British Journal of Haemarology, 1978, 40, 265-276
Iron Deficiency Anaemia in Newborn sla Mice: a Genetic Defect of Placental Iron Transport P. J. KINGSTON, CATHERINE E. M. UANNERMAN A N D R. M. BANNERMAN
Medical Genetics U n i t , Department ofhledicine, State University of New York at Bujfaalo, and Buffalo General Hospital, Buffalo, N . Y . , U.S.A. (Received z December 1977; acceptedfor publication 20 Fcbruary 1978) SUMMARY.Newborn mice with X-linked anaemia (gene symbol sla) have lower haemoglobin levels a t birth than normal and carrier mice but there is considerable overlap. Serial observations showed that the haemoglobin values of segregating male mice separate into a bimodal distribution by 42 d of age, and 50 d values were used to assign genotypes retrospectively. The anaemia in newborn sla mice is attributable to iron deficiency, since their total body iron is lower than in normal newborn mice, while their birth weights are almost identical. Haemoglobin levels a t birth in normal, anaemic and carrier mice are also influenced by the mother’s genotype and phenotype, and the haemoglobin value was progressively lower according to the sla gene dose of the mother. Materno-fetal iron transfer was examined by labelling pregnant carrier females with radioiron in various ways. When given as single or intermittent doses by injection no clearcut differences emerged in apparent iron transfer to anaemic as compared to non-anaemic fetuses. However, when radioiron was administered continuously in food a significant reduction in iron transfer to anaemic fetuses was demonstrated. The sla gene is already known to have a major effect in reducing iron transport in the small intestine. The present studies provide evidence of an analogous defect in placental iron transport. X-linked anaemia in the mouse (gene symbol sla) is a recessively inherited trait (Falconer & Isaacson, 1962) associated with marked hypochromic anaemia (Grewal, I 962; Bannerman & Pinkerton, 1967).In adult mice anaemia is secondary to a specific dcfect in intestinal absorption of iron (Pinkerton & Bannerman, 1967), associated with the accumulation of iron in the duodenal mucosal cells (Pinkerton, 1968). Newborn sla mice were observed to be anaemic a t birth (Falconer & Issacson, 1962; Grewal, 1962). This suggested to Dancis & Jansen (1970) that there might be an analogous defect in placental iron transport, but they were unable to demonstrate such a defect in the tranxfer of radioiron in acute experiments. In this laboratory, preliminary observations on newborn male mice from male segregating litters showed a bimodal distribution of haemoglobin level but Correspondence: Dr P. J. Kingston, Department of Haematology, St Thomas’ Hospital, London, S.E.I. 0007- 1o48/78/1o00-0265$02.00
197X Blackwell Scientific Publications
265
P.J. Kingston et a1
266
there was considerable overlap (Uannerman et al, 1973). The finding that the correct genotypes cannot be reliably determined a t birth complicates the investigation of a hypothetical placental lesion. We therefore carried out serial studies on sla anaemic mice and Littermate controls from birth to over 50 d of age, the results of which allow the retrospective assignment of genotypes a t birth. Using these data, total iron content of newborn normal and anaemic mice was determined. Finally, experiments with radioiron repeatedly administered to pregnant sla carrier females, and subsequent observation of their segregating offspring, provide direct evidence of a defect in placental iron transfer in sla anaemic mice. MATERIALS AND METHODS
Animals. The sla gene is maintained in this laboratory in mice of the C57BL/6J strain into which it was transferred from the original mixed stock (Bannerman & Pinkerton, 1967),and animals of the ninth to eleventh backcross generations were used in these experiments. The following conventions have been used for recording genotypes: ‘sla’ represents the gene for anaemia carried on the X-chromosome; represents the normal wild-type allele; and ‘Y’ indicates the absence of either allele on the Y chromosome. Thus ‘+I+’ indicates a normal female, ‘sla/+’ a heterozygous carrier female, ‘rla/sla’ a homozygous anaemic female, ‘ +/Y’ a normal male, and ‘sla/Y’ a hemizygous anaemic male. The usual mating was between a carrier female (sla/+) and a normal (+/Y) or affected ( s l a / Y ) male, to obtain equal numbers of affected and normal male littermates for comparison. Mice were kept in plastic cages with chromium-plated covers. As in previous studies, they were maintained on Rockland Mouse/Rat diet (complete) containing approximately 0.02 I YO Fe, except for breeding pairs which received Rockland Mouse Breeder Diet (0.023% Fc); the pups had access to this diet from about the loth day of life until the time of weaning. O n the 21st day weanlings were segregated by sex and fed Mouse/Rat diet (complete). Tapwater was given ad libitum. Body weights were recorded to the nearest 0.01g a t birth and regularly thereafter to the nearest 0.1g. Haematol9gy. Blood was taken from the retro-orbital venous sinus, or by puncture of the jugular vein in newborn or fetal mice, and collected into heparinized capillary tubes. Haemoglobin was determined as cyanmethaemoglobin using 10 pl blood diluted in 2.5 ml Drabkin’s solution and read a t 540 mm in a Coleman Junior spectrophotometer. Chemical determination of iron. The total body iron content of mice was determined by a wet ashing method. The weighed carcass was dissolved with heating in a mixture of equal volumes of concentrated sulphuric and nitric acids. For newborn mice 5 ml of acid mixture was sufficient; appropriate volumes were used for larger carcasses. Hydrogen peroxide was added with reheating to clear the solution, followed by sodium metabisulphite. After cooling, dilution and neutralization, iron content was determined colorimetrically using orthophenanthroline (Pinkerton et al, 1970). Placental transfer experiments. All observations were carried out on pregnancies or offspring of sla/ females mated to da/Y males to produce ‘male/female’ segregating litters. Gestational age was determined by examination for copulatory plugs approximately 12 h after the male was introduced into the breeding cage; spontaneous delivery usually occurred 20 d after
‘+’
+
Iron Dejciency Anaemia in Newborn sla Mice
267
observation of the plug. A small number of litters were delivered by caesarian section after administration of a single dose of radioiron. Tracer radioiron as [59Fe]ferrouscitrate (approximately 25 pCi/pg Fe) was administered to the pregnant female mice by one of the following schedules: (a) 3-5 pCi administered as a single intravenous or intraperitoneal injection on the 18th or 19th day ofgestation; (b) 3-4 pCi administered by repeated intraperitoneal injections of 0.4 pCi given over a 5 d period during pregnancy; (c) continuous feeding of the pregnant female on an artificial milk diet (McCall et al, 1962) with ferrous citrate added to a final concentration of 266 pg elemental iron and 0.6 pCi per gram of diet. Newborn mice were separated from their mothers immediately after delivery, and before feeding. They were held in a special incubator before counting in prewarmed tubes in an automatic gamma scintillation well counter (Nuclear-Chicago Model 1085). Duplicate I min counts were made and these were within I o/‘ in every case. After counting, the haemoglobin level was determined, and then the pups were returned to their mother to be raised in the usual fashion for determination of genotype a t 50 d of age.
RESULTS Newborn male mice from ‘male-segregating’ litters were studied serially from birth to after 50 d of age. Body weights and haemoglobin lcvels were determined within 24 h of birth (usually within 12 h), and at 21,42 and 50 d. Abdominal haematomata were noted to be responsiblc for neonatal anaemia in a small number of mice. This is a little-known cause of anaemia, but it is important to be aware of it, and these mice were excluded from subsequent analysis, though they were followed and their genotypes could ultimately be determined. A total of 283 males were available for study. Haemoglobin levels in newborn mice ranged from 5 to 15 g/dl; although the pattern of distribution is bimodal there is no clear separation. A t 21 d the overlap is even greater. A t 42 and 50 d of age segregation is apparent, and two groups of equal numbers emerged, so that genotypes could then be assigned unequivocally. Their validity was confirmed by previous studies with test matings (Bannerman & Pinkerton, 1967), reconfirmed by further test matings from representative mice in the present investigation. Retrospective analysis of the accumulated data after assignation of genotypes show the overlap in values which obscures an underlying difference at birth, as illustrated in Fig i ( a ) . Body weights are almost identical in newborn normal and anaemic males, but a difference begins to emerge at 21 d of age and anaemics are visibly smaller after the age of 50 d (Fig Ib).
Body Iron at Birth Haemoglobin values were determined on fiz newborn male litter mates from ‘male-segregating’ litters which were then killed by ether anaesthesia, and the iron content of the carcasses determined as described above. An appropriate correction was added for the blood withdrawn to determine the haemoglobin level. There is a direct correlation between haemoglobin level and total iron (Fig 2a) and a more marked correlation between haemoglobin level and iron concentration expressed as pg/g body weight (Fig 2b). These results are derived from a mixed group, comprising mice of two genotypes. For
P. J . Kingston et
26s
111
FIG I . Haemoglobin levels (a) and body weights (b) in a group of sla anaemic (shaded) and normal male littermates (unshaded) at birth, 2 1 and 50 d. Genotypes have been assigned retrospectively from the segregation observed at 50 d.
Newborn
d Mice ( s l a / Y and + / Y l
8or
..
N e w b o r n d M i c e ( s / a / Y and t / Y )
om
r.048 ( P (0011
r = 0.57 ( p (0.01)
-1-1-1-1- I 3 Za
I < 6
9
Haemoglobin y / d l
12
15
L4-L 3
18 Zb
6
1-1-1-1 9
I2
15
Haemoglo bi n y /dl
FK 2 . (a) Total body iron (pg) and (b) body iron concentration ( p g / g )compared with haemoglobin levels in newborn male mice of genotypes sla/Y and + / Y born to s h / + carrier females.
18
Iron Deficiency Anaemia in Newborn sla Mice
269
TABLE I. Haemoglobin levels, body weights and total body iron of selected male littermates at birth. Mean values, 2 standard deviation.
+ / Y (a8 mice) Haemoglobin (g/dl) Weight (9) Total body iron (pg) Body iron concentration (pg/g)
sla/Y (26 mice) (P
Iron Deficiency Anaemia in Newborn sla Mice: a Genetic Defect of Placental Iron Transport P. J. KINGSTON, CATHERINE E. M. UANNERMAN A N D R. M. BANNERMAN
Medical Genetics U n i t , Department ofhledicine, State University of New York at Bujfaalo, and Buffalo General Hospital, Buffalo, N . Y . , U.S.A. (Received z December 1977; acceptedfor publication 20 Fcbruary 1978) SUMMARY.Newborn mice with X-linked anaemia (gene symbol sla) have lower haemoglobin levels a t birth than normal and carrier mice but there is considerable overlap. Serial observations showed that the haemoglobin values of segregating male mice separate into a bimodal distribution by 42 d of age, and 50 d values were used to assign genotypes retrospectively. The anaemia in newborn sla mice is attributable to iron deficiency, since their total body iron is lower than in normal newborn mice, while their birth weights are almost identical. Haemoglobin levels a t birth in normal, anaemic and carrier mice are also influenced by the mother’s genotype and phenotype, and the haemoglobin value was progressively lower according to the sla gene dose of the mother. Materno-fetal iron transfer was examined by labelling pregnant carrier females with radioiron in various ways. When given as single or intermittent doses by injection no clearcut differences emerged in apparent iron transfer to anaemic as compared to non-anaemic fetuses. However, when radioiron was administered continuously in food a significant reduction in iron transfer to anaemic fetuses was demonstrated. The sla gene is already known to have a major effect in reducing iron transport in the small intestine. The present studies provide evidence of an analogous defect in placental iron transport. X-linked anaemia in the mouse (gene symbol sla) is a recessively inherited trait (Falconer & Isaacson, 1962) associated with marked hypochromic anaemia (Grewal, I 962; Bannerman & Pinkerton, 1967).In adult mice anaemia is secondary to a specific dcfect in intestinal absorption of iron (Pinkerton & Bannerman, 1967), associated with the accumulation of iron in the duodenal mucosal cells (Pinkerton, 1968). Newborn sla mice were observed to be anaemic a t birth (Falconer & Issacson, 1962; Grewal, 1962). This suggested to Dancis & Jansen (1970) that there might be an analogous defect in placental iron transport, but they were unable to demonstrate such a defect in the tranxfer of radioiron in acute experiments. In this laboratory, preliminary observations on newborn male mice from male segregating litters showed a bimodal distribution of haemoglobin level but Correspondence: Dr P. J. Kingston, Department of Haematology, St Thomas’ Hospital, London, S.E.I. 0007- 1o48/78/1o00-0265$02.00
197X Blackwell Scientific Publications
265
P.J. Kingston et a1
266
there was considerable overlap (Uannerman et al, 1973). The finding that the correct genotypes cannot be reliably determined a t birth complicates the investigation of a hypothetical placental lesion. We therefore carried out serial studies on sla anaemic mice and Littermate controls from birth to over 50 d of age, the results of which allow the retrospective assignment of genotypes a t birth. Using these data, total iron content of newborn normal and anaemic mice was determined. Finally, experiments with radioiron repeatedly administered to pregnant sla carrier females, and subsequent observation of their segregating offspring, provide direct evidence of a defect in placental iron transfer in sla anaemic mice. MATERIALS AND METHODS
Animals. The sla gene is maintained in this laboratory in mice of the C57BL/6J strain into which it was transferred from the original mixed stock (Bannerman & Pinkerton, 1967),and animals of the ninth to eleventh backcross generations were used in these experiments. The following conventions have been used for recording genotypes: ‘sla’ represents the gene for anaemia carried on the X-chromosome; represents the normal wild-type allele; and ‘Y’ indicates the absence of either allele on the Y chromosome. Thus ‘+I+’ indicates a normal female, ‘sla/+’ a heterozygous carrier female, ‘rla/sla’ a homozygous anaemic female, ‘ +/Y’ a normal male, and ‘sla/Y’ a hemizygous anaemic male. The usual mating was between a carrier female (sla/+) and a normal (+/Y) or affected ( s l a / Y ) male, to obtain equal numbers of affected and normal male littermates for comparison. Mice were kept in plastic cages with chromium-plated covers. As in previous studies, they were maintained on Rockland Mouse/Rat diet (complete) containing approximately 0.02 I YO Fe, except for breeding pairs which received Rockland Mouse Breeder Diet (0.023% Fc); the pups had access to this diet from about the loth day of life until the time of weaning. O n the 21st day weanlings were segregated by sex and fed Mouse/Rat diet (complete). Tapwater was given ad libitum. Body weights were recorded to the nearest 0.01g a t birth and regularly thereafter to the nearest 0.1g. Haematol9gy. Blood was taken from the retro-orbital venous sinus, or by puncture of the jugular vein in newborn or fetal mice, and collected into heparinized capillary tubes. Haemoglobin was determined as cyanmethaemoglobin using 10 pl blood diluted in 2.5 ml Drabkin’s solution and read a t 540 mm in a Coleman Junior spectrophotometer. Chemical determination of iron. The total body iron content of mice was determined by a wet ashing method. The weighed carcass was dissolved with heating in a mixture of equal volumes of concentrated sulphuric and nitric acids. For newborn mice 5 ml of acid mixture was sufficient; appropriate volumes were used for larger carcasses. Hydrogen peroxide was added with reheating to clear the solution, followed by sodium metabisulphite. After cooling, dilution and neutralization, iron content was determined colorimetrically using orthophenanthroline (Pinkerton et al, 1970). Placental transfer experiments. All observations were carried out on pregnancies or offspring of sla/ females mated to da/Y males to produce ‘male/female’ segregating litters. Gestational age was determined by examination for copulatory plugs approximately 12 h after the male was introduced into the breeding cage; spontaneous delivery usually occurred 20 d after
‘+’
+
Iron Dejciency Anaemia in Newborn sla Mice
267
observation of the plug. A small number of litters were delivered by caesarian section after administration of a single dose of radioiron. Tracer radioiron as [59Fe]ferrouscitrate (approximately 25 pCi/pg Fe) was administered to the pregnant female mice by one of the following schedules: (a) 3-5 pCi administered as a single intravenous or intraperitoneal injection on the 18th or 19th day ofgestation; (b) 3-4 pCi administered by repeated intraperitoneal injections of 0.4 pCi given over a 5 d period during pregnancy; (c) continuous feeding of the pregnant female on an artificial milk diet (McCall et al, 1962) with ferrous citrate added to a final concentration of 266 pg elemental iron and 0.6 pCi per gram of diet. Newborn mice were separated from their mothers immediately after delivery, and before feeding. They were held in a special incubator before counting in prewarmed tubes in an automatic gamma scintillation well counter (Nuclear-Chicago Model 1085). Duplicate I min counts were made and these were within I o/‘ in every case. After counting, the haemoglobin level was determined, and then the pups were returned to their mother to be raised in the usual fashion for determination of genotype a t 50 d of age.
RESULTS Newborn male mice from ‘male-segregating’ litters were studied serially from birth to after 50 d of age. Body weights and haemoglobin lcvels were determined within 24 h of birth (usually within 12 h), and at 21,42 and 50 d. Abdominal haematomata were noted to be responsiblc for neonatal anaemia in a small number of mice. This is a little-known cause of anaemia, but it is important to be aware of it, and these mice were excluded from subsequent analysis, though they were followed and their genotypes could ultimately be determined. A total of 283 males were available for study. Haemoglobin levels in newborn mice ranged from 5 to 15 g/dl; although the pattern of distribution is bimodal there is no clear separation. A t 21 d the overlap is even greater. A t 42 and 50 d of age segregation is apparent, and two groups of equal numbers emerged, so that genotypes could then be assigned unequivocally. Their validity was confirmed by previous studies with test matings (Bannerman & Pinkerton, 1967), reconfirmed by further test matings from representative mice in the present investigation. Retrospective analysis of the accumulated data after assignation of genotypes show the overlap in values which obscures an underlying difference at birth, as illustrated in Fig i ( a ) . Body weights are almost identical in newborn normal and anaemic males, but a difference begins to emerge at 21 d of age and anaemics are visibly smaller after the age of 50 d (Fig Ib).
Body Iron at Birth Haemoglobin values were determined on fiz newborn male litter mates from ‘male-segregating’ litters which were then killed by ether anaesthesia, and the iron content of the carcasses determined as described above. An appropriate correction was added for the blood withdrawn to determine the haemoglobin level. There is a direct correlation between haemoglobin level and total iron (Fig 2a) and a more marked correlation between haemoglobin level and iron concentration expressed as pg/g body weight (Fig 2b). These results are derived from a mixed group, comprising mice of two genotypes. For
P. J . Kingston et
26s
111
FIG I . Haemoglobin levels (a) and body weights (b) in a group of sla anaemic (shaded) and normal male littermates (unshaded) at birth, 2 1 and 50 d. Genotypes have been assigned retrospectively from the segregation observed at 50 d.
Newborn
d Mice ( s l a / Y and + / Y l
8or
..
N e w b o r n d M i c e ( s / a / Y and t / Y )
om
r.048 ( P (0011
r = 0.57 ( p (0.01)
-1-1-1-1- I 3 Za
I < 6
9
Haemoglobin y / d l
12
15
L4-L 3
18 Zb
6
1-1-1-1 9
I2
15
Haemoglo bi n y /dl
FK 2 . (a) Total body iron (pg) and (b) body iron concentration ( p g / g )compared with haemoglobin levels in newborn male mice of genotypes sla/Y and + / Y born to s h / + carrier females.
18
Iron Deficiency Anaemia in Newborn sla Mice
269
TABLE I. Haemoglobin levels, body weights and total body iron of selected male littermates at birth. Mean values, 2 standard deviation.
+ / Y (a8 mice) Haemoglobin (g/dl) Weight (9) Total body iron (pg) Body iron concentration (pg/g)
sla/Y (26 mice) (P
