Hemochromatosis in India: First Report of Whole Exome Sequencing With Review of the Literature

Abstract

Background

Primary hemochromatosis is unusual in India. The homeostatic iron regulator (HFE) gene C282Y mutation, a common cause for hemochromatosis in Europe, is considered almost nonexistent in India. We are reporting a case of hemochromatosis with the HFE gene C282Y mutation and two other adult cases with a novel hemojuvelin (HJV) mutation from Kerala.

Methods

Of 434 cases with chronic liver disease, 3 cases were identified with the serum ferritin level of more than 1000 ng/mL and primary hemochromatosis after excluding secondary causes. Whole exome sequencing, including genes HFE, HJV, SLC40A1, TFR2, FTH1, HAMP, SKIV2L, TTC37, and BMP2, was performed for blood samples in all 3 cases.

Results

One patient with hemochromatosis had a homozygous HFE gene C282Y mutation, and two other adult cases had a novel homozygous HJV D355Y mutation. This is the first report of hemochromatosis associated with the HFE C282Y mutation from Kerala and the second report in India. This is the second report of hemochromatosis associated with an HJV mutation from India.

Conclusion

HJV mutations may explain some of the adult onset primary hemochromatosis in India.

Primary hemochromatosis is unusual in India. The homeostatic iron regulator (HFE) gene C282Y mutation, a common cause for hemochromatosis in Europe, is considered almost nonexistent in India. Although there are several reports of primary hemochromatosis from India, it has not been possible to identify the associated gene. Specifically, the HFE gene C282Y mutation has not been found except in one family from Andhra Pradesh. Only one recent study from Chandigarh investigated the hemojuvelin (HJV) gene in primary hemochromatosis, and 4 unrelated cases with significant mutations were found. We are reporting a case of hemochromatosis with the HFE gene C282Y mutation and two other adult cases with a novel HJV mutation from Kerala.

Patients and methods

Between 2017 and 2018, 434 patients with chronic liver disease were evaluated in our department, including one patient already diagnosed as having hemochromatosis and being treated with phlebotomy. Three of these patients had serum ferritin more than 1000 ng/mL with no secondary cause for iron overload (Figure 1). Patients with a previous history of blood transfusion, hematologic disorder, renal failure, or significant (>20 g/day) alcohol intake were excluded. Whole exome sequencing was carried out in all 3 of these patients.

Figure 1.

Patient cohort with cirrhosis and high serum ferritin in whom the genetic test was performed. HFE, homeostatic iron regulator. WES, whole exome sequencing.

Whole Exome Sequencing

The sample was processed in Centogene laboratory (Rostock, Germany), adhering to American College of Medical Genetics (ACMG) guidelines.¹ DNA was extracted from blood, and the exome was captured using the Illumina Nextera Rapid Capture Exome Kit (Illumina, Inc., San Diego, CA, USA) and sequenced on HiSeq4000 (Illumina, Inc.). Variant annotation and filtering was carried out as described previously.¹

Statistics

The characteristics of the patients were reported with the use of descriptive statistics.

Results

Clinical Features of Patients

The clinical features of the patients are given in Table 1.

Table 1.

Clinical Features of the Patients.

Age	Sex	Diabetes	TS, %	Ferritin, ng/mL	HFE	HJV
39	M	No	83	3540	C282Ya	Normal
32	M	No	97	5537	No	D355Ya
42	M	Yes	90	8690	No	D355Ya

TS, transferrin saturation; HFE, homeostatic iron regulator gene mutation; HJV, hemojuvelin BMP co-receptor gene mutation.

^aHomozygous.

Case 1

S.A.R., a 39-year-old man presented in June 2009 with history of elevated transaminases since January 2004 and darkening of face since 2005. On examination, he had dark complexion. No other abnormality was detected. Specifically, there was no clinical evidence of cardiac involvement or hypogonadism. Investigations showed the following values: hemoglobin, 15.5 g/dL; total leukocytes, 5900/μL; platelets, 214,000/μL; serum iron, 249 μg/dL; total iron-binding capacity, 300 μg/dL; transferrin saturation, 83%; serum ferritin, 3540 ng/mL; total protein, 7.1 g/dL; albumin, 4.1 g/dL; total bilirubin, 2.7 mg/dL; direct bilirubin, 0.3 mg/dL; serum glutamic-oxaloacetic transaminase (SGOT), 80 U/mL; serum glutamic-pyruvic transaminase (SGPT), 140 U/L; alkaline phosphatase, 49 U/L (normal 35–140 U/L); Gamma-glutamyl transferase (GGT), 65 U/L (normal <85 U/L); international normalised ratio (INR), 1.1; serum creatinine, 0.9 mg/dL; serum sodium, 140 mMol/L; fasting blood sugar, 93 mg/dL; and glycated hemoglobin (HbA1c), 5.3%. Liver biopsy showed grade 4 siderosis in hepatocytes with bridging fibrosis and pigment in the bile duct epithelium. The HFE gene test showed homozygous C282Y and homozygous normal H63=. Subsequently, whole exome sequencing confirmed the homozygous HFE variant, Cys282Tyr, characteristic of hemochromatosis. No other known pathogenic or expected pathogenic variant was seen in genes associated with hemochromatosis (HFE, HJV, SLC40A1, TFR2, FTH1, HAMP, SKIV2L, TTC37, and BMP2). No pathogenic or expected pathogenic variant was seen in the 59 genes listed by ACMG SF v2.0 for mandatory reporting. He has 2 sisters and one brother. There was no family history of liver disease. Despite intensive search, no family connection to European or Andhra Pradesh ancestry could be found. He was started on phlebotomy, initially every 2 weeks, then every 4, 8, or 12 weeks, as shown in Figure 2. There was no fall in hemoglobin concentration, but that of serum ferritin gradually became normal. He has remained asymptomatic and well over the 10-year follow-up.

Figure 2.

Course of parameters in case 1 over 10 years.

Case 2

S.V.S., a 42-year-old man presented in May 2018 with a history of detection of hemochromatosis and diabetes mellitus on family screening in 2011. His sister was diagnosed to have hemochromatosis in October 2011 and died in 2012 (Figure 3). He mentioned a history of poor semen discharge since 2012, suggestive of hypogonadism. On examination, no abnormality was detected. He was having phlebotomy every 4 months since 2011. Blood sugar was controlled with insulin. Investigations in October 2011 had shown the concentration of serum ferritin to be 8690 ng/mL and SGOT to be 109 U/mL. Investigations in May 2018 showed the following values: hemoglobin, 12.8 g/dL; total leukocytes, 7880/μL; platelets, 255,000/μL; serum iron, 170 μg/dL; total iron-binding capacity, 335 μg/dL; transferrin saturation, 51%; serum ferritin, 43 ng/mL; total protein, 7.2 g/dL; albumin, 4.4 g/dL; total bilirubin, 0.4 mg/dL; direct bilirubin, 0.1 mg/dL; SGOT, 25 U/mL; SGPT, 28 U/L; alkaline phosphatase, 104 U/L (normal 35–140 U/L); GGT, 36 U/L (normal <85 U/L); INR, 1.0; serum creatinine, 0.6 mg/dL; and serum sodium, 137 mMol/L. Hepatitis B surface antigen (HBsAg) and Hepatitis C antibody test (anti-HCV) were negative. The ultrasound of the abdomen showed hyperechogenicity and coarse echostructure of the liver and a gall bladder stone. The echocardiogram was normal.

Figure 3.

Family tree of patient 2.II.3, the proband.

Whole exome sequencing showed no known pathogenic variants in genes associated with hemochromatosis (HFE, HJV, SLC40A1, TFR2, FTH1, HAMP, SKIV2L, TTC37, and BMP2). A homozygous variant, of uncertain significance, Asp355Tyr in the HJV gene, associated with juvenile hemochromatosis was found.

A homozygous variant, Ile148Met in the PNPLA3 gene, associated with nonalcoholic steatohepatitis (NASH) was found. Heterozygous variants Asp905Tyr in the PPP1R3A gene associated with diabetes mellitus and Leu446Pro in the GCKR gene associated with fasting blood sugar quantitative trait were found. In addition, a heterozygous variant Arg818Cys of conflicting interpretation of pathogenicity in the WFS1 gene associated with diabetes mellitus was also found. No pathogenic or expected pathogenic variant was seen in the 59 genes listed by ACMG SF v2.0, for mandatory reporting. He has one living sister who is normal.

Case 3

R.R., a 32-year-old man presented in September 2018 with history of elevated transaminases since 2014. He was detected to have diabetes mellitus in 2011 and was on insulin. On examination, no abnormality was detected. Specifically, there was no clinical evidence of cardiac involvement or hypogonadism. Investigations showed the following values: hemoglobin, 15.2 g/dL; reticulocyte count, 0.5%; total leukocytes, 8080/μL; platelets, 180,000/μL; serum iron, 269 μg/dL; total iron-binding capacity, 278 μg/dL; transferrin saturation, 97%; serum ferritin, 5447 ng/mL; total protein, 7.6 g/dL; albumin, 4.7 g/dL; total bilirubin, 0.5 mg/dL; direct bilirubin, 0.2 mg/dL; SGOT, 122 U/mL; SGPT, 142 U/L; alkaline phosphatase, 218 U/L (normal 35–140 U/L); GGT, 138 U/L (normal <85 U/L); serum creatinine, 0.6 mg/dL; serum sodium 133 mMol/L; and HbA1c, 9.1%. HBsAg was negative. The ultrasound of the abdomen showed hyperechogenic liver but no evidence of portal hypertension or cirrhosis. Magnetic resonance imaging (MRI) hepatic iron quantification using T2 star multi-echo technique showed moderate iron deposition. The calculated iron concentration was 8.8 mg/g liver (normal less than 2 mg/g). The pancreas showed mild iron deposition.

Whole exome sequencing showed no known pathogenic variants in genes associated with hemochromatosis (HFE, HJV, SLC40A1, TFR2, FTH1, HAMP, SKIV2L, TTC37, and BMP2). A homozygous variant, of uncertain significance, Asp355Tyr in the HJV gene, associated with juvenile hemochromatosis was found. Heterozygous variants, Gly169Ser in the CEL gene, associated with diabetes mellitus and, Asp905Tyr in the PPP1R3A gene, associated with insulin resistance were also found. No pathogenic or expected pathogenic variant was seen in the 59 genes listed by ACMG SF v2.0, for mandatory reporting. He had only one sibling who died in a road traffic accident at the age of 12 years. There was no family history of liver disease. He was started on phlebotomy.

Discussion

The present study is the first to report comprehensive whole exome sequencing for primary hemochromatosis in India. HFE, HJV, SLC40A1, TFR2, FTH1, HAMP, SKIV2L, TTC37, and BMP2 genes were screened.

Primary hemochromatosis is unusual in India. In a series of 236 patients with chronic liver disease from Chandigarh, 17 patients showed clinical and biochemical features of primary iron overload. Liver biopsy performed in 12 of 17 of these patients showed grade 3 to 4 parenchymal iron overload consistent with hemochromatosis.² In another series of 249 patients with biopsy-proven nonalcoholic cirrhosis from New Delhi, only 24 patients had transferrin saturation more than 60% and none had grade 3 to 4 parenchymal iron in the liver.³ Yet, in another series of 496 patients with cirrhosis from Lucknow, 13 patients with primary iron overload were found.⁴

Several attempts have been made to find the genetic cause of primary hemochromatosis in India. Table 5 lists several studies from India in which a search for the homozygous HFE C282Y mutation was made but not found. The first and only report of detection of the homozygous C282Y mutation associated with hemochromatosis in India was from Vellore in a family from Andhra Pradesh.⁵ They reported all 5 children, male, 4 and female, 1, from a consanguineous marriage that probably had hemochromatosis. Three boys and one girl were tested and found to be C282Y homozygous. The boys had serum ferritin ranging from 652 to 1388 mg/mL. The proband had an MRI report that is suggestive of hemochromatosis. They also reported finding heterozygous C282Y in a presumably unrelated patient with HbE-β-thalassemia. Other C282Y heterozygotes have been reported from patients with or without primary iron overload.^{4, 5, 6, 7} Systematic search for other genes that may be involved in primary hemochromatosis had not been made in India till Dhillon et al. reported a search of the entire exons of HFE, HJV, HAMP, TFR2, and SLC40A1 in 19 unrelated north Indian patients with chronic liver disease and serum ferritin >1000 μg/L.⁸ Secondary causes of iron overload were excluded. Twelve patients had liver biopsy showing iron overload, and 3 had MRI evidence of liver iron overload. Six of these 19 patients had significant mutations in the HJV gene. The significant HJV mutations were Gly336Ter, c.-36 G > A, and c.-358 G > A.

Table 5.

Previous Reports of Search for HFE Mutations in Patients With Suspected Primary Iron Overload in India.

Author	Year	N	HFE allele				Ref
			C282Y		H63D
			Homozygous	Heterozygous	Homozygous	Heterozygous
Panigrahi	2006	31	0	0	1	6	22
Shukla	2006	5	0	0	1	0	23
Dhillon	2007	17	0	0	0	2	2
Jain	2011	13	0	1	0	2	4
Athiyarath	2013	4	4	0	0	0	5
Sharma	2014	1	0	0	0	0	24
Noor	2016	20	0	0	0	2	25
Dhillon	2018	19	0	0	0	3	8

HFE, homeostatic iron regulator.

Our patient, with the HFE C282Y mutation, had classical HFE hemochromatosis. The mutation is rare in India, but this second report from India confirms that HFE C282Y mutation–associated hemochromatosis does exist at least in South India. Detection of these patients before development of cirrhosis is important because patients without cirrhosis on regular phlebotomy have a normal life span, whereas those undetected are at risk of developing cirrhosis and hepatocellular carcinoma and having a reduced life span.^{9, 10, 11, 12}

Our two patients with HJV mutations had the same homozygous Asp355Tyr mutation in exon 4. This was the only variation found in the HJV gene. The D355Y (Asp355Tyr) mutation is highly conserved through different species (Table 2). The population frequency of this variant in Exome Aggregation Consortium (ExAC) and Genome Aggregation Database (gnomAD) is extremely low (Table 3). Only a single heterozygote has been reported in ExAC (1/121,368), from a South Asian region but not from any other region. Only 3 heterozygotes have been reported in gnomAD (3/246,028), 2 South Asian and 1 Latino. No homozygote has been reported. The D355Y variant is predicted to be damaging by both Sorting Intolerant from Tolerant (https://sift.bii.a-star.edu.sg/www/Extended_SIFT_chr_coords_submit.html [1,145416718,1,G/T]) and Polymorphism Phenotyping v2 (http://genetics.bwh.harvard.edu/pph2/[Q6ZVN8, 355, D, Y]). The amino acid aspartic acid has very different physicochemical properties compared with tyrosine, and this substitution is expected to change the three dimensional structure, function, and interactions of the protein. Owing to all these factors, the D355Y variant should be considered as the cause of the iron overload in these 2 patients.

Table 2.

Conservation of HJV D355 Amino Acid (AA) in Different Species.¹³

Species	AA position									↓
Human	347	C	K	E	G	L	P	V	E	D	A	Y	F	H	S	C	V	F	D	V
Rat	343	C	K	E	G	L	P	V	E	D	A	Y	F	Q	S	C	V	F	D	V
Mouse	340	C	K	E	G	L	P	V	E	D	A	Y	F	Q	S	C	V	F	D	V
Zebra fish	329	C	S	S	L	F	P	G	R	D	V	Y	F	Q	A	C	L	F	D	V

HJV, hemojuvelin.

HJV D355 equivalents shown in bold.

Table 3.

Frequency of HJV D355Y Mutation (1:145416718 G/T). http://exac.broadinstitute.org/variant/1-145416718-G-T, rs782371683, Accessed 02/02/2019.

Mutation	Coveragea	ExAC All	ExAC SA	gnomAD	SIFT prediction	Polyphen2 prediction (score)
D355Y	204/169	0.000008239	0.00006057	0.00001219	Damaging	Probably damaging (1.0)

ExAC, Exome Aggregation Consortium; gnomAD, Genome Aggregation Database; Polyphen2, Polymorphism Phenotyping v2; SA, South Asia; SIFT, Sorting Intolerant from Tolerant.

http://exac.broadinstitute.org/variant/1-145416718-G-T, rs782371683, Accessed 02/02/2019.

^aDepth of coverage in case 1/case 2.

HJV binds to bone morphogenetic protein (BMP) receptors, HFE and TRF2, to form a cell membrane complex.¹³ The binding of BMP6 to this complex promotes downstream signaling, resulting in transcriptional activation of hepcidin (HAMP) promoter.¹³ Hepcidin inhibits iron absorption from the intestine. Therefore, HJV binding to BMP inhibits iron absorption. Poor binding of HJV to BMP would result in excess iron absorption and iron overload.¹³ The D355Y variant is expected to alter the binding of HJV to BMP and cause hemochromatosis.

Hemochromatosis due to the HJV mutation classically occurs before the age of 30 years. However, several cases presented after the age of 30 years have recently been described (Table 4). Although the disease is generally severe and cardiac involvement is frequent, the phenotype is very variable. Mutations found in patients presenting after the age of 30 years are highlighted with * in Figure 4. Because the commonest HJV genotype, G320V, is found in young and older patients, factors other than the specific HJV mutation may be playing a significant role in determining the age, organ involvement, and severity at presentation.

Table 4.

Previous Reports of Patients With HJV Mutations Presenting Above the Age of 30 Years.

Year	Age	Sex	Diabetes	Cardiac	TS, %	Ferritin, ng/mL	HFE	HJV	Ref
2005	48	M	Yes	AF	95	6115	No	D249H	14
	51	M	Yes	CCF	96	2280	No	Q312X
	51	F	Yes	?	96	4278	No	Q312X
2009	32	M	No	?	100	3756	No	L194P	15
2010	31	F	?	No	85	3542	No	G320V	16
2010	58	F	?	?	60	443	H63Da	A310Ga	17
	46	F	?	?	95	2794	No	c.-36G > C
2013	39	F	?	CCF	100	116,070	No	G320V	18
2014	37	M	Yes	No	60	>2000	No	Q6Ha C321Xa	19
	47	M	Yes	?	40	>2000	No
2018	45	M	Yes	No	94	3136	H63Da	G336X	7
	49	F	No	No	100	1700	No	G336X
	38	F	Yes	No	69	9400	No	G336X
	47	M	Yes	Yes	91	12,433	No	G336X
	43	M	Yes	No	79	2800	No	c.-36G > A c.-358G > A
	32	F	Yes	Yes	68	4250	No
2018	68	F	Yes	Yes	99	3987	No	C317S	20
2018	53	M	No	?	86	4220		R41Pa R176Ca	21
	31	M	No	?	91	3700		G320V
	31	M	No	?	100	2300		L101Pa H180Ra
	60	F	Yes	?	86	4800		L101Pa A384Va
	32	M	No	?	90	4959		R288W

AF, atrial fibrillation; CCF, congestive cardiac failure; HFE, homeostatic iron regulator; HJV, hemojuvelin; TS, transferrin saturation; ?, Not specified.

^aHeterozygous.

Figure 4.

Cartoon of HJV 1D structure with known mutations.^{7, 13, 14, 15, 16, 17, 18, 19} 1D, single-dimensional; HJV, hemojuvelin.

This study has several limitations. Functional studies have not been carried out to confirm the pathogenicity of the D355Y mutation in HJV. The study does not reveal the frequency of hemochromatosis or the frequency of the HJV mutation in India. The report of one new patient with the HFE mutation and 2 new patients with the HJV mutation in adults does not indicate the frequency of these mutations in adult patients with hemochromatosis in India. Furthermore, it does not exclude the possibility that other unknown iron overload–related genes may be important causes of primary iron overload in India.

The HFE C282Y mutation does exist in South India. HJV mutations may explain some of the primary adult onset hemochromatosis in India.

Conflicts of interest

The authors have none to declare.