Loss-of-Function Variants in SUPT5H as Modifying Factors in Beta-Thalassemia

Abstract

It is well known that modifiers play a role in ameliorating or exacerbating disease phenotypes in patients and carriers of recessively inherited disorders such as sickle cell disease and thalassemia. Here, we give an overview of the literature concerning a recently described association in carriers of SUPT5H Loss-of-Function variants with a beta-thalassemia-like phenotype including the characteristic elevated levels of HbA₂. That SUPT5H acts as modifier in beta-thalassemia carriers became evident from three reported cases in whom combined heterozygosity of SUPT5H and HBB gene variants was observed to resemble a mild beta-thalassemia intermedia phenotype. The different SUPT5H variants and hematologic parameters reported are collected and reviewed to provide insight into the possible effects on hematologic expression, as well as potential disease mechanisms in carriers and patients.

1. Introduction

Hemoglobinopathies are the most common monogenic disorders in the world, with an ever-increasing global disease burden each year [1,2]. The thalassemias are characterized by reduced synthesis of the globin chains of hemoglobin, specifically beta-globin chains in beta-thalassemia. As most hemoglobinopathies show recessive inheritance, carriers are usually clinically silent, although they can be identified based on hematological features including microcytic hypochromic erythrocytes and elevated levels of the minor adult hemoglobin known as HbA₂. Regular hemoglobinopathy diagnostics as well as pre-marital, pre-conceptional and/or antenatal thalassemia screening programs occasionally identify individuals with microcytic hypochromic parameters and elevated HbA₂ in whom no beta-thalassemia variants in the HBB gene can be identified [3,4]. In very rare cases, haplotype analysis may reveal a pattern of inheritance that does not segregate with the locus of the HBB gene located within the so-called beta-globin gene cluster (chromosome 11 p15), so-called unlinked beta-thalassemia [3]. This can be caused by other genes in the human genome involved in regulating beta-globin gene expression. Several erythroid-specific genes have been reported to carry variants causing elevation of the percentage of HbA₂, such as GTF2E2, GATA1, ASH1L and KLF1 [5,6,7,8,9], although not necessarily with beta-thalassemia traits like red cell indices.

Recently, Whole Exome Sequencing (WES) identified Loss-of-Function (LoF) variants in SUPT5H (NM_001111020.3) in several members of two independent Dutch families showing microcytic hypochromia and elevated HbA₂ but no beta-thalassemia variants in the HBB genes [3]. The discovery of splicing variants (including both donor- and acceptor variants) in SUPT5H in these two independent families and subsequent RNA sequencing confirmed an altered splice pattern leading to intron retention and a premature stop codon.

Additional LoF variants in SUPT5H were found in several other individuals of Dutch, French and Italian ancestry, all of whom presented hematological characteristics consistent with beta-thalassemia traits but were negative for variants in the HBB genes [3,10,11]. Three individuals from two independent Greek families expressing a non-transfusion-dependent beta-thalassemia intermedia phenotype were characterized to have co-inherited a known Mediterranean beta-thalassemia variant and an additional variant in SUPT5H. These findings indicated that SUPT5H may play a role in regulating beta-globin gene expression. Elevated HbA₂ in the absence of a beta-thalassemia variant in the HBB gene is one of the key hematological findings that may suggest the presence of a variant in SUPT5H. More cases have been reported with unknown SUPT5H variants expressing beta-thalassemia traits after the first publication by Achour et al. in 2020 [3].

2. SUPT5H Gene and hSpt5 Protein: Structure and Function

Promoter proximal pausing is an important regulatory step in eukaryotic transcription catalyzed by the enzyme RNA polymerase II (RNA Pol II). Participation in the general transcription factor DSIF is required for normal RNA synthesis. Human DSIF is a heterodimer composed of two subunits, hSpt4 and hSpt5. From the N-terminus, the hSpt5 protein contains an acidic domain, a region that is homologous to the bacterial transcription factor NusG (NGN domain), which interacts with Spt4h, followed by five Kyrpides–Ouzounis–Woese (KOW) domains, two C-terminal repeat regions and another two C-terminal KOW domains (Figure 1). The two C-terminal repeat regions (CTR-1 and CTR-2) can be phosphorylated by the positive transcription elongation factor pTEFb activating RNA Pol II elongation [12]. The C-terminal region of hSpt5 includes the tandem KOW domain, designated KOW6 and 7 in Figure 1, which is near the exiting RNA and might play a role in recruiting factors for RNA capping and in 3′RNA processing [13,14]. Zuber et al. (2018) [12] demonstrated a tight domain interaction between KOW4, KOW6 and KOW7, a typical beta-barrel fold that possibly enlarges the structure of the basic KOW fold compared to other KOW domains of hSpt5. This extended structure might be of importance to present a larger binding surface for additional molecular interactions. The KOW6-7 domains were shown to interact exclusively with protein factors necessary for RNA elongation and/or processing during transcription termination.

If an SUPT5H LoF variant leads to a truncated Spt5 protein, parts of the C-terminal domains are predicted to be missing. Depending on the domains that are missing, important regions include the NGN region (critical for Spt4 dimerization), the CTR-1 and 2 regions (sensitive to phosphorylation and, therefore, activation of the DSIF complex) and the KOW6-7 domains (necessary for binding protein factors involved in RNA elongation, processing and termination). It is currently unclear whether a non-functional truncated protein product is formed or whether the transcript is degraded by Nonsense-Mediated Decay (NMD). In any case, there is no apparent correlation between the location of the premature stop position and the hematologic indices observed amongst carriers of the diverse SUPT5H variants. For example, heterozygotes for the nonsense variants c.193C>T and c.142G>T, both predicted to lead to a truncated protein missing all functional domains, present with hematologic parameters comparable to those in heterozygotes for the most terminal c.3032_3033delTG frameshift variant. The latter codes for an almost intact Spt5 protein in which only the last KOW domain and the C-terminus are missing [4,15]. These findings could support the hypothesis of NMD; however, more studies are necessary to shed light on how the reduction of Spt5 levels affects the beta-globin LCR recruitment of transcription factors to eventually reduce the expression of the HBB genes.

3. Molecular SUPT5H Variants Reported

A total of 28 SUPT5H variants have been reported to date. In the first report describing SUPT5H as a gene associated with a beta-thalassemia trait, eight different variants were observed [3]. Subsequent papers have reported an additional 20 variants. Heterozygotes may present either near-to-normal hematology or microcytic hypochromic parameters, but all were observed to have elevated HbA₂ levels comparable to those in beta-thalassemia traits [4,10,11,15,16,17]. To date, there are no further reports of cases with double heterozygosity for variants in SUPT5H and HBB, and the only reason SUPT5H variants were suspected was due to elevated HbA₂ levels with or without microcytic hypochromic indices in the absence of variants in the HBB gene.

The SUPT5H variants presently known are listed in

Table 1. Variants in SUPT5H (NM_001111020.3). * Delta score [0–1.0] generated by SpliceAI # GnomAD v4.0.0.

Nr.	HGVS Annotation	Molecular Effect	In Silico Prediction *	Description	Etnic Origin	GnomAD Allele Frequency #	Reference
1	c.142G>T	p.Glu48*		nonsense	Chinese		[4]
2	c.193C>T	p.Arg65*		nonsense	Chinese	7.15 × 10−7	[15]
3	c.397C>T	p.Arg133*		nonsense	Chinese	3.05 × 10−6	[17]
4	c.458+1G>A		1.00	Donor loss	Dutch		[3]
5	c.631C>T	p.Gln211*		nonsense	Chinese		[4]
6	c.817delC	p.Leu273*		nonsense	French		[3]
7	c.1195C>T	p.Gln399*		nonsense	Chinese		[16]
8	c.1374+2T>C		0.71	Donor loss	Greek		[3]
9	c.1405delA	p.Arg469Glufs*10		Frameshift	Chinese		[4]
10	c.1741_1744dupAACC	p.Arg582Glnfs*21		Frameshift	Greek		[3]
11	c.1782_1785delCCAT	p.Ile594Metfs*2		Frameshift	Chinese?		[3]
12	c.1789A>T	p.Leu597*		nonsense	?		[11]
13	c.1855C>T	p.Arg619*		nonsense	?		[4]
14	c.1901dupG	p.Met635Hisfs*19		Frameshift	Chinese		[4]
15	c.1979delG	p.Gly660Valfs*6		Frameshift	Dutch		[3]
16	c.1993_1994delTA	p.Met665Glufs*19		Frameshift	?		[11]
17	c.2245C>T	p.Arg749Trp	0.69	Donor gain	?	2.40 × 10−6	[11]
18	c.2247delG	p.Leu750Serfs*11		Frameshift	?		[11]
19	c.2259-3C>A		0.09	-	Dutch		[3]
20	c.2368C>T	p.Gln790*		nonsense	French		[10]
21	c.2507A>G	p.Tyr836Cys		Alternative splicing	?		[11]
22	c.2725delC	p.Gln909Argfs*45		Frameshift	Dutch		[3]
23	c.2823_2850dup	p.Ser951*		nonsense	Chinese		[4]
24	c.3032_3033delTG	p.Met1011Metfs*9		Frameshift	Chinese		[17]
25	c.3034T>A	p.Cys1012Ser	0.02	-	?		[11]
26		EXON 21 -2 A>G		Splice acceptor	?		[11]
27		p.Glu455Aspfs*23		Frameshift	?		[11]
28		del(chr19:39936531-40030719)		?	Chinese		[4]

and shown in Figure 1. The majority are either nonsense or frameshift variants leading to a premature stop codon. Of four variants reported which alter the splice consensus sequence, one splice donor variant and one splice acceptor variant (c.458+1G>A and c.2259-3C>A, respectively) have been studied at the RNA level, demonstrating intron retention and an altered reading frame leading to a premature stop codon. Of the other splicing variants, two have not been examined at the RNA level, but in silico prediction software suggests an effect on splicing. The in silico prediction of two missense variants, SUPT5H:c. 2245C>T, p.(Arg749Trp) and c.2507A>G, p.(Tyr836Cys), also indicates alternative splicing, a subsequent altered reading frame and a premature stop codon. All variants, therefore, are consistent with the synthesis of a truncated protein, missing the 3′carboxy-terminus. The only exception is the variant SUPT5H: c.3034T>A, p.(Cys1012Ser), a missense variant in the penultimate exon of the gene without a clear in silico prediction in favor of alternative splicing.

As almost the entire spectrum of SUPT5H variants in individuals expressing an unlinked beta-thalassemia phenotype from different geographic areas lead to an LoF, this suggests that an intact C-terminus is essential for the correct function of the Spt5 protein. The haplo-insufficiency of Spt5, or a truncated Spt5 which misses the carboxy-terminus, is apparently the mechanism which gives rise to the typical phenotype of beta-thalassemia traits with the characteristically elevated levels of HbA₂.

4. In Vitro Studies of SUPT5H LoF

The SUPT5H gene codes for human Spt5, a protein known to play a role in controlling the release of promoter-proximal pausing of RNA Pol II, which is crucial for gene regulation and cellular differentiation. That RNA pol II pausing plays a role in modulating hematopoietic stem cell emergence was already shown previously in zebrafish [18,19]. During erythropoiesis, cell cycle genes are largely paused as cells transition from progenitors to precursors. RNA Pol II pauses ~30 to 50 bp downstream of promoters and its release into productive elongation is a highly regulated process [20,21,22,23]. One of the factors involved in the convergence from paused to released RNA Pol II is the DRB (5,6,-dichloro-1-beta-D-ribofuranosylbenzimidazole) sensitivity-inducing factor (DSIF complex), composed of Spt5 and Spt4. As Spt5 is an essential and highly conserved elongation factor, depletion often leads to cell death and its role in cellular differentiation is therefore difficult to study [24,25,26,27,28]. The identification of LoF mutations in SUPT5H in patients with a beta-thalassemia trait-like phenotype in the absence of HBB mutations created a unique opportunity to study RNA Pol II pausing [11]. By perturbing SUPT5H in human hematopoietic stem and progenitor cells (HSPCs) using CRISPR-Cas9 to generate a heterozygous population of cells with a 50% reduced level of Spt5 protein levels, it was possible to study the effect on globin expression and cell differentiation.

Martell et al. (2023) [11] observed that RNA Pol II pause release, which facilitates productive elongation, was hindered in human hematopoiesis with disrupted SUPT5H expression. As these cells began transitioning from progenitors to precursors, both cell cycle kinetics and the onset of erythroid gene expression programs were delayed, although cells still underwent terminal differentiation. This suggests that RNA Pol II pausing plays a role in the regulation of cell cycle progression during erythroid differentiation as cells begin transitioning from progenitors into more mature erythroid precursors. The authors also examined the level of non-coding enhancer-derived RNAs (eRNAs) in the regulatory regions of both the beta- and alpha-globin gene clusters, as eRNA transcription is considered a reliable marker of enhancer activity [29,30]. The reduced levels of eRNA within the beta-globin Locus Control Region (beta-LCR) observed in SUPT5H-edited cells suggests that this may contribute to the reduced HBB expression in carriers of SUPT5H LoF variants. HBB expression is more susceptible than HBA expression to transcription perturbations, particularly at a later stage of erythropoiesis, when there is a higher demand for globin expression. It is not clear why HBB expression is affected more than HBA expression, but as the modes of regulation of gene expression in the beta- and alpha-globin gene clusters are different, it is possible that the effect of perturbation of RNA Pol II pausing may also be different. As Martell et al. (2023) [11] clearly state, patients harboring these mutations are relatively healthy, exhibiting subtle but consistent phenotypes characterized by mild or no anemia but clear signs of beta-/alpha-globin chain imbalance. This was confirmed by the imbalance in globin chain-synthesis results in one of the SUPT5H carriers reported by Achour et al. (2020) [3] and supports a down-regulation of HBB expression but not of HBA1/2 expression, or at least not to the same extent. The fact that HBB expression is regulated through the recruitment of Spt5 via the beta-LCR suggests that this may be an explanation for the reduced HBB expression observed in carriers of SUPT5H variants. In contrast, HBA enhancer sites were largely unaffected in SUPT5H-edited cells, with the exception of one site which was significantly up-regulated, which suggests that LoF variants can have differential effects on enhancer transcription across these different co-expressed genes [31].

5. Hematologic Phenotype of Carriers of LoF SUPT5H Variants

How do carriers of SUPT5H variants compare to carriers of known beta-thalassemia variants at the hematological level? The mean and range of hemoglobin concentration (Hb g/L), Mean Cellular Volume (MCV fL), Mean Corpuscular Concentration of Hemoglobin (MCH pg) and HbA₂ values were collected from 47 SUPT5H heterozygotes reported in the literature (Table S1) and compared to the mean values and ranges for Hb, MCV, MCH and HbA₂ levels for the normal [32] and beta-thalassemia carriers [1] as known from the literature. For SUPT5H/HBB double heterozygotes, the numbers were too limited to determine a mean or range; they are depicted as values in the graph (Figure 2).

The comparison between the four groups, consisting of normal, beta-thalassemia carriers, SUPT5H carriers and double heterozygotes for SUPT5H and HBB variants, shows a decrease in Hb, MCV and MCH values between the normal and the beta-thalassemia carriers (Figure 3a–c). The comparison of hematologic indices (Hb, MCV and MCH) between the SUPT5H carriers, the normal range and the beta-thalassemia carrier range demonstrates that SUPT5H carriers show anemia due to a decrease in hemoglobin concentration equivalent to that seen in beta-thalassemia carriers. As far as MCV and MCH are concerned, the mean value for the SUPT5H carriers is between the normal and beta-thalassemia carriers, while the range overlaps between normal and beta-thalassemia carriers; this is a finding consistent with the relatively mild effect on the bi-allelic reduction in beta-gene expression of the SUPT5H carriers. Only three cases of double heterozygosity for an SUPT5H variant and a beta-thalassemia determinant in the HBB gene have been reported, showing Hb levels falling within the lower range of beta-thalassemia carriers. The MCV and MCH values seem to be below average compared to the beta-thalassemia group. Even though the number of double heterozygotes is not enough to calculate reliable statistical differences, these findings suggest that there is a modifying effect induced by SUPT5H haplo-insufficiency that down-regulates the beta-globin gene expression in beta-thalassemia heterozygotes. As the SUPT5H protein Spt5 acts transiently, it seems most likely that HBB expression on both alleles is down-regulated, enhancing the imbalance between alpha- and beta-globin polypeptide synthesis. Considering that carriers of SUPT5H variants have a hematological phenotype comparable to that of beta-thalassemia carriers, the expected phenotypic expression of double heterozygotes is more severe, which seems to be confirmed by the hematologic data.

No significant differences are seen in the percentage of HbA₂ in beta and SUPT5H carriers (Figure 3d). However, HbA₂ levels were significantly elevated in double heterozygous SUPT5H/beta-thalassemia carriers. As there are only three reported cases of combined beta-thalassemia and SUPT5H LoF variants, it is difficult to determine if this is due to the specific feature of the beta-thalassemia variant or the effect of the interaction of the two variant alleles. Two of the cases described to date are double heterozygotes for HBB:c.118C>T and SUPT5H:c.1374+2T>C (with HbA₂ levels of 8.5% and 11.1%, respectively). Heterozygotes for HBB:c.118C>T usually present with relatively lower HbA₂ levels of 4.89 ± 0.84%. The third case is a double heterozygote for HBB:c.92+1G>A and SUPT5H:c.1741_1744dup (HbA₂ 12.4%). The heterozygote HBB:c.92+1G>A presents with a higher HbA₂ of approximately 9–10%. The extra-ordinary elevation in HbA₂ percentage seen in two double heterozygotes, HBB:c.118C>T and SUPT5H:c.1374+2T>C, cannot be attributed to the HBB variant. In beta-thalassemia carriers expressing a more severe clinical phenotype than expected from beta-thalassemia traits, such elevated HbA₂ percentages might be an indication for a co-inherited SUPT5H variant.

6. Conclusions

LoF variants in SUPT5H are associated with a beta-thalassemia-like phenotype in so-called unlinked beta-thalassemia cases with typically elevated HbA₂ and, in most cases, microcytic hypochromic erythrocyte parameters. SUPT5H is a modifying factor which may play a role in beta-thalassemia carriers who express a more severe phenotype than is seen in simple beta-thalassemia heterozygotes. However, examples from the literature are apparently scarce, with only three reported cases to date.

Recent studies have demonstrated that diminished SUPT5H expression levels can induce a stage-specific delay in erythroid differentiation [11]. In spite of extensive analysis of the Spt5 mutations in human HSPCs, we do not fully understand the mechanism with which Spt5 acts on beta-globin gene expression. It is also not clear why Spt5, which seems to play such an important role in RNase II polymerase pausing and elongation, has no detectable effect on alpha-globin gene expression in these cell lines, nor gives rise to a more severe pathological phenotype in carriers with abnormal development in other tissues.

The comparison of red cell indices amongst different groups of reference samples, beta-thalassemia carriers and carriers of SUPT5H variants and compound heterozygotes provided more insight into the clinical phenotype of reduced SUPT5H expression. The elevated levels of HbA₂ in the absence of a beta-thalassemia trait causing variants in the HBB gene is one of the most suggestive elements to identify variants in the SUPT5H gene. Carriers of SUPT5H LoF variants express a mild beta-thalassemia trait, while double heterozygotes for SUPT5H and a beta-thalassemia variant present with a mild beta-thalassemia intermedia phenotype. For the latter, elevated HbA₂ levels of 10.7% (SD = 2.0%) potentially implicate a diagnosis of combined SUPT5H/HBB in beta-thalassemia carriers expressing beta-thalassemia intermedia. However, little is known about the clinical impact of double heterozygosity as detailed information about increased transfusion need, splenomegaly, manifestations of extramedullary hematopoiesis, etc., is missing, with only three cases reported in the literature. The clinical severity may also impact genetic counseling when couples are identified as being carriers of an SUPT5H LoF variant and a regular beta-thalassemia trait variant in HBB, respectively. As only a few cases are known from the literature, it remains difficult to predict the severity of the beta-thalassemia intermedia phenotype in an affected double heterozygous child. If both parents are carriers of SUPT5H LoF variants, 25% of pregnancies will not lead to a viable embryo as the total lack of SUPT5H expression is likely incompatible with life, with hSpt5 playing such an essential role in the biologically essential mechanism of RNA Pol II pausing and elongation.

As the modifying effect of SUPT5H LoF is relatively subtle, it might be underdiagnosed amongst beta-thalassemia carriers with a more severe clinical expression. The extraordinarily elevated HbA₂ levels may be a biomarker for the presence of an additional SUPT5H variant in cases in which a single HBB variant does not explain the clinical phenotype. Next Generation Sequencing (NGS) to investigate the SUPT5H gene would be the next diagnostic step or, if not available, contacting local, more specialized clinical genetic laboratories for further testing.

More studies are necessary to investigate the mechanism underlying reduced beta-gene expression in carriers of SUPT5H variants. The hemoglobinopathy network INHERENT (URL https://www.inherentnetwork.org (accessed on 14 August 2024)) is instrumental in investigating this topic further. International collaboration between diagnostic laboratory scientists and hematologists is essential in identifying cases where the clinical severity is not fully explained by the genotype or the elevated HbA₂ is not explained by variants in the HBB gene.