Downloaded from http://adc.bmj.com/ on June 16, 2017 - Published by group.bmj.com
Arch. Dis. Childh., 1967, 42, 259.
Immunoglobulins in Umbilical Cord Plasma
I: Healthy Infants
HAZEL THOM, ERIC McKAY, and DEREK GRAY
From the Department of Child Health, University of Aberdeen
Methods
Ig G is selectively transferred across the placenta
from mother to foetus (Gitlin, Kumate, Urrusti, and
Collection of specimens. Umbilical venous blood
Morales, 1963, 1964a, b), and both by electro- was collected via a wide-bore needle into heparinized
phoretic study of y-globulins (Stemnberg, Dagenais- tubes during the third stage of labour after the umbilical
Perusse, and Dreyfuss, 1956) and by specific cord had been clamped. The blood was stored at 40 C.
estimation of Ig G by immunological methods for up to 24 hours, when the plasma was separated and
at approximately - 200 C. until tested. Specimens
(Kohler and Farr, 1966), it appears that this protein kept
were obtained from 23 normal full-term infants and 19
is normally in higher concentration in foetal than apparently
premature infants of birthweight less
maternal serum at term. Only very small amounts than 2500 g.healthy
All infants with congenital abnormalities
of other plasma proteins are transferred from mother or haemolytic disease of the newborn were excluded from
to foetus, and of the large molecules, such as the series, as were stillbirths and premature infants who
fibrinogen and the macroglobulins, little or none is died within 24 hours of birth or who later developed
transferred (Gitlin, 1964). Ig A was at first respiratory distress syndrome. The results in those
considered to be completely absent from umbilical groups will be reported later. Also excluded were
cord serum (Hitzig, 1957; Scheidegger and Martin infants born of mothers who had a threatened abortion in
du Pan, 1957; Roth, 1962; West, Hong, and the early weeks of pregnancy, but those from pregnancies
Holland, 1962), but small amounts have been characterized by toxaemia were included in the series.
demonstrated in many sera (de Muralt and Roulet,
of immunoglobulins Ig G, Ig A, and
1962; Haworth, Norris, and DiUing, 1965; Stiehm Ig Quantitation
and
M,
albumin.
Individual proteins were estimaand Fudenberg, 1966). Early tests for Ig M like- ted by single radial diffusion
in agar using a modification
wise failed to show it in umbilical cord serum of the method described by Fahey
and McKelvey (1965).
(Hitzig, 1957; Scheidegger and Martin du Pan, The agar was prepared as described by these authors
1957), though it is now known to be present in with antiserum added to give final concentrations of 1 in
concentrations of from 5 to 25% of normal adult 40 for albumin, 1 in 40 for Ig G, 1 in 100 for Ig M, and
values (Franklin and Kunkel, 1958; West et al., 1 in 100 for Ig A. Antiserum from Hyland Laboratories
(Los Angeles, California) was used for Ig M and Ig A
1962; Roth, 1962; Stiehm and Fudenberg, 1966).
Although the above series have differentiated estimations, and rabbit antiserum prepared in this
laboratory for albumin and Ig G. The agar was poured
between mature and premature infants by the on
perspex plate to gave a film measuring 95 x 95 x
dividing standard of 2500 g. birthweight, no account 1 8a mm.
This
36 wells which were of
has been taken of birthweight differences within the 2 3 mm. diameteraccommodated
for the Ig M estimations and 1 * 7 mm.
groups, except that Roth (1962) divided the pre- diameter for the other proteins.
mature infants into two subgroups, and in later
All diffusion plates contained 5 duplicate sets of
studies of Ig A, Haworth and Dilling (1966) divided standard solutions made as dilutions in saline of a
them into three subgroups. The present study was reference plasma which was stored at - 20° C. and
carried out in order to investigate the relation consisted of pooled plasma from 12 healthy adults.
between the concentrations of individual immuno- Standard dilutions offrom 30% to 7 5 %, reference plasma
used for Ig G estimations, 20% to 5% for Ig M,
globulins in umbilical cord blood and birthweight were
to
20%
2-5% for Ig A, and 25 % to 7 5% for albumin.
of the infant, and the interrelationships between the The unknown
plasma specimens were used neat for
concentrations of the separate immunoglobulins.
estimations of Ig A and Ig M, but diluted 1 in 5 in saline
-
for Ig G and albumin and, if necessary, repeat estimations
Received July 20, 1966.
were made with further dilution of the specimens.
259
Downloaded from http://adc.bmj.com/ on June 16, 2017 - Published by group.bmj.com
260
Thom, McKay, and Gray
25
Results
I
Ig A. The concentration of Ig A in all specimens
of umbilical venous plasma from healthy infants was
below the critical testing level of 2 6 mg./100 ml.
20
8
-SE.
*-
I
5
xc
10-
except in one full-term infant where the concentration was 23-3 mg./100 ml.
0
.
Ig M. The infant noted above to have a high
Ig A concentration also had an Ig M concentration
clearly above that in all other infants in the group
0
p
e
ea
* e
(Fig. 1). These differences were so definite that it
decided to remove the case from the series of
healthy
infants. His results are shown on the
0
graphs for individual protein concentrations but are
not used for statistical analysis including mean
1000 1500
values.
Birthweight (9.)
Although the concentration of Ig M in umbilical
FIG. l.-Ig M concentration in umbilical venous plasma cord plasma rises from a mean of 9 1 mg./100 ml. in
according to birthweight, with mean concentration and 95% the premature infants to 10O4 mg./100 ml. in the
confidence limits. , Case not included in statistical full-term
infants, there is not a statistically signifianalysis (see text).
cant correlation between Ig M concentration and
0181; p > 0 1). The over-all
The agar plates were incubated for 24 hours at 25° C. birthweight (r
in a humid chamber before readings were made by view- mean is 9 * 8 mg./100 ml. for the 22 full-term and 19
ing the diffusion circles by dark field illumination and premature infants in the series, with a fairly wide
measuring circle diameters using either a lens and ruler scatter (standard deviation 3 6 mg./100 ml.).
with 0 5 mm. scale or a microscope and Vernier scale. Individual results are recorded in Fig. 1.
E
0~~~~~
was
5.
2000
2500
3000
3500
4000
=
-
Readings estimated to the first decimal point were made
by two observers and rarely differed by more than 0 1
mm. The mean of the two readings was used to calculate results.
The concentrations of Ig G, Ig A, and Ig M in the
reference plasma were estimated using Hyland immunoplates and found to be 1670 mg./100 ml., 204 mg./100 ml.,
and 101 mg./100 ml., respectively. The concentrations
of the immunoglobulins in the umbilical cord plasma
specimens were first expressed as percentages of their
concentrations in the reference plasma and then converted to mg./100 ml. on the basis of the Hyland immunoplate readings for the reference plasma. Albumin has
been expressed as a percentage of the concentration
present in the reference standard plasma.
The accuracy ofthe method was assessed by calculating
the coefficient of variation Standard deviation x 100
mean
For albumin, Ig G, and Ig A, this value was calculated
from single diffusion plates consisting of 6 dilutions of
the reference plasma each applied six times, and for Ig M
it was calculated from the standards applied to several
agar plates, in all, 6 dilutions each 6 times. The
coefficients of variation were Ig G-5 * 0%, Ig A-3 *5%,
Ig M-7-8%, and albumin-6 - 6%.
It was expected that most specimens of umbilical cord
plasma would contain very little Ig A, and preliminary
double diffusion in agar tests were done on all specimens.
Quantitative single radial diffusion tests were done only
on specimens giving precipitin lines on the double
diffusion test which was sensitive down to a concentration of 1-25% of the reference plasma (2.6 mg./100 ml.).
Ig G. The mean concentration of Ig G in the
remaining 22 healthy full-term infants was 1759 mg./
100 ml., and that in the 19 healthy premature
infants, 1317 mg./100 ml. Fairly close correlation
exists between Ig G concentration and birthweight
(r =0 518, p < 0 001), Ig G increasing with rising
birthweight as expressed in the regression equation
Y = 0*421X + 484, where X = birthweight in
grams and Y = Ig G concentration in mg./100 ml.
Taking the birthweight of the average full-term
infant to be 3250 g., the mean Ig G concentration at
term calculated from the regression equation is
1852 mg./100 ml. or 111% of the adult reference
plasma. The scatter of individual results in the
series may be seen in Fig. 2.
Albumin. Umbilical cord albumin concentration is closely correlated to birthweight (r = 0 * 627;
p < 0 001), and the concentration shows a steady
rise with increasing birthweight (Fig. 3) as expressed
in the regression equation, Y = 0 -0146X + 35*6,
where X = birthweight in grams, and Y = albumin
concentration expressed as a percentage of the
concentration present in the reference plasma used.
This gives a calculated concentration of 83 0% of
the reference plasma albumin concentration per 100
ml. plasma for an infant of 3250 g. birthweight.
The actual mean concentration for the 22 infants of
Downloaded from http://adc.bmj.com/ on June 16, 2017 - Published by group.bmj.com
Immunoglobulins in Umbilical Cord Plasma
261
3000
2 500
/
E
8 2000
0
,
E
51000
500
IqG
I 11-11'
.-
20'00
1000
30'00
Z100
FIG. 4.-Relationships between the concentrations of Ig M,
Ig G, and albumin in umbilical venous plasma, and birthweight of the infant.
Birthweiqht (9.)
FIG. 2.-Regression of umbilical venous plasma Ig G
concentration on birthweight, with 95°0 confidence limits
for Ig G concentration. A Case not included in statistical
analysis (see text).
birthweight over 2500 grams was 82 4% of the
reference plasma albumin concentration, and for the
0 of the reference
19 infants under 2500 g. was 63 1%
plasma albumin concentration.
Relationships between individual immunoglobulin concentrations. Fig. 4 illustrates the
interrelationships between the concentrations of the
individual plasma proteins studied. The concentra12 5
Albumin
p<0001
r=+0792
tions of albumin and Ig G are closely correlated, and
both correlate with birthweight, but Ig M concentration is not significantly related to birthweight.
There is, however, a significant relation between the
concentration of both Ig G and albumin and that of
Ig M. Fig. 5 and 6 shows the individual relations
and the regression lines and 95% confidence limits
relating Ig M and Ig G and albumin. The graphs
provide a useful means of showing discrepancies
between the concentrations of the proteins in
pathological states.
2 5-
-
20-
E
100
8
_
SO
0
.5
0~~~~~~~~~~~
C C7 5
*
* 0~~~~~~~
0
50
@ @@ ~~~S
c
a.
0
00
:*
10
E
00
0
0
Sol
*
0
@.0
*,0
0
0~~~~~~
0~~~
0
0
5
:
0
25
0
1000
2000
3000
4000
Birthweight (q.)
FIG. 3.-Regression of umbilical venous plasma albumin
concentration on birthweight, with 95% confidence limits
for albumin concentration. V Case not included in
statistical analysis (see text).
500
1000
1500
2000
2500
Plasma IqG conc. (mq./ 100ml.)
FIG. 5.-Regression of Ig M concentration on Ig G
concentration in umbilical venous plasma, with 95%
confidence limits for Ig M concentration. , Case not
included in statistical analysis (see text).
Downloaded from http://adc.bmj.com/ on June 16, 2017 - Published by group.bmj.com
Thom, McKay, and Gray
262
25
E
8
-
20
-
'5
-
0
I
10
0
C
5
*
0
0
0~~~
- o @-.@-~~
*.
0~~~~~~
*0-
E
0
*
*I*
-
IL
0
f
100
75
S0
Plasma albumin conc. (a/oconc. in reference plasma)
FIG. 6.-Regression of Ig M concentration on albumin
concentration in umbilical venous plasma, with 95%
confidence limits for Ig M concentration. % Case not
included in statistical analysis (see text).
Discussion
As in several previous studies which have been
referred to in the introductory paragraphs, Ig A was
not detectable by the method used in most specimens of umbilical cord plasma in the present series.
Only one infant gave a positive test, with a concentration of 23-3 mg. Ig A/100 ml. plasma. Detailed
comparison of the case notes of this mother and
infant with those of the others in the 'normal' group
showed no abnormality to account for the result.
Estimation of other plasma proteins on the specimen
showed: albumin: 75*0 % of the normal reference
plasma concentration; Ig G: 2100 mg./100 ml.;
Ig M: 26-0 mg./100 ml. Because the Ig A and
10
9
-
8
-
7
7 C
6 5-
6
4 -
6
z
3 2-
[FL
I
Il-HI]
300
100
200
Plasma IqA conc. (mq./lOOml-)
n
I.
n
L
400
FIG. 7.-Concentration of Ig A in maternal venous plasma
at delivery.
Ig M results were clearly above all others in the
group, the case was excluded from the statistical
analysis. However, it was of interest to find that a
specimen of the mother's plasma taken at the time of
delivery showed an Ig A concentration of 416 mg./
100 ml., which was 204% of the concentration in the
pooled reference plasma. It was decided to estimate
the concentration of Ig A in a number of mothers at
term to discover whether the result in this mother
was outside the normal range. In 55 mothers
studied (including the case under discussion) the
mean concentration of Ig A was 206 mg./100 ml.,
with a range from 59 to 530 mg./100 ml. and a
standard deviation of 86 mg./100 ml. Fig. 7,
however, shows that all but two of the results lay
between 59 and 347 mg./100 ml. The mother
under discussion had the second highest result in
the series. The infant of the mother with the
highest value of 530 mg./100 ml. had a normal Ig A
concentration of below 2-6 mg./100 ml., and it is
generally agreed that Ig A does not normally cross
the placenta from mother to foetus in appreciable
amounts. The reason for the high Ig A concentration in this one child and his mother is not clear.
In comparable studies of serum Ig A concentration in mothers at time of delivery, Haworth and
Dilling (1966) found mean concentrations of from
215 3 to 289 8 mg./100 ml. in five groups based on
birthweight of the infant and growth retardation.
There was no significant difference between the
results in the five groups.
The minimal sensitivity of the method used for
the estimation of Ig A in umbilical cord plasma
specimens was a concentration of 2 6 mg./100 ml.,
which is just above the mean of 1 0 mg./100 ml.
found by Stiehm and Fudenberg (1966) in their
series of normal infants, and it seems likely that the
techniques used in the present series would readily
detect pathological increase in Ig A concentration
such as has been recorded in the rubella syndrome
(Alford, 1965; Bellanti, Artenstein, Olson, Buescher,
Luhrs, and Milstead, 1965; Soothill, Hayes, and
Dudgeon, 1966) and in infection by cytomegalic
inclusion virus (McCracken and Shinefield, 1965).
The concentration of Ig M in umbilical cord
blood does not rise significantly over the weight
range studied-from 1450 to 3760 g.-and it is,
therefore, permissible to use the mean of the results
in a series of normals as a standard for comparison
with results in pathological states. In similar
comparisons of Ig G, mean values cannot be used
unless the groups are carefully controlled for
weight, since Ig G concentration rises steeply with
increasing birthweight, an increase of 500 g. (1 1 lb.)
in birthweight being associated with a rise in Ig G of
Downloaded from http://adc.bmj.com/ on June 16, 2017 - Published by group.bmj.com
Immunoglobulins in Umbilical Cord Plasma
211 mg./100 ml. Albumin shows a similar steep
rise in concentration with increasing birthweight.
The fluid balance of the infant is labile and readily
upset both after and before birth as, for example, in
the infants of mothers given large volumes of
hypotonic solution (Altstatt, 1965). It is, therefore,
readily possible for true increase in total circulating
immunoglobulins in the infant to be masked by
haemodilution or simulated by haemoconcentration.
A high concentration of immunoglobulins in the
infant's plasma may represent a true increase or
haemoconcentration, and normal or low concentrations may be found when increased total circulating
immunoglobulins and haemodilution occur together. It is, therefore, necessary to have some
criterion of haemoconcentration and haemodilution.
Haematocrit readings may be used for this purpose,
but in this and later studies albumin concentration
has been utilized since it is convenient to measure by
the technique used for immunoglobulins, and the
test can be postponed until a convenient time after
collection of the specimen. By studying together
the concentration of the four plasma proteins
estimated, it may be possible to differentiate between
alterations in fluid balance and changes specifically
affecting the immunoglobulins of foetal or maternal
origin.
Summary
Ig G, Ig A, Ig M, and albumin concentrations
were estimated on umbilical venous plasma from 42
healthy infants, 19 of whom were under 2500 g.
birthweight. Ig G and albumin concentrations
were positively related to birthweight, whereas Ig M
showed an insignificant rise over the weight range
studied. A significant positive relation existed
between Ig M concentration and both Ig G and
albumin concentrations. Ig A concentration was
below 2 * 6 mg./100 ml. in all but one infant, in whom
the concomitant presence of high Ig M levels
suggested that he should not be considered to be
normal.
The concentrations of Ig G, Ig M, and albumin
in infants weighing less than 2500 g. at birth were
1317 mg./100 ml., 9 1 mg./100 ml., and 63 1% of
the concentration in pooled adult reference plasma
respectively, with corresponding values for infants
weighing over 2500 g. at birth of 1759 mg./100 ml.,
10*4 mg./100 ml., and 82*4% adult reference
plasma concentration.
We are indebted to Professor R. G. Mitchell and
Professor I. MacGillivray for permission to study
263
patients under their care, and to the nursing staff of the
maternity hospital for their co-operation in the collection
of blood specimens. We thank Dr. Gillian Fowler for
advice regarding statistical analysis.
REFERENCES
Alford, C. A., Jr. (1965). Studies on antibody in congenital rubella
infections. I. Physiochemical and immunologic investigations of rubella neutralizing antibody. Amer. J. Dis. Child.,
110, 455.
Altstatt, L. B. (1965). Transplacental hyponatremia in the newborn infant. J. Pediat., 66, 985.
Bellanti, J. A., Artenstein, M. S., Olson, L. C., Buescher, E. L.,
Luhrs, C. E., and Milstead, K. L. (1965). Congenital rubella.
Clinocopathologic, virologic and immunologic studies. Amer.
J. Dis. Child., 110, 464.
Fahey, J. L., and McKelvey, E. M. (1965). Quantitative determination of serum immunoglobulins in antibody-agar plates.
J. Immunol., 94, 84.
Franklin, E. C., and Kunkel, H. G. (1958). Comparative levels of
high molecular weight (19S) gamma globulin in maternal and
umbilical cord sera. J. Lab. clin. Med., 52, 724.
Gitlin, D. (1964). Protein metabolism, cell formation, and immunity. Borden Award Address. Pediatrics, 34, 198.
-, Kumate, J., Urrusti, J., and Morales, C. (1963). The marked
selectivity of the human placenta in the transfer of proteins from
mother to fetus. J. Pediat., 63, 870.
-, -, -, and - (1964a). Selective and directional
transfer of 7S gamma2-globulin across the human placenta.
Nature (Lond.), 203, 86.
and - (1964b). The selectivity of the human
placenta in the transfer of plasma proteins from mother to fetus.
J. clin. Invest., 43, 1938.
Haworth, J. C., and Dilling, L. (1966). Concentration of gammaA
globulin in serum, saliva, and nasopharyngeal secretions of
infants and children. J. Lab. clin. Med., 67, 922.
, Norris, M., and Dilling, L. (1965). A study of the immunoglobulins in premature infants. Arch. Dis. Childh., 40, 243.
Hitzig, W. H. (1957). Die physiologische Entwicklung der
'Immunglobuline' (gamma und beta2 globuline). Helv.
paediat. Acta, 12, 596.
Kohler, P. F., and Farr, R. S. (1966). Elevation of cord over
maternal IgG immunoglobulin: Evidence for an active placental
IgG transport. Nature (Lond.), 210, 1070.
McCracken, G. H., Jr., and Shinefield, H. R. (1965). Immunoglobulin concentrations in newborn infants with congenital
cytomegalic inclusion disease. Pediatrics, 36, 933.
de Muralt, G., and Roulet, D. L. A. (1962). Recherches immunologiques sur les proteines foetales, en particulier sur une
globuline specifique du foetus humain. In Protides of the
Biological Fluids: Proceedings of the 9th Colloquium, Bruges, 1961,
ed. H. Peters, p. 297. Elsevier, Amsterdam.
Roth, N. (1962). Zur semiquantitativen Erfassung der beiden
Serum-Immun-Globuline beta2A und beta2x im Neugeborenen-und Kindesalter. Ann. paediat. (Basel), 199, 548.
Scheidegger, J. J., and Martin du Pan, R. (1957). Etude immuno6lectrophoretique des proteines s6riques du nouveau-n6 et du
nourrisson. gtud. neo-natal., 6, 135.
Soothill, J. F., Hayes, K., and Dudgeon, J. A. (1966). The immunoglobulins in congenital rubella. Lancet, 1, 1385.
Stemnberg, J., Dagenais-Perusse, P., and Dreyfuss, M. (1956).
Serum proteins in parturient mother and newborn: an electrophoretic study. Canad. med. Ass. J., 74, 49.
Stiehm, E. R., and Fudenberg, H. H. (1966). Serum levels of
immune globulins in health and disease: a survey. Pediatrics,
37, 715.
West, C. D., Hong, R., and Holland, N. H. (1962). Immunoglobulin levels from the newborn period to adulthood and in
immunoglobulin deficiency states. J. clin. Invest., 41, 2054.
Downloaded from http://adc.bmj.com/ on June 16, 2017 - Published by group.bmj.com
Immunoglobulins in umbilical
cord plasma. I. Healthy infants.
H. Thom, E. McKay and D. Gray
Arch Dis Child 1967 42: 259-263
doi: 10.1136/adc.42.223.259
Updated information and services can be found at:
http://adc.bmj.com/content/42/223/259.citation
These include:
Email alerting
service
Receive free email alerts when new articles cite this
article. Sign up in the box at the top right corner of the
online article.
Notes
To request permissions go to:
http://group.bmj.com/group/rights-licensing/permissions
To order reprints go to:
http://journals.bmj.com/cgi/reprintform
To subscribe to BMJ go to:
http://group.bmj.com/subscribe/