Note : Les descriptions sont présentées dans la langue officielle dans laquelle elles ont été soumises.
. 7 ~
The invention relates to a sample preparatory
agent for lipoprotein containing samples and for
immunoassaying an antigen as well as a kit for the
preparation of a sample preparatory agent. Likewise,
the invention relates to a method of determining an
antigen by an immunoassay as well as a method of
reducing or eliminating turbidities in a biological
liquid, e.g., in a blood or plasma sample.
A lipoprotein is a particle that has a core of ;~
hydrophobic lipids (cholesterol, cholesterol ester,
phosphatides, triglycerides) surrounded by a sheath of
polar lipids and apopro-teins A-1, A-2, A-4, s-48, B-
100, C and E. The lipoproteins, on the one hand,
solubilize highly hydrophobic lipids and, on the other
hand, contain signals regulating the transfer of
certain lipids on specific cells and tissues.
Lipoproteins may be subdivided into different
density classes:
Chylomicrons, VLDL (very low density lipoproteins),
IDL (intermediate density lipoproteins), LDL (low
density lipoproteins) and HDL (high densi-ty
lipoproteins).
Depending on the protein portion, lipoproteins in
electrophoresis exhibit different mobilities and are
classified into beta-lipoproteins, pre-beta-
lipoproteins and alpha-lipoproteins. LP-X is a
pathological lipoprotein consisting of phosphatides and
-- 1 --
~.i7~ 3
cholesterol ester in a high por-tion.
Lp(a) is another lipoprotein involved in the
arteriosclerotic process. The structure of Lp(a) may be
derived from the structure of the LDL particle. Lp(a)
is an LDL particle that is connected with the highly -
glycosilated apo-lipoprotein (a) via disulfide bridges.
Lp(a) exhibits both inter- and intra-individual ~ '~
heterogenity in its lipid and protein composi-tion. It
constitutes a genetically determined independent risk
factor for arteriosclerosis.
The determination of lipoproteins in a biological
liquid, such as a blood or plasma sample, has increased
in importance in the course of preventive medical
examinations for determining the risk of heart and
circulatory diseases and also within the scope of
therapies for controlling the lipoprotein content. In
general, lipoproteins are determined by immunoassays.
To this end, antibodies against the lipopro-tein to be
determined are used and the immune comple~es formed are
assayed. Detection of the immune complexes formed in
this case is effected by optical methods, for instance,
by detecting the increase in -the turbidity of a sample.
A sample solution as clear as possible is the
prerequisite for the optical tests required therefor in
order to maximize the sensitivity of the assay.
The turbidity inherent in sera cons-titu-tes a
problem primarily with lipemic or hyperlipoproteinemic
~ 5 7~
sera. For, such turbidity may render impossible the
detection of slight amounts of lipoproteins even at a
low degree of lipemia.
Thus, for instance, the detection of immuno-
precipitates which are measured nephelometrically or
turbidime-trically are considerably dis-turbed by the
turbidity inherent in -the sample material.
The turbidimetrical determination of Lp(a), in
particular, again and again encounters difficulties
because of the size of the particle and its structural
heterogenity, especially with samples having high
inherent turbidities (chylomicrons, triglycerides) and
samples that had been frozen and thawed again. With
such samples, falsely high Lp(a) concentrations are
fre~uently found. Similar difficulties arise with the
nephelometrical determination of Lp(a).
Conse~uently, the elimination of turbidities in a
serum is of extreme importance for clinical analysis,
in particular, for the de-termination of lipoproteins.
A common measure for the turbidity exhibited by a
solution is LSU (light scattering unit). In general, a
solution is considered as -turbld if it has a measured
signal of more than 70 LSU when measured in the
nephelometer a-t 340 nm and a path length of 1 cm.
Measurement of an antigen-antibody reac-tion at 70 LSU
and thereabove is clearly falsified. For clinical
analysis, samples should have a maximum turbidity
~ 7 7 ;~ ~j
preferably ranging between 30 and 40 LSU.
To eliminate disturbing turbidities in llpemic
samples, treatment with lipases or high speed
centrifugation is, for instance, recommended in Clinica
Chimica Acta (135 (1983) 203-208). However, it is
decisively advised against using de-tergents, since ;
these do reduce the turbidity inherent in a sample, yet
also result in a strong reduction of the immunochemical
response. Polyethylene glycol is proposed as a reaction
accelerator for the immunochemical reaction.
In the lyophilization of control sera -turbidities
occur, in particular, with lipemic samples on account
of the procedures of lyophilization and subsequent
reconstitution of the originally clear sera. As a~ -
precautionary measure in the prevention of -turbidities,
EP-0 141 922 proposes the addition of a detergent and
of proline prior to lyophilization of the control
serum. ~he sera are mixed with proline and sodium-
desoxycholate and are lyophilized. After reconstitution
of the lyophilisate it could be seen that the clear
sera had not become turbid. Since it is always departed
from clear sera when stabilizing control sera,
clarification of the solution, i.e., reduction of its
turbidi-ty by detergen-t/proline could not be effected.
EP-0 058 959 also mentions various additions for
the preparation of control sera that are to remain
clear after lyophilization and reconstitution. Organic
- 4 -
: ''
~ ~ ~ 7 ~
non-sugar-like substances, such as methanol, alanine,
triethylene glycol, valine, acetate, lactate or sodium-
2-hydroxymethylbutyrate, for instance, are suggested as
agents for the prevention of turbidities caused by
changes in the state of aggregation. But not even EP-0
058 959 discloses a way of treating a sample -to be
assayed in a manner that undisturbed determination of
the lipoprotein content is ~easible.
The invention has as its object to provide a
technology by which, on the one hand, possibly present
turbidies inherent in biological liquids can be reduced
or eliminated and, on the other hand, analysis of an
antigen in a sample as unfalsified as possible is
feasible.
In accordance with the invention, this object is
achieved by an agent whi.ch, for sample treatment,
contains an amino acid reducing or eliminating
turbid.ity in a sample while laaving unaffected the
immunochemical determination of an antigen, preferably
of a lipoprotein, by aid of an antibody contained in
the agent and directed against the antigen to be
determined. The agent according to -the invention
advantageously contains a reaction accelerator, for
instance, polyethylene glycol. The amino acid is
present in a buffered medium, preferably at a
concentration of from 0.05 to 3 M, most preferred from
0.1 to 0.5 M. I-t has been shown that proline, above ~.
- 5 - .
~ -~ i 7 7 ~
all, is excellently suited to eliminate the turbidity
in a lipemic or turbid sample irrespec-tive of the
pretreatment of the sample.
A preferred embodiment of -the agent contains
antibodies directed against a lipoprotein and is used
for determining this lipoprotein by an immunoassay.
For instance, the agent of the invention may be
used for assaying Lp(a), the agent in such a case
containing Lp(a) antibodies, preferably an Lp(a)
antiserum. The agent of the invention can be prepared
in a simple manner by modification of commercially
available reagents or reagent components.
The agent according to the invention enables the
undisturbed determination preferably of lipoproteins in
a sample irrespective of whether the sample has been
previously deepfrozen, stored in the cooled state or
freshly drawn. As such a sample, a biological liquid,
such as a blood, serum or plasma sample, which is
optically turhid due to its lipoprotein content may be
particularly applied.
Surprisingly, the agent according to the invention
substantially increases the sensitivity of the
determination of, for instance, Lp(a), wherein it can
simultaneously be proved by a reference method that the
results obtained are of high relevance. The addition of
proline to a sample preparatory agent clearly lowers
the blank value of a serum sample and increases the
7 7 ~
nephelometrically determined difference in absorbance
from the time of the addition of antibody.
The advantageous effect of amino acids, in
particular, of proline, in the agent according to the
invention is the more surprising as amino acids
generally have been proposed in connection with
detergents only for preventing -the formation of
turbidity in a lyophilized control serum, yet such
admixtures (detergents) to a sample are considered as
disturbing in an immunoassay. Moreover, such samples
are not lyophilized, whence there has not been any need
to prevent the formation of turbidities during
lyophilization and reconstitution by such admixtures.
It is advantageous also with the agent according to
the invention that the sample is not admixed with a
tenside in order to render feasible -the undisturbed
immunochemical determina~ion of an antigen. The agent
according to the invention or the kit for preparing
such agent as well as the components o such a kit
advantageously are substantially f ee of tensides.
Solutions are regarded as "substantially" free of
tensides if they have a tenside conten-t of less than
about 0.01 %.
An imunoassay in a solution having a tenside
content of more than 0.05 % is clearly falsified by the
tenside such that the measured results obtained are no
longer relevant.
- 7 -
. "' .
~ ~ . 1 7 J j
It has been shown that it is advantageous to
provide the agent according to the invention as a kit
of at least two components. Thus, the amino acid and,
if desired, a reaction accelerator may be made
available in the first component in the presence of
buffer substances, the second component containing
antibodies directed against the antigen to be assayed.
According to a further aspect, the invention
relates to a method of determining an antigen by an
immunoassay, wherein an amino acid is added to a sample
proposed for determining by an immunoassay, antibodies
agains-t the an-tigen to be assayed are introduced into
the sample subsequently or simultaneously and the ;~
amount of antibody-antigen complex formed is
determined.
Another aspect of the invention relates to a method
of reducing or eliminating turbidities in a turbid
biological liquid, which method is characterized in
that an amino acid is added to said -turbid biological
liquid. Thus, a method of treating a sample liquid is
provided at the same time.
With the methods according to the invention,
proline preferably is used as the amino acid.
The methods according -to the invention preferably
are employed with blood or plasma samples as well as
with turbid lipoprotein containing samples of an
antigen provided for determining by an immunoassay.
~ ~ '3" ~ 7 ~, ,r,3
Another aspect of the invention is related to the
use of an amino acid, in particular, proline for
reducing or eliminating turbidities in a biological
liquid, preferably in a turbid lipoprotein containing
sample of an antigen provided for determining by an
immunoassay, as well as the use of an amino acid, in
particular, proline, for treating lipoprotein
containing samples, preferably in a sample provided for
determining an antigen by an immunoassay.
The invention will be explained in even more detail ~
by the following examples. The examples illustrate the -
determination of Lp(a) in serum samples by aid of
appropriately modified commercially available
individual reagents of Immuno AG according to a valid
working prescription (anti-human Lp(a) of sheep, ~-
reference standard Lp(a) human, norm control Lp(a)
human). The difference in absorbance is detected by aid
of a photometer or an automatic analyzer, in the
examples indicated by aid of a COBAMSIRA device
(Hoffmann-La Roche).
Exam~e 1: Absorbance course of calibration curve
The absorbance was measured, and a calibration
curve was established, of a series of Lp(a) solutions
having different concentrations of Lp(a). To one Lp(a)
containing sample dilution buffer containing 1 x PBS (1
x P8S: 10 mM Na/K phosphate, 0.8 % NaCl, 0.2 % HCl, pH
7.2) and 4 % PEG with or without addition of O.1 mol/l
.,'. : :.
~ 7 7 .~ ~
proline and, after 5 min, an Lp( a) antibody solution
were added. In Fig. lA the absorbance course of the
calibration curve with -the addition of 0.1 mol/l
proline in dilution medium and in Fig. lB the same
course without addition are represented.
As is apparent, the calibrator having -the highest
Lp(a) concentration, i.e., 108 mg/dl, has an absorbance
of 0.09 RATES as compared to 0.125 RATES of the same
calibrator in dilution medium with 0.1 mol/l proline. ~-
By the addition of proline, the reaction signal is, :
thus, intensified. RATES in this case correspond to the
measured signal of the antigen-antibody reaction after
subtraction of the measured signal that occurred upon
admixture of dilution buffer as compared -to the
original solution.
Furthermore, it is apparent from the course of the
calibration curve that the calibration curve, without
addition in the dilution medium, flattens towards -~
higher aoncentrations (Fig. lB), whereas the curve
shown in Fig. lA has a linear course.
E~am~le 2: Lp(a) determination with pre-incubation
A serum sample having a high inherent -turbidity is
incubated for 5 min with dilution buffer containing 0.1
mol/l proline. After this, Lp(a) an-tibodies are added.
The absorbance is measured at a wave length of 340 nm
and plotted against -the -time unit (Fig. 2A). The same
assay was carried out also without (Fig. 2B) addition
- 10-
i 7 7 !1 ;3
of proline in dilu-tion buffer. From this, it is
apparent that the absorbance corresponding to the blank
value of the serum sample could be lowered from 0.36
(without addition) to 0.24 (with proline).
The absorbance difference from the time of addition
of the antibodies in the reaction course of the mixture
with proline (Fig. 2A) is greater by 200 ~ than in the
mixture without proline (Fig. 2s). Hence an Lp(a)
concentration of 103.3 mg/dl in the mixture with
proline, yet, at the same time, a falsely low value of
19.2 mg/dl in the mixture without proline are
calcula-ted. In general, an Lp(a) concentration of below
30 is considered as normal, values thereabove being
regarded as pathological. In the instant case, the
sample, thus, would have been regarded as "normal"
(19.2 mg/dl) on grounds of conventional measuring
methods, although the correct concen-tra-tion (determined
by the reerence method) clearly constitutes a
pathological value.
Example 3: Lp(a) determination with pre-incubation
Analogous to Example 2, the Lp(a) content o a
sample was turbidimetrically determined by the addition
of Lp(a) antibodies. The same reaction course as in
:, , .
Figs. 2A and 2B is represented in Figs. 3A and 3B with ~;
the calibrator of the highest Lp( a) concentra-tion (=
108 mg/dl). In this example, the reaction signal
(without addition, Example 3B) could be intensified by
,~
~ ~ i ' 7 ~ ~3
~ .
50 ~ by the addition of 0.1 mol/l proline.
Example 4: Correlation between fresh and thawed
samples
As mentioned above, discrepancies occurred in the
turbidime-trical determination of Lp(a) with sera that
had been frozen and rethawed as compared to freshly
tested samples because of the turbidity of the thawed ~;~
sample material.
The improvement obtained by the admixture of
proline to the agent is demonstrated by the following
example:
A panel of 25 serum samples was aliquoted, one
portion having been assayed on the same day the blood
had been drawn, the other portion having been frozen.
The latter aliquot was thawed on the next day and
assayed afterwards.
The resul-ts of the assays without proline are
represented in Fi~. 4A. The correlation between the
freshly tested and the thawed samples was R = 0.908 and
the linear equation was y = 1.01 + 0.88x. In a parallel
assay, the thawed samples with proline were determined.
Comparison to the freshly -tested samples is illustrated ~ -
in Fig. 4B. The correlation was improved to R = 0.972,
also the linear equation with y= 3.52 -~ l.O9x showed a
marked improvemen-t in the consistency with the ideal
values having the gradient 1.
Example 5: Lp(a) determination by electroimmuno-
- 12 -
~'
' 7 -~
diffusion
Electroimmunodiffusion serves as a reference method
of immunologic Lp(a) determination.
The frozen and thawed samples used in Example 4
were tested in a parallel assay also in electro-
immunodiffusion. (Method according to Laurell, Anal.
Biochem. 15 (1966), 45).
In Fig. 5A, the resul-ts of the turbidimetric
determination without proline are compared to those
determined by elec-troimmunodiffusion. The correlation
was R = 0.918 and the linear equation for the
elec-troimmunodiffusion curve was y = 1.95 + 0.54x. (y =
1.01 + 0.88x for turbidimetric determina-tion: Fig. 4A).
Thus, there is a distinct difference between the two
methods in respect of the results obtained. `
By adding proline, not only the correlation
coefficient could be raised to R = 0.989, but also the
linear equation y = 3.34 + 0.92x implied an almost
perfect consistency of the methods (Fig. 5B); y = 3.52
~ l.O9x for the turbidimitrical determination in the
presence of proline, Fig. 4B).
~"'~ `'~'"
- 13 - ~