Introduction

Iron deficiency anaemia (IDA), which is unresponsive to oral iron and shows partial responsiveness to parenteral iron, is suggestive of iron resistance iron deficiency anaemia (IRIDA). It is an inherited iron metabolism disorder and a rare autosomal recessive disorder [1] caused due to mutations in the TMPRSS6 gene. The IRIDA phenotype is characterized by hypochromic microcytic anaemia with low normal to normal ferritin levels, very low levels of serum iron and transferrin saturation (TSAT) and inappropriately high levels of hepcidin, in contrast to classical IDA [2, 3]. Symptoms of IRIDA include tiredness, weakness, pale skin and exercise-induced dyspnoea, mostly pronounced during childhood. TMPRSS6 is located on chromosome 22 (22q12.3–13.2) [4] which encodes matriptase-2 (MTP2), a member of the type 2 serine protease family expressed in the liver. TMPRSS6 spans 18 exons and comprises 51,125 base pairs, encoding a protein of 802 amino acids. TMPRSS6 causes down-regulation of hepcidin, the key regulator of iron homeostasis through the cleavage of the cell surface haemojuvelin (HJV) BMP co-receptor, an activator of hepcidin expression [5, 6]. Hepcidin is a circulatory hormone that inhibits duodenal iron absorption and macrophage iron recycling when body iron is repleted [7]. Loss of function mutation in TMPRSS6 leads to elevated levels of hepcidin, which ultimately hampers iron absorption. In this systematic review, we aim to unravel the enigma of the single nucleotide polymorphism (SNPs) residing within the TMPRSS6 gene and explore the significant ethnic differences in the prevalence and distribution of TMPRSS6 mutations and SNPs among the included studies indicating the influence of genetic variations across different ethnic populations. Multiple studies have reported geographical variations in iron status at population level. These disparities have led researchers to propose the hypothesis that differences in genetic mutations among various ethnicities might be a contributing factor to these variations in iron status. This review highlights the pathogenic mutations intricately linked to IRIDA and sheds light on the profound impact of pathogenic mutations/SNPs on vital parameters, such as haemoglobin (Hb) and iron panel. Our review will focus on providing a comprehensive understanding of how specific genetic variations in TMPRSS6 contribute to the development and progression of IRIDA and the role of ethnicity as an explanatory factor across diverse populations.

Methods

Information sources and search strategy

This systematic literature search adhered to the standard criteria Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) [8, 9]. We systematically reviewed the literature to investigate the association between TMPRSS6 variants and IRIDA. A search of relevant electronic databases was conducted, including Scopus, PubMed, Web of Science and Embase/Medline from 2008 to 2023, and the last search was conducted until June 24, 2023. Pathogenic variant data were also extracted from the NCBI dbSNP dataset (https://www.ncbi.nlm.nih.gov/snp/). A combination of relevant keywords and MeSH terms was used to search the studies. The search terms used were TMPRSS6, Transmembrane serine protease 6, matriptase-2, IRIDA, Iron-resistance iron-deficiency Anaemia and iron-resistance iron-deficiency Anaemia.

Eligibility criteria and study selection

The initial search yielded a total of 538 potentially relevant studies. The authors (A.S. and A.K.) systematically screened the articles using web-based filters that adhered to specific eligibility criteria. Studies were included if they met the following criteria: (1) published in English, (2) conducted on human subjects, (3) included patients with IRIDA, (4) reported the association between the TMPRSS6 gene variant and IRIDA and (5) used genetic or molecular biology techniques to measure the TMPRSS6 gene variant. Studies were excluded if they were (1) animal studies; (2) reviews, meta-analyses, and non-English articles; (3) studies without focus on the TMPRSS6 gene variant; (4) anaemia due to other causes; and (5) studies without a measure of IRIDA. Following the inclusion and exclusion criteria, 281 studies were excluded. Among 257 studies, 152 duplicates were identified and screened across all databases, leaving 105 studies for the first screening. Two independent reviewers (A.S. and A.K.) screened the titles and abstracts of the search results and full-text articles were retrieved for potentially relevant studies. After evaluating the titles and abstracts of these studies, 74 were excluded based on predefined criteria and 31 were identified as potentially relevant for the second screening. The remaining studies underwent full-text review, and ultimately, 25 studies were included in the final analysis. A PRISMA flow diagram (Fig. 1) provides a summary of the search process and results.

Fig. 1
figure 1

Algorithm for database search and article selection 

Full size image

Quality assessment and data extraction

Data extraction was conducted by one author (A.S.) and repeated by another author (A.K.). We developed a data extraction form that included the following elements: study design, sample size, patient characteristics, genotype and allele frequencies and outcomes related to IRIDA. We used the JBI critical appraisal tool to assess the quality of the included studies. The form was pilot-tested by two reviewers and refined before being used to extract data from all 25 included studies. Discrepancies were resolved through discussions and consensus.

Quality assessment of the included studies was conducted using the JBI critical appraisal guidelines. These guidelines provide a structured approach to evaluate the methodological rigour and overall quality of each study. A set of ten customized questions was formulated based on the JBI guidelines, with each question representing one star. These questions assessed various aspects of the studies, such as their design, sampling methods, data collection procedures, outcomes and statistical analysis techniques.

Following the assessment, five studies received a score of 5 stars or less, indicating potential methodological limitations or shortcomings. In contrast, an impressive number of 21 studies scored 6 stars or higher, indicating a satisfactory level of methodological rigour and quality. These studies met the predetermined criteria for inclusion and demonstrated robust research designs, comprehensive data collection processes and rigorous analysis techniques.

Our systematic review has been officially registered on the OSF (Open Science Framework) registries, with a link https://osf.io/73am4. The OSF project is currently under an embargoed period, and access to the details will be restricted until March 13, 2024.

Results

Characteristics of the studies

A comprehensive search was conducted in PubMed, Google Scholar, Web of Science and Embase/MEDLINE to identify relevant published articles, whereas NCBI was used to search for the pathogenic variant of TMPRSS6. The search was conducted until June 24, 2023, and provided 538 relevant articles. A total of 25 articles met the final inclusion criteria and were included in the analysis. The collected data included variables such as identification number, gender, population, haemoglobin (Hb) levels, mean corpuscular volume (MCV), mean corpuscular haemoglobin (MCH), mean corpuscular haemoglobin concentration (MCHC), ferritin levels, transferrin saturation, serum iron levels, serum hepcidin levels, peripheral blood smear results and ethnicity, as well as information on single nucleotide polymorphisms (SNPs), likely pathogenic and pathogenic mutations, and other details are represented in Fig. 2, Tables 1 and 2.

Fig. 2
figure 2

This diagram illustrates the general structure of the TMPRSS6 gene, highlighting its various domains, exons and key mutations associated with likely pathogenic effects. The gene consists of multiple exons (1–18) that encode functional domains, including transmembrane domain, serine glutamic acid and alanine domain, complement C1r/C1s, Uegf and Bmp1 domain. Notable pathogenic mutations, identified from a rigorous analysis of the research articles and the NCBI database, are marked in the diagram. These mutations are of significant clinical interest

Full size image
Table 1 Demographic and haematological parameters of participants of the included studies
Full size table
Table 2 TMPRSS6 SNPs associated with iron resistance iron deficiency anaemia (IRIDA)
Full size table

The studies encompassed case studies and case series from multiple countries across different continents, ensuring a broad representation of various genetic backgrounds and regional influences on TMPRSS6 genetic polymorphism. Additionally, the studies included in the analysis exhibited significant variations in sample sizes, highlighting the diversity of research methodologies and participant recruitment approaches. Certain studies had larger sample sizes, surpassing 21 patients, whereas others had smaller sample sizes, typically around one participant. Moreover, the included studies encompassed a wide array of ethnic populations, including South American, Middle Eastern, sub-Saharan African, East Asian, South Asian, Iberian, Hellenic, Arab and Caucasian populations. Overall, the diverse range of ethnic populations and the variation in sample sizes enhance the robustness and generalizability of the findings and allow for a comprehensive examination of the interethnic differences in TMPRSS6 genetic polymorphism and their association with IRIDA.

Single nucleotide polymorphisms (SNPs) in TMPRSS6

TMPRSS6 is known to harbour numerous SNPs. These SNPs offer valuable insights into the genetic variations within TMPRSS6 that are associated with IRIDA. Table 2 lists all SNPs identified in the included studies, along with their functional implications. Among these SNPs, rs1373272804, rs1430692214 and rs855791 were the most frequently reported, appearing in three studies. They are located in 8th, 4th and 18th exons at the genome positions 22:37,482,412–37,482,412, 22:37,492,697–37,492,697 and 22:37,462,936–37,462,936, respectively. The functional implications of these SNPs were obtained from relevant articles and databases, such as dbSNP (https://www.ncbi.nlm.nih.gov/snp), ClinVar (https://www.ncbi.nlm.nih.gov/clinvar) and UniProt (https://www.uniprot.org/uniprotkb/Q8IU80/variant-viewer). This rs1373272804 variant is associated with a reduction in HJV-mediated inhibition of HAMP transcription, loss of catalytic activity and impaired autoproteolytic and HJV processing. This rs1430692214 is associated with reduced HJV-mediated inhibition of HAMP transcription, absence of proteolytic processing, impaired localization to the cell membrane and the ability to interact with HJV (as stated in UniProt). Recent studies investigated the functional role of the rs855791 (V736A) in regulating hepcidin expression, which plays a crucial role in iron metabolism. rs855791 (V736A) variant is linked to decreased serum iron and ferritin levels, lower haemoglobin concentration and an increased risk of IDA and IRIDA. Conversely, these variants are associated with a reduced risk of iron overload.

Furthermore, other common SNPs, such as rs4820268, rs2235324 and rs776877803, have also been reported, but are less significant than the above-mentioned SNPs. The rs776877803 variant reduced HJV-mediated inhibition of HAMP transcription, reduced localization to the cell membrane and had no effect on catalytic activity. Additionally, other SNPs were also identified in the included studies, although they were less common than the above-mentioned SNPs. These may be less prevalent genetic variations within the TMPRSS6 gene but could still have significant implications for IRIDA. The presence of these SNPs suggests that different populations or individuals may harbour unique genetic variants that contribute to the phenotypic heterogeneity observed in IRIDA.

Our comprehensive analysis revealed the presence of single nucleotide polymorphisms (SNPs) within TMPRSS6 gene. When scrutinizing the SNP pattern and profile, our review revealed a distinct susceptibility of certain domains to genetic variations. Among the domains examined, CUB-1, LDRA-2 and Peptidase S1 domains exhibited the highest predilection for SNPs. Approximately 27% of the studies observed SNPs in the Peptidase S1 domain, followed by 23% in CUB-1 and 7% in LDRA-2, making them highly prone domains for SNP variations. Following closely, the CUB-2, SEA, TM and LDRA-3 domains also display a considerable presence of SNPs. Interestingly, the LDRA-1 domain emerged as an exception as no SNPs were detected across the included studies. Moreover, our scrutiny extends beyond the domain level and comprises both the intronic and exonic regions. Various SNPs have been detected within intronic regions, suggesting their relevance in genetic variation. In exonic regions, exons 5, 7, 13 and 15 were found to be major hotspots for SNPs, highlighting their susceptibility to genetic alterations. Exons 9, 11, 16 and 17 also exhibited a notable number of SNPs. In contrast, exons 1, 3, 4, 6, 8, 10, 14 and 18 demonstrated a relatively lower frequency of SNPs. Notably, exon 12 stands out by virtue of the absence of any SNP in the included studies. Among the various types of genetic variations, missense variants took the top spot in terms of frequency, followed by frameshift variants. Silent and synonymous SNPs also exhibited a significant number of SNPs. In contrast, splice variants and splice acceptor variants were comparatively less prevalent in the included studies.

Likely pathogenic and pathogenic mutations observed in TMPRSS6

In our review, we conducted an analysis that assessed the pathogenic impact of mutations identified in various studies using web-based bioinformatics tools, including PolyPhen-2 (http://genetics.bwh.harvard.edu/pph2/), SIFT (https://sift.bii.a-star.edu.sg/) and MutationTaster (https://www.mutationtaster.org/). These tools were employed to predict the functional effects of SNP variants and classify them as either pathogenic or non-pathogenic. We specifically focused on mutations that exhibited deleterious effects, intolerance and damage using all web-based tools. These mutations were then compiled and organized in Table 3, providing details such as rsIDs, CDS position, amino acid changes, genomic and cDNA positions, consequences, genome assembly, transcript information, global minor allele frequencies (MAF) and functional impact. Furthermore, we gathered additional data on the pathogenic variants from the NCBI dbSNP database. Our analysis of the included studies revealed a substantial number of likely pathogenic mutations in TMPRSS6, as predicted by the mutation prediction tools mentioned earlier. Among the pathogenic mutations, missense mutations were the most frequently reported, followed by frameshift, synonymous and stop-gained variants. Notably, certain variants exhibit functional consequences, further highlighting their potential significance. Additionally, allele frequencies were available for variants associated with the rsIDs. In the data extracted from the NCBI dbSNP database, most of the pathogenic mutations were frameshift mutations, whereas a few were deletion and splice site mutations. According to the variant effect predictor tool (VEP) of the ensemble (https://grch37.ensembl.org/Homo_sapiens/Tools/VEP?db=core;expand_form=true;tl=1vYytJJPpDOJMP8r-9296474), all the analysed mutations were classified depending on their functional impact like high, moderate and modifier. In Table 3, most of the mutations had moderate functional impact, except c.76_80delGGTGA (p.G26Wfs ∗ 14), c.188delT (p.L63Pfs ∗ 14, rs1438085143) and c.2172_2173insCCCC (p.P686fs*), which had high impact, and there were few more high-impact mutations that were already reported in the NCBI data, so we mentioned them later in the data extracted from NCBI.

Table 3 Comprehensive overview of predicted (likely pathogenic) mutations in the TMPRSS6 gene associated with IRIDA
Full size table

The mutation c.76_80delGGTGA in the GRCh37.p13 assembly of the genome, located at position 22:37,499,404–37499409, results in a frameshift mutation. This mutation causes deletion of the nucleotide GGTGA at positions 76–80, leading to a frameshift in the coding sequence. Consequently, the amino acid sequence was altered and a premature termination codon (stop codon) was introduced at position 26 (p.G26Wfs*14). The mutation rs1438085143, identified as c.188delT, is a frameshift mutation located at position 22:37,499,296–37,499,297 in the GRCh37.p13 genome assembly. This mutation involves the deletion of a single nucleotide, T, at position 188, resulting in a frameshift in the coding sequence. Consequently, the amino acid sequence was altered and a premature termination codon (stop codon) was introduced at position 63 (p.L63Pfs*14). This frameshift mutation observed in the ENST00000346753.9 transcript is classified as a pathogenic mutation, leading to a truncated protein with a significant functional impact. The allele frequency of this mutation is reported to be T = 0.000004, indicating a low prevalence in the population. The c. 2172_2173insCCCC mutation is a frameshift mutation located at position 22:37,462,970–37462970 in the GRCh37.p13 genome assembly. This mutation involves the insertion of four consecutive C nucleotides (CCCC) at positions 2172 and 2173 within the coding sequence. As a result, the reading frame was disrupted, leading to a frameshift in the protein sequence. The consequence of this frameshift mutation was the introduction of a premature termination codon (stop codon) at position 686 (p.P686fs*). This premature termination codon leads to the formation of a truncated protein, resulting in a shorter and more likely non-functional form. The variant was identified in the ENST00000346753.9 transcript and is classified as a frameshift variant with a high possibility of pathogenicity.

Other information about the pathogenic variants already reported in the NCBI database is listed in Table 4. Almost all pathogenic mutations have a strong functional impact. Here are the details of only those mutations in the NCBI table that have all the possible details: rs786205059 is a splice variant without a specified location in the genome. It had a minor allele frequency (MAF) of G = 0.000008 in the TOPMED database. The mutation rs786205060 is a frameshift variant with a duplication at position c.2055_2058, resulting in a protein change in Val687fs. Its location in the genome is 22:37,465,195–37,465,195, and the corresponding cDNA position is 2172–2175. It is present in the 16th exon. This mutation has a MAF of dupGC = 0.000004 in the TOPMED database. The mutation rs869320724 is a frameshift variant caused by duplication at position c.1904_1905, leading to a protein change in Lys636fs. It is located at 22:37,465,348–37,465,348 in the genome, with a cDNA position of 2021–2022. This mutation occurs in the 16th exon. Its MAF was dupGC = 0.000004 in the TOPMED database. The mutation rs767094129 is a frameshift mutation resulting from a deletion at position c.1812_1813, causing a protein change in Ala605fs. Its location in the genome is 22:37,466,579–37,466,581, and its corresponding cDNA position is 1924–1930. The mutation was found in the 15th exon. It has a MAF of dupC = 0.000004 in the TOPMED database and C = 0.0002 17 in the ExAC database. Another mutation, rs767094129, involved a deletion at position c.1813, leading to a protein change in Ala605fs. Its location in the genome was 22:37,466,578–37,466,579, and the corresponding cDNA position was 1930. This mutation was found in the 15th exon. Its MAF was C = 0.000008/2 in the TOPMED database and C = 0.000217 in the ExAC database. rs137853123 is a stop-gained mutation resulting from the substitution of T in place of C at position 1795, leading to an amino acid change in Arg599Ter. Its genomic location and corresponding cDNA positions were 22:37,466,597–37,466,597 and 1912, respectively. It was identified in the 15th exon and had a minor allele frequency (MAF) of A = 0.000023 in the TOPMED database and A = 0.000014 in the GnomAD database. rs786205058 is a splice site donor variant present at genomic location 22:37,469,571–37,469,571 with a MAF T = 0.000004 in the genome-exome database. rs1384933966 is a deletion mutation that arises from the deletion of the nucleotide at position 1382, leading to a change in the frame and causing amino acid changes in Glu461fs. The genomic location and cDNA position of this mutation were 22:37,470,735–37470736 and 1499, respectively. This mutation was found in the 12th exon with MAF delT = 0.000008 in the TOPMED database and delT = 0.000014 in the GnomAD database. The mutation rs137853122 is a stop-gained mutation found in exon 10. This occurs due to a substitution of a T nucleotide for a G nucleotide at CDS position 1179, resulting in an amino acid change in Tyr393Ter. The genomic position of this mutation was 22:37,480,379–37480379, and its corresponding cDNA position was 1296. Similarly, rs137853121 is another stop-gain mutation found in the 9th exon, resulting from the substitution of a C nucleotide for an A nucleotide at position 1065, causing an amino acid change in Tyr355Ter. The genomic location of this mutation was 22:37,480,815–37480815 and its cDNA position was 1182. It has a minor allele frequency (MAF) of T = 0.000011 in the TOPMED database. C.2028_2031dup and c.1877_1878dup both are frameshift mutations resulting from the duplication of nucleotides at 2028–2031 and 1877_1878 positions, causing a protein change of V678fs and K636fs, respectively. The genomic location and cDNA position of these mutations were 22:37,465,224–37,465,224, 22:37,465,375–37,465,375 and 2145–2146, 1994–1995, respectively. Both mutations belong to the 16th exon. C.1786del causes a frameshift mutation resulting from the deletion of the nucleotide at position 1786, resulting in a protein change in R590*. The genomic location and cDNA position of this mutation were 22:37,466,605–37466606 and 1903, respectively. This mutation was present in the 15th exon. rs137853122 is a stop-gain mutation resulting from the substitution of T in place of G at position 1152, causing a protein change in Y384*. The genomic location and cDNA position of these mutations were 22:37,084,339–37084339 and 1293, respectively. This mutation was present in the 10th exon.

Table 4 Pathogenic mutations in the TMPRSS6 gene associated with iron resistance iron deficiency anaemia (IRIDA), extracted from the NCBI database
Full size table

Interethnic differences associated with TMPRSS6

Interethnic differences in TMPRSS6 pathogenic mutations were explored based on data available from various ethnic groups. A literature review provides insights into these mutations, their distribution and their functional impacts.

European populations revealed a relatively higher number of reported TMPRSS6 mutations than other ethnic groups. Among these populations, various mutations were identified in different countries, including Switzerland, Spain, Italy, France, Ireland, Belgium and Austria. It is worth noting that Switzerland and Austria do not exhibit any pathogenic mutations in TMPRSS6. In Spain, three mutations have been identified: c.1817 T > G (p.L606R), c.1562A > G (p.D521G) and c.76_80delGGTGA (p.G26Wfs ∗ 14). Italy showed mutations, such as c.1768 T > C (p.W590R), which had a global MAF of G = 0.000004 in the GnomAD dataset and G = 0.000008 in the ExAC dataset. The mutation c.188delT (p.L63Pfs ∗ 14) had an MAF of T = 0.000004 in the GnomAD dataset and T = 0.00008 in other sources. The mutation c.1025C > T (p.S304L) had functional implications as it was found to impair matriptase-2 autoproteolytic activation and decrease the ability to cleave membrane HJV, which finally inhibits HJV-dependent hepcidin activation. Another mutation identified in Italy is the IVS6 + 1G > C variant. Additionally, this c.911C > T mutation had a MAF of A = 0.000056 in the ALFA dataset and A = 0.000004 in the GNOMEXD dataset, and was associated with functional impact, specifically disrupting normal splicing sites between exons 15 and 16. Furthermore, the c.1213C > T mutation has been identified in an Italian population. In France, several mutations were identified: c.2293C > G, c.2172_2173insCCCC, c.1561G > A (with MAF T = 0.000053 in TOPMED and T = 0.000028 in genome-exome datasets) and c.1564G > A (with MAF T = 0.000030 in TOPMED and T = 0.000016 in genome-exome datasets). Ireland exhibited the c.1813del mutation, which had an MAF of C = 0.000008 in the TOPMED dataset and C = 0.000217 in the ExAC dataset.

Because the minor allele frequency (MAF) for all the mentioned TMPRSS6 gene mutations is not provided in the dataset, it is impossible to definitively predict which mutation has a higher or lower allele frequency. However, based on the available data, it appears that the c.1813del mutation in Ireland may have a relatively higher allele frequency than other mutations mentioned in studies from European populations.

In South America, c. 860C > A mutation was identified in Venezuela and Colombia, with a minor allele frequency (MAF) of T = 0.000004 in the TOPMED dataset and T = 0.000014 in the GnomAD dataset. Moving to Middle Eastern populations, Turkey exhibited the mutation c.1528 T > C, which has functional implications associated with iron resistance iron deficiency anaemia (IRIDA). This mutation is known to result in reduced HJV-mediated inhibition of HAMP transcription, loss of catalytic activity and impaired autoproteolytic and HJV processing, as documented in UniProt. Mutations such as c.1229G > C, c.2066 T > C and IVS10 + 1 G > A were also identified in this population.

Several mutations have been identified in East Asian populations, specifically in Korea. These included the mutation c.1807G > C, which had a minor allele frequency (MAF) of A = 0.00004/1 in the ALFA dataset and G = 0.00004/1 in the TOMMO dataset, and the mutation c.863þ1G > T, which was reported twice in this population. In contrast, studies from China and Japan included SNPs; however, no pathogenic mutations were found in these studies. Pathogenic mutations were not detected in the Iberian and Hellenic ethnicities. Saudi Arabia exhibited the c.1813delG mutation (p.P686fs*), which is associated with a frameshift variant leading to a truncated protein. Furthermore, the c. 2383 G > A mutation was identified in the Caucasian group.

Interestingly, in the studies reviewed, the mutation c.1025C > T appeared in three studies, two of which were conducted in the Italian population (ID2 and ID4) and the third in the Netherlands (ID21). Another notable finding was the presence of the c.1813delG mutation in two distinct populations: Ireland (ID2) and Saudi Arabia (ID11). These findings shed light on the widespread occurrence of these mutations and emphasize their potential impact on iron metabolism disorders.

Impact of mutations on haematological and biochemical parameters

The impact of TMPRSS6 mutations on haematological and biochemical parameters, including Hb, MCV, MCH, MCHC, serum ferritin levels, transferrin saturation and serum hepcidin, was evaluated across the included studies. In previous studies (ID2, ID3, ID4, ID6, ID7, ID8, ID10, ID11, ID12, ID15, ID18 and ID19), specific mutations were observed in various ethnic groups, and they may have potential implications in the altered levels of haematological and biochemical parameters.

Among the included studies, ID2 focused on the paediatric population and revealed several mutations. However, the mutations c.1768 T > C (p.W590R, rs770897887) and c.1528 T > C (p.C510R) were predicted to be pathogenic and likely exhibited a significant impact on haematological and biochemical parameters. Specifically, the first mutation was found in a proband who exhibited characteristics associated with IRIDA having Hb levels of 6.1 g/dL, MCV of 55 fL, ferritin levels of 23 ng/mL, TSAT of 2.3% and serum iron levels of 7 µg/dL, and the second mutation was found in proband having Hb levels of 6.1 g/dL, MCV of 47.8 fL, ferritin levels of 22.8 ng/mL, TSAT of − 1%, serum iron levels of 6 µg/dL and serum hepcidin levels of 137.6 ng/mL respectively.

Study ID3 also focused on the paediatric population and disclosed many mutations, among which c.749 T > C (p.I212T) and c.1904_1905dupCG (K636fs) were predicted to be pathogenic and found in proband with Hb levels of 8.6 g/dL, MCV of 63 fL, MCH of 18 pg, MCHC of 29 g/dL, ferritin levels of 36.1 ng/mL, TSAT of 3.5%, serum iron levels of 13 µg/dL and serum hepcidin levels of 11.6 nmol/L. Another mutation, c.2066 T > C (L689P), was found in proband having Hb levels of 8.4 g/dL, MCV of 51 fL, MCH of 16 pg, MCHC of 32 g/dL, ferritin levels of 75.8 ng/mL, TSAT of − 4.11%, serum iron levels of 6 µg/dL and serum hepcidin levels of 14.5 nmol/L. Both mutations have a significant impact on the IRIDA phenotype.

Similarly, study ID4 focused on a paediatric population showing a likely pathogenic missense mutation c.536A > G (p.Y141C, rs1430692214) found in a proband with an Hb level of 9.1 g/dL, MCV of 60 fL, MCH of 17 pg, MCHC of 29 g/dL, ferritin levels of 26 µg/L, TSAT of 3.7% and hepcidin levels of 9.78 nM.

The adult population examined in study ID6 revealed the presence of a splice site mutation (IVS6 + 1G > C) in the proband. This individual exhibited significant impacts on both haematological and biochemical parameters, as evidenced by a haemoglobin level of 9.3 g/dL, MCV of 59 fL, MCH of 19.1 pg, ferritin level of 234 ng/mL, TSAT of 6%, serum iron level of 16 µg/dL and serum hepcidin level of 450 ng/mL.

Study ID7 involved a paediatric population showing the presence of missense and splice mutation c.1807G > C (p.G603R) and c.863þ1G > T (IVS6þ1G > T), respectively, in a subject having Hb levels of 10.2 g/dL, MCV of 66.6 fL, MCHC of 30.1 pg, ferritin levels of 56 ng/mL, TSAT of 3.5%, serum iron levels of 15 µg/dL and hepcidin levels of 50.3 ng/mL which were normal according to reference range but were inappropriately high with regard to the above referred low level of Hb and serum iron.

Study ID8 focused on the adult population, likely showing missense and synonymous mutations c.1565C > T (p.D521D) and c.2210 T > Cp.V736A) in probands with Hb level of 9.1 g/dL, MCV of 60 fL, MCH of 17 pg, MCHC of 29 g/dL, ferritin levels of 26 ng/L, TSAT of 3.7% and hepcidin levels of 9.78 nM.

Study ID10 was a case–control study that focused on a mixed population. The proband of case one showed a missense mutation c.911C > T (p.S304L, rs1373272804) having Hb levels of 7.7 g/dL, MCV of 5.6 fL, MCHC of 29.3 pg, ferritin levels of 46 mg/mL, TSAT of 3%, serum iron levels of 1.8 mmol/L and hepcidin levels of 12.2 ng/mL which was quite significant for IRIDA phenotype.

In study ID12, the researchers investigated a group of children and identified a pathogenic splice site mutation, IVS10 + 1G > A (p.C510R), in the proband. This proband exhibited noteworthy haematological and biochemical characteristics associated with the IRIDA phenotype, including Hb levels of 8.1 g/dL, MCV of 48 fL, MCHC of 30.1 pg, ferritin level of 23 ng/mL, transferrin saturation (TSAT) < 2%, serum iron level of 6 µg/dL and hepcidin level of 2.7 nmol/L. These findings have significant implications in the manifestation of the IRIDA phenotype.

Study ID15 was a case–control study focused on a paediatric population showing a pathogenic missense mutation c.2383G > A in a proband with Hb levels of 5.2 g/dL, MCV of 49 fL, MCH of 12.6 pg, MCHC of 25.8 g/dL, ferritin levels of 15 ng/mL, TSAT of 3.2%, serum iron levels of 12 µg/mL and hepcidin levels of 106.1 ng/mL. Study ID18 focused on the adult population showing a pathogenic frameshift mutation c.2172_2173insCCCC (p.P686fs*) in a proband having Hb level of 9.1 g/dL, MCV of 65 fL, MCH of 19.6 pg, MCHC of 25.8 g/dL, ferritin levels of 123 µg/L, TSAT of 3.2%, serum iron levels of 3 µmol/L and hepcidin levels of ng/mL. In study ID19, a pathogenic frameshift mutation c.1905Del (p.L636Rfs*16) was found in a proband of the paediatric population having Hb levels of 8.1 g/dL, MCV of 56.2 fL, MCHC of 27.5 g/dL, ferritin levels of 29.6 ng/mL, TSAT of 6%, serum iron levels of 13 mcg/dL and hepcidin levels of 90.8 ng/mL, which were closely associated with the IRIDA phenotype.

Diagnosis and management of IRIDA

Diagnosing IRIDA requires a comprehensive approach that involves clinical evaluation, laboratory tests and consideration of patient’s medical history. IRIDA is a complex form of iron deficiency anaemia characterized by oral iron refractoriness with partial response to parenteral iron [18]. Clinically, it begins by assessing the patient’s symptoms which may include fatigue, weakness and pallor along with other signs of anaemia. Laboratory investigations play a crucial role in confirming the diagnosis with tests such as haematological parameters (complete blood count, CBC), peripheral blood smear (PBS), iron panel study (serum iron, serum ferritin, total iron binding capacity, transferrin saturation) and certain tests like LFT, RFT, occult blood loss and any haemoglobinopathies to rule out other causes of anaemia. Along with this, biochemical parameters like serum folate, vitamin B12 and anti-tissue transglutaminase (for celiac disease/malabsorption) are also assessed. Following all these investigations and ruling out all other causes of anaemia, patients satisfying the key abnormalities like microcytic hypochromic anaemia with low levels of serum iron and serum ferritin and very low transferrin saturation (TSAT) with inappropriately high levels of hepcidin are labelled with IRIDA phenotype and are advised for genetic analysis for further confirmation.

Traditional oral iron supplementation is generally ineffective in IRIDA [24]. Instead, intravenous iron therapy is considered to bypass the absorption issues. Iron infusions can help rapidly replenish iron stores and improve haemoglobin levels. Further, collaborative efforts between clinicians and haematologists are essential to accurately diagnose IRIDA, enabling the implementation of targeted treatment strategies like intravenous iron therapy and erythropoiesis-stimulating agents to address this challenging condition.

Clinical implications

The analysis of TMPRSS6 genetic polymorphisms and their association with IRIDA holds significant clinical implications, especially for haematologists. Firstly, the study identifies specific SNPs associated with IRIDA, such as rs855791 and rs1373272804, with reported functional implications [28]. These SNPs show a correlation with low haemoglobin levels and mean corpuscular volume (MCV). Understanding these genetic variations can aid clinicians in identifying individuals at a higher risk of IRIDA and potentially modifying treatment strategies based on genetic profiles.

Moreover, the exploration of TMPRSS6 mutations across different ethnic populations reveals variations in the prevalence of specific mutations [11]. This insight is crucial for haematologists dealing with diverse patient populations, enabling them to consider ethnic-specific genetic factors in diagnosis and treatment planning [19]. The study’s detailed examination of TMPRSS6 domains and exonic regions prone to genetic alterations provides valuable information on potential genetic hotspots. Haematologists can leverage this knowledge to deepen their understanding of how these genetic variations might impact iron metabolism and related disorders.

Additionally, the analysis of pathogenic mutations highlights specific variants that significantly affect haematological and biochemical parameters associated with IRIDA. Clinicians can utilize this information to better interpret laboratory results and guide interventions for patients with these mutations. In summary, this study’s clinical implications lie in its potential to inform haematologists about genetic factors contributing to IRIDA, guide personalized treatment approaches and enhance the understanding of the molecular mechanisms underlying this complex haematological disorder. Further research is warranted to establish allele frequencies conclusively and unravel the links between genetic variations and the observed phenotypic heterogeneity in IRIDA.

Discussion

Interpretation and conclusion

This is the most comprehensive and rigorous systematic review to date regarding the analysis of interethnic differences in TMPRSS6 genetic polymorphism and its association with the IRIDA phenotype. The detailed analysis revealed a diverse range of ethnic groups and sample size variations, enhancing the resilience and applicability of the findings. The review’s notable strength lies in its comprehensive inclusion of 25 studies and over 100 SNPs of the TMPRSS6 gene from various global ethnic populations. This extensive data integration allows for robust analysis of genetic variations and their potential implications.

This review recognized several SNPs within the TMPRSS6 gene associated with IRIDA. Among the reported SNPs, rs1373272804, rs1430692214 and rs855791 were the most frequently noted and were associated with functional implications. Our findings were in accordance with the other results [34]. In 2009, genome-wide association studies (GWAS) conducted on individuals of Indian Asian and European ancestry identified rs855791, situated on exon 17, as the most strongly associated SNP with low Hb levels [34]. Additionally, a robust correlation was observed between rs855791 and low Hb levels and mean MCV. One more study reported a remarkably notable rise in the frequency of SNP rs855791 among individuals with IRIDA compared to healthy control subjects [35]. The functional impacts of these SNPs were obtained from relevant articles and databases such as dbSNP, ClinVar and UniProt. Other TMPRSS6 SNPs linked to iron deficiency biomarkers, such as rs4820268, rs2235324 and rs776877803, were also identified, suggesting the presence of unique genetic variants in different populations that contribute to the phenotypic heterogeneity observed in IRIDA. In conclusion, rs855791 and rs4820268 SNPs were found to be significantly associated with elevated hepcidin levels in individuals with IRIDA compared to the IDA group. According to Gichohi-Wainaina et al., the MAF of rs855791 is higher in the Asian population compared to the Caucasian population [36]. These SNPs are commonly found in African, European, Caucasian and Asian people.

Furthermore, this review extended beyond the overall examination of TMPRSS6 and delved into a detailed exploration of the distribution of SNPs within specific domains and exonic regions of this gene. By analysing these specific regions, researchers aimed to uncover potential genetic hotspots and gain a deeper understanding of gene variability. Among the domains investigated, CUB-1, LDRA-2 and Peptidase S1 were particularly susceptible to genetic variations, suggesting that alterations in these regions may have significant functional implications for TMPRSS6. Moreover, this review identified several exons that have emerged as substantial hotspots for SNPs. Notably, exons 5, 7, 13 and 15 exhibited a high frequency of genetic variation, indicating their critical role in shaping the genetic landscape. Interestingly, in contrast to other domains, the LDRA-1 domain did not contain any SNP in the studies considered for analysis. The apparent stability of this domain in terms of genetic variation may imply specific functional importance, warranting further investigation to better understand its role within the TMPRSS6 gene. By understanding the specific domains and exons prone to genetic alterations, researchers can gain valuable insights into the structure and function of genes and their potential impact on iron metabolism and related disorders.

This review comprehensively analysed pathogenic mutations in TMPRSS6 and their potential impact on IRIDA. Among the pathogenic mutations identified and extracted from the dbSNP database, missense (rs770897887, s1438085143, rs1373272804, rs199474805, rs855791 and rs1449962575) and frameshift variants (rs1438085143, rs786205060, rs869320724, rs767094129, rs1384933966, rs869320724 and rs137853123) were the most commonly reported, followed by stop-gain variants (CM1411671, rs137853123 and rs137853122) and synonymous variants (rs4820268). Specifically, these variants were found to impair matriptase-2 autoproteolytic activation, reduce the ability to cleave membrane HJV, inhibit HJV-dependent hepcidin activation and slightly reduce the HJV-mediated inhibition of HAMP transcription and aberrant protein production. These findings shed light on how these genetic variations play a crucial role in perturbing iron homeostasis and contribute to the development of IRIDA.

Based on this information, we comprehensively examined interethnic differences in TMPRSS6 pathogenic mutations using data from various ethnic groups. European populations have a relatively higher number of reported TMPRSS6 gene mutations (p.L606R, p.D521G, rs770897887, rs1438085143, rs1373272804, p.S304L and p.Q405 ∗), indicating the prevalence of various mutations across countries such as Switzerland, Spain, Italy, France, Ireland, Belgium and Austria. South American (rs1449962575, IVS10 + 1 G > A, p.L689P), Middle Eastern (rs767094129), East Asian, Iberian, Hellenic, Arab and Caucasian populations also exhibited specific mutations. The absence of MAF data for TMPRSS6 gene mutations in the dataset makes it challenging to definitively determine their relative allele frequencies. However, from the available information, it seems that the c.1813del mutation has a higher allele frequency than other mutations reported in studies from European populations.

Additionally, the mutations that appeared to have the most significant impact on haematological and biochemical parameters were rs855791 and rs1373272804. These findings suggest that rs855791 is associated with altered iron metabolism. In contrast, rs1373272804, as reported in UniProt, is associated with a reduction in the HJV-mediated inhibition of HAMP transcription, loss of catalytic activity and impaired autoproteolytic and HJV processing. Both rs855791 and rs1373272804 mutations appear to affect haematological parameters such as haemoglobin Hb levels, MCV, MCH and MCHC and biochemical parameters such as hepcidin levels. It is worth noting that other mutations, such as rs4820268, rs1430692214 and rs776877803, were also reported in these studies, but they were considered less significant than rs855791 and rs1373272804 in terms of their impact on haematological and biochemical parameters.

In conclusion, this review provides a comprehensive analysis of the interethnic differences in TMPRSS6 genetic polymorphisms and their association with IRIDA. These findings highlight the presence of specific SNPs and pathogenic mutations in TMPRSS6 and their functional implications. The results also indicated the impact of these mutations on haematological and biochemical parameters associated with IRIDA. However, further research is necessary to accurately determine the allele frequencies of these mutations and to explore the underlying mechanisms linking genetic variations to the phenotypic heterogeneity observed in IRIDA.

Future perspective

This systematic review has provided valuable insights into various SNPs and pathogenic mutations in TMPRSS6 gene; however, future research shall be conducted for better understanding of the mechanistic basis of these variations. Additionally, exploring the molecular pathways may uncover the potential therapeutic targets or interventions for iron resistance iron deficiency anaemia. Further study could also help to establish genotype–phenotype correlations and their clinical outcomes which will ultimately lead to the development of personalized medicine. Moreover, genetic data can be integrated with other data like transcriptomics, proteomics and metabolomics to further develop novel biomarkers and potential therapeutic avenues.

Study limitations

Future studies with larger sample sizes and more extensive genetic data are crucial to better understand the prevalence and frequency of these mutations within different populations and their contribution to the phenotypic variability of the disease. Due to the lack of minor allele frequency data among ethnic groups, we were unable to statistically quantify our results. Further, this review emphasizes the need to conduct more studies with larger sample sizes to establish the allele frequencies of these mutations in diverse populations.