Neonatal Hemochromatosis - Europe PMC

12 downloads 0 Views 1MB Size Report
Neonatal hemochromatosis (NH) is an entity of unknown cause that is characterized clinically by intrauterine growth retardation and preterm birth, presentation ...
American Journal ofPathology, Vol. 134, No. 2, February 1989 Copyright K American Association ofPathologists

Neonatal Hemochromatosis The Regulation of Transferrin-Receptor and Ferritin Synthesis by Iron in Cultured Fibroblastic-Line Cells

A. S. Knisely,* Joe B. Harford,t Richard D. Klausner,t and Suzanne R. Taylort From the Program in Developmental Pathology, Brown University, and the Department ofPathology and Laboratory Medicine, Women & Infants'Hospital ofRhode Island, Providence, Rhode Island,* the Cell Biology and Metabolism Branch, National Institute of Child Health and Human Development, National Institutes ofHealth, Bethesda, Maryland,t and the Department ofPathology, Children's Hospital ofPittsburgh, Pittsburgh,

Pennsylvania* The authors have investigated the hypothesis that neonatal hemochromatosis (NH), a generallyfatal disease ofinfancy, is due to abnormalities in cellular response to ambient levels of iron. The clinical and necropsy findings in two infants with NH, the results of evaluations for iron-storage disease in theirfirst-degree relatives, and the results of the authors' studies of ferritin and transferrin-receptor (TJR) synthesis in NH and normal fibroblasts are presented. No differences between cultured skin fibroblasts from a normal infant and similar cells from the two infants with NH were seen with respect to TfR andferritin synthesis rates or their modulation by iron. NH and adult idiopathic hemochromatosis (AH) share a pattern of siderosis in which epithelial and mesenchymal elements contain large quantities ofstainable iron, while reticuloendothelial elements contain almost none. Although nofamilial correlation between NH and AH has been established, and none appeared to exist in these two families, the authors' results parallel those ofprevious studies of various cell types from persons with AH. The abnormalities in cellular iron handling, undefined at present, that are associated with the phenotype common to NH and AH do not appear primarily to involve the regulation by iron of rates of TfR and ferritin synthesis. (Am J Pathol 1989,

growth retardation and preterm birth, presentation in hepatic failure at only hours of postnatal age, rapid progression to death, and recurrence in sibships.1' 2 It is characterized histopathologically by diffuse hepatic fibrosis with varying degrees of nodular regeneration,2 and by an abundance of stainable iron in the liver, many epithelia, and cardiac myocytes, with marked sparing of the spleen, lymph nodes, and bone marrow.' 2 This combination of liver disease with a distinctive pattern of siderosis, also typical of adult idiopathic hemochromatosis (AH),3 has given NH its name.4 Marked hyperferritinemia25-7 has been identified in cases of NH, but no investigations of cellular iron handling in NH have been reported. In normal cells, levels of iron in cytoplasmic "regulatory pool(s)" influence the rates at which ferritin and the transferrin receptor (TfR) are synthesized.8 These rates can be influenced experimentally by supplementing growth media with sources of iron or agents that chelate iron. Although connective tissue is not a major site of iron deposition in NH, we speculated that studies of iron handling in cultured skin fibroblasts ("fibroblasts") from patients with NH might yield information on the nature of any metabolic abnormality underlying the phenotypic disorder. To evaluate the possibility that abnormalities in cellular response to iron play a role in NH, we investigated the synthesis of ferritin and TfR by fibroblasts from two infants with NH and by fibroblasts from a normal infant. We also manipulated the levels of iron available to these cells and examined the resulting changes in ferritin and TfR synthesis rates. This report presents the clinical and necropsy findings in the two infants with NH, the results of evaluations for iron-storage disease in their firstdegree relatives, and the results of our studies of ferritin and TfR synthesis in NH and normal fibroblasts.

Clinical Histories Patient 1

134:439-445)

A male white infant was born at term to a 25-year-old. The pregnancy had been uncomplicated. Both parents, of

Neonatal hemochromatosis (NH) is an entity of unknown cause that is characterized clinically by intrauterine

Accepted for publication October 20, 1988. Address reprnt requests to A. S. Knisely, MD, Department of Pathology, Children's Hospital of Philadelphia, 34th Street and Civic Center Boulevard, Philadelphia, PA 19104.

439

440

Knisely et al

AJP February 1989, Vol. 134, No. 2

Table 1 Patient 1, Parents, and Brother: PartialHLA Haplotypes and Serum Concentrations ofIron and Iron-Binding Proteins Iron, Iron-binding capacity, mmol/I mmol/I HLA haplotype

Al Bw57/A2 B51 Patient 1 Al Bw57/A2 B51 Brother Al Bw57/A2 B44; Cw5, Cw6; DR2, DR7, DQwl, DQwx Mother A2 B35/A2 B51; Cw4; DR2, DR5, DQwl, DQw3 Father * Abnormal value.

13 10 15 11

39 57 44 37

Ferritin, mg/I 2000* 19 50 116

French-Canadian origin, were well. Consanguinity was denied. Their firstborn child, a son, had required treatment for transient neonatal hypoglycemia (1.39 mmol/l [normal, 3.9-6.1]) and idiopathic neonatal hyperbilirubinemia prolonged over several weeks (peak: total, 308,umol/l [normal, 2-18]), but was well at 46 months of age, without evident sequelae. Birth weight of Patient 1 was 2660 g (expected,9 3480 ± 460 g); Apgar scores at 1 and 5 minutes were 8 and 9 respectively. The infant was not dysmorphic, and the placenta was not examined. At 24 hours of age lethargy, irritability, and hypothermia (36 C) were noted. The leukocyte count was 17.6 X 109/l, with 74% neutrophilic-line cells. The platelet count was 95 x 109/l. Acanthocytes were present; the hematocrit was 0.64. There was no serologic evidence for maternal-fetal blood group incompatibility or intrauterine infection. Initial bacterial cultures of blood and cerebrospinal fluid (CSF) and viral cultures of urine, CSF, and ascitic fluid yielded no growth. Laboratory studies documented hypocoagulability (prothrombin time > 40 seconds [normal, 10-13]; partial thromboplastin time > 100 seconds [normal, 24-33]), hypoalbuminemia (25 g/l [normal, 35-50]), elevated serum concentrations of alkaline phosphatase (12.00 ,gkat/l [normal, 0.331.67]) and aspartate and alanine aminotransferase (aspartate, 7.63 Mkat/l [normal, 0.083-0.42]; alanine, 1.48 ,ukat/l [normal, 0-0.40]) activity, hyperammonemia (86 ,gmol/l [normal, 38-79]), and hyperbilirubinemia (total bilirubin, 468 Mmol/l [normal, 2-18]; conjugated, 227 MAmol/l [normal, 0-4]); bile was present in the feces. Amino acid chromatograms of serum and urine showed nonspecific changes of liver disease. Serum concentrations of iron, iron-binding capacity, and ferritin were determined on the sixth postnatal day (Table 1). Liver biopsy was contemplated, but was deferred when splenomegaly, a caput medusae, and ascites appeared and hypocoagulability could not be corrected. The infant developed Escherichia coli peritonitis and septicemia, and died on the 20th postnatal day.

were well. Consanguinity was denied. Their firstborn child, a son, had been born at term and died in hepatic failure at 38 days postnatal age. No etiology of liver disease was established antemortem; necropsy findings were consistent with NH. The pregnancy for Patient 2 was complicated by decreased fetal movement from 32 weeks onward. Cesarean section was performed for late decelerations. The fetal membranes were meconium-stained; there was moderate placental villous edema without siderosis. Birth weight was 1955 g (expected,9 2345 ± 315 g), and Apgar scores at 1 and 5 minutes were 2 and 7 respectively. The infant was not dysmorphic, but slight hypotonia was apparent and ventilatory assistance was required in the delivery room. The leukocyte count was 10.9 X 109/l, with 47% neutrophilic-line cells. The hematocrit was 0.50. There was no serologic evidence for maternal-fetal blood group incompatibility or intrauterine infection. Initial bacterial cultures of blood yielded no growth. Laboratory studies documented hypocoagulability (prothrombin time > 40 seconds; partial thromboplastin time > 100 seconds), hypoalbuminemia (21 g/l), elevated serum concentrations of alkaline phosphatase (17.35 ,ukat/l) and aspartate and alanine aminotransferase (aspartate, 2.60 ,ukat/l; alanine, 0.65 ,kat/l) activity, hyperammonemia (179 ,umol/l), and hyperbilirubinemia (total bilirubin, 337 ,umol/l; conjugated, 113 Amol/l); bile was present in the feces. Amino acid chromatograms of serum and urine showed nonspecific changes of liver disease. On the tenth postnatal day serum concentrations of iron, ironbinding capacity, and ferritin were determined (Table 2). Splenomegaly, a caput medusae, and ascites developed. Findings on magnetic resonance imaging (MRI) of the abdominal viscera on the 17th postnatal day were consistent with siderosis of the liver and pancreas. Siderosis of the spleen was not observed. Hepatic transplantation was considered, but no donor was found. The infant developed Enterococcus sp. septicemia, and died on the 72nd postnatal day.

Patient 2

Necropsy Findings and Family Studies Patient 1

A female white infant was born at 35 weeks gestational age to a 26-year-old. Both parents, of Portuguese origin,

The body was icteric. Ascites, splenomegaly (27 g; expected weight,'° 11.2 ± 4.1 g), and widespread evidence

Iron as Regulator in Neonatal Hemochromatosis

441

AJP February 1989, Vol. 134, No. 2

Table 2. Patient 2, Parents, and Brother: PartialHLA Haplotypes and Serum Concentrations ofIron and Iron-Binding Proteins HLA haplotype

mmol/I

Iron-binding capacity, mmol/I

A2 B7/A3 B44 Al B27/Al B35 A3 B44/A1 B27; Cw2; DR2, DR4, DQwl, DQw3 A2 B7/Al B35; Cw4, Cw7; DR3, DRw8, DQw2

11 ND 7

24 ND

Iron, Patient 2 Brother Mother Father

16

48 40

Ferritin, mg/l 769* ND 55 144

ND, not determined. * Abnormal value.

of a hemorrhagic diathesis were evident. The liver was small (67.5 g; expected,10 149 ± 35 g), dark green-brown, and very firm, without apparent nodularity. Microscopy showed extensive pseudocholangiolar transformation of hepatocytes and scaftered foci of giant-cell change, with marked diffuse fibrosis that was at least as prominent in central as in portal regions (Figure 1). Nodular regeneration was not present. Subcapsular hepatic vessels were ectatic, the lumen of the umbilical vein was recanalized, and there were esophageal varices. Pancreatic islets appeared increased in size. Stainable iron was conspicuous in hepatocytes and Kupffer cells, many epithelia, cardiac myocytes, and osteoclasts. There was no stainable iron elsewhere in the bone marrow, in lymph nodes, or in the spleen. Although gastric chief cells and renal proximal tubular epithelium were siderotic, cell blocks prepared from gastric contents and bladder urine did not include cells containing stainable iron. Other findings included nephromegaly (combined weight, 66.9 g; expected,10 28.4 ± 7.6 g) and diffuse Alzheimer type 11 astrogliosis. Fibroblasts cultured from anterior thoracic skin (#GM 09564; National Institute of General Medical Sciences Human Genetic Mutant Cell Repository [NIGMS Repository], Camden, NJ) showed a 46,XY karyotype. Human leukocyte antigen (HLA) typing was performed on fibroblasts from the infant (Organogenesis, Inc., Cambridge, MA) and on peripheral-blood lymphocytes from the brother Figure 1. The liver from Patient 1 shows difuse fibrosis, more prominent in central (arrow) than portal (arrowhead) regions. Pseudocholangiolar transformation and bile plugging are apparent. Hematoxylin and eosin, X56

(Department of Hematology, School of Medicine, University of Utah; Salt Lake City, UT) and parents (Rhode Island Blood Center; Providence, RI). Two months after the infant's death, serum concentrations of iron and iron-binding proteins were determined for the brother and parents. These results are shown in Table 1. Hematocrits and erythrocytic indices for the brother and parents were within normal ranges, as were concentrations of transferrin in the parents' sera. Aspartate aminotransferase activity in serum from the brother was elevated (0.97 ,kat/1).

Patient 2 The body was icteric. Ascites, splenomegaly (42 g; expected weight,11 13 g), and widespread evidence of a hemorrhagic diathesis were present. The liver was small, dark green, and firm, and contained tan and dark green nodules up to 0.5 cm in diameter. Microscopy showed the nodules to consist of densely bile-stained regenerating hepatocytes without atypia. In the internodular regions, there was extensive pseudocholangiolar transformation of hepatocytes, with focal bile plugging. As in Patient 1, diffuse fibrosis was present, and was at least as prominent in central as in portal regions. The lumina of some central venules were obliterated, and phlebosclerosis also was seen in the juxtahepatic inferior vena cava. Pan-

442

Knisely et al

AJP February 1989, Vol. 134, No. 2

line CCd-325k; American Type Culture Collection, Rockville, MD) were maintained at 37 C under 5% CO2 in Dulbecco's modified Eagle medium supplemented with 10% fetal bovine serum and containing 100 U/ml penicillin and 0.1 mg/ml streptomycin (growth medium). Experiments were performed before the tenth passage of each cell line.

Figure 2. Knee synovium (arrow) from Patient 2 is siderotic. Prussian blue and nuclearfast red (Gomori iron stain), X 180.

creatic islets appeared increased in size. Stainable iron was conspicuous in hepatocytes and Kupffer cells, many epithelia, laryngeal chondrocytes, and cardiac myocytes. Knee synovium was siderotic (Figure 2). There was no stainable iron in the bone marrow, lymph nodes, or spleen. Other findings included villous transformation of the gastric mucosa, nephromegaly (combined weight, 97 g; expected,11 37 g), and multicystic encephalomalacia. Nephromegaly, similar hepatic abnormalities, and an identical pattern of siderosis were found when necropsy material from the brother was reviewed. Fibroblasts cultured from anterior thoracic skin (#GM 09811; NIGMS Repository) showed a 46,XX karyotype. Fibroblasts from the brother (#GM 09894; NIGMS Repository) showed a 46,XY karyotype. HLA typing was performed on fibroblasts from both these lines (Organogenesis, Inc.) and on peripheral-blood lymphocytes from the parents (Rhode Island Blood Center). Three months after the patient's death, serum concentrations of iron and ironbinding proteins were determined for the parents. These results are shown in Table 2. Parental hematocrits, erythrocytic indices, and serum transferrin concentrations were within normal ranges.

Studies of Cultured Fibroblasts Materials and Methods Cell Lines and Culture Conditions Fibroblasts from both index cases (#GM 09564 and #GM 09811) and from skin of a healthy male neonate (cell

Radiolabeling, Immunoprecipitation, Ge/ Electrophoresis, and Autoradiography Cells were radiolabeled with 100 uCi/ml of [IS] Translabel (ICN Radiochemicals, Irvine, CA) for 3 hours in methionine-free growth medium. Cell lysates were prepared, TfR was immunoprecipated with B3/25 monoclonal antibody against human transferrin receptor (BoehringerMannheim, Indianapolis, IN) and Protein A-agarose (Bethesda Research Laboratories, Bethesda, MD), and samples were analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), autoradiography, and scanning densitometry with computer-aided peak integration, as described previously.12 Analyses for ferritin were performed similarly using a rabbit antibody against human ferritin (Boehringer-Mannheim). When both TfR and ferritin were immunoprecipitated from the same sample of a cell lysate (Figure 3A, lanes 1 and 2), the respective antibodies were mixed with separate aliquots of Protein A-agarose. The resins were then washed with lysis buffer and added to the cell lysate. Iron Supplementation and Chelation Levels of available iron were manipulated before radiolabeling by treatment for 16 hours with either hemin (ferric protoporphyrin IX [Sigma Chemical Co., St. Louis, MO]), a source of supplemental iron, or deferoxamine (CibaGeigy, Summit, NJ), an iron chelator, in 50,uM concentrations.

Results

Qualitative Characterization of TfR and Ferritin The TfR and ferritin synthesized by fibroblasts cultured from two patients with NH were qualitatively indistinguishable from those synthesized by fibroblasts from a normal neonate. On SDS-PAGE analysis, the mobilities of both ferritin and TfR polypeptides were the same for all three cell lines (Figure 3A).

Quantitation of Rates of TfR and Ferritin Synthesis When NH fibroblasts and normal fibroblasts were maintained in the same growth medium, rates of synthesis for

Iron as Regulator in Neonatal Hemochromatosis

443

AJP February 1989, Vol. 134, No. 2

both TfR and ferritin were similar for both cell types (Figure 3A, lanes 1 and 2). When hemin was added to the growth medium, the rate of TfR synthesis fell and the rate of ferritin synthesis rose for both cell types (Figure 3A, lanes 3 and 5). When deferoxamine was added to the growth medium, the rate of TfR synthesis rose and the rate of ferritin synthesis fell for both cell types (Figure 3A, lanes 4 and 6). Ferritin synthesis by NH fibroblasts was assessed as 65 times greater following hemin treatment than following deferoxamine treatment (Figure 3B, hatched bars). TfR synthesis by NH fibroblasts was assessed as 24 times less following hemin treatment than following deferoxamine treatment (Figure 3C, hatched bars). When normal fibroblasts were subjected to identical manipulations of iron availability, the rates at which they synthesized ferritin varied between extremes by a factor of 62 (Figure 3B, open bars), and the rates at which they synthesized TfR varied between extremes by a factor of 23 (Figure 3C, open bars). These responses were similar to those of NH fibroblasts. When the concentration of hemin added as an iron source was varied over a range from 2 to 50 gM, no significant differences between the two cell types were observed in either ferritin or TfR synthesis rates (data not shown).

0 ._

D 40

1F

0

mc

.-

*_

Discussion ._a

Clinical Courses and Necropsy Findings c 40

m

D

C

H

Fe Treatment Figure 3A. Composite of autoradiographs of immunoprecipitated radiolabeled TfRandferritin (Fer) after SDS-PAGE analysis. Lanes 1, 3, and 4 contain materialfrom lysates of normal fibroblasts; lanes 2, 5, and 6 contain materialfrom lysates of fibroblastsfrom Patient 1. Autoradiographs (not shown) of materialfrom Patient2 were indistinguishablefrom those of identically treated materialfrom Patient 1. Cells were either untreated (lanes 1 and 2), pretreated with 50uM hemin (lanes 3 and 5), orpretreated with 50juM deferoxamine (lanes 4 and 6). Lanes 1 and 2 represent experiments in which TfR andferritin were simultaneously immunoprecipitatedfrom single aliquots of cell lysates (see text). Lanes 3-6 represent experiments in which TfR and Fer were independently immunoprecipitatedfrom separate aliquots of cell lysates. Synthesis of TfR and ferritin by NHfibroblasts does not differ from synthesis of TfR andferritin by normalfibroblasts. B: Quantitation offerritin synthesis by fibroblasts from a normal infant (open bars) andfrom Patient 1 (hatched bars), with and without manipulation oflevels ofavailable iron. Cells treated with either deferoxamine (D) or hemin (H) are compared with untreated (C) cells. Ferritin regions in gels like those shown in panel A were quantitated by scanning densitometry with computer-aided integration. Relative levels offerritin synthesis are expressed in arbitrary units ofpeak area. No differences between cell types

The two patients reported here meet the clinical and histopathologic criteria for NH. The findings of villous edema without siderosis in the placenta of Patient 2 are characteristic.1'2 Synovial siderosis and villous transformation of gastric mucosa in NH have not previously been described. Nephromegaly was present in both patients and, by history, in the brother of Patient 2; its etiology and significance are not clear. Of clinical note is that in Patient 1, cytologic preparations of gastric contents and bladder urine obtained at necropsy did not contain siderotic epithelial cells, and that in Patient 2, MRI findings supported the diagnosis of NH. The expected utility of urinary cytology in the antemortem diagnosis of NH has been discussed.1 MRI studies are known to permit diagnosis of AH.13 They may also be of value when NH is suspected. are seen in responses to alterations in available iron. C: Quantitation of Tfl? synthesis byfibroblastsfrom a normal infant (open bars) andfrom Patient 1 (hatched bars), with and without manipulation of levels of available iron. Designations D, H, and C correspond to those in panel B. Gels were analyzed as describedfor panel B, but Tfl? rather thanferritin regions were quantitated. Again, no differences between cell types are seen in responses to alterations in available iron.

444

Knisely et al

AJP February 1989, Vol. 134, No. 2

Iron, Iron-Binding Proteins, and Family Studies Although a serum concentration of aminotransferase activity was elevated, liver disease was not clinically apparent in the brother of Patient 1. NH is characterized by recurrence in sibships, as seen in Patient 2 and her brother. Survival without clinically evident sequelae has been reported in one instance of biopsy-documented NH.7 A parallel case without biopsy confirmation of the diagnosis also has been reported.14 It can be speculated that transient neonatal hepatic dysfunction in the brother of Patient 1 was related to the process that, at its most severe, leads to fatal NH. Measurements of serum concentrations of iron and iron-binding proteins (Tables 1 and 2) as well as hematologic studies in the surviving first-degree relatives of Patients 1 and 2 found no abnormalities. In NH this is the rule2'4'6'7'15-17 rather than the exception.15'14'18 The high concentrations of ferritin in the serum of Patients 1 and 2 would be consistent, in adults, with either liver disease19 or iron overload:20 both were present in these patients. The significance of the relatively low saturation of ironbinding capacity in both patients, in determinations performed after repeated transfusions of blood products, is not clear. HLA types associated with AH have occurred in the parents of patients with NH2'5'7'17 and in the patients themselves,27 but determinations of HLA types in Patients 1 and 2 and in their first-degree relatives (Tables 1 and 2) did not provide evidence that the strong association with HLA A3, B7, and B14 present in AH20 21 is seen in NH.

Iron Regulation of TfR and Ferritin Synthesis: General Conclusions Cells take up iron through receptor-mediated endocytosis of transferrin, a serum protein, and sequester iron within cytoplasmic ferritin.8 The synthesis of both TfR and ferritin is modulated by iron. The availability of iron can be experimentally manipulated in cultured cells using hemin as an iron source or deferoxamine as an iron chelator. These two agents appear to exert opposing influences on the iron content of a common regulatory pool.12 In all cell types examined to date, ferritin is synthesized at a relatively higher rate and TfR is synthesized at a relatively lower rate when iron is abundant. Under similar growth conditions, normal and NH fibroblasts synthesized ferritin and TfR at similar rates. When available iron was augmented or reduced, the responses of NH fibroblasts were similar to those of normal fibroblasts: inicreases or decreases in rates of ferritin and TfR synthesis by the two cell types were quantitatively alike. These findings support the conclusion that the pro-

cess by which iron regulates cellular expression of ferritin and TfR is not altered in NH fibroblasts. In addition, no significant differences in rates of ferritin or TfR synthesis by the two cell types were observed when varying concentrations of hemin were added to growth medium. This permits the inference that the intracellular process by which iron is liberated from hemin to enter the regulatory iron pool is unaffected in NH fibroblasts. Because both normal and NH fibroblasts could be shown to respond to changes in their iron status, the similarities in their levels of TfR and ferritin synthesis under normal growth conditions permit the inference that both types of cells acquire similar amounts of iron from the growth medium. In this connection, studies of fibroblasts from another infant with NH5 have shown that values for both transferrin binding and iron uptake could not be distinguished from those seen in normal fibroblasts (Kaplan J, personal communication,

1988). These findings are consistent with those of studies of TfR regulation and function in cultured fibroblasts and lymphocytes from individuals with AH, which identified no abnormalities.22 However, they do not completely exclude the possibility that the primary disorder of iron handling in NH involves the regulation of ferritin synthesis, TfR synthesis, or both by ambient concentrations of iron. If such a disorder were not expressed by fibroblasts, but only by hepatocytes or trophoblast, excess accumulation of iron in the liver or excess maternofetal transport of iron might result.' It should be noted that in individuals with AH, hepatocytes show normal regulation of TfR by iron in viVo,23'24 and that TfR receptor numbers in duodenal mucosa, the postnatal analogue to trophoblast in terms of iron uptake, are not increased AH.2425 Postulating a hepatocyte- or trophoblast-specific disorder in NH also leaves open the question of why reticuloendothelial elements fail to accumulate stainable iron. It is intriguing that the pattern of parenchymal iron deposition with reticuloendothelial sparing that characterizes both AH and NH is duplicated in congenital26 and acquired27 hypotransferrinemia. Mice of a newly identified hypotransferrinemic mutant strain also gradually develop the "hemochromatotic" phenotype.28 It has been speculated that in these mice, with the effective ablation of transferrin-mediated iron transport, the tissue distribution of iron reflects the relative degrees of activity of a nontransferrin iron transport system.289 Candidates for the role of iron carrier in such a system have been identified.20 Dysfunction of this postulated system may be common to hypotransferrinemia, AH, and NH.

References 1. Knisely AS, Magid MS, Dische MR, Cutz E: Neonatal hemochromatosis. Birth Defects 1987, 23(1):75-102

Iron as Regulator in Neonatal Hemochromatosis

445

AJP February 1989, Vol. 134, No. 2

2. Silver MM, Beverley DW, Valberg LS, Cutz E, Phillips MJ, Shaheed WA: Perinatal hemochromatosis: Clinical, morphologic, and quantitative iron studies. Am J Pathol 1987, 128: 538-554 3. Sheldon J: Haemochromatosis. London, Oxford University Press, 1935 4. Cottier H: Uber ein der Hamochromatose vergleichbares Krankheitsbild bei Neugeborenen. Schweiz Med Wochenschr 1957, 37:39-43 5. Glista BA, Bautista A, Prudencio M, Desposito F: Neonatal iron storage disease (Abstr). Pediatr Res 1986, 20:410A 6. Jonas MM, Kaweblum YA, Fojaco R: Neonatal hemochromatosis: Failure of deferoxamine therapy. J Pediatr Gastroenterol Nutr 1987, 6:984-988 7. Colletti RB, Clemmons JJW: Familial neonatal hemochromatosis with survival. J Pediatr Gastroenterol Nutr 1988, 7: 39-45 8. Seligman PA, Klausner RD, Heubers HA: Molecular mechanisms of iron metabolism, The Molecular Basis of Blood Diseases. Edited by G Stamatoyannopoulos, AW Nienhaus, P Leder, PW Majerus. Philadelphia, WB Saunders, 1987, pp 219-244 9. Usher R, McLean F: Intrauterine growth of live-born Caucasian infants at sea level: Standards obtained from measurements in 7 dimensions of infants born between 25 and 44 weeks of gestation. J Pediatr 1969, 74:901 -910 10. Gruenwald P, Minh HN: Evaluation of body and organ weights in perinatal pathology: I. Normal standards derived from autopsies. Am J Clin Pathol 1960, 34:247-253 11. Coppoletta JM, Wolbach SB: Body length and organ weights of infants and children. Am J Pathol 1933, 9:55-70 12. Rouault T, Rao K, Harford J, Mattia E, Klausner RD. Hemin, chelatable iron, and the regulation of transferrin receptor. J Biol Chem 1985,260:14862-14866 13. Stark DD, Moss AA, Goldberg HI: Nuclear magnetic resonance of the liver, spleen, and pancreas. Cardiovasc Intervent Radiol 1986, 8:329-341 14. Jacknow G, Johnson D, Freese D, Smith C, Burke B: Idiopathic neonatal iron storage disease (Abstr). Lab Invest 1983, 48:7P 15. Fienberg R: Perinatal idiopathic hemochromatosis: Giant cell hepatitis interpreted as an inborn error of metabolism. Am J Clin Pathol 1960,33:480-491 16. Goldfischer S, Grotsky HW, Chang C-H, Berman EL, Richert RR, Karmarkar SD, Roskamp JO, Morecki R: Idiopathic neonatal iron storage involving the liver, pancreas, heart, and endocrine and exocrine glands. Hepatology 1981,1:58-64 17. Drut RM, Itarte H, Drut R: Hematocromatosis neonatal idiopatica. Arch Arg Pediatr 1987, 85:31-35 18. Laurendeau T, Hill JE, Manning GB: Idiopathic neonatal hemochromatosis in siblings: An inborn error of metabolism. Arch Pathol 1961, 72:410-423 19. Prieto J, Barry M, Sherlock S: Serum ferritin in patients with iron overload and with acute and chronic liver diseases. Gastroenterology 1975, 68:525-533. 20. Cartwright GE, Edwards CQ, Kravitz K, Skolnick M, Amos

DB, Johnson A, Buskjaer L: Hereditary hemochromatosis: Phenotypic expression of the disease. N Engi J Med 1979,

301:175-179 21. Simon M, Alexandre J-L, Rauchet R, Genetet B, Bourel M: The genetics of hemochromatosis, Genetics of Gastrointestinal Disease. Prog Med Genet (new series). Vol. 4. Edited by AG Steinberg, AG Bearn, AG Motulsky, B Childs. Philadelphia, WB Saunders, 1980, pp 135-168 22. Ward JH, Kushner JP, Ray FA, Kaplan J: Transferrin receptor function in hereditary hemochromatosis. J Lab Clin Med

1984,103:246-254 23. Sciot R, Paterson AC, Van Den Oord JJ, Desmet VJ: Lack of hepatic transferrin receptor function in hemochromatosis. Hepatology 1987, 7:831-837 24. Lombard M, Bomford A, Polson R, Williams R: Differential expression of the transferrin receptor in liver and gut in hereditary hemochromatosis (HH) (Abstr). Gut 1988, 29:A725 25. Banerjee D, Flanagan PR, Cluett J, Valberg LS: Transferrin receptors in the human gastrointestinal tract: Relationship to body iron stores. Gastroenterology 1986, 91:861-869 26. Heilmeyer L, Keller W, Vivell 0, Keiderling W, Betke K, Wohler F, Schultze HE: Kongenitale Atransferrinamie bei einem sieben Jahre alten Kind. Dtsch Med Wochenschr 1961, 86:1745-1751,1755 27. Westerhausen M, Meuret G: Transferrin-immune complex disease. Acta Haematol 1977, 57:96-101 28. Craven CM, Alexander J, Eldridge M, Kushner JP, Bernstein S, Kaplan J: Tissue distribution and clearance kinetics of non-transferrin-bound iron in the hypotransferrinemic mouse: A rodent model for hemochromatosis. Proc Natl Acad Sci USA 1987, 84:3457-3461 29. Bernstein SE: Hereditary hypotransferrinemia with hemosiderosis, a murine disorder resembling human atransferrinemia. J Lab Clin Med 1987,110:690-705 30. Jones RL, Grady RW, Sorette MP, Cerami A: Host-associated iron transfer factor in normal humans and patients with transfusion siderosis. J Lab Clin Med 1986,107:431-438

Acknowledgment The authors thank G. F. Vawter, MD, for permission to report portions of the clinical and necropsy findings in the brother of Patient 2; E. Mroczek, MD, for performing the necropsy examination of Patient 2; F. Kaplan, MS, MT (ASCP) SBB, for HLA studies of cultured fibroblasts; J. Kaplan, PhD, for permission to cite unpublished results, and for access to HLA studies of peripheral blood lymphocytes from the brother of Patient 1; and M. Kimball, MS, MT (ASCP) SBB, for HLA studies of parental peripheral blood

lymphocytes.

Note Added in Proof After this manuscript was submitted, a preliminary description of Patient 2 and her family was published by other investigators (Driscoll SG, Hayes AM, Levy HC: Neonatal hemochromatosis: Evidence for autosomal recessive transmission [abstr]. Am J Hum Genet 1988, 43:A232).