Abstract
CD36, also known as glycoprotein IV, is classified into two distinct subgroups based on the presence or absence of its expression on monocytes. The CD36 gene spans approximately 50,000 base pairs. Historically, research has focused on identifying CD36 mutations through Sanger sequencing and next-generation sequencing (NGS), with limited exploration of haplotypes. In this study, we collected blood samples from donors with type I and type II CD36 deficiencies as well as from healthy controls, and employed single-molecule long-read sequencing (also known as Third-Generation Sequencing) of genomic DNA to analyze the genetic basis of CD36. The study identified 180 genetic variants, 12 of which were found to alter the amino acid sequence. Notably, four of these mutations (c.220 C > T; c.329_330delAC; c.430-1 G > C; c.1006 + 2 T > G) are premature termination mutations that lead to protein truncation. Using Fisher’s exact test, we statistically analyzed a specific haplotype, c.-132A > C and c.329_330delAC, along with their clinical phenotypes, revealing a strong association between these variants in the 5’ block and type I CD36 deficiency. We analyzed the CD36 gene sequences in platelet donors and patients with PTR (platelet transfusion refractoriness) and FNAIT (fetal and neonatal alloimmune thrombocytopenia), conducting a detailed haplotype analysis associated with type I CD36 deficiency and FNAIT.
Supplementary Information
The online version contains supplementary material available at 10.1038/s41598-024-74917-0.
Keywords: Type I CD36 deficiency, Haplotypes, Long-read sequencing, Fetal and neonatal alloimmune thrombocytopenia
Subject terms: Mutagenesis, Disease genetics
Introduction
CD36, an 88-kDa integral membrane glycoprotein, is expressed on various cell types, including megakaryocytes, monocytes, erythrocytes, microvascular endothelial cells, and mammary epithelial cells, though less on red blood cells (RBCs)1,2. Also known as glycoprotein IV, CD36 was first identified as a platelet membrane component in 19763. CD36 deficiency has two subgroups: Type I (total lack of expression) and Type II (absent on platelets but present on monocytes). Individuals with Type I deficiency are generally healthy but may develop anti-CD36 antibodies after blood transfusions or during pregnancy4–6. In Japan, a Philadelphia-positive acute lymphoblastic leukemia patient developed anti-CD36 and multiple-specificity HLA antibodies and was managed with HLA-compatible, CD36-negative platelets. In China, a male fetus with severe FNAIT due to anti-CD36 antibodies was linked to a 329_330delAC mutation in the CD36 gene7,8.
The CD36 gene on chromosome 7q11.2 comprises 15 exons, 12 of which are coding. Exons 3 and 14 encode the N-terminal and C-terminal domains of CD369. Exon 3 includes the final 89 nucleotides of the 5’-untranslated region (UTR) and encodes the N-terminal cytoplasmic and transmembrane domains. It also incorporates two noncoding exons from the 5’-flanking region. The 3’-UTR is exclusively to exon 14 or spans both exons 14 and 15 depending on the splicing variant10. Over 20 mutations within the coding sequence of CD36 cause its deficiency, but the mechanisms of Type I and Type II deficiencies are largely uncharacterized11,12.
CD36 is highly polymorphic, with over 500 variants in the dbSNP database, each with a global minor allele frequency exceeding 0.005 (as of December 27, 2023). More than 20 variant sites associated with Type I deficiency have been identified10,13–16. Most variants lead to frameshift mutations or changes in post-translational modifications and surface trafficking. Linkage disequilibrium analysis showed the 5’ upstream region of the CD36 gene is tightly linked17. CD36 deficiency results from combined effects of alterations in the coding region and cis-regulatory elements in the 5’ UTR.
Research focused on identifying mutations through Sanger and next-generation sequencing, but associations among CD36 gene variants, especially haplotypes, are not well understood. We recruited 22 donors with CD36 deficiency, 8 patients with CD36 deficiency (4 FNAIT patients and 4 PTR patients) and 7 normal donors (control). Long-read sequencing was used to phase CD36 haplotypes and explore linkages between genotypes and phenotypes. We also reconstructed the phylogeny of CD36 haplotypes and assessed the impact of variants in the 5’ upstream region.
Materials and methods
Study population
All donors and patients were of Chinese origin. Informed consent was obtained from all participants. Nine type I CD36-deficient blood samples and thirteen type II CD36-deficient blood samples from healthy blood donors, comprising 76.19% men and 23.81% women with an average age of 34.19 ± 7.13 years (range, 18–55 years), were randomly collected from the Guangzhou Blood Center CD36 Deficiency Bank. Seven normal blood samples from healthy blood donors (71.43% men and 28.57% women with an average age of 35.57 ± 7.51 years, range 18–55 years), were also randomly collected at the Guangzhou Blood Center. Four blood samples from patients with type I CD36 deficiency were collected randomly from FNAIT patients or PTR patients with anti-CD36 antibodies between 2015 and 2022 at the Clinical Testing Service Department of the Guangzhou Blood Center (Table 1)8,18,19. The samples were screened for antibodies against erythrocytes using the Coombs test, and all results were negative.
Table 1.
FNAIT and PTR cases associated with anti-CD36 antibodies.
| Patient | Age | Sex | Disease | CD36 deficiency (I or II) |
Antibodies | Clinical Symptoms | Published |
|---|---|---|---|---|---|---|---|
| FNAIT-1 | 29 | F | Adverse history of pregnancy and childbirth | I | Anti-CD36 antibodies, anti-HLA I antibodies |
G4P1A3 Fetus hydrops, anemia |
Current report |
| FNAIT-2 | 30 | F | Adverse history of pregnancy and childbirth | I | Anti-CD36 antibodies |
G2P1A1 Fetus hydrops, anemia |
Current report |
| FNAIT-3 | 33 | F | Adverse history of pregnancy and childbirth | I | Anti-CD36 antibodies, anti-HLA I antibodies |
G2P0A6 Fetus hydrops, anemia |
8 |
| FNAIT-4 | 32 | F | Adverse history of pregnancy and childbirth | I | Anti-CD36 antibodies, anti-HLA I antibodies |
G7P2A5 Fetus hydrops, anemia |
19 |
| PTR-1 | 35 | F |
Acute leukemia (AL) |
I | Anti-CD36 antibodies |
CCI < 7000, PPR < 20% |
18 |
| PTR-2 | 30 | F | Acute leukemia(AL), G6PD | I | Anti-CD36 antibodies, anti-HLA I antibodies |
CCI < 7000, PPR < 20% |
Current report |
| PTR-3 | 48 | F | ITP, Acute hemorrhage of upper digestive tract | I | Anti-CD36 antibodies, anti-HLA I antibodies |
CCI < 7000, PPR < 20% |
Current report |
| PTR-4 | 46 | F | Adenocarcinoma of endometrium | I | Anti-CD36 antibodies |
CCI < 7000, PPR < 20% |
Current report |
Primer design, amplification, library preparation, and SMRT sequencing
The primer panel was designed to amplify regions from 5’ UTR to the 3’ UTR of CD36 genes. Sequences from intronic regions were also included. Seven sets of primers were designed to amplify amplicons with more than 1 kb overlap between each amplicon, and all the amplicons are listed in Fig. 1. All primers used in this study are listed in Supplementary Material 1. Long-range polymerase chain reaction (LR-PCR) was performed with KOD FX Neo DNA polymerase (TOYOBO) according to the manufacturer’s instructions, using PCR cycling conditions consisting of a two-step cycle of 10 min each for 30 cycles. PCR products were detected by agarose gel electrophoresis and then purified with 0.6× Ampure PB beads (Pacific Biosciences).
Fig. 1.
Genetic structure of the CD36 gene locus. This delineates the human CD36 gene’s exon-intron structure, with long-read PCR amplicon locations and the exonic locations. Untranslated regions are in red, and coding sequences in blue.
Single-molecule real-time (SMRT) libraries were prepared using a one-step method according to the manufacturer’s instructions. In this method, DNA damage repair, end repair, and adapter ligation were performed simultaneously to produce pre-sequencing libraries containing unique barcode adapters. The final library was annealed with sequencing enzymes and primers using the Sequel II Binding Kit 2.2 (Pacific Biosciences) and the DNA Internal Control Kit 1.0 (Pacific Biosciences). Then, 150 pM DNA polymerase complexes were loaded and sequenced using the Sequel II platform (Pacific Biosciences) with a 20-hour run time.
Data analysis, DNA variant calling, and haplotype phasing
SMRT Link software (v10.1.0, Pacific Biosciences) was primarily used to analyze the output data. Raw reads were first demultiplexed and barcoded automatically at the end of the sequencing runs, and then subreads were analyzed to generate CCS reads using the ccs software v.3.0.0 (https://github.com/pacificbiosciences/unanimity/). Filtered CCS reads were aligned to the human reference genome (GRCh38) using pbmm2 (https://github.com/PacificBiosciences/pbmm2). The target CCS reads were realigned to the reference genome (GRCh38) using pbmm2. Variant calling was conducted with Google DeepVariant (v1.2.0) to identify SNVs and small indels. Finally, amplicon contig sequences were phased into individual haplotypes using the Flye assembler. Finally, the output data were compiled, and individual haplotypes, including gene crossover region haplotypes, were generated using an in-house sciprt module.
Linkage disequilibrium analysis of CD36, phylogenetic tree construction, and variant association analysis
To visualize pairwise linkage disequilibrium between genetic variants and identify linkage groups within CD36, we generated LD heatmaps using LDBlockShow (v1.40) with minor allele frequencies < 0.01. The maximum likelihood phylogenetic tree of haplotypes in LD blocks was constructed using MEGA11 (v11.0.13) with 1000 bootstrap repeats. The types of CD36 deficiency in the donors were classified as 0, 1, and 2. The associations between the genotypes of each SNV and the phenotypes were tested using Fisher’s exact test. A P value < 0.005 indicated a significant difference, and the odds ratios between variants and the CD36 phenotype were calculated.
Luciferase assay
A luciferase reporter assay was conducted to investigate the potential regulatory role of the c.-132A > C mutation in the 5’ UTR on CD36 transcriptional activity. The transcriptional activity of CD36 was assessed using luciferase activity driven by the c.-132 A > C mutation. HEK293T cells were cotransfected with pGL3-basic or pGL3-CC and empty pcDNA3.1 vector plasmids in 24-well plates using Lipofectamine 2000 (Invitrogen) according to the manufacturer’s instructions. Renilla luciferase activity was measured using the Luciferase Reporter Assay System (Promega) and a BioTek Synergy H1 microplate reader (Agilent) to determine the firefly-to-Renilla luciferase ratio.
FNAIT 1
At 23 weeks of gestation, a Chinese fetus was diagnosed with hydrops fetalis (HF) by ultrasound. An increase in the middle cerebral artery peak systolic velocity (MCA-PSV) to 58 cm/s was detected in the fetus. His mother, a 29-year-old (Luo, the number is 23N00901, Supplementary Material 2) with no history of blood transfusion or transplantation, had experienced three pregnancy losses due to intrauterine fetal death (IUD) or hydrops fetalis (HF). Direct and indirect antiglobulin assays were used to test for irregular antibodies against erythrocytes in the maternal serum. Targeted analysis test results for pathogenic variants in the thalassemia genes (HBA1, HBA2, HBB) were also negative. The long-read sequencing of maternal CD36 gene reveals that both of her alleles carried nonsense SNVs or in/dels (c.329_330delAC on one allele and c.1006 + 2 T > G on the other). Both alleles also carried the c.-132 A > C mutation in the 5’-UTR. FACS analysis of CD36 surface antigen expression on platelets and monocytes revealed type I CD36 deficiency in this patient (Table 2).
Table 2.
Clinical characteristics before and after two intrauterine interventions in two FNAIT patients with the c.-132 C-c.329_330delAC haplotype.
| Case | CD36 gene (Mother) | Antibodies (Mother) |
Antibodies (Fetus) |
GA (weeks + day) |
FBS | Intra-uterine transfusion | Before Intra-uterine transfusion | After Intra-uterine transfusion | ||
|---|---|---|---|---|---|---|---|---|---|---|
| RBC(ml) | Hb(g L-1) | Hct(%) | Hb(g L-1) | Hct(%) | ||||||
| FNAIT-1 |
Haplotype 1: c.-132 A > C- c.329_330delAC Haplotype 2: No mutations in coding region |
Anti-CD36 antibodies Anti-HLA(62GE) |
Anti-CD36 antibodies | 23 | √ | 50 | 29 | 9.3 | 142 | 45 |
| 24 | √ | 40 | 74 | 21.8 | 148 | 44 | ||||
| 27 + 6 | √ | 40 | 66 | 19.2 | 142 | 41 | ||||
| 32 + 4 | √ | 30 | 103 | 30 | - | - | ||||
| FNAIT-2 |
Haplotype 1: c.-132 A > C- c.329_330delAC Haplotype 2: No mutations in coding region |
Anti-CD36 antibodies | Anti-CD36 antibodies | 28 + 1 | √ | 40 | 32 | 9.0 | 134 | 43 |
FNAIT 2
The 27-year-old mother with type I CD36 deficiency had no history of blood transfusions or transplants and experienced a miscarriage detected by ultrasound at 23 weeks of gestation due to HF. The middle cerebral artery peak systolic velocity (MCA-PSV) at 22 weeks of gestation during the second pregnancy was 60 cm/sec. Both direct and indirect antiglobulin assay results were negative for abnormal antibodies against red blood cells in the maternal serum. Targeted analysis test results for pathogenic variants in the thalassemia genes (HBA1, HBA2, and HBB) were also negative. Each allele of her CD36 gene carried a nonsense SNVs or in/dels (c.329_330delAC on one and c.430-1G > C on the other one ). Both alleles also carried the c.-132 A > C mutation. Anti-CD36 antibodies were detected in both fetal and maternal blood serum (Table 2). The patient gave birth to a premature baby girl at 32 weeks of gestation.
Flow cytometry analysis of anti-CD36 antibodies from fetal serum using transfected HEK293T cells
The Flow cytometry analysis assay was conducted utilizing transfected HEK293T cells as previously described4. Aliquots comprising 10^5 cells and 25 µL of fetal serum were incubated at 37℃ for 30 min. Unbound antigen-antibody complexes were eliminated through washing, Then, 50 µL of anti-human IgG (or IgM or IgG + IgM + IgA)-FITC was added and analyzed by Flow cytometry.
Ethics approval and consent to participate
Informed consent was obtained from all enrolled subjects and protocol was approved by the Ethics Committee of Guangzhou Blood Center (ethics approval No. GZBC-2022-052). All experiments were performed in accordance with relevant guidelines and regulations.
Results
CD36 gene and haplotype variants
We identified 180 variants within the CD36 gene, spanning from the 5’UTR to the 3’UTR, in 37 donors and patients. The amino acid sequence was altered in 12 of these variants. Among these, four variants (c.220 C > T; c.329_330delAC; c.430-1 G > C; c.1006 + 2 T > G) were classified as premature termination mutations, and seven others were nonsynonymous. A 12-base pair deletion (TATTGGTCAAGC) from positions 1228 to 1239 in exon 12 resulted in the loss of Leu-Leu-Val-Lys at amino acid positions 391–394. Additionally, a variant (c.-132 A > C) upstream of the start codon, which is reported to reduce CD36 protein expression, was detected19. The characteristics of the identified variants in the CD36 gene, as referenced from the CD36 RefSeqGene, are shown in Table 3. These variants were validated through the dbSNP (https://www.ncbi.nlm.nih.gov/snp/). Haplotype phasing was performed to align the variants to two distinct chromosomes, as shown in Supplementary Material 2. Among the 37 haplotypes from individuals classified as type I, 29 contained at least one mutation that altered the amino acid sequence. Notably, 18 of these haplotypes harbored a 5’ UTR mutation c.-132 C, which is prevalent at significantly greater frequencies in type I individuals than in type II (4/26) and CD36 + individuals (0/14).
Table 3.
Characteristics of the polymorphisms of the CD36 gene.
| CD36 RefSeqGene Position (genbank ID: NG_008192.1) | Region | Variation ID | Amino Acid change | HGVS |
|---|---|---|---|---|
| 48,952 | 5’UTR | rs1049654 | / | c.-132 A > C |
| 59,432 | CDS2 | rs556181210 | Ile67Asn | c.200T > A |
| 59,452 | CDS2 | rs545489204 | Gln74* | c.220 C > T |
| 63,881 | CDS3 | rs70961715 | Arg96Pro | c.287 G > C |
| 63,922 | CDS3 | rs572295823 | Thr111frameshift | c.329_330delAC |
| 63,974 | CDS3 | rs201765331 | Ser127Leu | c.380 C > T |
| 65,802 | Intron3 | rs199530093 | Splice Acceptor Variant | c.430-1 G > C |
| 65,911 | CDS4 | rs201759307 | Trp180Arg | c.538 T > C |
| 73,979 | Intron8 | rs555480623 | Splice Donor Variant | c.1006 + 2 T > G |
| 75,613 | CDS10 | rs148910227 | Arg386Trp | c.1156 C > T |
| 75,620 | CDS10 | rs201355711 | Gln388Leu | c.1163 A > T |
| 76,194 | CDS11 | rs550565800 | 410–413 del Ile-Val-Pro-Ile | c.1228_1239del ATTGTGGCCTATT |
| 76,950 | CDS12 | rs200771788 | Thr470Ile | c.1409 C > T |
Haplotype blocks and their association with CD36 deficiency
Variants detected in the haplotype sequences were used to define the haplotype blocks of the CD36 gene. The approximately 35 kb region of the CD36 genomic sequence is divided into two main blocks: the 5’ block and the 3’ block. These blocks are separated by a breakpoint around intron 3 to intron 4, specifically between positions 65,436 and 65,601 (Fig. 2). Fisher’s exact test was used to assess the associations between genetic variants and the observed phenotypes. Consistent patterns of haplotype blocks were observed in groups with significant associations. Notably, a stronger association was found between the variants in the 5’ block and type I deficiency (Table 4). In contrast, the significance of associations with type II classifications varied, indicating that these variants are more specifically linked to type I deficiency rather than to type I and II deficiency both type I and type II deficiencies. Phylogenetic analysis of the 5’ block sequences revealed that the haplotypes can be categorized into four distinct evolutionary lineages, as shown in Fig. 3. Notably, most of the haplotypes associated with type I individuals are grouped within the lineage that harbors the nucleotide c.-132 A and other significant amino acid-altering mutations, including the c.329_330delAC mutation.
Fig. 2.
Pairwise linkage disequilibrium between variants and linkage groups in CD36 with LD patterns. The approximately 35 kb region of the CD36 gene, spanning from the 5’ UTR to the 3’ UTR, is divided into the 5’ block (front part) and the 3’ block (tail part). These blocks are separated by a breakpoint around intron 3 to intron 4. The above points indicated the Fisher’s exact test between genetic variants and observed phenotypes (Type I/II or healthy). A stronger association was found between the variants in the 5’ block and type I deficiency.
Table 4.
Fisher’s exact test was used to calculate the associations between the variants and the phenotypes.
| Pos. | Anont. | REF | ALT | Fisher’s exact test (-log10 (P value)) | Odds ratio for * |
|||
|---|---|---|---|---|---|---|---|---|
| CD36 normal VS deficiency I & II; |
CD36 normal & type II VS deficiency I |
|||||||
| REF carrier vs. noncarriers |
ALT carrier vs. noncarriers |
REF carrier vs. noncarriers |
ALT carrier vs. noncarriers |
|||||
| 48,299 | c.1-1255 | A | G | 0.000 | 0.393 | 0.724 | 2.535 * | 9.75 |
| 48,952 | c.1-132 | A | C | 0.380 | 0.515 | 1.298 | 2.724 * | Inf |
| 51,134 | c.120 + 1462 | T | Del | 0.189 | 0.000 | 3.232 * | 0.630 | 0.00 |
| 51,487 | c.120 + 1814 | G | A | 0.393 | 0.000 | 2.535 * | 0.724 | 0.10 |
| 51,962 | c.120 + 2289 | A | C | 0.000 | 0.393 | 0.724 | 2.535 * | 9.75 |
| 52,060 | c.120 + 2387 | A | G | 0.380 | 0.515 | 1.298 | 2.724 * | Inf |
| 52,123 | c.120 + 2450 | A | G | 0.000 | 0.393 | 0.724 | 2.535 * | 9.75 |
| 52,329 | c.120 + 2656 | A | T | 0.189 | 0.000 | 3.232 * | 0.630 | 0.00 |
| 52,833 | c.120 + 3160 | G | C | 0.189 | 0.000 | 3.232 * | 0.630 | 0.00 |
| 53,250 | c.120 + 3577 | T | C | 0.189 | 0.000 | 3.232 * | 0.630 | 0.00 |
| 53,614 | c.120 + 3941 | A | C | 0.000 | 0.393 | 0.724 | 2.535 * | 9.75 |
| 53,959 | c.120 + 4286 | G | A | 0.512 | 0.409 | 3.201 * | 0.761 | 0.00 |
| 54,262 | c.120 + 4589 | C | A | 0.189 | 0.000 | 3.232 * | 0.630 | 0.00 |
| 54,363 | c.120 + 4690 | G | T | 0.189 | 0.000 | 3.232 * | 0.630 | 0.00 |
| 54,676 | c.121–4677 | C | A | 1.099 | 0.000 | 3.177 * | 1.058 | 27.14 |
| 54,797 | c.121–4556 | T | C | 0.189 | 0.000 | 3.232 * | 0.630 | 0.00 |
| 55,588 | c.121–3764 | G | Del | 0.189 | 0.000 | 3.232 * | 0.630 | 0.00 |
| 55,591 | c.121–3762 | A | T | 0.189 | 0.000 | 3.232 * | 0.630 | 0.00 |
| 56,167 | c.121–3186 | T | C | 0.189 | 0.000 | 3.232 * | 0.630 | 0.00 |
| 56,396 | c.121–2957 | C | A | 0.189 | 0.000 | 3.232 * | 0.630 | 0.00 |
| 56,820 | c.121–2533 | A | G | 0.512 | 0.409 | 3.201 * | 0.761 | 0.00 |
| 57,572 | c.121–1781 | T | C | 0.189 | 0.000 | 3.232 * | 0.630 | 0.00 |
| 59,767 | c.281 + 254 | C | G | 0.189 | 0.000 | 3.232 * | 0.630 | 0.00 |
| 61,882 | c.282–1994 | G | A | 0.809 | 0.788 | 2.703 * | 2.363 | 0.05 |
| 62,019 | c.282–1857 | C | T | 0.189 | 0.000 | 3.232 * | 0.630 | 0.00 |
| 62,540 | c.282–1336 | A | T | 0.225 | 0.000 | 2.026 | 2.535 * | 9.75 |
| 63,922 | c.329_330 | AC | Del | 0.809 | 0.000 | 2.703 * | 0.690 | 21.38 |
| 65,436 | c.430 − 367 | T | C | 0.189 | 0.000 | 3.232 * | 0.310 | 0.00 |
| 65,601 | c.430 − 202 | C | T | 0.183 | 0.000 | 2.811 * | 0.310 | 0.00 |
| 66,538 | c.609 + 556 | G | C | 0.000 | 0.000 | 2.410 * | 0.000 | 0.00 |
| 66,710 | c.610 − 509 | A | G | 0.000 | 0.000 | 2.410 * | 0.000 | 0.00 |
| 67,653 | c.701 + 344 | A | Del | 0.000 | 0.000 | 2.410 * | 0.000 | 0.00 |
| 68,101 | c.701 + 791 | A | G | 0.000 | 0.000 | 2.410 * | 0.000 | 0.00 |
| 68,106 | c.701 + 796 | C | A | 0.000 | 0.000 | 2.410 * | 0.000 | 0.00 |
| 72,384 | c.749 − 382 | T | C | 1.511 | 0.967 | 2.563 * | 1.011 | 0.10 |
| 72,454 | c.749 − 312 | G | A | 1.511 | 0.806 | 2.563 * | 1.318 | 0.10 |
| 72,610 | c.749 − 156 | G | A | 1.511 | 0.806 | 2.563 * | 1.318 | 0.10 |
| 75,433 | c.1126 − 150 | C | A | 1.511 | 0.806 | 2.563 * | 1.318 | 0.10 |
| 79,768 | c.1419 + 2808 | G | A | 1.511 | 0.485 | 2.563 * | 1.645 | 0.10 |
| 79,847 | c.1419 + 2887 | C | G | 1.511 | 0.485 | 2.563 * | 1.645 | 0.10 |
| 81,668 | c.1419 + 4708 | C | A | 1.511 | 0.225 | 2.563 * | 1.011 | 0.10 |
Fig. 3.
Phylogenetic tree of 5’ block rigion (from position 47,000 to position 64,000 according the LD analyses). The colors of the circles indicate the haplotypes from different individuals (red for Type I; orange for Type II; and green for healthy). The “Mut” in the sequence name indicates that the haplotype carries at least one mutation leading to an amino acid change or the-132A>C mutation; “wild” indicates none have been found. Phylogenetic analysis revealed four distinct evolutionary lineages. Notably, most of the haplotypes associated with Type I individuals are grouped within the lineage that harbors the nucleotide c.-132A and other significant amino acid-altering mutations, such as the c.329_330delAC mutation.
Expression study using vectors containing CD36 c.-132 C or CD36 c.-132 A
To confirm whether the c.-132 A > C substitution significantly diminishes CD36 expression, we constructed expression vectors harboring either the wild-type CD36 c.-132 A [denoted as CD36 (WT)] or the mutant CD36 c.-132 C [denoted as CD36 (Mut)]. These vectors were transiently expressed in HEK293T cells. The transfection efficiencies for CD36 (WT) and CD36 (Mut) were validated through equivalent Renilla luciferase activity, as shown in Fig. 4. The results indicated that the c.-132 A > C substitution significantly decreases CD36 expression. Additionally, to investigate anti-CD36 isoantibodies in fetuses, we examined fetal sera via flow cytometry using CD36-transfected HEK293T cells. Notably, as depicted in Fig. 5 (left), fetal serum specifically reacted with CD36-transfected cells but not with nontransfected (mock) HEK293T cells.
Fig. 4.
The c.132A>C mutation leads to a marked reduction in CD36 expression, as determined by luciferase activity. Luciferase reporter assay was performed to explore the regulatory effect of the c.-132A>C mutation in the 5’ UTR on CD36 transcriptional activity. HEK293T cells were cotransfected with pGL3-CD36-WT or pGL3-CD36-MUT and control vector plasmids, and luciferase activity was measured to assess CD36 transcriptional activity.
Fig. 5.
Analysis of anti-CD36 antibodies by flow cytometry in Fetal serum. CD36-transfected HEK293T cells were incubated with fetal serum as indicated (red fluorograms). After washing, bound antibodies were detected with FITC-labeled antibodies specific for human IgG and analyzed by flow cytometry.
Discussion
CD36 is a B-type scavenger receptor expressed in various cell types. Over 20 variants in the coding sequence of CD36 have been identified, predominantly leading to type I receptor deficiency due to truncated protein translation12,20,21. The study of CD36 deficiency has intensified due to its association with clinical conditions such as the Naka- phenotype22. However, current genotyping methods, such as whole-genome sequencing with short reads or Sanger sequencing of PCR products, have significant limitations; they provide minimal haplotype information at the individual genome level23,24. In this study, we utilized long-read sequencing for comprehensive genomic analysis, clearly distinguishing between maternal and paternal sequences. We systematically cataloged and characterized CD36 haplotypes that potentially influence protein functionality. Although prior studies primarily associated type I deficiency with homozygous or compound heterozygous mutations, our analysis highlights transconfiguration mutations on both chromosomes that could impact both gene alleles. The 329_330delAC mutation, identified as the predominant mutation in CD36-deficient Chinese populations, leads to an amino acid change that potentially affects CD36 protein expression19,25. It is crucial to recognize that human CD36 gene has two distinct promoters, and coding region mutations alone do not explain this specific deficiency. Further investigations are necessary to explore whether mutations in one of the CD36 gene promoters might be responsible for CD36 deficiency. Notably, a variant (c.-132 A > C) located upstream of the start codon has been shown to decrease CD36 protein levels19,25,26. This study highlights the role of a SNP at position c.-132 in CD36 deficiency, found either as homozygous C/C or heterozygous A/C in 16 of 22 individuals with type I or II deficiency. None of the control subjects exhibited this SNP. The c.-132 SNP likely reduces promoter activity, as suggested by its impact in a luciferase assay, and is specifically linked to CD36 deficiency. Additionally, several potential cis-regulatory elements within the 5’-UTR could influence transcriptional regulation, contributing to this deficiency. Intriguingly, a patient with type I CD36 deficiency was found to be heterozygous at the DNA level (with both wild-type and mutant alleles present), and neither monocytes nor platelets exhibited CD36 expression, as confirmed by FACS. During pregnancy, the production of anti-CD36 isoantibodies is associated with HF. This study utilized single-molecule long-read sequencing with genomic DNA as the template. We observed that the c.-132 C and c.329_330delAC haplotypes directly cause type I CD36 deficiency. Furthermore, haplotypes comprising c.-132 C and EXON 5 329_330delAC are particularly prone to causing CD36 deficiency. Notably, four cases of FNAIT combined with hydatidiform moles all exhibited this critical haplotype. In two of these patients, one allele was normal, while the other carried the c.-132 C and EXON 5 329_330delAC haplotypes. The importance of the c.-132 C and EXON 5 329_330delAC haplotypes in the pathogenesis of FNAIT with hydatidiform moles is emphasized. Haplotype analysis is invaluable for identifying disease susceptibility genes and elucidating the functional consequences of variations linked to haplotypes27,28. This study’s approach to sequencing and allele phasing enhances the efficacy of association testing at a reduced cost.
In conclusion, we utilized single-molecule long-read sequencing to analyze the CD36 gene sequences in platelet donors and patients with PTR or FNAIT. This technology facilitated a detailed haplotype analysis associated with type I CD36 deficiency and FNAIT. Moreover, molecular genetic analysis indicated that mutations in the 5’-UTR might play a significant role in the pathogenesis of CD36 deficiency.
Electronic supplementary material
Below is the link to the electronic supplementary material.
Acknowledgements
This study was supported by grants from Natural Science Foundation of Guangdong Province of China (grant No. 2024A1515012652), Science and Technology Projects in Guangzhou (No.2023A03J0549, 202201010013, 2023A03J0997, 2024A03J0375) and Shenzhen Science and Technology Program (JCYJ20210324115011030).
Author contributions
XL, HL and YF contributed to the study conception and design. WX wrote the main manuscript. DC, XL, JL, XX, XY, JD, HD, HR, and YC prepared figures and datas. All authors reviewed the manuscript.
Data availability
The data reported in this paper have been deposited in the GenBase in National Genomics Data Center, Beijing Institute of Genomics, Chinese Academy of Sciences/China National Center for Bioinformation, under accession number from C_AA085919.1 to C_AA085992.1 and are publicly accessible at https://ngdc.cncb.ac.cn/genbase/. Additionally, the data are also available in the GenBank repository under accession numbers from PQ390750 to PQ390823.
Declarations
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Wenjie Xia, Dawei Chen and Xinnian Li contributed equally to this work.
Contributor Information
Huaqin Liang, Email: lianghuaqin11@163.com.
Xingqiang Lai, Email: laixq11@163.com.
Yongshui Fu, Email: fuyongshui@sina.com.
References
- 1.Caruso, J. A. et al. Loss of PPARγ activity characterizes early protumorigenic stromal reprogramming and dictates the therapeutic window of opportunity. Proc. Natl. Acad. Sci. U S A. 120, e2303774120. 10.1073/pnas.2303774120 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.van Schravendijk, M. R., Handunnetti, S. M., Barnwell, J. W. & Howard, R. J. Normal human erythrocytes express CD36, an adhesion molecule of monocytes, platelets, and endothelial cells. Blood. 80, 2105–2114 (1992). [PubMed] [Google Scholar]
- 3.Tian, J. et al. Molecular cloning and gene/protein expression of FAT/CD36 from grass carp (Ctenopharyngodon idella) and the regulation of its expression by dietary energy. Fish. Physiol. Biochem. 43, 875–888. 10.1007/s10695-017-0342-7 (2017). [DOI] [PubMed] [Google Scholar]
- 4.Xu, X. et al. Improvement of Anti-CD36 antibody detection via monoclonal antibody immobilization of platelet antigens assay by using selected monoclonal antibodies. Ann. Lab. Med. 43, 86–91. 10.3343/alm.2023.43.1.86 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Khatri, S. S., Curtis, B. R. & Yamada, C. A case of platelet transfusion refractoriness due to anti-CD36 with a successful treatment outcome. Immunohematology. 35, 139–144 (2019). [PubMed] [Google Scholar]
- 6.Wu, Y. et al. Hydrops fetalis associated with anti-CD36 antibodies in fetal and neonatal alloimmune thrombocytopenia: possible underlying mechanism. Transfus. Med. 30, 361–368. 10.1111/tme.12705 (2020). [DOI] [PubMed] [Google Scholar]
- 7.Matsui, M. et al. A case of Philadelphia chromosome-positive acute lymphocytic leukaemia with type I CD36 deficiency. Vox Sang. 117, 128–132. 10.1111/vox.13119 (2022). [DOI] [PubMed] [Google Scholar]
- 8.Xu, X. et al. Successful management of a hydropic fetus with severe anemia and thrombocytopenia caused by anti-CD36 antibody. Int. J. Hematol. 107, 251–256. 10.1007/s12185-017-2310-5 (2018). [DOI] [PubMed] [Google Scholar]
- 9.Rać, M. E., Safranow, K. & Poncyljusz, W. Molecular basis of human CD36 gene mutations. Mol. Med. 13, 288–296. (2007). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Taylor, K. T., Tang, Y., Sobieski, D. A. & Lipsky, R. H. Characterization of two alternatively spliced 5’-untranslated exons of the human CD36 gene in different cell types. Gene. 133, 205–212. 10.1016/0378-1119(93)90639-k (1993). [DOI] [PubMed] [Google Scholar]
- 11.Li, R., Qiao, Z., Ling, B., Lu, P. & Zhu, Z. Incidence and molecular basis of CD36 deficiency in Shanghai population. Transfusion. 55, 666–673. 10.1111/trf.12890 (2015). [DOI] [PubMed] [Google Scholar]
- 12.Masuda, Y. et al. Diverse CD36 expression among Japanese population: defective CD36 mutations cause platelet and monocyte CD36 reductions in not only deficient but also normal phenotype subjects. Thromb. Res. 135, 951–957. 10.1016/j.thromres.2015.03.002 (2015). [DOI] [PubMed] [Google Scholar]
- 13.Kashiwagi, H. et al. Identification of molecular defects in a subject with type I CD36 deficiency. Blood. 83, 3545–3552 (1994). [PubMed] [Google Scholar]
- 14.Kashiwagi, H. et al. Molecular basis of CD36 deficiency. Evidence that a 478C–>T substitution (proline90–>serine) in CD36 cDNA accounts for CD36 deficiency. J. Clin. Invest. 95, 1040–1046. 10.1172/jci117749 (1995). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Kashiwagi, H. et al. A single nucleotide insertion in codon 317 of the CD36 gene leads to CD36 deficiency. Arterioscler. Thromb. Vasc Biol. 16, 1026–1032. 10.1161/01.atv.16.8.1026 (1996). [DOI] [PubMed] [Google Scholar]
- 16.Kashiwagi, H. et al. Family studies of type II CD36 deficient subjects: linkage of a CD36 allele to a platelet-specific mRNA expression defect(s) causing type II CD36 deficiency. Thromb. Haemost. 74, 758–763 (1995). [PubMed] [Google Scholar]
- 17.Madan, N. et al. Functionalization of CD36 cardiovascular disease and expression associated variants by interdisciplinary high throughput analysis. PLoS Genet. 15, e1008287. 10.1371/journal.pgen.1008287 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Xia, W. et al. Two cases of platelet transfusion refractoriness and one case of possible FNAIT caused by antibodies against CD36 in China. Transfus. Med. 24, 254–256. 10.1111/tme.12137 (2014). [DOI] [PubMed] [Google Scholar]
- 19.Xu, X. et al. Variants of CD36 gene and their association with CD36 protein expression in platelets. Blood Transfus. 12, 557–564. 10.2450/2014.0209-13 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Wang, W. et al. Scavenger receptor class B, type 1 facilitates cellular fatty acid uptake. Biochim. Biophys. Acta Mol. Cell. Biol. Lipids. 1865, 158554. 10.1016/j.bbalip.2019.158554 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Kashiwagi, H. et al. Analyses of genetic abnormalities in type I CD36 deficiency in Japan: identification and cell biological characterization of two novel mutations that cause CD36 deficiency in man. Hum. Genet. 108, 459–466. 10.1007/s004390100525 (2001). [DOI] [PubMed] [Google Scholar]
- 22.Okuyama, S. et al. Successful allogeneic hematopoietic stem cell transplantation in a patient with type I CD36 deficiency: a case study and literature review. Int. J. Hematol. 118, 656–660. 10.1007/s12185-023-03637-4 (2023). [DOI] [PubMed] [Google Scholar]
- 23.Zhang, Z. et al. Accurate long-read sequencing allows assembly of the duplicated RHD and RHCE genes harboring variants relevant to blood transfusion. Am. J. Hum. Genet. 109, 180–191. 10.1016/j.ajhg.2021.12.003 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Tiruvayipati, S., Tang, W. Y., Barkham, T. M. S. & Chen, S. L. GBS-SBG - GBS serotyping by genome sequencing. Microb. Genom. 7. 10.1099/mgen.0.000688 (2021). [DOI] [PMC free article] [PubMed]
- 25.Lyu, Q. et al. Frequency and molecular basis of CD36 deficiency among platelet donors in Kunming, China. Platelets. 34, 2176168. 10.1080/09537104.2023.2176168 (2023). [DOI] [PubMed] [Google Scholar]
- 26.Kuriki, C., Tanaka, T., Fukui, Y., Sato, O. & Motojima, K. Structural and functional analysis of a new upstream promoter of the human FAT/CD36 gene. Biol. Pharm. Bull. 25, 1476–1478. 10.1248/bpb.25.1476 (2002). [DOI] [PubMed] [Google Scholar]
- 27.Guo, Z., Hood, L., Malkki, M. & Petersdorf, E. W. Long-range multilocus haplotype phasing of the MHC. Proc. Natl. Acad. Sci. U S A. 103, 6964–6969. 10.1073/pnas.0602286103 (2006). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Goodin, D. S., Khankhanian, P., Gourraud, P. A. & Vince, N. Genetic susceptibility to multiple sclerosis: interactions between conserved extended haplotypes of the MHC and other susceptibility regions. BMC Med. Genomics. 14, 183. 10.1186/s12920-021-01018-6 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data reported in this paper have been deposited in the GenBase in National Genomics Data Center, Beijing Institute of Genomics, Chinese Academy of Sciences/China National Center for Bioinformation, under accession number from C_AA085919.1 to C_AA085992.1 and are publicly accessible at https://ngdc.cncb.ac.cn/genbase/. Additionally, the data are also available in the GenBank repository under accession numbers from PQ390750 to PQ390823.