- Email: [email protected]
Differential HFE allele expression in hemochromatosis heterozygotes
1308
CORRESPONDENCE
induced inflammatory response through a CD11b dependent mechanism. Gut 2000;47:88-96.
Differential HFE Allele Expression in Hemochromatosis Heterozygotes Dear SIr: Ro~morduc et al. 1 recently desctlbed differential expressIOn of HFE alleles m subJecrs who are hererozygous for HFE mutations associated with hereditary hemochromatosis, and showed rhat transctlpts carrying the H63D mutarlOn predommate compared With wild-type tranSCtlpts, whereas transcripts With the C282Y mutation are underrepresented. These results are in contrast with our recent observatIOns showing that heterozygotes for HFE mutations express similar levels of notmal and mutated transcripts. 2 Both studies are based on the creation or the elimination of restrictIOn enzyme recognition sites by the mutations; enzymatIC digestIOn of the reverse-transcriptIOn polymerase cham reaction (RT-PCR) product IS therefore used to disctlmmate and quantifY the 2 alleles. We thmk that the results of expressIOn studies like these must be mterpreted With cautIOn, because artifacts can occur. In particular, pamng of wild-type helices With complementary helices carrying a pomt mutatIOn, during the last cycles of PCR, creates heteroduplex molecules that are not digested by restriction enzymes. ThiS will increase the intensity of bands correspondmg to alleles that miss recognition sites for restrictIOn enzymes. In the case of HFE, thiS will cause a predommance of the wild-type transctlpt m those heterozygous for the C282Y mutatIOn, because the mutatIOn mtroduces a new Rsal site, and of the H63D transctlpt in both H63D heterozygotes and compound heterozygotes, because thiS mutatIOn elimmates a DpnII or an Mbol site. We have found thiS particularly evident with the C282 mutation. We therefore chose to perform RT-PCR usmg ptlmer pairs that gave products Similar m size to those obtamed With DNA amplification and performed parallel amplificatIOns of DNA (in which the 2 alleles are equally represented) and complementary DNA (cDNA) (Figure 2 m Vercesl et aJ.2). The ratIOs between bands derived from mutated and Wild-type alleles were Similar m both DNA and cDNA PCR products, suggesting that Similar amounts of the 2 transctlpts were present m the origmal samples. Rosmorduc et al. report consistently reprodUCible ratIOs between the more represented and the less represented alleles (of ~70% and 30%, respectively), with very small standard deViatIOns or standard errors. Because the C282Y allele IS less represented than the normal one and the normal allele is less represented than the one carrymg the H63D mutatIOn, compound heterozygotes would be expected to show a more strlkmg difference between the 2 alleles. The constant difference between alleles, even m compound heterozygotes, IS therefore suggestive of a technical problem. Interpretation ofHFE protem quantificatIOn in Epstein-Barr virusImmortalized cell lines is also problematic. An accurate evaluatIOn of HFE mutations' effects on protem levels would probably reqUire analysis of several cell hnes; the Western blot analySIS by Rosmorduc et al. I shows that differences in HFE protein levels between cell lines are not impressive and, if both the 45- and 43-kilodalton bands are considered, differences could even disappear. In addition, little is known about the regulatIOn of HFE expression at both messenger RNA (mRNA) and protein levels; differences m protem levels could be an effect of altered iron metabolism on mRNA expressIOn or of the ammo acid substitutIOn on protein stability. In fact, available eVIdence indicates that the protein carrymg the C282Y ammo aCid substitutIOn assoCiates poorly, or not at all, With I3z-mlcroglobulin and the transferrin receptor and IS subject to accelerated degrada-
GASTROENTEROLOGY Vol. 120, No. 5
tlOn. 3 .4 Therefore, the most convmcmg eVidence of differential HFE allele expression IS probably represented by cDNA sequencmg m the subject carrymg 2 mutatIOns ar nucleotide 845 (Figure 6 in Rosmorduc et al. I ). We thmk that the problem of the differential expressIOn of HFE alleles IS still open and deserves further investigation. However, one point seems to be clear: if HFE alleles can be expressed at different levels, this must be related to differences in mRNA stability because a transctlptional mechanism would affect the levels of different alleles m a more vatlable way. This confirms our hypotheSIS that CIS-acting elements regulatmg HFE transctlptlOn are not altered m primary Iron overload. GAETANO BERGAMASCHI VALERIA ROLANDI MARIO CAZZOLA Department of Internal Medmne Policfzmco San Matteo Pavza, Italy
1. Rosmorduc 0, Poupon R, Nlon I, et al. Differential HFE allele expression in hemochromatosis heterozygotes. Gastroenterology 2000;119:1075-1086. 2. Vercesi E, Ceranl P, Rolandi V, et al. Abnormal regulation of HFE mRNA expression does not contribute to primary iron overload. Haematologica 2000;85:787-791. 3. Waheed A, Parkkila S, Zhou XV, et al. Hereditary hemochromatosis: effects of C282Y and H63D mutations on association with J3Tmlcroglobulin, intracellular processing, and cell surface expression of the HFE protein in COS-7 cells. Proc Natl Acad SCI USA 1997;94:12384-12389. 4. Feder IN, Penny DM, Irrinki A, et al. The hemochromatosis gene product complexes with the transfernn receptor and lowers its affinity for ligand binding. Proc Natl Acad Sci USA 1998;95: 1472-1477. doi:10.1053/gast.2001.23644
Reply. We appreCiate the mterest of Bergamaschl et al. in our study. We cannot rule out the presence of HFE heteroduplexes, but we assume that they did not Significantly interfere With the quantificatIOn of HFE transctlpts for the following reasons. First, we performed an amplification of different ratIOs of mixed genomic DNA from C282Y and H63D homozygotes (I.e., 25% from C282Y and 75% from H63D; 50% from both C282Y and H63D; and 75% from C282Y and 25% from H63D) as descnbed in the Patients and Methods sectIOn, followed by resttlctlOn enzyme digestion and direct quantification of the amplified products using the GeneScan software of an automated DNA sequencer In all cases, we observed that the tatlO of amplified (845GA; C282Y) and (187CG; H63D) mutated DNA was Similar to the mltial ratIO of genomic mutated DNA. Second, we performed a parallel amplificatIOn of genomic DNA and cDNA around nucleotides 845 and 187, followed by direct sequencing, m the heterozygous patients from 3 families (BI, Le, Be). In all cases, we found a double peak of the same size after amplification of genomic DNA, correspondmg to the presence of a similar amount of mutated and unmutated alleles, but a smgle peak after amplification of the cDNA, correspondmg to the presence of a predommant RNA (as illustrated for patient Be2 m study; see Figure 6). These results were m keepmg With the restrictIOn enzyme digestion data and confirmed the HFE differential allele expressIOn. Similar conclUSIOns were proVided by the Western blot data. However, we could not precisely evaluate the relative importance of the different mechanisms that could modulate the final amount of the different Isoforms of the
CORRESPONDENCE
induced inflammatory response through a CD11b dependent mechanism. Gut 2000;47:88-96.
Differential HFE Allele Expression in Hemochromatosis Heterozygotes Dear SIr: Ro~morduc et al. 1 recently desctlbed differential expressIOn of HFE alleles m subJecrs who are hererozygous for HFE mutations associated with hereditary hemochromatosis, and showed rhat transctlpts carrying the H63D mutarlOn predommate compared With wild-type tranSCtlpts, whereas transcripts With the C282Y mutation are underrepresented. These results are in contrast with our recent observatIOns showing that heterozygotes for HFE mutations express similar levels of notmal and mutated transcripts. 2 Both studies are based on the creation or the elimination of restrictIOn enzyme recognition sites by the mutations; enzymatIC digestIOn of the reverse-transcriptIOn polymerase cham reaction (RT-PCR) product IS therefore used to disctlmmate and quantifY the 2 alleles. We thmk that the results of expressIOn studies like these must be mterpreted With cautIOn, because artifacts can occur. In particular, pamng of wild-type helices With complementary helices carrying a pomt mutatIOn, during the last cycles of PCR, creates heteroduplex molecules that are not digested by restriction enzymes. ThiS will increase the intensity of bands correspondmg to alleles that miss recognition sites for restrictIOn enzymes. In the case of HFE, thiS will cause a predommance of the wild-type transctlpt m those heterozygous for the C282Y mutatIOn, because the mutatIOn mtroduces a new Rsal site, and of the H63D transctlpt in both H63D heterozygotes and compound heterozygotes, because thiS mutatIOn elimmates a DpnII or an Mbol site. We have found thiS particularly evident with the C282 mutation. We therefore chose to perform RT-PCR usmg ptlmer pairs that gave products Similar m size to those obtamed With DNA amplification and performed parallel amplificatIOns of DNA (in which the 2 alleles are equally represented) and complementary DNA (cDNA) (Figure 2 m Vercesl et aJ.2). The ratIOs between bands derived from mutated and Wild-type alleles were Similar m both DNA and cDNA PCR products, suggesting that Similar amounts of the 2 transctlpts were present m the origmal samples. Rosmorduc et al. report consistently reprodUCible ratIOs between the more represented and the less represented alleles (of ~70% and 30%, respectively), with very small standard deViatIOns or standard errors. Because the C282Y allele IS less represented than the normal one and the normal allele is less represented than the one carrymg the H63D mutatIOn, compound heterozygotes would be expected to show a more strlkmg difference between the 2 alleles. The constant difference between alleles, even m compound heterozygotes, IS therefore suggestive of a technical problem. Interpretation ofHFE protem quantificatIOn in Epstein-Barr virusImmortalized cell lines is also problematic. An accurate evaluatIOn of HFE mutations' effects on protem levels would probably reqUire analysis of several cell hnes; the Western blot analySIS by Rosmorduc et al. I shows that differences in HFE protein levels between cell lines are not impressive and, if both the 45- and 43-kilodalton bands are considered, differences could even disappear. In addition, little is known about the regulatIOn of HFE expression at both messenger RNA (mRNA) and protein levels; differences m protem levels could be an effect of altered iron metabolism on mRNA expressIOn or of the ammo acid substitutIOn on protein stability. In fact, available eVIdence indicates that the protein carrymg the C282Y ammo aCid substitutIOn assoCiates poorly, or not at all, With I3z-mlcroglobulin and the transferrin receptor and IS subject to accelerated degrada-
GASTROENTEROLOGY Vol. 120, No. 5
tlOn. 3 .4 Therefore, the most convmcmg eVidence of differential HFE allele expression IS probably represented by cDNA sequencmg m the subject carrymg 2 mutatIOns ar nucleotide 845 (Figure 6 in Rosmorduc et al. I ). We thmk that the problem of the differential expressIOn of HFE alleles IS still open and deserves further investigation. However, one point seems to be clear: if HFE alleles can be expressed at different levels, this must be related to differences in mRNA stability because a transctlptional mechanism would affect the levels of different alleles m a more vatlable way. This confirms our hypotheSIS that CIS-acting elements regulatmg HFE transctlptlOn are not altered m primary Iron overload. GAETANO BERGAMASCHI VALERIA ROLANDI MARIO CAZZOLA Department of Internal Medmne Policfzmco San Matteo Pavza, Italy
1. Rosmorduc 0, Poupon R, Nlon I, et al. Differential HFE allele expression in hemochromatosis heterozygotes. Gastroenterology 2000;119:1075-1086. 2. Vercesi E, Ceranl P, Rolandi V, et al. Abnormal regulation of HFE mRNA expression does not contribute to primary iron overload. Haematologica 2000;85:787-791. 3. Waheed A, Parkkila S, Zhou XV, et al. Hereditary hemochromatosis: effects of C282Y and H63D mutations on association with J3Tmlcroglobulin, intracellular processing, and cell surface expression of the HFE protein in COS-7 cells. Proc Natl Acad SCI USA 1997;94:12384-12389. 4. Feder IN, Penny DM, Irrinki A, et al. The hemochromatosis gene product complexes with the transfernn receptor and lowers its affinity for ligand binding. Proc Natl Acad Sci USA 1998;95: 1472-1477. doi:10.1053/gast.2001.23644
Reply. We appreCiate the mterest of Bergamaschl et al. in our study. We cannot rule out the presence of HFE heteroduplexes, but we assume that they did not Significantly interfere With the quantificatIOn of HFE transctlpts for the following reasons. First, we performed an amplification of different ratIOs of mixed genomic DNA from C282Y and H63D homozygotes (I.e., 25% from C282Y and 75% from H63D; 50% from both C282Y and H63D; and 75% from C282Y and 25% from H63D) as descnbed in the Patients and Methods sectIOn, followed by resttlctlOn enzyme digestion and direct quantification of the amplified products using the GeneScan software of an automated DNA sequencer In all cases, we observed that the tatlO of amplified (845GA; C282Y) and (187CG; H63D) mutated DNA was Similar to the mltial ratIO of genomic mutated DNA. Second, we performed a parallel amplificatIOn of genomic DNA and cDNA around nucleotides 845 and 187, followed by direct sequencing, m the heterozygous patients from 3 families (BI, Le, Be). In all cases, we found a double peak of the same size after amplification of genomic DNA, correspondmg to the presence of a similar amount of mutated and unmutated alleles, but a smgle peak after amplification of the cDNA, correspondmg to the presence of a predommant RNA (as illustrated for patient Be2 m study; see Figure 6). These results were m keepmg With the restrictIOn enzyme digestion data and confirmed the HFE differential allele expressIOn. Similar conclUSIOns were proVided by the Western blot data. However, we could not precisely evaluate the relative importance of the different mechanisms that could modulate the final amount of the different Isoforms of the