- Email: [email protected]
The determination of serum immunoglobulin concentrations on the basis of their light-chain antigenic properties
343
Clinica Chimica Acta, 78 (1977) 343-349 @ Elsevier/North-HolIand Biomedical Press
CCA8717
THE DETERMINATION OF SERUM IMMUNOGLOBULIN CONCENTRATIONS ON THE BASIS OF THEIR LIGHT-CHAIN ANTIGENIC PROPERTIES
K. EICKHOFF
and R. HEIPERTZ
Neurochemical Laboratory, Department Strasse 5, D-34 Giittingen (G.F.R.) (Received
of Neurology,
University of Giittingen, uon Siebold
March 4th, 1977)
Summary A method for the measurement of light-chain determined immunoglobulin concentrations (Ig-type K and Ig-type h) in serum is described. The method is based on the principle of radial immunodiffusion. The results obtained after calibration with standarized human serum and with isolated Bence-Jones protein are closely correlated. The values for Ig-type K, Ig-type x and the K/A ratio from sera with normal concentrations of IgG, IgA and IgM are presented and some problems of calibration discussed.
Introduction All classes of human immunoglobulins contain light chains of either the K- or h-type [ 1,2]. Because of their antigenic properties immunoglobulin concentrations in blood serum can be quantitatively determined as Ig K (or Ig-k) and Ig X (or Ig-1). To date little is known of the immunoglobulin light-chain concentrations of serum in neurological disease, although an altered K/A ratio has been found in the CSF of patients with multiple sclerosis [ 31. Materials and methods Sera obtained from 54 neurological patients (aged 18 to 67 years) were examined. Concentrations of IgG, IgA and IgM were estimated for each serum Marburg). The specific light-chain immuno(Tri-Partigen, Behringwerke, globulin concentrations were determined by radial immunodiffusion according to the following approach: Antiserum (from rabbit) containing either antiimmunoglobulin against light chain type k or antiimmunoglobulin against light chain type 1 (Behringwerke). Barbitone buffer at pH 8.6, I = 0.02.
344
Agarose solution containing 2% agarose (Behringwerke) with 1% polyethylene glycol and 0.1% sodium azide in buffer. 0.3 ml antiserum were mixed with 8 ml liquid agarose solution (50°C) and poured into a level plastic dish (Behringwerke). After the gel had solidified the dishes were covered and stored in the fridge at 4°C overnight. The next day 12 antigen wells of 4 mm diameter were cut into the gel layer and filled with 20 ~1 of sample each, 6 holes were used for the calibration standards and 6 holes for the serum samples diluted 1 : 100. The diffusion process lasted for three days at room temperature after which the diameter of the precipitate was measured with a magnified scale. The estimation of the protein concentrations as well as the statistical analyses were performed on a programmed calculator (Wang 600). Two different calibration systems were used on each diffusion-dish: (a) Three dilutions of a standard human serum for Ig-k and Ig-1 (Nr. 974, Behringwerke) with a defined concentration of Ig-k and Ig-1 in the undiluted serum of 1000 U/l. The dilution of 1 : 200,l : 100 and 1 : 51 corresponded to 5, 10 and 19.6 U/l Ig-k and Ig-1, respectively. (b) Three dilutions of Bence-Jones protein control serum, type k and type 1 (Behringwerke). Concentration of Bence-Jones protein in the undiluted serum 2.5 mg/ml for both Ig-k and Ig-1. The dilutions of 1 : 200, 1 : 51 and 1 : 21 corresponded to 12.5, 49 and 119 mg/l for each. Results 42 of the sera had normal immunoglobulin concentrations, mean values and standard deviation of these are shown in Table I. When calibration is performed TABLE
I
RESULTS
OF
NORMAL
IgG.
Results globulin
are
QUANTITATIVE AND
expressed
type
Bence-Jones __-
THE IgA
A: HS.
IgM
as mean value
IMMUNOGLOBULIN
DETERMINATION
IN
(X)
refers
and to
standard
deviation
calibration
with
(s).
human
Ig-k,
immunoglobulin
serum;
IgG
BJP,
value
type refers
K; Ig-I. immunocalibration
__ 12.49
IgA
g/l
2.57
0.87
IgM
g/l
1.82
0.51
Ig-kHS
U/1
870
Ig-IHS
U/l
870
Ig-kHS
to
with
s
gil
Ig-IHS
WITH
~~. _~__ _. _-__
Ig-kHS
SERA
protein.
r
Quotient
42
CONCENTRATIONS
k/IHS + Ig-IHS (k + 1 = 100) (k + I = 100)
2.43
190 240 1.07
U/I
1740
0.40 350
%
50
7.8
%
49
7.8
lg-kBJP
g/l
5.01
1.44
Ig-lBJP Quotient
g/l
2.92
0.75
1.78
0.55
Ig-kB Jp + Ig-IB Jp
g/l
7.93
1.87
Ig-kgJp (k + 1 = 100) I!Z-lBJp (k + 1 = 100) Ig-kBJp (IgG + A + M = 100) &-lBJp (IgG + A + M = 100) Ig-kBJp + Ig-lBJp (IgG + A + M = 100)
7% 9%
63
7.1
37
7.1
%
30
7.2
%
18
4.5
%
47
9.5
k/lgJp
._
345
n i = 63 s = 7.1
Ig
(k) EJP
I5
n ii = 37 s = 7.1
Fig. 1. Distribution curves of relative Ig-k and Ig-1 concentrations of 42 sera with normal IgG, IgA and IgM concentrations. The individual results are expressed as percent of total light chain concentration (Ig-k + Ig-I = 100%). Calibration with Bence-Jones protein.
TABLE II RESULTS OF THE LINEAR GLOBULIN CONCENTRATION
REGRESSION ANALYSIS, OF ALL 54 SERA
CORRELATIONS
BETWEEN
IMMUNO-
For abbreviations see Table I. r, correlation coefficient; a, y-axis intercept: b, gradient: syx, deviation. x
Ig-kH,y
I&-IS
Y
r
Ig-kBJP Ig-lBJP
0.890
Ig-kHs + IdHS
k-kBJp
%-Is
kflBJP Ig-kB Jp + I&B
IgG IgA IgM
h3-kB Jp + Ig-lBJp
IgG + IgA + IgM
+ Ig-lBJp Jp
h3-kBJp
+ I%+,Jp
k+BJp
+ Ig-lBJp
0.699 0.815 0.718 0.678 0.633 0.551 0.051
0
-0.61 1.10 0.30 0.7 1.51 2.47 5.06 7.84
b
SY.X
6.5 2.1 4.4 1.1 0.4 0.4 1.2 0.2
0.80 0.52 1.28 0.4 1.62 1.71 1.84 2.20
346
o*
5oIg (I),,
DO
Isa
” /dl
Fig. 2. Correlation between results obtained after calibration with standwised human serum (HS) and Bence-Jones protein (BJP) for Ig-k and E-1. yk = -61.1 + 6.5 xk, rk = 0.890: Y1 = 109.6 + 2.1 xl, q = 0.699.
with human serum relative concentrations of Ig-k and Ig-1 are practically equal (Ig-k + Ig-1 = lOO%), the quotient is 1.07. When Bence-Jones protein is used for calibration the results can be expressed as mg/l and 63% of total light chain concentration consists of Ig-k, 37% of Ig-1 (Fig. l), the quotient is 1.78. Only 47% of total immunoglobulins (IgG, IgA and IgM) can be typified by their light chains when values are expressed as mg/l. Table II shows the results of the linear regression analysis on all 54 sera. This demonstrates that there is a highly significant correlation between values
341
. . .
.
/ .
.
/ . .
. .
/
.
.
l
l
.
..
. .
l . . .
.
/
. /*
.
.
.
.
Fig. 3. Correlation between total immunoglobulin concentration concentrations (Ig-k + Ig-1).
(IgG + IgA + IgM) and sum of light chain
Y 20
L
I5 U/I
0
I
1
I
I
L
20
LO
60
60
I
100
I
1
120
X
mg/l
Fig. 4. Relationship between light chain concentrations Ig-k and Ig-1 of the two calibration standards: Standarised Human Serum (y-axis) and Bence-Jones Protein Control (x-axis). The conversion of U/l into mg/l and vice versa is possible according to the following criteria: Ig-k. y = 1.0 + 0.16 x (r = 0.999): Ig-I. Y = -0.34 + 0.31 x (r = 0.997).
348
obtained with both methods of calibration (Fig. 2). A significant correlation also exists between total immunoglobulin concentrations (IgG, IgA and IgM) and total light-chain dependent immunoglobulin concentrations (Ig-k and Ig-1) (Fig. 3). There is no correlation between IgM and light chain concentration. The statistical relationship permits conversion of data from U/l into mg/l and vice versa (Fig. 4). Discussion The occurrence of blocked antigenic sites on the light chains of intact immunoglobulins [4] indicates a comparison of two calibration methods for the quantitative light-chain determination in serum. The approach with BenceJones protein as a calibration standard is derived from Iwashita et al. [5]. The close correlation between the two calibration methods is an indication that the portion of the light chain are antigenic determinants of the “constant” responsible for the quantitative measurement. Reactions with the “hidden determinants” can only be expected with increased concentrations of free light chains as for instance in uremia and multiple myeloma with Bence-Jones protein [6]. In contrast to Fahey [7] we were only able to measure 47% of total immunoglobulins as type k and type 1 on calibration with Bence-Jones protein, this could be attributed to the concentration of the antiserum used in our experiment. On the other hand the relative distribution with 63% for Ig-k and 37% for Ig-1 is in good agreement with the results of this author. Also the k/l quotient with 1.78 is within the range of values quoted from the literature: 1.86 [8,2], 1.68 and 1.91 [9] and 2.0 [lo]. While isolated antibodies can have a defined light-chain type [l] little has been reported of the distribution of Ig-k and Ig-1 concentrations in serum or of any alterations of the k/l ratio. It has been demonstrated that the k/l ratio shows some dependance on age [9]. Skvaril et al. [ 111 have described a wide scatter of values for the k/l ratio in patients with antibody deficiency syndrome and hypergammaglobulinemia whereas other authors have described abnormal findings only in cases of multiple myeloma, macroglobulinemia and lupus erythematosus [2,12]. With quantitation of the results as mg/l (calibration with Bence-Jones protein) a more precise analysis of light chain determined immunoglobulin concentrations seems possible. Acknowledgements The authors thank Miss A. Kliippner for her technical assistance. This research was financially supported by the Deutsche Forschungsgemeinschaft (Dr. Fischer-Bosch Stiftung). References 1 2 3 4
Franklin, E.C. and Fudenberg, H.H. (1964) Arch. Biochem. Biophys. 104.433 McKelvey. E.M. and Fahey, J.L. (1965) J. Clin. Invest. 44,1778 Link, H. and Zettervall. 0. (1970) Clin. EXP. Immunol. 6, 435 Epstein. W.V.. Tan, M. and Gross, D. (1964) Nature 202. 1175
349
5 Iwashita.
H., Grunwald,
6 Epstein, W.V..
Gulyassy,
F. and Bauer. H. (1974)
J. Neural.
P.F., Tan. M. and Rae, A.I. (1968)
7 Fahey, J.L. (1963) J. Immunol. 91.438 8 Fahey. J.L. and McKelvey, E.M. (1965)
J. lmmunol.
207. 45 Ann. Int. Med. 68.48
94, 84
9 Skvaril, F.. Barandun. S., Morel& A., Kuffer, F. and Probst. M. (1976) Blut 33. 281 10 Von Drach, J. (1975) Untersuchunpen iiber Kappa-Lambda-Relationen und Nachweis freier L-Ketten im Liquor cerebrospinalis unter besonderer Beriicksichtigung der Multiplen Sklerose. Dissertation. Gottingen 11 Skvaril, F., Barandun, S., Morell, A., Kuffer, F. and Probst. M. (1975) Prot. Biol. Fluids. 23rd Coll. p. 415, Pergamon Press, New York 12 Epstein, W.V. and Tan, M. (1966) Arthritis Rheum. 9, 713
Clinica Chimica Acta, 78 (1977) 343-349 @ Elsevier/North-HolIand Biomedical Press
CCA8717
THE DETERMINATION OF SERUM IMMUNOGLOBULIN CONCENTRATIONS ON THE BASIS OF THEIR LIGHT-CHAIN ANTIGENIC PROPERTIES
K. EICKHOFF
and R. HEIPERTZ
Neurochemical Laboratory, Department Strasse 5, D-34 Giittingen (G.F.R.) (Received
of Neurology,
University of Giittingen, uon Siebold
March 4th, 1977)
Summary A method for the measurement of light-chain determined immunoglobulin concentrations (Ig-type K and Ig-type h) in serum is described. The method is based on the principle of radial immunodiffusion. The results obtained after calibration with standarized human serum and with isolated Bence-Jones protein are closely correlated. The values for Ig-type K, Ig-type x and the K/A ratio from sera with normal concentrations of IgG, IgA and IgM are presented and some problems of calibration discussed.
Introduction All classes of human immunoglobulins contain light chains of either the K- or h-type [ 1,2]. Because of their antigenic properties immunoglobulin concentrations in blood serum can be quantitatively determined as Ig K (or Ig-k) and Ig X (or Ig-1). To date little is known of the immunoglobulin light-chain concentrations of serum in neurological disease, although an altered K/A ratio has been found in the CSF of patients with multiple sclerosis [ 31. Materials and methods Sera obtained from 54 neurological patients (aged 18 to 67 years) were examined. Concentrations of IgG, IgA and IgM were estimated for each serum Marburg). The specific light-chain immuno(Tri-Partigen, Behringwerke, globulin concentrations were determined by radial immunodiffusion according to the following approach: Antiserum (from rabbit) containing either antiimmunoglobulin against light chain type k or antiimmunoglobulin against light chain type 1 (Behringwerke). Barbitone buffer at pH 8.6, I = 0.02.
344
Agarose solution containing 2% agarose (Behringwerke) with 1% polyethylene glycol and 0.1% sodium azide in buffer. 0.3 ml antiserum were mixed with 8 ml liquid agarose solution (50°C) and poured into a level plastic dish (Behringwerke). After the gel had solidified the dishes were covered and stored in the fridge at 4°C overnight. The next day 12 antigen wells of 4 mm diameter were cut into the gel layer and filled with 20 ~1 of sample each, 6 holes were used for the calibration standards and 6 holes for the serum samples diluted 1 : 100. The diffusion process lasted for three days at room temperature after which the diameter of the precipitate was measured with a magnified scale. The estimation of the protein concentrations as well as the statistical analyses were performed on a programmed calculator (Wang 600). Two different calibration systems were used on each diffusion-dish: (a) Three dilutions of a standard human serum for Ig-k and Ig-1 (Nr. 974, Behringwerke) with a defined concentration of Ig-k and Ig-1 in the undiluted serum of 1000 U/l. The dilution of 1 : 200,l : 100 and 1 : 51 corresponded to 5, 10 and 19.6 U/l Ig-k and Ig-1, respectively. (b) Three dilutions of Bence-Jones protein control serum, type k and type 1 (Behringwerke). Concentration of Bence-Jones protein in the undiluted serum 2.5 mg/ml for both Ig-k and Ig-1. The dilutions of 1 : 200, 1 : 51 and 1 : 21 corresponded to 12.5, 49 and 119 mg/l for each. Results 42 of the sera had normal immunoglobulin concentrations, mean values and standard deviation of these are shown in Table I. When calibration is performed TABLE
I
RESULTS
OF
NORMAL
IgG.
Results globulin
are
QUANTITATIVE AND
expressed
type
Bence-Jones __-
THE IgA
A: HS.
IgM
as mean value
IMMUNOGLOBULIN
DETERMINATION
IN
(X)
refers
and to
standard
deviation
calibration
with
(s).
human
Ig-k,
immunoglobulin
serum;
IgG
BJP,
value
type refers
K; Ig-I. immunocalibration
__ 12.49
IgA
g/l
2.57
0.87
IgM
g/l
1.82
0.51
Ig-kHS
U/1
870
Ig-IHS
U/l
870
Ig-kHS
to
with
s
gil
Ig-IHS
WITH
~~. _~__ _. _-__
Ig-kHS
SERA
protein.
r
Quotient
42
CONCENTRATIONS
k/IHS + Ig-IHS (k + 1 = 100) (k + I = 100)
2.43
190 240 1.07
U/I
1740
0.40 350
%
50
7.8
%
49
7.8
lg-kBJP
g/l
5.01
1.44
Ig-lBJP Quotient
g/l
2.92
0.75
1.78
0.55
Ig-kB Jp + Ig-IB Jp
g/l
7.93
1.87
Ig-kgJp (k + 1 = 100) I!Z-lBJp (k + 1 = 100) Ig-kBJp (IgG + A + M = 100) &-lBJp (IgG + A + M = 100) Ig-kBJp + Ig-lBJp (IgG + A + M = 100)
7% 9%
63
7.1
37
7.1
%
30
7.2
%
18
4.5
%
47
9.5
k/lgJp
._
345
n i = 63 s = 7.1
Ig
(k) EJP
I5
n ii = 37 s = 7.1
Fig. 1. Distribution curves of relative Ig-k and Ig-1 concentrations of 42 sera with normal IgG, IgA and IgM concentrations. The individual results are expressed as percent of total light chain concentration (Ig-k + Ig-I = 100%). Calibration with Bence-Jones protein.
TABLE II RESULTS OF THE LINEAR GLOBULIN CONCENTRATION
REGRESSION ANALYSIS, OF ALL 54 SERA
CORRELATIONS
BETWEEN
IMMUNO-
For abbreviations see Table I. r, correlation coefficient; a, y-axis intercept: b, gradient: syx, deviation. x
Ig-kH,y
I&-IS
Y
r
Ig-kBJP Ig-lBJP
0.890
Ig-kHs + IdHS
k-kBJp
%-Is
kflBJP Ig-kB Jp + I&B
IgG IgA IgM
h3-kB Jp + Ig-lBJp
IgG + IgA + IgM
+ Ig-lBJp Jp
h3-kBJp
+ I%+,Jp
k+BJp
+ Ig-lBJp
0.699 0.815 0.718 0.678 0.633 0.551 0.051
0
-0.61 1.10 0.30 0.7 1.51 2.47 5.06 7.84
b
SY.X
6.5 2.1 4.4 1.1 0.4 0.4 1.2 0.2
0.80 0.52 1.28 0.4 1.62 1.71 1.84 2.20
346
o*
5oIg (I),,
DO
Isa
” /dl
Fig. 2. Correlation between results obtained after calibration with standwised human serum (HS) and Bence-Jones protein (BJP) for Ig-k and E-1. yk = -61.1 + 6.5 xk, rk = 0.890: Y1 = 109.6 + 2.1 xl, q = 0.699.
with human serum relative concentrations of Ig-k and Ig-1 are practically equal (Ig-k + Ig-1 = lOO%), the quotient is 1.07. When Bence-Jones protein is used for calibration the results can be expressed as mg/l and 63% of total light chain concentration consists of Ig-k, 37% of Ig-1 (Fig. l), the quotient is 1.78. Only 47% of total immunoglobulins (IgG, IgA and IgM) can be typified by their light chains when values are expressed as mg/l. Table II shows the results of the linear regression analysis on all 54 sera. This demonstrates that there is a highly significant correlation between values
341
. . .
.
/ .
.
/ . .
. .
/
.
.
l
l
.
..
. .
l . . .
.
/
. /*
.
.
.
.
Fig. 3. Correlation between total immunoglobulin concentration concentrations (Ig-k + Ig-1).
(IgG + IgA + IgM) and sum of light chain
Y 20
L
I5 U/I
0
I
1
I
I
L
20
LO
60
60
I
100
I
1
120
X
mg/l
Fig. 4. Relationship between light chain concentrations Ig-k and Ig-1 of the two calibration standards: Standarised Human Serum (y-axis) and Bence-Jones Protein Control (x-axis). The conversion of U/l into mg/l and vice versa is possible according to the following criteria: Ig-k. y = 1.0 + 0.16 x (r = 0.999): Ig-I. Y = -0.34 + 0.31 x (r = 0.997).
348
obtained with both methods of calibration (Fig. 2). A significant correlation also exists between total immunoglobulin concentrations (IgG, IgA and IgM) and total light-chain dependent immunoglobulin concentrations (Ig-k and Ig-1) (Fig. 3). There is no correlation between IgM and light chain concentration. The statistical relationship permits conversion of data from U/l into mg/l and vice versa (Fig. 4). Discussion The occurrence of blocked antigenic sites on the light chains of intact immunoglobulins [4] indicates a comparison of two calibration methods for the quantitative light-chain determination in serum. The approach with BenceJones protein as a calibration standard is derived from Iwashita et al. [5]. The close correlation between the two calibration methods is an indication that the portion of the light chain are antigenic determinants of the “constant” responsible for the quantitative measurement. Reactions with the “hidden determinants” can only be expected with increased concentrations of free light chains as for instance in uremia and multiple myeloma with Bence-Jones protein [6]. In contrast to Fahey [7] we were only able to measure 47% of total immunoglobulins as type k and type 1 on calibration with Bence-Jones protein, this could be attributed to the concentration of the antiserum used in our experiment. On the other hand the relative distribution with 63% for Ig-k and 37% for Ig-1 is in good agreement with the results of this author. Also the k/l quotient with 1.78 is within the range of values quoted from the literature: 1.86 [8,2], 1.68 and 1.91 [9] and 2.0 [lo]. While isolated antibodies can have a defined light-chain type [l] little has been reported of the distribution of Ig-k and Ig-1 concentrations in serum or of any alterations of the k/l ratio. It has been demonstrated that the k/l ratio shows some dependance on age [9]. Skvaril et al. [ 111 have described a wide scatter of values for the k/l ratio in patients with antibody deficiency syndrome and hypergammaglobulinemia whereas other authors have described abnormal findings only in cases of multiple myeloma, macroglobulinemia and lupus erythematosus [2,12]. With quantitation of the results as mg/l (calibration with Bence-Jones protein) a more precise analysis of light chain determined immunoglobulin concentrations seems possible. Acknowledgements The authors thank Miss A. Kliippner for her technical assistance. This research was financially supported by the Deutsche Forschungsgemeinschaft (Dr. Fischer-Bosch Stiftung). References 1 2 3 4
Franklin, E.C. and Fudenberg, H.H. (1964) Arch. Biochem. Biophys. 104.433 McKelvey. E.M. and Fahey, J.L. (1965) J. Clin. Invest. 44,1778 Link, H. and Zettervall. 0. (1970) Clin. EXP. Immunol. 6, 435 Epstein. W.V.. Tan, M. and Gross, D. (1964) Nature 202. 1175
349
5 Iwashita.
H., Grunwald,
6 Epstein, W.V..
Gulyassy,
F. and Bauer. H. (1974)
J. Neural.
P.F., Tan. M. and Rae, A.I. (1968)
7 Fahey, J.L. (1963) J. Immunol. 91.438 8 Fahey. J.L. and McKelvey, E.M. (1965)
J. lmmunol.
207. 45 Ann. Int. Med. 68.48
94, 84
9 Skvaril, F.. Barandun. S., Morel& A., Kuffer, F. and Probst. M. (1976) Blut 33. 281 10 Von Drach, J. (1975) Untersuchunpen iiber Kappa-Lambda-Relationen und Nachweis freier L-Ketten im Liquor cerebrospinalis unter besonderer Beriicksichtigung der Multiplen Sklerose. Dissertation. Gottingen 11 Skvaril, F., Barandun, S., Morell, A., Kuffer, F. and Probst. M. (1975) Prot. Biol. Fluids. 23rd Coll. p. 415, Pergamon Press, New York 12 Epstein, W.V. and Tan, M. (1966) Arthritis Rheum. 9, 713