Human participant statement
All methods were carried out in accordance with relevant guidelines and regulations. Serologic typing and antibody identification had been previously performed as part of routine clinical testing. All other experiential protocols including targeted sequencing and WGS were performed on archived samples with approval from the Mass General Brigham HealthCare Human Research Committee (IRB), which is the umbrella organization IRB that oversees the Brigham and Women’s Hospital. The IRB approved access to previous clinical results and for new experiments on archived samples under an excess clinical sample protocol, which was deemed to be minimal risk given no direct patient involvement and thus was exempt from obtaining informed consent.
Samples and serologic reactivity of anti-Emm
Blood samples were collected in EDTA. RBC antigen typing and antibody identification was performed by standard tube methods12. Genomic DNA was isolated by standard methods (QIAamp, QIAGEN, Inc. Valencia, CA) from fresh samples and frozen RBC samples.
Briefly, interaction between antibody and red cells is observed as agglutination and often requires use of a secondary anti-human globulin reagent. Serum or plasma containing the antibody and red cells of known phenotypes are incubated at 37 °C, centrifuged, and examined for agglutination. For the indirect antiglobulin test (IAT), following incubation the cells are washed with phosphate buffered saline to remove unbound immunoglobulins, and antihuman globulin is added, centrifuged, and examined for agglutination. For in vitro enhancement of the interaction of red cell antigens and antibodies, low ionic strength solution (LISS) or Polyethylene glycol (PEG) are added prior to incubation. Treating the test cells with proteolytic enzymes such as papain, ficin, or trypsin can be used to enhance antibody-antigen reactions or, in addition to dithiothreitol (DTT), can be used to identify the specific antibody target based on known sensitivity or cleavage pattern of the red cell protein.
Proband 1, a 65-year-old group AB, D+ Indian male with heart disease, presented with an antibody that reacted with all panel cells tested but did not react with autologous RBCs. The reactivity in saline at 4 °C and by the IAT with LISS enhancement was 2+, with 3+ reactivity by IgG gel test. The antigen being detected was resistant to treatment with papain, trypsin α-chymotrypsin and dithiothreitol (DTT); papain treatment enhanced the reactivity. Testing of the RBCs with antibodies to high prevalence antigens from our collections showed that his RBCs were Emm–. His plasma was nonreactive with three examples of Emm– RBCs from our collections: two Caucasian U.S. patients4 and one from the Japanese Red Cross5. Testing of the family revealed a compatible younger brother whose RBCs were Emm– and although he too had never been transfused, his plasma contained anti-Emm (1+ by IgG gel). RBCs of the proband’s daughter, his son, and wife were Emm+ and incompatible when tested with the proband’s plasma.
DNA was extracted from stored RBCs of Proband 2, who was previously reported in abstract form in 20135. This 58-year-old Japanese man with total blindness and renal carcinoma, who had never been previously transfused, was in urgent need of transfusion due to massive bleeding. An antibody reacting 2+ with all cells tested was detected in the plasma, but his clinical condition required transfusion of crossmatch-incompatible blood. He experienced an acute hemolytic transfusion reaction (HTR); his RBCs reacted in the DAT: 1+ on day 1, 2+ on day 3, and negative on day 7, suggesting complete removal of the transfused RBCs from circulation. The antibody was identified as anti-Emm and was shown to have both IgG1 and IgG3 components and to fix complement. The titer of the antibody increased from 16 (saline-IAT) pre-transfusion to 128 by day 10.
DNA was extracted from stored RBCs of Proband 3, who was previously reported by our laboratory in an abstract in 19984. This 70-year-old untransfused male of European ancestry was admitted for transurethral resection of the prostate (TURP) and, like other Emm– probands presented with an antibody that was reactive with all cells tested but nonreactive with autologous RBCs. The antibody reactivity was strong: 3+ to 4+ by albumin and PEG IAT and 4+ when tested against enzyme treated and DTT treated RBCs.
Proband 4 and her family are of North African origin6. In 2012 this 26-year-old female presented at week 25 of her first pregnancy with an antibody against a high frequency antigen resistant to trypsin, ficin, DTT, chymotrypsin and AET. The antibody reacted 2+ in the PEG IAT, and weakly positive in the LISS gel technique, with a negative auto control. The antibody was identified as anti-Emm and her RBCs were Emm−. She had never received a blood transfusion. Her plasma was incompatible with RBCs from her husband, parents, and six of her seven siblings. One sister was found to be compatible and was also Emm−.
Proband 5 and her brother are of Palestine origin, previously reported to have a loss of function PIGG variant and nonprogressive severe generalized ataxia and tonic clonic seizures with moderate delayed development11, were also investigated. Neither child had ever been transfused. Their RBCs typed as Emm−. Anti-Emm was not detected in the plasma.
Illumina short read WGS
PCR free Illumina paired-end short read WGS was performed using standard methods. Briefly, genomic DNA (350 ng in 50µL) was fragmented acoustically using a Covaris focused-ultrasonicator (385 bp fragments). Additional size selection was performed using solid-phase reversible immobilization. Library preparation was performed using a commercially available kit (KAPA Biosystems, Wilmington, MA. Hyper Prep, product KK8505), and with palindromic forked adapters with unique 8-base index sequences embedded within the adapter (Integrated DNA Technologies, Coralville, IA). Libraries were quantified using quantitative PCR (KAPA Biosystems) using Agilent’s automated Bravo liquid handling platform. Libraries were normalized to 1.7 nM and pooled into 24-plexes. Sample pools were combined with HiSeqX Cluster Amp Reagents EPX1, EPX2 and EPX3 into single wells on a strip tube using the Hamilton Starlet Liquid Handling system. Cluster amplification was performed according to the manufacturer’s protocol (Illumina, San Diego, CA) with the Illumina cBot. Flow cells were sequenced on Illumina HiSeqX utilizing sequencing-by-synthesis kits to produce 151 bp paired-end reads. Output from Illumina software was processed by the Picard data-processing pipeline13 to yield CRAM files14 containing demultiplexed, aggregated aligned reads referenced to GRCh38/hg38. The Integrative Genomics Viewer (IGV)15 was used as needed to verify sequence identity and depth of coverage in the CRAM file (i.e. the number of times a specific site was sequenced). The CRAM files were genotyped using Genome Analysis Tool Kit (GATK) HaplotypeCaller v3.5-0-g36282e416 to create variant calling format (VCF) files containing germline short variant calls (SNVs [single nucleotide variations] and indels [insertions and deletions]) found in the samples.
Identification of potential variants
The VCF files were annotated with gnomAD v317 allele frequencies using vcfanno v0.3.218 (variants not present in gnomAD were inferred to have an allele frequency of 0). A custom program was written in Go v1.15.219 to filter the annotated VCF to only keep homozygous variants found in the Emm− proband and his brother, but not homozygous in the proband’s two children. The VCF file was then filtered to only include variants with a gnomAD allele frequency < 0.001. The potential impact of the remaining variants were scored by Ensembl Variant Effect Predictor (VEP) v10120 (–sift b –regulatory –hgvs options). The following PIGG reference sequences were used throughout the manuscript: NG_051621.1 (genomic) and NM_001127178.3 (transcript).
Sanger sequencing
PIGG Exons 1–13 were amplified (HotStarTaq Master Mix Qiagen, Hilden, Germany) using primers (Table S1) with the following PCR profile; 15 min at 95 °C, 35 cycles of 30 s at 94 °C, 30 s at 58 °C and 30 s at 72 °C, and a final elongation of 10 min at 72 °C. Long range amplification was performed (PrimeSTAR GXL, Takara Bio, Madison, WI) with 30 cycles of 10 s at 98 °C and 3 min at 68 °C for PIGG Exons 1 to 4 and 35 cycles of 10 s at 98 °C and 4 min at 68 °C for PIGG Exons 6–10. Products were visualized by agarose gel electrophoresis and ethidium bromide staining and purified with ExoSAP-IT (Applied Biosystems, Foster City, CA). Sanger sequencing was performed and analyzed with ClustalX.
Targeted next generation sequencing (NGS)
Long range amplified products (8 ng at 1.6 ng/uL) were subjected to NGS library preparation using the Illumina Nextera XT kit (Illumina, San Diego, CA) according to the manufacturer’s instructions. Sequencing was performed on the Illumina MiSeq platform (Illumina, San Diego, CA) to generate 150 bp paired-end sequence reads using a nano flow cell (Illumina, San Diego, CA). Sequencing reads were aligned to GRCh38/hg38 human reference genome using BWA-MEM v0.7.12-r103921. The resulting SAM file was then transformed through the use of a series of steps: samblaster v0.1.2422 to addMateTags and removeDups, samtools v1.723 to convert to BAM and sort file by genomic coordinates, Picard v2.5.013 to AddOrReplaceReadGroups, and then samtools v1.723 to create BAM index file. The Integrative Genomics Viewer (IGV)15 was used to view the sequence depth of coverage and paired sequence reads.
