|
|
|
|||||||
|
|||||||||
(Received for publication, August 3, 1994; and in revised form, November 18, 1994) From the
We have used differential polarization imaging microscopy to
measure the amount and orientation of aligned sickle hemoglobin polymer
in quickly deoxygenated sickle red blood cells. Images of the angular
orientation of the aligned polymer at each point in the cell allowed
for determination of the inclination of individual domains, providing
detailed information regarding the polymerization and elongation of
sickle hemoglobin polymers ex vivo. We found that the number
of aligned polymer domains increased with increasing mean cell
hemoglobin concentration. Sickle and holly leaf-shaped cells contained
single or few domains of aligned polymer, while more compact cells such
as irreversibly sickled cells contained many domains. A new class of
cells was discovered by examination of images of the angular
orientation of aligned polymer, which contained a single central
nucleation site, with growth of polymer occurring outward in all
directions in a spherulite-like domain. The fundamental abnormality of sickle hemoglobin (HbS) ( Hemoglobin
preferentially absorbs light polarized parallel to the plane of the
heme(24) . In sickle hemoglobin polymers, the average
orientation of the hemes is perpendicular to the polymer long
axis(25, 26) . Light polarized perpendicular to the
polymer axis will be absorbed more strongly than light polarized
parallel to it. Utilizing this phenomenon, imaging of the differential
absorption of two light beams polarized at right angles to each other
provides information about the quantity and orientation of aligned
sickle hemoglobin polymer (AHP) in sickle cells. We have developed the
technique of differential polarization imaging microscopy to measure
the amount and distribution of AHP at each point within individual
sickle erythrocytes(7, 27, 28, 29) .
(This method does underestimate the AHP that is aligned parallel to the
optical axis of the microscope because the linear dichroism of AHP
aligned in this way is axially symmetric.) These measurements allow the
calculation of the ratio of AHP to total Hb at each point, thereby
minimizing artifacts induced by light scattering, thus providing a more
accurate measurement of the fraction of AHP than birefringence
dependent methods, which are influenced by light scattering. A thorough
mathematical treatment of this subject has previously been presented by
Kim et al.(30, 31, 32) The AHP
images obtained in our lab are in qualitative agreement with the
findings of Beach et al.(33) using substantially
different instrumentation. It must also be pointed out that our
method measures the quantity of aligned hemoglobin polymer,
where polymer randomly distributed in a given area would not be
detected. In solution studies, the initial rise in light scattering
associated with the onset of polymerization is followed by a later rise
in birefringence(34, 35) . This indicates that the
initial polymer in the given area is partially randomly arranged and
that individual polymer fibers within that area subsequently coalesce
and become more aligned. We allowed a 30-min period following
deoxygenation prior to fixation to assure that as many polymer fibers
became aligned as possible. Our goal is to determine how various
physical parameters affect the amount and distribution of AHP within
individual deoxygenated sickle red blood cells (SRBC). Using this
technique, we demonstrated previously that the amount of AHP within
deoxygenated sickle cells may have had little relationship to the
external cellular morphology but was related to the number of
identifiable individual domains of AHP within the cell(27) . A
specific example of the lack of correlation between amount of AHP and
external cellular morphology is the homokentrocyte, a rare type of
deoxygenated SRBC that has a normal biconcave disc morphology but a
large amount of AHP arranged concentrically within(27) . We
also showed that the speed of deoxygenation has only a minor effect on
the number of AHP domains in SRBC(7) . Since the previous
report(7) , we have extended our morphological classification
system, which is based on AHP arrangement within the cell and other
unique visual characteristics, to include new, more detailed classes.
We also present images that show not only the amount of AHP but also
its orientation at each point within the cell. We utilized this
classification system to assess the effect mean cell hemoglobin
concentration (MCHC) had on the nucleation and distribution of AHP in
intact cells after deoxygenation. Cell populations with increasing
average MCHC, which were prepared by fractionating SRBC by density on
discontinuous Stractan density gradients, showed a concomitant increase
in the average number of AHP domains/cell after deoxygenation. After obtaining informed consent, blood samples from two
subjects from the Northern California Comprehensive Sickle Cell Center
at San Francisco General Hospital were drawn by venipuncture into
heparin. Both subjects were homozygous for sickle cell anemia and had
fetal hemoglobin levels below 6%. All cell manipulations were performed
within 6 h of venipuncture, and fixed cells were imaged within 2 days.
Cells were attached to microscope slides with poly-L-lysine
(DP = 66) prior to imaging to restrict the motion of the fixed
RBC during the course of the measurement.
The array
measures transmitted intensities of two orthogonally polarized beams,
arbitrarily defined as parallel (par) and perpendicular (perp). The
absorption (A) of these beams can be related to the
differences in transmitted intensities divided by their sum as shown
below (for small differential signals) and described
previously(27, 29) .
Combining this measurement with data obtained with each
polarization rotated by 45° allowed for the unambiguous
determination of the relative amount, distribution, and orientation of
AHP in the cells. The images presented here consist of an absorption
image, a differential absorption image, and an angle image. These are
proportional to total Hb, AHP, and the orientation of the AHP,
respectively.
Figure 1:
Images of Hb, AHP, and the angular
orientation of AHP in sickled red blood cells. The lowerright-hand image represents absorption, with
gray scale intensity proportional to total Hb. The lowerleft-hand image is the same field of cells with
gray scale intensity representing the amount of AHP. The gray scale
coding for absorbance of Hb varies from 0 (black) to 0.5 (white) and for the absorbance of AHP from 0 (black)
to 0.05 (white) for most images. In the colored image (topleft), color represents the angular orientation of AHP at
each point within the cell and represents the same field of cells. Cell 5 is labeled in all three images to illustrate the
relationship between them. The color wheel in the upper right is used
to determine the relationship between color and angle. The angular
orientation of a color within the color wheel represents the
orientation of the AHP for any pixels in the angle image with that
color displayed. For example, green areas in the angle image
represent AHP oriented with its long axis left to right and red areas represent AHP oriented top to bottom. This image contains examples of several classes of
cells: 1 is a single-domain cell, 2 is a three-domain
cell, 3 is a multiple domain cell, 4 is a zero-domain
cell, 5 and 6 are central constriction cells, and 7 is a spherulite cell. The angle image of the spherulite cell
clearly shows that each AHP domain is arranged radially from a central
point within the cell and may have resulted from a single nucleation
event.
Figure 2:
Images of
Hb, AHP, and the angular orientation of AHP in sickled red blood cells.
As explained in Fig. 1, the lowerright-hand image represents absorption, and the lowerleft-hand image is the same field of
cells with gray scale intensity representing the amount of AHP. The
colored image (topleft) displays the angular
orientation of AHP at each point within the cell and represents the
same field of cells. Cell 1 is a classic, sickle shaped,
single-domain cell, and cells 2 and 3 are multiple
domain cells. Of note is the continuous AHP domain in cell 1 where, by utilizing the angle image, the AHP can be seen to bend
gently with the curvature of the cell.
Figure 4:
Images representing three of the RBC
densities studied including MCHC of 31.2 (A), 38.7 (B), and 42+ g/dl (C). The cells were separated
by density on discontinuous Stractan gradients. The right-hand images, which represent Hb level, reveal
an increase in hemoglobin concentration from density A to C as indicated by the increase in the average gray level in
the cells (angle images have been omitted for clarity). Cells in A had the largest amount of AHP/cell and contained only a few
domains. The number of domains of AHP increased in B and C as hemoglobin concentration increased. This AHP image in C is displayed at a 2-fold higher intensity relative to the others
to ensure visibility. As a result, the large number of domains may be
difficult to discern as reproduced here.
The color portion of the image (topleft) in Fig. 1and Fig. 2represents the angle of orientation of the AHP at each
point within the cells. These images are of the same field of cells as
the Hb and AHP images. The colorwheel in the upperright can be used to determine the relationship
between color and angle of alignment. The angular orientation of a
color within the color wheel represents the orientation of the AHP for
pixels in the angle image with that color displayed. For example, green areas in the angle image represent AHP oriented with its
long axis left to right and red areas
represent AHP oriented top to bottom. This
allows the determination of the two-dimensional orientation of the AHP
within individual AHP domains. Only pixels with a higher level of AHP
than that measured in control (nonsickle) cells have a color displayed.
This ensures that spurious angles representing edge polarization or
other artifacts are not represented. Depictions of the arrangement
of intracellular AHP for each class of cells are shown in Fig. 3. The lines within the cell represent AHP
segments. Experimentally, the intracellular AHP arrangement of each
cell imaged was deduced from visual inspection of images.
Figure 3:
Depictions of the arrangement of
hemoglobin polymers within sickled RBC. The intracellular AHP
arrangements for the various morphological classes shown here in
cartoon form have been deduced from inspection of images of the cells.
The AHP and angle images provided unambiguous determination of AHP
distribution and orientation for well separated AHP domains. The lines within the cell represent AHP. Distributions are shown
for a classic sickled shaped single-domain cell (A), a single
nucleation site central constriction cell (B), a single
nucleation site spherulite cell (C), a three-domain cell (D), and a multiple domain cell (E).
Two subsets of the
one-three-domain class, central constriction cells and spherulite
cells, appeared to contain only a single nucleation site and exhibited
unusual AHP distributions.
Fig. 4includes representative images
from three of the RBC densities studied with MCHC of 31.2, 38.7, and
42+ g/dl. The top pair of images (31.2 g/dl, A) exhibited
the lowest hemoglobin concentration as indicated by the low gray scale
intensity in the absorbance (right-hand) image and a
relatively large amount of AHP represented by the high gray scale
intensity on the left. In the middle images (B), the
hemoglobin concentration was higher (38.7 g/dl), and the number of
domains had increased significantly. Fig. 4C, with an
MCHC of 42+ g/dl, exhibited the highest Hb absorbance and a large
number of domains. This AHP image of Fig. 4C is
displayed at a 2-fold higher intensity relative to the others, and the
large number of domains may be difficult to discern as reproduced here. The percentage of AHP decreased with increasing cell density when
cells from all classes were averaged (Table 1). This is somewhat
counterintuitive, because with more hemoglobin available for
polymerization, a higher level of AHP would be expected. We believe
that there was an increase in the total amount of AHP with increasing
cell density but that it was masked by the crisscrossing of AHP
domains, reducing the total linear dichroism signal. This was supported
by the presence of the highest %AHP in single-domain cells and the
lowest in myriad domain cells, regardless of cell fraction involved (Table 1). It was also possible that the alignment that occurred
following the initial polymerization phase(34, 35) ,
which is responsible for the sizable dichroic signals we measure, was
not able to proceed fully in the higher density cells. This was
supported by the decrease in the %AHP within each class as MCHC
increased (Table 1). As the number of domains reached the extreme
present in myriad domain cells, the measured signal was nearly
extinguished. %AHP was therefore inversely representative of the number
of domains (except in the zero domain case). Histograms of the
percentage of AHP/cell for each MCHC clearly indicated a shift toward
lower percentages as MCHC was increased (Fig. 5).
Figure 5:
Histograms representing the distribution
of the percentage of AHP per cell at 31.2 g/dl (A), 34.5 g/dl (B), 38.7 g/dl (C), and 42+ g/dl (D).
The average percentage of AHP/cell decreased as density increased.
Cells with %AHP greater than 6% are very scarce in both C and D. The overlapping of the AHP domains, or the inability of the
AHP to align in higher MCHC cells, was responsible for the reduction in
the linear dichroism signal at higher MCHC. This would have limited the
number of multiple or myriad domain cells with a large percentage of
AHP. Cells were separated by density on discontinuous Stractan
gradients.
Another
influence on the length of AHP fibers and the size of domains is the
shear stress to which AHP is subjected during
deoxygenation(40, 41) . We have not yet examined this
variable. Higher MCHC resulted in changes in morphological
presentation. As the average MCHC increased, the percentage of
one-three-domain cells decreased (Fig. 6A).
Conversely, the percentage of multiple domain cells over the same range
in MCHC increased (Fig. 6B). For the most dense
fraction, which contained the largest percentage of irreversibly
sickled cells, there was a smaller percentage of multiple domain cells
compared with the second densest fraction and a larger number of myriad
domain cells. This supports the evidence that the number of polymer
domains increases with increasing MCHC as predicted by supersaturation
behavior(16, 17, 42) .
Figure 6:
The fraction of total cells in the
one-three-domain (A) and multiple (
Intracellular AHP domains probably play an important role in
the flow of sickled RBC through the vasculature(43) . To better
understand the factors regulating domain formation, we measured the
number of AHP domains in fixed, sickled RBC populations with various
intracellular HbS concentrations. In these experiments, SRBC were
separated into narrow density fractions and then deoxygenated in about
4.5 s to approximate conditions in the human circulation. The number of
AHP domains, the amounts of Hb and AHP, and the orientation of AHP were
determined for each cell. The number and arrangement of AHP in the
different density cells provided support that our morphological
classification scheme was based on nucleation and growth
phenomena(7, 37, 43) , as predicted by
current theories(16, 17, 18, 42) . The average number of domains present in fully deoxygenated sickled
SRBC was found to rise in concert with cellular hemoglobin
concentration, while the percentage of Hb existing as AHP decreased.
The large number of domains seen in higher density cells implied that
overlapping of the AHP domains, or the inability of the AHP to align in
higher MCHC cells, was responsible for the reduction in the linear
dichroism signal. This would have limited the number of multiple or
myriad domain cells with a large percentage of aligned HbS. This
reversal in expected distribution was previously calculated and found
to fit the experimental distribution(28) . The supersaturation
theory for polymer nucleation, which describes the fraction of Hb that
is insoluble compared with the total amount of Hb, predicts an increase
in domain number with increasing
MCHC(16, 17, 18, 42) . This theory
describes two distinct polymerization phases: homogeneous nucleation,
where a small number of Hb molecules form a stable cluster that
undergoes subsequent elongation, and heterogeneous nucleation, where
nucleation occurs on the surface of existing polymer fibers.
Heterogeneous nucleation proceeds more rapidly than homogeneous
nucleation and can therefore lead to explosive polymer growth. The
dumbbell-shaped domains of central constriction cells and the radial
domains of spherulite cells may illustrate this process. It appears
that the central constriction point may have been a single homogeneous
nucleation site and that the domains of polymer grew by heterogeneous
nucleation. Simulations by Dou and Ferrone (38) have shown that
this type of 2-fold symmetry predominates initially with radial or
spherulite symmetry dominating at later times. Of the two, central
constriction cells predominated here, probably because the cellular Hb
was exhausted before full spherulite formation could occur. The
striking example of a spherulite cell (cell 7 of Fig. 1) shows that symmetry had begun to approach circular, but
again the cellular Hb pool appeared to have become exhausted before a
full radial domain was formed. Rare examples of spherulite cells were
seen where the radial domain was fully formed (data not shown). Single domain cells exhibited the highest measured %AHP in these
studies partly because a single continuous domain generates a large
signal with no loss of signal from overlap with other domains. The
average %AHP for the 2 densest fractions were similar, but the fraction
of myriad domain cells was higher in the most dense fraction. This
suggested that cells with MCHC above 38.7 g/dl contained a very large
number of AHP domains. The difficulty in determining %AHP accurately
for the denser fractions was exemplified by the small differences in
the %AHP between myriad-domain cells (those that contain greater than
10 AHP domains) and zero-domain cells (Table 1). For this reason,
at higher MCHC, the only observed effect of increasing MCHC (from 38.7
to 42+ g/dl) was an increase in the number of myriad domain cells.
This morphological shift was indicative of an increase in the number of
AHP domains but was not significantly represented by a large change in
the %AHP. The plateau in the percentages of multiple + myriad
domain cells at high MCHC could also be representative of a saturation
in the number of nucleii formed due to limitations in oxygen diffusion
out of the cells(48) . Upon closer inspection of cell 2 in Fig. 2, it is possible to imagine that the bottom and
the two rightmost pairs of domains were each individual central
constrictions with the leftmost quartet either two central
constrictions or a near-spherulite domain. All of these apparent
domains could then have resulted from only three individual nucleation
sites instead of many more. For this reason, our assignment of the
number of domains in cells with many domains may be an overestimation
of the number of nucleation sites. As MCHC increases, the
supersaturation ratio rises as well (up to 70% with fast
deoxygenation(44) ), and this increases the probability of
homogeneous nucleation, generating more individual domains of polymer.
In accordance with this theory, we saw an increase in the number of
domains with increasing MCHC. A rigorous comparison between MCHC and
the number of AHP domains would require a knowledge of the exact number
of domains/cell, which is not possible with this technique. The
persistence of preformed nucleation sites at full oxygenation (7) would also complicate such a calculation. Sickle (Fig. 3A) and holly leaf (Fig. 3B)
shaped cells were found to contain mostly single domains of AHP and are
known to be highly undeformable in the deoxygenated
state(43, 45) . These rigid cells may be important in
vascular occlusion due to their inability to traverse the narrow
confines of the microcirculation. It has been shown that reversibly
sickled cells and other poorly deformable sickle cells are important to
the rheologic impairment that results in vascular
occlusions(43) . Flow studies ex vivo have shown that
vaso-occlusion is initiated by adherence of low density sickle cells to
vascular endothelium and propagated by the trapping of poorly
deformable cells behind these niduses(46) . It is important to
remember that such cells often result from RBC with low intracellular
Hb concentrations and thus long delay times would be predicted. These
cells would not be expected to sickle during the time spent in the
microcirculation(47) . Uncertainties about full reoxygenation
in the lungs and the possibility of preformed nucleation sites existing
upon the cells reentry into the microcirculation complicate such
predictions.
Volume 270,
Number 6,
Issue of February 10, 1995 pp. 2708-2715
©1995 by The American Society for Biochemistry and Molecular Biology, Inc.
ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
FOOTNOTES
ACKNOWLEDGEMENTS
REFERENCES
)responsible for the severe clinical disease sickle cell
anemia is its low solubility when deoxygenated(1) . The
substitution of valine for glutamic acid as the sixth amino acid of
mutant 
-globin chains (2) results in changes
in the structure of the deoxy form of the HbS molecule, which
diminishes its solubility. As sickle erythrocytes give up oxygen,
poorly soluble deoxy HbS within these cells rapidly aggregates and
polymerizes(3, 4) . Deoxygenated HbS polymer within
red cells results in their deformation and
rigidity(5, 6) . Those cells with the greatest degree
of deformation (sickling) contain a large amount of aligned Hb polymer (7) . Although the exact event that initiates microvascular
occlusion remains controversial, the presence in the circulation of
cells containing polymerized Hb contributes to this process and,
thereby, to the resultant episodic painful crises and the chronic and
acute organ damage(8, 9, 10) . Methods of
quantifying HbS polymer that employ centrifugation(11) ,
nuclear magnetic resonance(12, 13, 14) , and
laser photolysis(15, 16, 17, 18, 19, 20, 21) do not
spatially distinguish the angular distribution of intracellular HbS.
Optical methods that rely on birefringence to measure hemoglobin
polymer are subject to artifactual variations related to the quantity
and alignment of polymer(22, 23) .
Stractan Density Separation
Discontinuous Stractan
density gradients (36) were used to separate sickle RBC into
six different density fractions. Five 2.0 ml aliquots of Stractan, with
densities of 1.085, 1.092, 1.101, 1.107, and 1.122 g/ml, were layered
on a cushion of density 1.150 g/ml in 15-ml ultracentrifuge tubes.
Packed RBC, previously washed 4 times in phosphate-buffered saline
(0.14 M NaCl, 5 mM PO
, pH 7.4, 10 mM glucose), were resuspended in five volumes of phosphate-buffered
saline, layered on the Stractan gradient, and centrifuged at 30,000 rpm
for 30 min at 4 °C in an SW40 rotor (Beckman Instruments, Palo
Alto, CA). Successive RBC fractions were harvested from the gradient
using a Pasteur pipette and washed 5 times in phosphate-buffered saline
with 1% bovine serum albumin. RBC from the 1.085-1.092,
1.092-1.101, 1.107-1.122, and 1.122-1.150 g/ml
interfaces were used in the imaging experiments. These fractions had
median MCHC values of 31.2, 34.5, 38.7, and 42+ g/dl(36) .
42+ g/dl represents all cells that had densities greater than or
equal to 42 g/dl. We observed that reversibly sickled cells were the
predominant sickle RBC type in the lightest fraction, a mixture of
reversibly sickled cells and irreversibly sickled cells were found in
the middle densities, and predominantly irreversibly sickled cells were
found in the densest fraction.Deoxygenation
4.5 ml of phosphate-buffered saline
were deoxygenated by passing humidified nitrogen over the solution for
30 min with stirring at 37 °C. 0.5 ml of a solution of
density-separated sickle RBC (50% hematocrit) at room oxygen tension
was injected into the deoxygenated solution, causing quick
deoxygenation of the cells (final hematocrit, 5%). Monitoring of
absorbance changes during separate direct mixing experiments (using
equivalent fractions of deoxygenated buffer and oxygenated cells) in
normal RBC showed that deoxygenation occurred in 4.5 ± 1 s (data
not shown). Deoxygenation in SRBC may occur slightly more rapidly or
more slowly, depending on MCHC(48) . The suspensions were then
gently swirled under nitrogen for 30 min and then fixed by mixing with
an equal volume of deoxygenated 3% glutaraldehyde in 50 mM sodium phosphate (pH 7.4) also at 37 °C, and the fixed cells
were swirled for a further 30 min. This procedure was performed on all
density fractions simultaneously.Imaging Microscope
The microscope employed in
these experiments was described in detail elsewhere(7) .
Briefly, it consists of a Zeiss microscope with strain-free 32
(N.A. = 0.4) Zeiss objectives used to focus the mercury light
source (filtered to 415 nm, 10-nm bandwidth) onto the sample slide and
the transmitted light onto a 1024-site Thompson CSF linear diode array.
A Pockels cell was modulated at 40 Hz to produce alternately
horizontally and vertically polarized light. The sample area imaged by
one photosite was approximately 0.3 0.3 µm. Since the
resolution of the optical system was about 0.6 µm, AHP that curved
significantly in 2 pixels or less would not be completely resolved and
would display a reduced signal. The array was mechanically scanned
across the image plane to produce two-dimensional images.
Morphological Classification
This classification
scheme is based on a scheme presented previously(7) , with the
addition of two new categories. In the images of red blood cells in Fig. 1, Fig. 2, and Fig. 4, there are two black
and white images. The right-hand image represents absorption, and we
used this to calculate the amount of hemoglobin in the cell. The
left-hand image is of the same field of cells but with the gray scale
representing AHP. The gray scale coding for absorbance of Hb varies
from 0 (black) to 0.5 absorbance units (white) and for the absorbance
of AHP from 0 (black) to 0.05 (white) for most images. A domain is
defined as an isolated area within an RBC where AHP exists in a single
continuous area. Visual inspection of the images of AHP determined the
number of domains visible within each cell. We and others previously
have shown that the number of domains of AHP is related directly to the
number of nucleation sites in certain classes of
cells(7, 37) .
One-Three-Domain Cells
These cells exhibit
continuous AHP in one, two, or three distinct domains. Cell 1 in Fig. 1and cell 1 in Fig. 2are examples
of single-domain cells. These cells comprise as much as 48% of the
total cells in the least dense (31.2 g/dl) RBC fraction. The angle
image of the classically sickle shaped cell 1 in Fig. 2shows that the AHP follows the curvature of the cell in a
single continuous domain, as diagrammed in cell A of Fig. 3. Cell 2 of Fig. 1has three distinct
regions of AHP and represents a three-domain cell as diagrammed in cell D of Fig. 3.Central Constriction Cells
These cells (cells 5 and 6 in Fig. 1) had characteristic sickle
cell shapes in the absorption image, but the arrangement of AHP was
unusual. The center of the cell contained very little AHP, but the
amount of AHP increased rapidly in the outer regions along a single
axis to form a dumbbell shape, as diagrammed in Cell B of Fig. 3. The central constriction appears to have been the
initial point of nucleation for AHP growth. This shape has been seen
during solution polymerization of HbS (34) and in simulations
of polymerization(38) . As many as 18% of the total cells in
the least dense fraction fall into this class.Spherulite Cells
These cells also appeared to have
a single central nucleation site with heterogeneous growth of AHP
occurring outward in many directions. They resembled the spherulite
domains created by the deoxygenation of purified Hb
solutions(22, 39) . The growth was often limited to
discreet arms of AHP. Cell 7 in Fig. 1is a spherulite
cell, and the angle image clearly shows that all of the polymer domains
are oriented radially from the center of the cell as diagrammed in cell C of Fig. 3. Approximately 1% of all cells in the least
dense fraction were spherulite cells.Multiple Domain Cells
This class includes cells
with more than three and less than approximately 10 domains. Cells 2 and 3 in Fig. 2are multiple domain cells,
and a typical domain pattern is diagrammed in cell E of Fig. 3. These RBC usually contained a smaller measured amount of
polymer than one-three-domain cells and were often lumpy and
deformed. Four domains was chosen as the cutoff between the
one-three and the multiple domain classes because the number of
domains in cells with three or fewer domains could usually be
determined exactly.Myriad Domain Cells
This class was represented by
cells that contained greater than 10 polymer domains. Careful visual
inspection was used to differentiate between this and the multiple
domain class. This extension of the multiple domain cell class was
chosen to provide a more accurate characterization of the cell
population. The measured quantity of polymer was often small in myriad
domain cells due to extensive crossing of polymer domains or the
inability of the polymer to align (see ''Imaging,``
below). The external morphology of most of these cells was similar to
that of an irreversibly sickled cell. Many of the cells in Fig. 4C are myriad domain cells.Zero Domain Cells
A zero-domain cell was one in
which the linear dichroism image exhibited no discernible polymer (cell 4 in Fig. 1). The small linear dichroism probably
resulted from reflection polarization and changes in the index of
refraction at the edges of the cell. A base line value of 1.8% AHP has
been determined for normal RBC resulting from these phenomena (7) and was similar to the value of 1.7-2.4% AHP found in
the zero-domain class (Table 1). The fraction of zero domain
cells was less than 9% for all cell densities, except for the least
dense fraction, where 20% were zero domain cells. In this fraction, we
assumed that the cells had polymerization delay times so long that none
or only a very small amount of polymer was formed.
Imaging
A total of 444 cells were analyzed, and
cell fields were selected only so that cells were well separated for
accurate analysis. Average MCHC of 31.2, 34.5, 38.7, and 42+ g/dl
were obtained by fractionation in Stractan density gradients. The data
presented were pooled from identical MCHC fractions of the two
subjects, thereby obviating differences in the subjects cell density
profiles. For each cell, percentages of AHP were calculated, and the
number of domains was determined by inspection. Data from each density
fraction were averaged.
) and myriad
(&cjs2113;) domain classes (B) as a function of MCHC. Higher
cell densities resulted in an increase in the number of multiple and
myriad domain cells, and a concomitant decrease in the number of
one-three-domain cells. Cells were separated by density on
discontinuous Stractan gradients and morphological classification was
accomplished by visual inspection of
images.
)
We thank the patients for their generous donations of
blood without which this research would not be possible. We also thank
Nacho Tinoco for support, Sarah Goolsby for excellent technical
assistance, and Dr. Steve Embury for the procurement of the blood
samples and excellent critical reading of the manuscript.
©1995 by The American Society for Biochemistry and Molecular Biology, Inc.
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  O. Galkin, K. Chen, R. L. Nagel, R. E. Hirsch, and P. G. Vekilov Liquid-liquid separation in solutions of normal and sickle cell hemoglobin PNAS, June 25, 2002; 99(13): 8479 - 8483. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| All ASBMB Journals | Molecular and Cellular Proteomics |
| Journal of Lipid Research | ASBMB Today |