Erik Cheng

Dino Redzic
2014-04-24

Characterization of a human recombinant antibody

fragment to be used in a diagnostic test: Evaluation by
SDS-PAGE, NanoDrop, ELISA, MASS-1 and antibody
microarray.

Supervisors:
Mattias Levin
Mattias Brofelth

CDR

Erik Cheng
VH
VL

Linker
His(6)-tag

IgG

scFv

Dino Redzic
2014-04-24
VL
VH

Introduction
The aim of this laboratory exercise was to characterize a human recombinant singlechain fragment variable (scFv). This was done through different common experiment
such as ELISA, SDS-PAGE, MASS-1 and NanoDrop. In the end as an application of
all experiments, Antibody Microarray was used to identify viral antigens in two
different complex samples.
More specifically the scFv F3 fragment will be analyzed, which binds to glycoprotein
B (gB). Below is a picture of the F3 fragment which is an affinity-matured variant of
the intact igG antibody. Understanding F3 fragment plays an important role in todays
society since it can bind and neutralize Cytomegalovirus which is a herpes virus.
Figure 1, IgG and the svFv fragment

Principle behind the SDS-PAGE is to detect the

purity of in our case scFv antibody fragments and compare the differences between
the scFv fragment and an intact antibody (igG). Our protein samples will be reduced
and denatured. Afterwards it will be applied to a polymerized acryl amide gel and run
through an electric current. The charged proteins will therefore wander through the
gel and the distance it wanders depends on proteins size.
The Concentration of a protein production is really important in order to plan the next
steps in a project. NanoDrop is a very common method using a spectrophotometer to
measure the amino acids with aromatic side chains absorbance at 280 nm and with
beers law calculate the concentration. Proteins will have different concentrations of

Erik Cheng
Dino Redzic
2014-04-24
amino acids and to correct this one have to choose a standard for the concentration for
amino acids in the program.
Beers law A=Ebc
Where A is absorbance, E is a protein specific extinction coefficient, b is the path
length of the light and c the concentration.
ELISA is a very common way to determine the concentration in a protein sample.
This experiment uses known concentrations to make a calibration curve and then the
unknown sample can be measured by comparison.
MASS-1 studies show the binding characteristics of protein-protein interaction. By
reflecting light on a gold plate with our sample on it, difference in resonance is
received because of the surface plasmon resonance. A biosensor shows different
binding characteristics depending on the different resonance and then affinity constant
is calculated by the biosensor.
a reaction A+B=AB
d [ AB]
K d[ AB ]=
dt
d [ AB]
K a[ A ][ B]=
dt
kd
K D=
ka
Microarray is a way to detect the differential expression patterns of proteins in a
healthy or affected person. By using fluorescence scanner, one can detect which
antigens are presence in the body and can therefore distinguish affected ones from
healthy ones.

Erik Cheng
Dino Redzic
2014-04-24

Material and methods

All the materials and methods used are well described with certain exceptions in the
compendium Characterization of a human recombinant antibody fragment to be used
in a diagnostic test: Evaluation by SDS-PAGE, Nanodrop, ELISA, MASS-1 and
antibody microarray- Mattias Levin and Mattias Brofelth, spring semester 2015
No duplicates are made in ELISA which resulted in 16 wells whereof eight was the
unknown sample, six positive control and two blanks.

Erik Cheng
Dino Redzic
2014-04-24

Results
SDS-PAGE
From the SDS-PAGE experiment figure 2 was obtained to be able to
determine the purity and estimate the molecular weight of the scFV
F3 fragment. In the figure, the lanes 1 and 7 are molecular weight
markers that are used for comparison, lane 2 is a positive control,
lane 5 is an intact IgG antibody and the rest are samples from the
lab groups. The molecular weights of lane 5 are presented in table 1
in ascending order of weight. For the Z2 sample in lane 6, the
molecular weight is 31kDa.
Lane 5, intact IgG
(kDa)
30
60
100
120
Table 1 Table showing estimated molecular weights of lane 2 (positive control) and lane 6
(intract IgG) from the SDS-PAGE experiment, which can be seen in figure 2,

Figure 2, SDS-PAGE: Lanes 1 and 7 are molecular weight markers (the ladder is shown in
figure 3), lane 2 is a positive control, lane 5 is an intact IgG antibody, and the rest are
group results. Z2 was done by this group.

Erik Cheng
Dino Redzic
2014-04-24

Figure 3 Pageruler showing molecular weight in kDa which is used as a reference in

determining molecular weights for Z2 in the SDS-PAGE experiment.

NanoDrop
To determine the concentration of the sample NanoDrop
measurements were made at wavelength 280nm, with the path
lenght b=1 cm and the protein specific extinction coefficient
1.76 L
=
; absorbance values 0.195, 0.422, 0.381, and 0.414 were
gcm
acquired. The first is an outlier and is neglected because it produces
negative concentration values when it is considered and does not
follow the trend displayed by the others. An average is calculated
g
and using equation 1 was determined to be 0.2305
.
L
ELISA
ELISA was also used to determine the concentration of the antibody
sample. A positive control series was made (second row table 2) as
well as a series for the unknown sample (first row of table 1) and
two blanks in well 7 and 8 on the second row of table 1. Duplicates
were omitted. The average of the two blanks was subtracted from
the absorbance and plotted against the concentration; the graph
can be seen in figure 4. A linear fit was made which yielded the
equation y=8.410 3x +0.015 . The absorbance values of the
unknown sample which fall into the absorbance range of the known
sample are inserted into the equation and the corresponding
concentrations were averaged (0.4710, 0.5610, and 0.3638 for well
5, 6, and 7; 0.1410 for well 8 was excluded because it is significantly
lower than the others and a result of procedure error) which yielded
the final concentration C=0.47 mg/ml.
2,157
0,891

1,899
0,529

1,204
0,391

0,987
0,177

0,565
0,118

0,365
0,079

0,166
0,056

0,089
0,055

Erik Cheng
Dino Redzic
2014-04-24
Table 2 Absorbance values from the ELISA experiment. The first row is the absorbance
values for the unknown sample and the second row is a positive control series. Wells 7 and
8 on the second row are blanks.

Figure 4 Plotting absorbance measured for the known sample

series in the ELISA experiment versus concentration and
showing the resulting equation for the linear fit. It is used in
determining the concentration of the unknown samples.
Microarray
To compare the different expression patterns of antigens in healthy and diseased blood
samples microarray measurements were made; the scFv F3 fragment array scheme
can be seen in figure 5 and the results from the fluorescence measurements can be
seen in figure 6. Two wells on the slide were assigned each group: the left was the
healthy sample and the right the diseased. Notably, the uppermost row was a positive
control as well as the lowermost row; the 11th row was a negative control. The PMT
gain was set to 50% and then at 60%. The positive controls both indicate binding to
an antigen present in all human blood (a-C1q), the negative control is dark which is
also expected (PBS does not bind antigens and shows that there is no unspecific
binding). The scFv spots bind the fluorophore and light up: in the diseased blood with
the cytomegalovirus, this is expected. The 1:1 F3 spot has a concentration of
0.09mg/ml and the rest are derived from this.

Erik Cheng
Dino Redzic
2014-04-24

PC = BSA-Biotin
F3 scFv 1:1
F3 scFv 1:2
F3 scFv 1:4
F3 scFv 1:1
F3 scFv 1:2
F3 scFv 1:4
F3 scFv 1:1
F3 scFv 1:2
F3 scFv 1:4
NC = PBS
PC = a-C1q scFv
Figur 5 The layout of the subarray of spots used in the microarray experiment showing the
two positive controls and one negative control and different antibody fragment spots.

Figur 6 Figures showing result from fluorescence measurements of the healthy (left
subarray) and diseased (right subarray) samples at 50% gain (left figure) and 60% gain
(right figure). Intensities for the different spot rows are written in the tables to the right.

Erik Cheng
Dino Redzic
2014-04-24

MASS-1
Immobilization of the antigen AD2 to the chip has to be done first before loading with
scFv samples. During loading of AD2 to the chip a coupling sensogram was printed
out with the different steps marked out (figure 7).

Figure 7, the coupling sensogram

After loading the chip, preparation of the scFv sample was diluted in a running buffer
to different concentration. This is made so we can get a more precise value of
dissociation constant An affinity sensogram is acquired from the biosensor between
scFv and AD2 (figure 8). The figure below show the different binding characteristics
at different concentrations. An average dissociation constant was calculated by the
biosensor to K D =9.025109 M .

Erik Cheng
Dino Redzic
2014-04-24

Figure 8, the affinity sensogram. K_a = association rate, k_d = dissosiatoin rate, K_D =
dissociation constant and X2 = standard deviation.

Erik Cheng
Dino Redzic
2014-04-24

Discussion
NanoDrop
The first absorbance value was omitted because it was significantly lower than the
others; the deviation is probably due to poor application of the drop onto the pedestal
on the spectrophotometer, which requires some precision on the part of the lab
worker. The reliability of the NanoDrop is reasonable, but one expects it to be less
precise than the ELISA measurements because measurements with a
spectrophotometer detect all potential particles and compounds which may contain
aromatic side chains; one expects there to be some sort of protein or cellular
contaminants because it is derived from a biological system. The handling of the
sample and the keeping of it on ice (higher temperatures make the samples prone to
degradation by proteases) could also affect the results, but the most probable cause of
imprecision is likely the application on and cleaning of the pedestal on the
spectrophotometer.
ELISA
The third absorbance value in table 2, 0.391, for the unknown sample was omitted
because it was an outlier and made the data points not linear: a line fit to a nonlinear
dataset would produce deviation when using the fitted equation or even negative
concentrations within the data range. Only absorbance values for the unknown sample
within the absorbance values of the standard curve were used because the certainty of
the relation that would follow would one extrapolate the relation over a larger range
cannot be assumed. Because ELISA relies on the specific detection of binding
between antigen and antibody, while NanoDrop indiscriminately measures
absorbance, it should be more accurate when determining the concentration. The
dilution steps in ELISA however may be prone to error, especially from inexperience
researchers, but because of the usual addition of duplicates this may be greatly
mitigated. NanoDrop measurements are performed several times so here also operator
error might be greatly mitigated. For these reasons, in principle, ELISA is more
accurate. In our experiments however, duplicates were not included and this makes
the values more prone to dilution errors and deviating data points so the results should
be less reliable than in normal procedures; considering though that one outlier was
removed and the rest of the of the data points were reasonably linear this should not
have too much of an effect. The agreement between the NanoDrop measurements,
0.2305mg/ml, and ELISA, 0.3842mg/ml, is in the same range and confirms that the
measurements are reasonable for the methods.
Microarray
The results from the microarray experiment are in accordance with what is expected
beforehand. The positive control a-C1q binds antigens from all blood and should
therefore fluoresce in the scanner, which it does. BSA-Biotin controls for the binding
of the used Alexa fluorophore and concordantly is seen. The negative control is
PBS and should not bind any antigen, which it does not as can be seen in figure 6; if it
did fluoresce, the results would be hard to trust. The signal intensity that is produced

Erik Cheng
Dino Redzic
2014-04-24
from different concentrations of the F3 sample (and other possible samples and
sample dilutions) should in the ideal case be linear but in reality depends on the PMT
gain, producing different relations depending on which is chosen. The optimal PMT
gain is thus the gain which gives the most linearly dependent signal for a set of
dilutions of a sample and this is true for the values in figure 6 which shows that 50%
gain is preferrable; 60% is further oversaturated at some of the spot rows. Setting a
too low PMT gain produces more noise in relation to signal and is thus not
preferrable; similarily, setting a too high gain oversaturates the spot and the data may
become more unreliable or muddled.
SDS-PAGE
The molecular weight markers in lane 1 and 7 are as expected; furthermore, the group
samples in lanes 3, 4 and 6 are clean looking and do not contain any impurities which
is desirable. Lane 2 is a positive control and confirms that the experiment works.
Overall, the results is exemplary. The 5th lane is the intact IgG antibody which shows
several bands higher up. When denaturing proteins, there is not always complete
fragmentation or even partial fragmentation so the band highest up may be the full,
intact IgG; those under may be the heavy and light chains of the IgG respectively.
There is also probably some fragmentation that is different than the distinct
immunoglobulin parts and may show up in the lower part.
MASS-1
If we take a brief look at our affinity sensogram in figure 8, we can see that all the
different concentrations follow a similar pattern. It starts with an increasing resonance
signal during protein association and followed by decreasing resonance signal in
dissociation. So all the curves follow this logic explanation, but the one with lowest
concentration might be hard to detect the pattern.
The fitting is pretty good and our values on dissociation-, association rate constant
and affinity constant are all within the typical range and therefore reasonable (table 2).
Since the affinity constant/equilibrium dissociation constant can be defined as
division of dissociation rate constant by association rate constant, we can tell that a
low dissociation rate constant is something to aim for. We got an affinity constant of
approximately 10-8 M and that is classified as normal-high affinity considering the
typical range.
Variable
ka
kd
KD

Typical range
10-3-107
10-1 5106
10-5 10-12

Table 3, the typical range for the different association and dissociation rate/constant.

Only looking at the affinity might be deceiving. Since affinity is a measurement of

how fast the compounds associate and dissociate. So two compounds can have the
same affinity but differs a lot in kinetics. For example one compound can have a very
good and fast association but dissociates easily and fast. While the other compound
have a little worse and slower association but have a harder and slower dissociation
which makes it much more effective. So kinetics is very important and we can see in

Erik Cheng
Dino Redzic
2014-04-24
our graph (figure 8) that we have a great balance between the association and
dissociation
.
Conclusion
Considering the different properties that were measured in this laboratory work, we
can clearly see that F3 is a very useful and good fragment to use. Especially the
affinity constant was really good and had some very good kinetics.