Whole exome sequencing and rare variant association study to identify genetic modifiers, KLF1 mutations, and a novel double mutation in Thai patients with hemoglobin E/beta-thalassemia

Figures & data

Figure 1. Diagram of the procedure for the 24 rare variant association studies (RVAS’s) between candidate variants and clinical severity in patients with Hb E/beta-thalassemia. Variants with passing the Variant Quality Recalibration Score (Pass) and/or Excess het filter (Excess het), different allele frequencies, and different predictions [high impact, moderate impact to protein by variant effect predictor (VEP), and moderate impact with deleterious effect by meta-logistic regression (meta-LR)] were grouped for analysis.

Table 1. Clinical and laboratory information of mild and severe patients were compared according to the scoring system.

Patient characteristics	Mild (n = 173)	Severe (n = 165)	P-value
Male (%)	61 (35.3)	82 (49.7)	0.007
Age, median, y (min, max)	37.0 (5.33, 75.0)	26.0 (11.0, 60.0)	<0.001
Age at presentation, median, y (min, max)	11.0 (0.0, 70.0)	2.0 (0.0, 10.0)	<0.001
Age at receiving first blood transfusion, median, y (min, max)	20.0 (3.0, 70.0)	3.0 (0.0, 34.0)	<0.001
Regular transfusion required (%)	9.0 (5.20)	148.0 (89.7)	<0.001
Normal growth and development (%)	116.0 (67.1)	43.0 (26.1)	<0.001
Spleen size, median, cm (min, max)	5.0 (0.0, 16.0)	12.0 (4.0, 25.0)	<0.001
Splenectomy (%)	12.0 (6.94)	122.0 (73.9)	<0.001
Hb at steady state, median, g/dL (min, max)	8.0 (5.0, 11.9)	6.2 (2.1, 10.6)	<0.001
Hb F(%)*, median (min, max)	40.9 (7.4-66.8) (n = 149)	35.2 (13.7-64.5) (n = 47)	0.999

*Only data of Hb F without interfering by blood transfusion was collected.

Figure 2. Sanger sequencing revealed in cis interaction of IVS I-7 (A > T) (β⁺) and codon 26 (G > A) (β^E) coinherited with in trans codon 26 (G > A) (β^E).

Figure 3. Diagram showing 53 out of 338 patients with phenotype of thalassemia but no thalassemia genotypes analyzed by whole exome sequencing (WES).

Figure 4. Whole exome sequencing showing the number of reads covered on IVS-II-654 and nt-28 mutations. The number of reads was too low. Therefore, the mutations could not be confidently called by whole exome sequencing. (A) The IVS-II-654 mutation was located in the second intron. (B) The nt-28 mutation was located in the promoter. Both mutations were confirmed by PCR.

Figure 5. The sample was called as homozygous Hb E by whole exome sequencing, but it was compound heterozygous mutation of 3.48 kb deletion and codon 26 (G > A) (β^E) by PCR. Codon 26 (G > A) (β^E) from the remaining chromosome was called from all reads.

Table 2. Results of 24 rare variant association studies between variants within the top six genes and clinical severity of patients with beta-thalassemia.

Gene name	Gene symbols	Observed p-values	Adjusted p-values (Bonferroni corrected)
Collagen type IV alpha 3 chain	COL4A3	<0.001–0.04	1.0
Delta-like non-canonical Notch ligand 1	DLK1	(2.19E−05)–(2.94E−05)	0.27–0.37
Family with sequence similarity 186 member A	FAM186A	0.002–0.42	1.0
PZP alpha-2-macroglobulin like	PZP	0.002–0.62	1.0
Thrombopoietin	THPO	2.68E−05	0.31–0.34
Tripartite motif-containing 51	TRIM51	0.002–0.03	1.0

Table 3. Number of cases of mild and severe phenotypes with variants in known modified genes of beta-thalassemia.

Modified genes of beta-thalassemia	Mild phenotype	Severe phenotype
Variant in MYB gene	0	0
Variant in BCL11A gene c.2036G > C, p.Arg679Pro	1	1
Variants in KLF1 gene	5	1
c.519_525dupCGGCGCC, p.Gly176ArgfsX179	3	0
c.629C > G, p.Pro210Arg	1	0
c.1057C > T, p.His353Tyr	1	0
c.809C > G, p.Ser270Trp	0	1

Note: The data are presented as n. Each patient had only one variant.

Table 4. Clinical and laboratory information of the six patients with heterozygous KLF1 variants.

Case	Phenotype	Gender	Age at presentation, y	Age at first blood transfusion, y	Blood transfusion frequency	Spleen, cm	Growth, percentile	Hb at steady state, g/dL	Hb E, %	Hb F, %	Absolute Hb F, g/dL	Αlpha genotype	Beta genotype	Nucleotide change
1	Mild	Female	12	Never	Never	0	>25	7.6	43.5	53.7	4.08	αα/αα	β^0/β^E	c.519_525dupCGGCGCC
2	Mild	Female	7	7	Two units per y	6	>25	9.1	11.3	55.2	5.02	αα/αα	β/β^E	c.519_525dupCGGCGCC
3	Mild	Female	40	40	One to two units per y	12	3–25	8.4	42.9	57.1	4.8	αα/αα	β^0/β^E	c.519_525dupCGGCGCC
4	Mild	Male	3	6	Once	8	>25	8.0	47	53	4.24	αα/αα	β^0/β^E	c.629C > G
5	Mild	Female	1	5	Once	0	3–25	8.3	41.9	51.2	4.25	αα/αα	β^0/β^E	c.1057C > T
6	Severe	Female	3	5	Two units per mo	11	<3	5.6	61.7	28.5	1.6	αα/αα	β^0/β^E	c.809C > G

Table 5. Predicted consequences and allele frequencies of the four KLF1 variants.

Nucleotide change	Amino-acid change	Consequence	Impact to proteins by VEP	SIFT	PolyPhen	gnomAD AF	gnomAD_EAS AF
c.519_525dupCGGCGCC	p.Gly176ArgfsX179	Frameshift_variant	High			8.104e−06	0.0001
c.629C > G	p.Pro210Arg	Missense_variant	Moderate	Deleterious (0)	Probably_damaging (0.928)	0.0005897	0
c.1057C > T	p.His353Tyr	Missense_variant	Moderate	Deleterious (0)	Probably_damaging (0.999)	3.985e−06	0
c.809C > G	p.Ser270Trp	Missense_variant	Moderate	Deleterious (0)	Possibly_damaging (0.899)	3.508e−05	0.000467

Abbreviation: VEP, variant effect predictor; meta-LR, meta-logistic regression; SIFT, sorting intolerant from tolerant; PolyPhen, polymorphism phenotyping; gnomAD, The Genome Aggregation Database; EAS, East Asian; AF, allele frequency.