoI and NotI and subsequently ligated into the parental vector.  (Fig 2). HEK, Human embryonic kidney; SSC-A, side scatter-area; WT, wild-type.   Local normal ranges for hematology and biochemistry tests (abnormal results indicated in boldface): White blood cells: Male/female: 6-10 K/mL. Hemoglobin: 14.5-15.7 g/dL for men and 13-14 g/dL for women. Platelets: 201-324 K/ml for men and 211-337 K/ml for women. Glucose