MUTATION REPORTS

Late-Onset Erythropoietic Porphyria Caused by aChromosome 18q Deletion in Erythroid Cells

Carolyn Aplin, Sharon D. Whatley,* Peter Thompson,² Terry Hoy,³ Paul Fisher,§ Charles Singer,¶Christopher R. Lovell, and George H. Elder*Department of Dermatology, Royal United Hospital, Bath; Departments of *Medical Biochemistry, ²Medical Genetics, ³Haematology, and

§Psychological Medicine, University of Wales College of Medicine and University Hospital of Wales, Cardiff; ¶Department of Haematology,

Royal United Hospital, Bath, U.K.

The erythropoietic porphyrias, erythropoietic proto-porphyria and congenital erythropoietic porphyria,result from germline mutations in the ferrochelatasegene and uroporphyrinogen III synthase gene,respectively. Both conditions normally present inchildhood but rare cases with onset past the age of40 y have been reported. Here we show that late-onset erythropoietic protoporphyria can be causedby deletion of the ferrochelatase gene in hematopoie-

tic cells with clonal expansion as part of the myelo-dysplastic process. This is the ®rst directdemonstration of porphyria produced by an acquiredmolecular defect restricted to one tissue. Some othercases of late-onset erythropoietic porphyria may beexplained by a similar mechanism. Key words: dele-tion/erythropoietic protoporphyria/FECH/ferrochelatase/myelodysplasia. J Invest Dermatol 117:1647±1649, 2001

The porphyrias are disorders that result from over-production of heme precursors caused by partialde®ciencies in enzymes of heme biosynthesis. In allporphyrias, apart from the sporadic form of porphyriacutanea tarda, the enzyme de®ciencies are present in

all tissues and caused by mutations in the corresponding genes thatare inherited through the germline to produce diseases withautosomal dominant or recessive patterns of inheritance (Kappas etal, 1995). In the erythropoietic porphyrias, congenital erythro-poietic porphyria (CEP) (Fritsch et al, 1997; Desnick et al, 1998)and erythropoietic protoporphyria (EPP) (Todd, 1994), the bonemarrow is the main site of overproduction of heme precursors.

Patients with CEP are homoallelic or heteroallelic for mutationsin the uroporphyrinogen III synthase (UROS) gene (Desnick et al,1998) whereas clinical manifestation of EPP, a low-penetranceautosomal dominant disorder (Went and Klasen, 1984), appears torequire a severe mutation in the ferrochelatase (FECH) gene onone allele and a low expression ferrochelatase allelic variant, whichis common in the general population, on the other (Gouya et al,1999). Together these decrease ferrochelatase activity to the 20%±30% of normal found in clinically overt EPP. Both conditionsnormally present in childhood with symptoms caused by photo-sensitization of the skin by porphyrins (Todd, 1994; Fritsch et al,1997) but rare, late-onset forms of each disorder in whichsymptoms do not appear until after the age of 40 y have beenreported (Murphy et al, 1985, 1995; Fallon et al, 1989; Lim et al,1992; Fritsch et al, 1997). In some of these patients, disorders

clinically and biochemically resembling EPP (Lim et al, 1992) orCEP (Yamauchi and Kushibi, 1992; Ibbotson et al, 1998) havedeveloped in association with the myelodysplastic syndrome(MDS). Molecular investigation of these cases has not beenreported. The 45 kb FECH gene is located at chromosome18q21.3 (Brenner et al, 1992). Here we show that a deletion of oneallele of this gene that is restricted to hematopoietic cells can causelate-onset EPP in MDS and suggest that somatic mutation inhematopoietic cell lineages may provide a complete or partialexplanation for the pathogenesis of some other cases of late-onseterythropoietic porphyrias.

MATERIALS AND METHODS

Patient A 51-y-old man presented with a 3 mo history ofphotosensitivity, dyspnoea, and lethargy. He described a severe burningsensation in the skin of his forearms and hands within minutes ofexposure to the sun with subsequent erythema, edema, and soreness thatlasted several days. His face was affected on one occasion only.Examination showed a resolving erythematous eruption and edema overthe dorsal aspects of both hands and forearms sparing the area covered byhis watch with diffuse thickening of the skin over the interphalangealjoints. He had a past history of tuberculosis as a child and alcohol abusebut had been teetotal for 15 y; there was no evidence for exposure toother toxic chemicals. Biochemical investigation showed increasederythrocyte protoporphyrin at 11.9 mmol per liter (normal, 0.4±1.7 mmolper liter) that was predominantly in the free form not complexed withzinc, increased plasma porphyrin (86 nmol per liter; normal less than10 nmol per liter), and normal urinary and fecal porphyrinconcentrations. This pattern of abnormalities con®rmed the clinicaldiagnosis of late-onset EPP. Histologic examination of a skin biopsytaken from the back of the hand showed extensive proliferation ofhyaline-like material positive to periodic acid±Schiff stain in and aroundsuper®cial dermal blood vessels, characteristic of cutaneous porphyria.Haematologic examination showed a peripheral blood pancytopenia(hemoglobin, 7.4 g per dl; leukocytes, 3.5 3 109 per liter; platelets,61 3 109 per liter) and a hypercellular bone marrow with excess

Manuscript received March 28, 2001; revised June 8, 2001; accepted forpublication August 15, 2001.

Reprint requests to: Professor G.H. Elder, Department of MedicalBiochemistry, University of Wales College of Medicine, Heath Park,Cardiff CF14 4XN, U.K. Email: elder@cardiff.ac.uk

Abbreviations: CEP, congenital erythropoietic porphyria; EPP, ery-thropoietic protoporphyria; FECH, ferrochelatase gene; FISH, ¯uores-cence in situ hybridization; MDS, myelodysplastic syndrome; UROS,uroporphyrinogen III synthase gene.

0022-202X/01/$15.00 ´ Copyright # 2001 by The Society for Investigative Dermatology, Inc.

1647

myeloblasts and ring sideroblasts consistent with a diagnosis of refractoryanemia with excess blasts in transformation.

Molecular investigation Genomic DNA from bone marrow,peripheral blood leukocytes, and skin was extracted by standardprocedures. All 11 exons of the FECH gene with 20±100 bp of ¯ankingsequence and 140 bp 5¢ to exon 1 were ampli®ed by the polymerasechain reaction (PCR) using previously described primers (Rufenacht etal, 1998) without G±C clamps. PCR-ampli®ed double-stranded DNAwas puri®ed and sequenced in forward and reverse orientations using anApplied Biosystems model 377 automated sequencer as describedpreviously (Whatley et al, 1999). Haplotyping, using single nucleotidepolymorphisms within the FECH gene, was carried out as describedpreviously (Gouya et al, 1999). Microsatellite allele ratios were measuredusing ¯uorescence-labeled PCR primers and an Applied Biosystemsmodel 373 automated sequencer with Genotyper software (PE AppliedBiosystems, Warrington, U.K.).

Other methods Bone marrow cytogenetic studies were carried outby standard procedures. Fluorescence in situ hybridization (FISH) withwhole chromosome 18 paint was carried out according to themanufacturer's instructions (Cambio, Cambridge, U.K.). Aspirated bonemarrow cells were sorted by ¯ow cytometry (FACS 440; BectonDickinson, Oxford, U.K.) according to the presence or absence ofporphyrin-induced auto¯uorescence, using excitation by a laserproducing beams at 351 nm and 363 nm and an emission wavelength ofgreater than 620 nm.

RESULTS

Identi®cation of abnormal bone marrow karyotype Bonemarrow cytogenetic studies showed a complex abnormalkaryotype: 44, XY, ±2, ±5, ±7, del(9) (q21q34), ±12, ±12, ±16, ±17, del(18) (q12q23 or q11.2q22), +5 mar (Fig 1). FISHcon®rmed an 18q deletion that included the region (18q21.3)containing the FECH gene (Fig 1). Chromosomes 1, 3, 9, 10, and11 that carry the other enzymes of heme biosynthesis appearednormal.

Demonstration of FECH haplode®ciency in bonemarrow Bone marrow cells were sorted by ¯ow cytometryinto major non¯uorescent and minor (21% of total cells) ¯uorescentpopulations. Morphologic examination of cytospin preparationsshowed that the minor fraction was enriched with late normoblasts.To demonstrate loss of the region of one FECH allele inhematopoietic but not other cells, the allele ratios of amicrosatellite marker (D18S858) adjacent to the FECH locus(Wang et al, 1999) were measured in genomic DNA from bonemarrow, ¯uorescent and non¯uorescent bone marrow fractions,peripheral blood leukocytes, and skin. Table I shows almostcomplete loss of heterozygosity in total and fractionated bonemarrow, indicating the presence of a single, dominant neoplasticclone containing the 18q deletion, and some loss of heterozygosityin peripheral blood leukocytes ± an observation that is consistentwith the expected noninvolvement of lymphoid cells in themyelodysplastic process (Heaney and Golde, 1999).

Demonstration that the remaining FECH allele in bonemarrow is normal Because one null FECH allele does notnormally lead to overproduction of suf®cient porphyrin byerythroid cells to cause photosensitivity (Gouya et al, 1999), weexamined the other allele for any point mutation that mightdecrease enzyme activity and for a haplotype de®ned by intragenicsingle nucleotide polymorphisms [±251G IVS1±23T] that isstrongly associated with a low expression ferrochelatase allele inFrance (Gouya et al, 1999) and was present in 97 of 98 unrelatedBritish patients with clinically overt EPP that we have haplotyped(Mason N, Whatley S, Elder G: unpublished data). The sequencewas identical to the reported sequence (HGMP GenBank accessionnumber AJ250235) except for one nucleotide in exon 8 (c798 G> C; P166P). Sequencing DNA from the patient's skin showed thathe was heterozygous at this site, thus allowing direct con®rmationof loss of one FECH allele in hematopoietic cells (Table I).Haplotyping of DNA from the patient's skin showed that he washomozygous for the haplotype [±251A IVS1±23C] that is notassociated with low expression of ferrochelatase (Gouya et al, 1999).Thus we could ®nd no evidence to indicate that the remainingFECH allele was abnormal.

DISCUSSION

Investigation of this patient shows that deletion of one FECH allelein hematopoietic cells can lead to the late onset of a form of EPPthat is otherwise clinically and biochemically indistinguishable from

Figure 1. Bone marrow cytogenetic studies reveal abnormalkaryotype with deletion of 18q. (a) Abnormal karyotype inhematopoietic cells. (b) FISH with chromosome 18 paint; arrows showintact and partially deleted 18q chromosomes.

Table I. Allele ratios show loss of heterozygositya

Tissue D18S858 FECH

Skin 1.0 1.0Peripheral blood leukocytes 0.62 NDBone marrow 0.15 NDNon-¯uorescent cell fraction 0.13 0.24Fluorescent cell fraction 0.06 0.12

aAllele ratios have been adjusted assuming that the ratio in genomic DNA fromskin is 1.0. Experimentally determined ratios were 1.3 for D18S858(207 bp:213 bp) and 2.2 for an intragenic FECH polymorphism (798C:798G).ND, not done.

1648 APLIN ET AL THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

the inherited disease. Overproduction of protoporphyrin fromerythroid cells in suf®cient quantity to produce photosensitizationdoes not normally occur in EPP in the absence of a low expressionferrochelatase allele trans to the null allele (Gouya et al, 1999). Inour patient this did not appear to be the case, though the presenceof a mutation leading to impaired expression, but lying outside theregions that we sequenced, cannot be excluded.

If the single FECH gene in our patient was expressed at normallevels, our ®ndings suggest either that haplode®ciency alone maylead to overproduction of protoporphyrin when erythropoiesis isabnormal as in MDS or that ferrochelatase activity is impaired in atleast some erythroid cells in MDS. Thus, the dysplastic erythropoi-esis with ring sideroblasts present in our patient may have decreasedferrochelatase activity in bone marrow below the 50% expectedfrom haplode®ciency alone; for example, damage by iron tomitochondrial membranes might impair ferrochelatase functioneither directly or indirectly by decreasing iron transfer from theinner membrane (Lange et al, 1999). Monosomy of chromosome18 and deletions of 18q are almost certainly more frequent in MDS(Berger et al, 1992) than protoporphyrin-induced photosensitivity,which suggests either that additional factors to decrease enzymeactivity are required, and are only present in a proportion of thosewith one FECH allele, or that increased erythrocyte freeprotoporphyrin concentrations, insuf®ciently high or persistent tocause overt photosensitivity, are more frequent in MDS thanappreciated.

Our ®ndings may provide a general explanation for theassociation with MDS of other late onset cases of EPP. In two of®ve other patients with EPP and MDS, bone marrow cytogeneticstudies showed deletions of chromosome 18q and, in one of these,allogeneic bone marrow transplantation returned the porphyrinpro®le to normal (Lim et al, 1992). The karyotype was eithernormal or not examined in the other three patients. Asmicrodeletions and other mutations in hematopoietic cells arecommon in MDS (Mufti, 1992), these patients may also haveacquired a null FECH allele. Any future patients with late onsetEPP should be screened for underlying hematologic disease andFECH mutations that are restricted to the bone marrow.

Whether the association of late-onset CEP with MDS can beexplained by acquired mutation of the UROS gene remains to bedetermined. Of the two cases reported, the bone marrow karyotypewas either not examined (Yamauchi and Kushibi, 1992) or showedno abnormalities (Ibbotson et al, 1998). The UROS gene is locatedon chromosome 10q, which is rarely deleted in MDS (Heim,1992). In order to reach the very low enzyme activities found inautosomal recessive CEP (Fontanellas et al, 1996), both UROSalleles would need to be abnormal in the hematopoietic cells ofpatients with CEP and MDS. Possible mechanisms include somaticmutation in the hematopoietic cells of an individual heterozygousfor a germline UROS mutation, analogous to the mechanismpostulated for late onset of another autosomal recessive porphyriaassociated with polycythaemia (Akagi et al, 2000); somatic mutationof both alleles with at least one mutation retaining some UROSfunction; or somatic mutation of one allele combined withchromosome 10 monosomy or disomy.

In addition to mutation of either the FECH or UROS gene inhematopoietic cells, expansion of the defective clone is required. InMDS, this occurs as part of the neoplastic process (Heaney and

Golde, 1999). At present, there is no evidence that somaticmutation in hematopoietic stem cells of any of the genes for hemebiosynthesis can initiate clonal expansion but analysis of the relevantgenes in genomic DNA from bone marrow would seem worth-while in any patient with late-onset erythropoietic porphyria inwhom germline mutation does not provide a complete explanationfor their disease.

REFERENCES

Akagi R, Nishitani C, Harigae H, et al: Molecular analysis of delta-aminolevulinatedehydratase de®ciency in a patient with an unusual late-onset porphyria. Blood96:3618±3623, 2000

Berger R, Le Coniat M, Derre J, Flexor MA, Hillion J: Abnormalities ofchromosome 18 in myelodysplastic syndromes and secondary leukemia. CancerGenet Cytogenet 63:97±99, 1992

Brenner DA, Didier JM, Frasier F, Christensen SR, Evans GA, Dailey HA: Amolecular defect in human protoporphyria. Am J Hum Genet 50:1203±1210,1992

Desnick RJ, Glass IA, Xu W, Solis C, Astrin KH: Molecular genetics of congenitalerythropoietic porphyria. Seminars Liver Dis 18:77±84, 1998

Fallon JD, Kvedar JC, Margolis RJ, Pathak MA: Erythropoietic protoporphyriapresenting in adulthood. Arch Dermatol 125:1286±1287, 1989

Fontanellas A, Bensidhoum M, Enriquez de Salamanca R, Moruno Tirado A, deVerneuil H, Ged C: A systematic analysis of the mutations of theuroporphyrinogen III synthase gene in congenital erythropoietic porphyria.Eur J Hum Genet 4:274±282, 1996

Fritsch C, Bolsen K, Ruzicka T, Goerz G: Congenital erythropoietic porphyria. J AmAcad Dermatol 36:594±610, 1997

Gouya L, Puy H, Lamoril J, Da Silva V, Grandchamp B, Nordmann Y, Deybach JC:Inheritance in erythropoietic protoporphyria: a common wild-typeferrochelatase allelic variant with low expression accounts for clinicalmanifestation. Blood 93:2105±2110, 1999

Heaney ML, Golde DW: Myelodysplasia. New Eng J Med 340:1649±1660, 1999Heim S: Cytogenetic ®ndings in primary and secondary MDS. Leukaemia Res 16:43±

46, 1992Ibbotson SH, Elder GH, Lawrence CM: Late-onset congenital erythropoietic

porphyria (Gunther's disease). Br J Dermatol 139:38, 1998Kappas A, Sassa S, Galbraith RA, Nordmann Y: The porphyrias. In: Scriver CR,

Beaudet AL, Sly WS, Valle D, eds. The Metabolic and Molecular Basis of InheritedDisease, 7th edn. New York: McGraw-Hill, 1995:pp 2103±2159

Lange H, Kispal G, Lill R: Mechanism of iron transport to the site of heme synthesisinside yeast mitochondria. J Biol Chem 274:18989±18996, 1999

Lim HW, Cooper D, Sassa S, Dosik H, Buchness MR, Soter NA: Photosensitivity,abnormal porphyrin pro®le, and sideroblastic anemia. J Am Acad Dermatol27:287±292, 1992

Mufti GJ: Chromosomal deletions in the myelodysplastic syndrome. Leukaemia Res16:35±41, 1992

Murphy A, Gibson G, Elder GH, Otridge BA, Murphy GM: Adult-onset congenitalerythropoietic porphyria (GuÈnther's disease) presenting withthrombocytopenia. J R Soc Med 88:357P±358P, 1995

Murphy GM, Hawk JL, Magnus IA: Late-onset erythropoietic protoporphyria withunusual cutaneous features. Arch Dermatol 121:1309±1312, 1985

Rufenacht UB, Gouya L, Schneider-Yin X, et al: Systematic analysis of moleculardefects in the ferrochelatase gene from patients with erythropoieticprotoporphyria. Am J Hum Genet 62:1341±1352, 1998

Todd DJ: Erythropoietic protoporphyria. Br J Dermatol 131:751±766, 1994Wang X, Yang L, Kurtz L, Lichtin A, DeLeo VA, Bloomer J, Poh-Fitzpatrick MB:

Haplotype analysis of families with erythropoietic protoporphyria and novelmutations of the ferrochelatase gene. J Invest Dermatol 113:87±92, 1999

Went LN, Klasen EC: Genetic aspects of erythropoietic protoporphyria. Ann HumGenet 48:105±117, 1984

Whatley SD, Puy H, Morgan RR, et al: Variegate porphyria in western Europe:identi®cation of PPOX gene mutations in 104 families, extent of allelicheterogeneity, and absence of correlation between phenotype and type ofmutation. Am J Hum Genet 65:984±994, 1999

Yamauchi K, Kushibi Y: Pyridoxal 5-phosphate therapy in a patient withmyelodysplastic syndrome and adult onset congenital erythropoieticporphyria. Br J Haematol 81:614±621, 1992

VOL. 117, NO. 6 DECEMBER 2001 LATE-ONSET ERYTHROPOIETIC PORPHYRIA 1649