вход по аккаунту


Molecular Aspects of Sickle Cell Disease.

код для вставкиСкачать
Molecular Aspects of Sickle Cell Disease
By Michael R. Waterman and G. Larry Cottam[*]
Sickle cell hemoglobin (hemoglobin S) contains valine instead of glutamic acid in position
6 of the 0-chain. Few detectable conformational differences exist between hemoglobin S and
normal adult hemoglobin (hemoglobin A). Following loss of oxygen, sickle cell hemoglobin
(hemoglobin S) self associates to form a liquid crystal phase which distorts the erythrocyte
into the sickle shape thereby resulting in the clinical symptomology associated with sickle
cell anemia. This gel contains deoxyhemoglobin S monomers and polymers in equilibrium.
The polymerization process is known to have a negative temperature coefficient, to be pH
dependent, and to be extremely dependent on hemoglobin S concentration. The polymerization
of deoxyhemoglobin S appears to be entropically driven and occurs in two kinetic phases,
a delay period and a rapid polymerization process. The polymer consists of tubules containing
six or eight strands of deoxyhemoglobin S tetramers which align with one another. Each
strand is wound around the tubule with a pitch of about 3 W A , but the precise nature
of the intermolecular hemoglobin S contacts is not known. Subsequent alignment of the tubules
occurs and results in the tactoids observed in (S/S) erythrocytes. While many of the details
of polymer formation and structure remain to be elucidated, several attempts to chemically
alter the sickling phenomenon have been carried out. As yet, however, no satisfactory chemical
treatment has been discovered.
1. Introduction
Sickle cell anemia was first reported in 1910 by Herrick['],
who described a patient that had
red blood cells. In an extensive study Nee!['] demonstrated
that sickle cell disease is inherited in a Mendelian fashion,
as previously suggested by Huckr3].Nee/ also explained the
existence of clinically benign individuals (sickle cell trait) as
being heterozygous for hemoglobin S and individuals with
sickle cell anemia as being homozygous for hemoglobin S.
However, the first chemical description of the disease and
the fact that it was the result of an abnormal hemoglobin
molecule awaited the application of electrophoretic techniques,
with which Pading et a!. in 1949 succeeded in separating
the sickle hemoglobin from the normal
In 1957, Ingram applied the techniques of peptide mapping
to determine the precise chemical differences between hemoglobin A and hemoglobin S. Normal adult hemoglobin (hemoglobin A) contains four polypeptide chains which consist of
two identical a-chains of 141 amino acid residues and two
identical 0-chains which contain 146 amino acid residues
Ingram was able to demonstrate that hemoglobin
S contained a single amino acid substitution different from
hemoglobin A. In the P-chain, the amino acid valine has
been substituted for glutamic acid six residues from the amino
terminus151.Except for this one amino acid, all other amino
acids in the a- and 0-chains of hemoglobin S are identical
to those found in hemoglobin A, Thus, hemoglobin S contains
two normal a-chains in addition to two abnormal P-chains
and is written as C I ~ P ' $ ~ ' " +
with sickle cell
trait have hemoglobin S and hemoglobin A, while individuals
with sickle cell anemia have only hemoglobin S.
The clinical manifestations of sickle cell anemia, i. e. the
anemia and the acute ischemia and infarction of tissues and,
therefore, the chronic loss of organ function, are a consequence
of the cell sickling phenomenon. This phenomenon, in turn,
has been shown to result from the polynierization of deoxy-
['I and Prof. Dr. G. L. Cottam
Department of Biochemistry,
Southwestern Medical School
The University of Texas Health Science Center
Dallas, Texas 75235 (USA)
Author to whom correspondence should be addressed.
[*] Prof. Dr. M. R. Waterman
[ '3
Angew. Chem. Inr. Ed. Engl. 1 Vol. 15 (1976) No. 12
Fig. I . Microscopic appearance of sickled erythrocytes. Oxygen was removed
from these erythrocytes chemically, using a 1 % solution of sodium metabisulfite.
hemoglobin S molecules[*]inside the erythrocyte. As the hemoglobin S molecules give up oxygen to surrounding tissues,
they aggregate into large fibers which distort the erythrocyte
membrane into the sickle shape (Fig. I). The polymerization
manifests itself as gel formation. This is in contrast to normal
adult hemoglobin which, upon deoxygenation, remains in solution. Sickled cells are less pliable than normal red cells and
tend to become trapped in capillary beds, resulting in an
occlusion of blood flow and ultimately tissue damage. An
enormous amount of clinical literature has appeared on sickle
cell anemia over the years and is beyond the scope of this
article; however, the reader's attention is drawn to a few
recent reviews which have been written on the subject['- 'I.
The number of studies on the biochemical and biophysical
aspects of sickle cell disease has increased explosively in recent
years, particularly in the United States. It is the purpose of
this article to review the most pertinent aspects of these works.
Due to the great complexity of the polymerization process,
much of the work has been phenomenological in nature. However, it is clear that a consistent pattern is developing from
several different types of experimental approaches and an
understanding of the fundamental structural, kinetic and equilibrium properties of this complicated interaction between
protein molecules is emerging. It is hoped that from this
understanding a useful clinical regimen will arise for the treatment of this molecular disease.
2. Early Studies of Gelation
Following the elucidation of the primary sequence of hemoglobin S, relatively few biochemical studies were undertaken
to determine the nature of the aggregation process until the
present decade. Since 1970, the aggregation process has been
studied extensively, using numerous physical methods to unravel its complex nature. Some crucial observations made
between 1950 and 1970, however, provided a basis for the
recent experiments.
The first important observation was made by Harris; he
found that the viscosity of red cell hemolysates from individuals
with sickle cell anemia increases markedly upon deoxygenation. Upon lysing the erythrocytes, he was able to show a
concentration-dependent viscosity change following oxygen
removal. At high concentrations of hemoglobin S (235 mg/ml),
he observed a semi-solid gel-like state which had an optical
birefringence similar to that observed in erythrocytes[' 'I. The
gelation, which can be made reversible by the reintroduction
ofoxygenfIZ1,occurs even in the presence of other hemoglobins
such as hemoglobin A['31. However, a minimal amount of
hemoglobin S must be present for gelation to occur (cf. also
Table 5).The diagnostic significance of the gelation for identification of hemoglobin S has also been considered.
Another important observation was made by Ifano. He
showed that deoxyhemoglobin S dissolves much more sparingly in 2.24 M phosphate buffer at pH 7.0 than does deoxyhemoglobin A['41. The reduced solubility of deoxyhemoglobin
S in concentrated salt solution has been exploited by several
investigators in studying the parameters of the polymerization
process[' "I. In addition, use has been made of this observa~~
The designation deoxyhemoglobin shall be used here whenever it should
be emphasized that deoxygenated hemoglobin is involved.
tion in thedevelopment ofa solubility method for the detection
of hemoglobin S in mass screening programs.
In 1957, Allison reinvestigated the reported viscosity changes
and showed that oxyhemoglobins S and A and deoxyhemoglobin A all had the same viscosity characteristics over a concentration range of 5 to 250 mg/ml and a temperature range
of 0 to 40°C. Deoxyhemoglobin S, on the other hand, behaved
quite differently. At concentrations in excess of 120 mg/ml,
a sharp increase in viscosity was observed as the temperature
of the solution was increased above about 15°C. On cooling,
the viscosity decreased again['81.The reversibility of the polymerization process by lowering the temperature is an extremely
important property which has been exploited in virtually every
study since its discovery. Conversely, at any set temperature
above ca. 15"C, there is a sharp increase in viscosity as the
hemoglobin S concentration is increased (see Table I).
Tdble I Specific viscosities q.p of aqueous solutions of deoxyhemoglobins
A dnd S ds a function of concentration c at 37°C [IS]
Deoxyheinoglobin A
[ms mil
Deoxyhemoglobin S
43 5
A valuable contribution made by Muragama was the concept
that the negative temperature coefficient of gelation is an
indication of the importance of hydrophobic interactions in
polymer formation["]. This is consistent with an amino acid
substitution of valine for glutamic acid, and several attempts
to design anti-sickling compounds have been based on this
concept. In summary, these early studies established that gelation is a highly concerted process dependent on temperature,
hemoglobin concentration and oxygen tension.
3. Differences between Hemoglobin S and A
Perhaps the most frustrating aspect of the study of the
aggregation of deoxyhemoglobin S is the fact that the physical
properties of the hemoglobin S molecules in the polymer
appear not to differ from those of the molecules in solution.
Careful studies, including the use of polarized absorption
spectroscopy[201,have shown that the optical properties of
the heme group do not change upon polymerization. Other
portions of the molecule have been probed by use of nitroxide
spin-labels, and no conformational changes could be observed
upon polymerization. The p93 cysteine residues have been
used as a site of attachment for spin-labels in hemoglobin
A to monitor by electron spin resonance the conformational
changes which take place during reversible ligand binding[z'.221.When spin-labels were attached to this site in hemoglobin S, the same oxy-deoxy transition as with hemoglobin
A could be observed, but no change was observed in going
from the deoxy-solution to deoxy-gel state[231.This result
is illustrated in Fig. 2. It follows that the region around
the p g 3 cysteine residues in deoxyhemoglobin S is not perturbed upon polymerization and, therefore, that some portions
of the surface of the protein remain unaltered. Preliminary
A I I ~ P I VChrrri.
l i i t . E d . Eiigl.
1 Vol. 1 5
( 1 9 7 6 ) No. I2
3235 G
Fig. 2. Electron spin resonance spectra of deoxyhemoglobin S spin-labeled
( l i at the Pg3
with 3-(2-~odoacetamido)-~,2,5.5-tetramethyl-l-pyrrolidinyloxyl
cysteine residues. a ) Sample at 53 mgjml (not a gell. b) sample at 283 m g m l
( a gel). The spectra were recorded at 23°C.
studies in which 3-{[2-(2-isothiocyanatoethoxy)ethyl]carbamoyl}-2,2,5,5-tetramethyl-l-pyrrolidinyloxyl (2) was employed for the spin labeling indicate that the N-terminal
residues of the 0-chains in hemoglobin S are also insensitive
to changes which may occur upon
of this p-chain was then isolated by affinity chromatography.
Cross reactivity with the 0-chain of hemoglobin A or hemoglobin r-chains was not observed[33'. This method has potential
utility in both the detection and quantification of hemoglobin
S, as well as in the study of conformational differences between
hemoglobin A and hemoglobin S.
The function of hemoglobin is sensitive to its conformation
and thus might provide a tool to distinguish small differences
between the two hemoglobins. However, the oxygen binding
parameters of dilute solutions of hemoglobin S are indistinguishable from those of hemoglobin A[291. Differences
between oxygen binding parameters of hemoglobin S and
hemoglobin A have been observed in whole blood[30. ' 1 . Blood
containing hemoglobin S is found to have a significantly lower
oxygen affinity; this is most likely due to the aggregation
of deoxyhemoglobin S which tends to stabilize the T (deoxy)
configuration, thereby shifting the equilibrium curve to the
right. The clinical implications of the relationship between
the extent of this shift and various red cell parameters such
as volume, hemoglobin concentration and 2,3-diphosphoglycerate concentration are not completely clear. The shifting
of the equilibrium curve may well have important consequences with respect to the severity of the disease.
4. Kinetic and Thermodynamic Studies
enthalpy change in the polymerization process has been estimated to be relatively small, a result which also indicates
no large conformational change in the deoxyhemoglobin S
molecules during g e l a t i ~ n [*'I.~ ~Needless
to say, this inability
to find aproperty of the hemoglobin S molecule which changes
upon polymerization has complicated the study of this process.
With respect to differences between hemoglobin S and
hemoglobin A, two interesting observations have recently been
made. A S U ~ ~et~ Y( I /N. observed that oxyhemoglobin S is very
much less stable toward mechanical shaking than is oxyhemoglobin A[*". This instability is unique for the oxygenated form
and, therefore, strongly suggests conformational differences
between oxyhemoglobin S and oxyhemoglobin A. Using highresolution 'H-NMR spectroscopy, H o e f t i / . have detected
small pK differences in three surface histidyl residues between
deoxyhemoglobin S and deoxyhemoglobin A['*]. The largest
of these differences is found to be 0.2 pK unit. The assignment
of these residues is not complete, but it is clear that these
results reflect localized conformational differences on the surface of the deoxyhemoglobin S molecule with respect to deoxyhemoglobin A. Therefore, it must be concluded that some
structural differences exist between hemoglobin S and hemoglobin A in solution, but these differences are subtle and thus
far have not been useful as probes for the study of the polymerization process. To date, it has been necessary to take advantage
of the changes in the physical characteristics of the solution
to study polymerization.
The techniques of immunology may also lead to a better
understanding of slight conformational differences between
hemoglobin S and hemoglobin A. Recently, the 0-chain of
hemoglobin S has been used to elicit
An antibody population with specificity for the N-terminal region
The kinetics of the sol-gel transition have recently been
studied by a variety of methods. Widely different techniques
such as viscosity r n e a ~ u r e m e n t s l351,
measurement of birefringen~e'~'],t~rbidity'~'',and nuclear
magnetic resonance water line
as well as transverse
water proton relaxation
have all shown that there
is a delay period prior to the onset of the polymerization
process on warming a solution of deoxyhemoglobin S from
4°C to ambient temperature or higher. The gel formation
itself follows the delay period and is rapid. The length of
the delay period is found to be dependent on the hemoglobin
Table 2. Summary of the kinetic delay times id in the polymerization of
deoxyhemoglobin S iis measured by different techniques.
Method. or
property measured
Optical birefringence
Water proton
NMR linewidths
Transverse water proton
relaxation times
Hemoglobin S
[mg m l l
~ 5 1
concentration and the temperature to which the system is
warmed. The delay time for the polymerization process has
been expressed as :
1 / t d = y S"
where fd is the delay time, S is the supersaturation ratio
(total hemoglobin S concentration divided by the minimum
concentration for gelation) while t7 and y are parameters independent of the experimental conditions[361.The value of the
constant y is IO-"s. The value of I I represents the number
of monomers in the nucleation species and is high, at least
30. A summary of the delay times observed by different
methods and under different conditions is given in Table
It is possible that different techniques are monitoring different aspects of the overall polymerization process, but a significant delay time is seen in all cases. The length of the delay
period is temperature dependent and apparently the reciprocal
of the delay time increases logarithmically with increasing
t e m p e r a t ~ r e ' ~On
~ ] . comparing the temperature dependence
of the delay time as determined from optical birefringence,
turbidity and reciprocal amplitude of the water proton magnetic resonance line, it was found that all three techniques gave
similar results[381.The length of thedelay period is also reciprocally related to a very high power ( 2 30) of the hemoglobin
concentration. E u m et a / . therefore proposed a model in
which the rate determining step is the formation of thermodynamically unstable nuclei of approximately 30 deoxyhemoglobin S molecules["]. The delay time has important implications clinically and will be discussed later (in Section 7).
From a kinetic standpoint the self association of deoxyhemoglobin S molecules is strongly dependent on hemoglobin
concentration and temperature and probably other parameters
such as pH. The concentration dependence seems to be manifested in two ways, one being a delay time before the onset
of polymerization and the other being the rate at which
polymer formation proceeds. Figure 3 shows the water proton
t [rntnl-
Fig. 3. Transverrc water proton relaxation times (7,)a s ii function of time
at two hernoglobin S concentrations (a: 172nig:ml: b: 20Xmg ml). The time
zero TZvalues were measured ;it 4°C and all other T2 \,dues bere measured
at 37°C. The pH i n both samples was 7.0: hemoglobin S w a s maintained
in the deoxy form by addition of N a 2 S 2 0 L [39].
relaxation times ( T2)as a function of time in the polymerization
of deoxyhemoglobin S. It can be seen that the delay time
rd decreases greatly with increasing hemoglobin concentration
(1 72 mg/ml: 20min, 208 mg/ml : 5 min), and the rate of polymer
formation increases. It is clear that this concentration dependence is not hemoglobin S specific. In Figure 4 it is seen
that by adding 40 mg/ml hemoglobin A to a solution of
170 mg/ml hemoglobin S, the delay time is similarly decreased
and the rate of polymerization is increased. The results resemble those obtained with a solution of 208 mg/ml hemoglobin
S (Fig. 3b)[391. The more subtle aspects of the kinetics of
hemoglobin S polymerization remain to be elucidated. In
particular, investigations on solutions in which hemoglobin
S is present in the same concentrations as in erythrocytes
have yet to be carried out.
a -
t Iminl-
Fig. 4. Traiwerse hater proton relaxiitioii times (71) as a function of time
i n a solution of hemoglobin containing 170mg ml hemoglobin S and 40
my ml hemoglobin A. Other conditiona were i i s described for Figure 3 [39].
As pointed out above, the size of the nucleation species
has been estimated to contain at least 30 hemoglobin molecules[2n1.Studies on the equilibrium properties of the deoxyhemoglobin S system have been largely concerned with the
question of whether small aggregates of deoxyhemoglobin
S, a few hemoglobin molecules in size, are present in the
sol-gel transition. From equilibrium centrifugation studies.
contradictory data has been obtained. Observations with a
schlieren optical system showed a pre-gelation transition zone
which could be ascribed to intermediate~[~"I.
On the other
hand, observations with a Rayleigh interferometer provided
no evidence for aggregates with less than 20 hemoglobin rnolecules; hence it is concluded that the polymerization process
is a very concerted
Light scattering studies have indicated that pre-gelation aggregates may exist in solution and
that the polymerization process is linear, rather than con~erted[~*].
From the measurement of spin-lattice relaxation rates of
solvent water protons, it has been estimated that about 90 '2)
of the deoxyhemoglobin S molecules remain in solution at
10°C while at 35°C virtually all are incorporated into the
Centrifugation studies indicate that the gel consists
of deoxyhemoglobin S monomers in equilibrium with a solid
polymer phase[2h.441.
Measurements oflongitudinal water proton relaxation times (TI)have suggested that interactions
between oxyhemoglobin S molecules may occur within erythr o c y t e ~ ~Determination
of transverse water proton relaxation times (T,) has been useful in the study of the deoxyhemoglobin S aggregation process and has recently shed some light
on the controversy surrounding pre-gelation aggregates[46-491.
In this technique. the T2 value is monitored as a function
of time (see Fig. 3 and 4). The technique is applicable to
both intact erythrocytes and isolated hemoglobin
I t has been used, for example, to study the influence of pH'481
and potential anti-sickling agents[501on the polymerization
process. In order to explain the observed drop in T2 the
three water environment model was proposed14": free water,
water of hydration, and irrotationally bound water. Table
3 summarizes the model in terms of the mol fractions of
water in each environment and its characteristic correlation
Table 3. Three Miller en\ironmcnts inside iiitiict erythrocytes [47].
Water tnvironment
Mol Fraction
Correlation Time
L [h]
Free Water ( B u l k Water)
Water or H q d f i i t i o n
Irrotation;illy Bound Water
0.013 0.015
3 x 10"
2 4x10
2 10
time T ~ .Upon polymerization the correlation time of the
“irrotationally” bound water fraction increases and gives rise
to the observed decrease in relaxation time T2. Thus the
reduced T2 value is a function of the degree of polymerization
in the deoxyhemoglobin S solution.
One must now ask how high the degree of polymerization
must be in order to yield the observed changes in T’? This
has been approached by assuming a linear polymer model
and calculating the expected correlation times for comparison
with those actually observed in solutions of deoxyhemoglobin
S[491.The correlation time of oxyhemoglobin S is 2.23 x lo-@s,
while that observed for polymerized deoxyhemoglobin S is
2 . 6 4 ~~ O - ’ S [ ~It~ can
~ . be seen from Table 4 that a linear
Table 4. Calculated correlation times for linear polymerization of deoxyhemoglobin S [49]. Calculations were made assuming hemoglobin a s a sphere
of 64A diameter. The degree of association gives the number of molecules
in the linear polymer and is also the axial ratio.
Degree of Association
In summary, the aggregation of deoxyhemoglobin S is a
concentration- and temperature-dependent process consisting
of several discernable steps whose driving force is the increase
in entropy. Kinetically, there is a delay period before the
polymerization begins. During this delay period there appears
to be no formation of stable small polymers. In contrast,
however, stable small polymers apparently are formed during
the aggregation phase of the overall process, and they serve
as precursors to the nucleation species which is larger than
30 hemoglobin molecules. In the light of what is known concerning the structure of the polymer described in the next
section, there must be kinetic processes such as the alignment
of tubules which remain to be sorted out from the overall
5. Structure of Crystals and Fibers
71 x 10-8
5.2ox 10-8
9 . 1 4 ~I O - ”
1 . 3 5 ~l o - ‘
1.81 x 10.’
2.26x lo-’
2.75 x 1 0 - 7
association of from one to seven hemoglobin S molecules
can yield the range of observed correlation times. Thus, the
measurement of transverse water proton relaxation rates is
sensitive to the initial events of polymerization. This result
is supported by the fact that with S/S erythrocytes, an 11-value
of 2.5 is obtained when the change in T2 is measured as
a function of oxyhemoglobin concentration and expressed
as a Hill plot. Therefore, in this special case the deoxygenation
of hemoglobin can be measured by this
the onset and rate of polymerization are very concentration
dependent, it may be assumed that intermediate states do
exist and an understanding of their formation and disappearance is essential to an overall understanding of the polymerization process.
Thermodynamic measurements d o not rule out the existence
of small polymers. The enthalpy of polymerization is 2-4
k ~ a l / m o l [*‘I~ ~and
. the increase in entropy of about 25 cal/K mol-’ at 20°C is considered to be the driving force for the
reaction[2h1.This entropy change is smaller than described
for some other biological polymerization processes and is
compatible with the existence of stable, small polymers[261.
Mirirori has proposed a two-step thermodynamic model
for the gelation of hemoglobin S[” -541. In this model, deoxyhemoglobin S molecules exist in the monomer form until
a critical concentration is attained. Above this concentration
rod-like microtubules form by a process which is thermodynamically equivalent to precipitation. The rod-like aggregates
then spontaneously align to form the gel when their volume
fraction exceeds a critical value. The critical volume fraction
is sensitive to the length of the rod-like forms in solution;
the longer the rod length, the lower the volume fraction necessary for alignment. This model is reported to be consistent
with experimental observations and has been applied to both
the chemical inhibition of gelation and the influence of oxyhemoglobin S on the gelation process.
Considerable interest has been focused on the structure
of the deoxyhemoglobin S polymer. It is anticipated that
by understanding the fiber structure, and in particular the
intermolecular amino acid contacts, it might be possible to
design a specific anti-sickling agent. Considerable detail concerning the fiber structure has been obtained from electron
microscopy (Fig. 5). This technique has shown that deoxyhemoglobin S in erythrocytes and in isolated solution forms
tubules having an outside diameter of about 170A[”. s s . s h J .
These studies show that fibers are formed by alignment of
the tubules with each other. Early studies by F i ~ i c hrr r r / . L s 7 1
and Edelsteiri rt ~ 1 / . [produced
evidence that the tubule
is made up of rings of six hemoglobin tetramers stacked
on top of each other with a helical pitch to the arrangement
of rings. In the two studies noted above, there were differences
in the helical pitch and the alignment of tubules, but the
model stemming from these studies showed that there are
three kinds of intermolecular hemoglobin contacts which result
in fiber formation. These are up-down within the tubule, resulting in the long axis of the tubule; side-side within the tubule,
resulting in the ring structure of the tubule: and interactions
between the tubules, resulting in alignment and therefore fiber
formation. Recent electron microscopy studies suggested that
the tubules are eight-stranded (Fig. 6), rather than six-stranded,
and that the adjacent strands in the tubule long axis are
staggered with respect to each other by half the repeat distance[”].
It is not known at present which amino acids are involved
in these interactions; however, some information is available
concerning contact points between deoxyhemoglobin S molecules. X-ray diffraction patterns have been obtained from
deoxyhemoglobin S gels[‘’’ and recently the 5 A and 3 A X-ray
structures of deoxyhemoglobin S crystals have been deterIt is interesting that much of the data from
the 3A X-ray study is in good agreement with the electron
microscopic data and other information known about fiber
structure. For example, in 1973 Hoj+chrer. rr crl. established,
by measuring absorption of polarized light, the maximum
angle (23”) which could exist between the tubular axis and
the molecular s
In the crystal, this angle is less than
10”. In addition, several of the known sites which affect the
gelation process are found to be intermolecular contacts in
the crystals. For example, one of the p6 residues is involved
in the side-side contacts, as is one of the p 7 3 residues. In
Fig. 5. Transmission electron micrograph of hemolysate from S/S erythrocytes upon deoxygenation showing both
the longitudinal and transverse sections of the fibers from Ref. [69].
hemoglobin C&lern, there is an Asp-Asn substitution at 873
in addition to the Glu-+Val substitution at position 8 6 . This
results in an increased concentration necessary for gelation
to occur[63]. Several other sites which have been suspected
of being involved in intermolecular contacts from studies with
abnormal hemoglobins are also implicated in the crystallographic data. Furthermore, the structure of deoxyhemoglobin
S does not differ significantly from that of deoxyhemoglobin
A in the p 6 region. From crystallographic data a model tubule
can be constructed which has many similarities with the eightstranded model described above. It is also suggested that
the tubule consists of double strands in which the molecules
are staggered by half the repeat distance[62!
Further studies on the mapping of the intermolecular contact points have been carried out using hybrid hemoglobins
consisting of mutant a-chains and hemoglobin S P-chains[l71.
In these studies, the solubilities of the hybrid hemoglobins
have been measured in order to assess the influence of the
a-chain abnormalities on the polymerization process. The
results appear to be consistent with those derived from X-ray
crystallography and electron microscopy. In addition, they
have been found, in part, to be consistent with the computer
modeling approach to the determination of the intermolecular
While many details of the deoxyhemoglobin S fiber remain
to be determined, it is clear that the fiber consists of tubules
made up of strands of deoxyhemoglobin S molecules. The
contacts between molecules and between strands probably
do not involve many amino acids so that much of the surface
of the hemoglobin molecule is free to the solvent. The forces
which result in tubule formation do not change the measurable
properties of the hemoglobin molecules. However, the tubular
fibers which are formed are sufficiently stable thermodynamically to force the erythrocyte membrane into the sickled shape.
6. Non-Gelling Hemoglobins
Fig. 6. Model of eight-stranded deoxyhemoglobin S fiber along the longitudinal
axis, as suggested from X-ray and electron microscopy studies [59, 621.
Each circle represents a single molecule of deoxyhemoglobin S.
It is well recognized that while homozygosity for hemoglobin
S (sickle cell anemia) is often associated with severe clinical
symptomology, heterozygosity for hemoglobin S (sickle cell
Angew. Chem. Int. Ed. Engl.
1 Vol. 15 (1976) No.
trait) is a clinically benign state. Therefore, non-gelling hemoglobins, such as hemoglobins A and F, have been extensively
studied for their effect on the polymerization process['3,
Expanding on the original work of Singer and Singer[13],
Bookchin and Nagel have developed a widely-used technique
for this type of study. In these experiments, the minimum
gelation concentration (MGC) is measured under a specific
set of conditions (0.15 M phosphate buffer, pH 7.35, 25"C),
the MGC being the concentration of hemoglobin below which
no gelation will occur. By using a constant set of conditions,
it is possible to study not only the effect ofnon-gelling hemoglobins on the MGC, but also potential anti-sickling agents
and mutations in the hemoglobin S molecule at positions
other than (j6.
Table 5 Effect of non-gelling hemoglobins ( H b ) on the MGC [60 611 of
deoxyhemoglobin S T h e normal hernoglobin concentrdtion in the erythrocytes is 320 t o 340 mg/ml
Hemoglobin composition
50 ",, HbS
353 -366
355- 361
[a] HbCH,,,,m contdins two substitutions in the P-chain
235 - 248
+ 50 ",, HbA
5 0 " ! HbS + 50% H b F
IOO",,HbCH,,l,m [d]
50 HbS + 50",, H b Korle-Bu [b]
- -
MCC [mgzml]
[b] Hb Korle-Bu contdins one substitution in the P-Chain (p%'*''')
In Table 5, the effect of non-gelling hemoglobins on the
MGC of hemoglobin S is shown. As can be seen, both hemoglobins A and F substantially increase the MGC, hemoglobin
F being more effective than hemoglobin A. In addition, the
substitution of Asp+Asn at p 7 3 also significantly increases
the MGC. An equal amount of hemoglobin A raises the
MGC to approximately the hemoglobin concentration inside
the erythrocyte, while hemoglobin F raises this concentration
above that found in erythrocytes. Therefore, while cells which
contain hemoglobins A and S in approximately equal concentration are known to sickle upon complete deoxygenation
in uitro, they probably never experience a low enough oxygen
concentration in uiuo to sickle. Hence, the normal sickle trait
hemoglobin profile (45 % hemoglobin S and 55 % hemoglobin
A) is a benign condition clinically. Another point to be drawn
from Table 5 is that an amino acid substitution far from
the Ph position can have a profound influence on the MGC.
This is due to the multiple amino acid contacts which are
necessary for fiber formation.
The mechanism by which non-gelling hemoglobins change
the MGC is unknown. However, hybrid tetramers (cr,pAps)
have been observed in mixtures of hemoglobin A and hemoglobin S[671,and it has been shown experimentally that under
conditions where hybrid tetramers can form, the MGC is
increased, but not to the level observed under conditions
where hybrid tetramers are not present[681.Therefore, non-gelling hemoglobins may be incorporated into the fiber structure
in two ways, as complete tetramers and as hybrid tetramers;
the latter combination results in an increase in the MGC
but still permits gelation to occur1691.Certainly, the possible
role of hybrid tetramers in gel formation in hemoglobin mixtures deserves further consideration.
However, the effect of non-gelling hemoglobins increasing
the MGC is not sufficient to explain the clinical differences
Anqew. Cliiwi. 1111. Ed.
Nil. I 5 ( I Y 7 6 ) No. 12
between sickle cell anemia and sickle cell trait. The hemoglobin
concentration in normal erythrocytes ranges between 320 and
340 mg/ml. From Table 5 it is clear that 50% hemoglobin
F instead of 50 % hemoglobin A would be very advantageous,
since it raises the MGC above the normal red cell hemoglobin
concentration. The occurrence of 50 % hemoglobin F, however, is extremely rare. On the other hand, 50 % hemoglobin
A also raises the MGC, but not above the normal red cell
hemoglobin concentration. It is estimated that between 8 and
10 % of the black population in the United States have about
45 % hemoglobin S and 55 % hemoglobin A (sickle cell trait),
and the erythrocytes from these individuals are known to
sickle following complete removal of oxygen. Therefore,
hemoglobin A must have another role in addition to raising
the MGC.
It has recently been shown using the measurement of water
proton relaxation times, that polymerization occurs immediately and concomitantly with deoxygenation in S/S erythrocyte~[~'].
In heterozygous erythrocytes (A/S or C/S) there is
a delay in the onset of polymerization until a critical level
of deoxyhemoglobin is achieved. The critical level of deoxyhemoglobin is dependent on the amount of hemoglobin S present
in the erythrocytes, the more hemoglobin S the lower the
level of deoxyhemoglobin necessary for the onset of polymerization.
Therefore, non-gelling hemoglobins play at least two important roles in the polymerization process which must, in part,
be responsible for the clinically benign state of sickle cell
trait. They increase the MGC and they increase the level
of deoxygenation which must be achieved before the onset
of the polymerization process. It is also anticipated that nongelling hemoglobins will prove to alter the kinetics of the
polymerization process in a favorable manner, i. e., slow down
the rate of polymerization.
7. Chemical Alteration of Gelation
An important goal in the study of the molecular aspects
of deoxyhemoglobin S polymerization is to find a chemical
solution to this serious and widespread disease. In view of
what is known about polymerization, several different compounds have been studied as potential anti-sickling agents.
The chemicals studied can be classified into two categories;
those which covalently modify hemoglobin S and those which
non-covalently alter the polymerization process. Examples
of the non-covalent modifiers are
organic solvent^^^'],
a l k y l ~ r e a s [ ' ~amino
~ , acids such as lysine, arginine, and aspartic
and 2,2-dimethyl-3,4-dihydro-2H-benzo[c,d]pyrene-6-butyric acid (3)[741.While these compounds, which do
C H, 13-C OOH
not form covalent derivatives of hemoglobin S, represent various degrees of potency as anti-sickling agents in uitro, there
is no conclusive evidence that any of them reduce the clinical
symptoms associated with the disease. In several cases, the
toxicity of the compound has yet to be determined.
The covalent modifiers which have been studied most extensively are ~ y a n a t e " ~771, carbamyl phosphate17'l and the
bifunctional cross-linking agent, dimethyladipimidate / 4 / [ " ] .
While the toxicity of the latter t w o compounds has yet to
be clearly established, the studies with cyanate are extensive
and illustrate the two major problems of covalent modification: it is difficult to covalently modify only hemoglobin iri
riro, and moreover to do so without impairing the oxygen
delivery function of hemoglobin. Therefore one may want
to modify only a few of the total number of amino acids
ofa specific type. For instance, in cyanate treatment of erythrocytes, cyanate was found to be incorporated to a large degree
into the hemoglobin, predominately at the amino terminal
valine residues. Yet a small amount of cyanate is bound to
other proteins resulting in deleterious clinical side effects,
particularly in the nervous system and the eye. As recently
discussed by M t ~ r i r i i r i g ~many
~ " ~ , principles have been derived
from studies on the carbamylation of hemoglobin S which
should also apply to other studies where covalent modification
of hemoglobin is attempted. Carbamylation of hemoglobin
S occurs primarily at the amino terminal valine residues with
the r-chain residues reacting faster than P-chain residues.
The anti-sickling effect of cyanate seems to result from two
quite different effects. Modification at the r-chain amino terminal residues produces a hemoglobin which has a higher
affinity for oxygen and therefore polymerization of this modified hemoglobin Swill only occur at lower oxygen tensionsr8'!
On the other hand, carbamylation of the p-chain amino terminal residues significantly increases the MGC of hemoglobin
S181'. Therefore, a covalent modification can have multiple
effects on the polymerization process.
To date, no chemical modifier of the polymerization process
has been found which would be useful in the general clinical
treatment of sickle cell anemia. This is not to say, however,
that such a compound does not exist. and the search for
an agent to specifically inhibit or slow down the gelation
phenomenon will continue at an accelerated rate as more
information becomes available about the biochemical aspects
of the process.
The above mentioned compounds have been studied because
they have the potential to prevent cell sickling under the
conditions of reduced oxygen tension. Errtori et ti/. have recently
presented the interesting hypothesis that one approach to
reducing the clinical severity of sickle cell anemia is to control
the delay time observed in kinetic
It is possible
that the delay time could be lengthened by some treatment
such that it would become longer than the time required
for red cell transit through the blood circulation ( :
15 s).
Under these circumstances, reoxygenation of hemoglobin S
would serve as the anti-sickling process. This hypothesis opens
up fruitful areas of both clinical and basic research as to
the role of the delay time in the clinical severity of the disease
and the perturbations of the hemoglobin environment which
are effective in lengthening the delay time.
The ultimate answer to the problem of sickle cell anemia
may lie in the elucidation of the mechanism by which the
switch from fetal hemoglobin (hemoglobin F) synthesis to
adult hemoglobin synthesis takes place. Since the capacity
to synthesize fetal hemoglobin is present in adults, i t may
be possible some day to turn on synthesis of this hemoglobin,
thereby producing heterozygous S/F trait individuals from
homozygous SjS individuals. Since sickle cell trait is a clinically
benign condition, considerable effort will be invested in this
direction in future years.
8. Conclusion
The authors hope that this rather selective review will provide chemists with an insight into the chemical events which
are responsible for the clinical symptoms associated with sickle
cell anemia. In addition to the problems remaining to be
solved, it is hoped that the reader has gained an appreciation
for the vigor with which these problems are being tackled
in many different laboratories. It is not overly optimistic to
anticipate that the structural, kinetic and equilibrium properties
of the self association of deoxyhemoglobin S will be much
better understood by the end of the present decade and that
a chemical solution to the problem will be in sight. In fact
it is likely that the treatment of the disease will go through
several different phases, each arising because of new advances
in the areas of biochemical and biophysical aspects of hemoglobin S.
The ~ r i i t l i o ricish
to u c k r i o ~ ~ l e r ltlie
~ e sirpport of Coritruct
N o . NOI-HB-2-2954 h!. the Sickle Cell D ~ . Y ~ uB. w
r r / / i ~ /01
the Nrrtioiitrl Hetrrt trriil Liiiig I r i s f i t i i r r , Nutioritrl 1ri.stifirte.Y
of' Hetrl/li. O w of' ( i s ( M R W ) trlso rrckrioicletlyes the s u p p o r t
of' Resrtrrch Grurit ROI-AMl6188,frorii t h e Nrrtioriirl Iiistitutes
of Health. n i e kiridriess of' Dr. Joliri F . Bertles iri prorirliiig
photoyrrrph irserl in Fi~girre5 is qreutl!. trpprvciutetl.
Receiicd. J u n e 15. 1976 [A 139 IE]
\ersioii: Ansc\r. Chem. ,YyS. X 2 I 119761
J . B I l w r i d . . Arch. Intern. Med. 6. 517 II9IOl.
J . I : V w l . Science 110. 64 1 1949).
J . 6 / / r i c k . Bull. J o h n s Hopkin\ Hosp. 3 4 . 335 (1923).
I. l J m i l i i i ~ ~fI. .A l i m o , S. J Siitcqer. and 1 . C . Wdl\. Science 1/11.
543 [ I949 ).
I.: \I /iii/rmt, Nnture /,YO. 376 11957).
I). ( 1 . .Viiiliiiit and I / A . P c ~ i r \Iiernato11,py
oflnkinc! and Childhood.
S;titnders. f'hil;tdelphia 1974. p. 419.
I.. 1. D. K ~ J i i ~ ~ r ~ , ~ - - 1 Arch.
l i ~ i l i il.i i t e r i i . Med. 1.13. 61 I 11974).
S. Cliwwlic,. Arch Intern. Med. 13.1. 69X 11974).
P. F. \ / i / i i i , r . Clinics iii Haematol. 3. 7x9 119741.
R. D. L c v c m : Sickle c'cll Hemoglobin and Other Hemoelohinopathies.
Academic Press. Neu Y o r k 1975.
.I. W! 111irri\. Proc. Soc. Exp. Btol. Med. 7.5, 197 (19501.
K . S i i i ~ / r i and
I . CItwiio//. Blood 7. 47 (1952).
K ~ S f i i q i vand L . .S,iiq~r,Blood 8. IOOX ( l Y S 3 ) .
H 4. / i m o . Arch. Biocliem. Biophy\. 47. I4X 11953).
6 . /.. C i i i i i i i i i snd \I. R . Wiiiwiim. Biochem. Biophys. Res. Cornmuti.
54. I157 11973).
R. B i w w It. R. E. BwvwIi.and S. Iiiii(1. Biochem. Biophys. Re\. Cornmuti.
55. 261 I 19731.
R. E. Bcwwli. S . liiiig. R B v i i w i l i . J . V d . and R. C. Sdtiieirlrr.
Nature 2611. 219 119761.
4. C. 4 I l i w i i . Biochem. .I.65. 212 (1957).
i f . \ / i i r i i ~ ~ iScience
153. 145 11966).
J . Ho/ricItrw. I). G. Hiwdri~kiv.; t i i d U: -1. € i i i , i i t . Proc. Natl. Acad.
Sci. I:SA 70. 3604 11973).
J. ( ' 1. B i w i w \ and H .\I. . M ~ C ~ i i t i i dProc.
Natl. Acad. Sci. U S A
5% 22 1 I96hl.
S. O q m u and H . M. \ / ~ C ~ J i i i i dProc.
Natl. Acad. Sci. U S A 58. 19
.\I. R . CWilwiiiiiii. K f i i i ~ t ~ i o L and
~ i . G.
Coiiiiiit. Bioclicm. Biophys.
Res. Coinmiin. 5s. 1058 (19741.
L/. R . I~V~ii~~riiiiii~.
K. ? m u o h i i . ;ind G I. Cofr[iiit Proceedings o l the
First N;ttion;tl Symposiiiiii oii Sickle Cell Disease. D H E W Pub. No.
75.723. Belliesd;~ 1974. p. 121.
P. D. R ~ J \ \J. . f i o / r i d i i e r . and U: 4. Eliioii. I . Mol. Biol. 96. 239 (1975)
B ~ / ~ i ~ / ~ l ~ ~ / / - ~CC:~ i .V.
i r i Poilloii.
~ l i i l ~ l T-1.
Li. ;tnd .I. F. Brrrlc\. Proc.
N a t l . Acad Sci U S A 7.1. 990 119761.
T ~4dtirii.T Oliit;\lii. S. F r i e d i i i ~ .and E S d m ~ i r i : .Proc. Natl Acad.
Sci lJSA 7 1 . I504 119741.
[?XI ('. f f 0 . L. kf'-.\f. ~ l l l i ~ / . C . 1.lll. G'. s ~ l l / ~ l l l \ kalld
K. J . b ! - ' ~ ~ ~ l / l l ~ / J J l l l l l .
I'roceedings of t l i e Sbniposium on the Molccu1;ir ;tiid C'ellular Aspects
of Sickle Cell Disease D H E W P u b No. 76-1007. Dallas 1976. p 65.
[19] I . Riq+ a n d .\I. MPI/\. Biochini. Biophys. Acta 50. 243 11961 I.
[30] .hi. S I ~ ; J I . \C:!. ,&. Gi/h\. P / . Mi/rrc~r. a n d J . F. Bwi/r,\. J. C'lin.
Iii\ 57. 412 (19731.
[ill R. CI. L ~ ' ~ J J $ / , I I \ . Proceedings of tlie Symposium o n the Molecular end
Cellular Aspects ofSickle Cell Disease. D H E W Pub. N o 76-1007. I)alla.;
1'176. p. 232.
[31] .I (; ( ' i d . D. /.idiiiq. and 1. ,V. .Si/ii411cr..I. Hiol. Chem. 251. 12x3
I I97(ll.
[ 3 3 ] ./. G. Cr~rd.,\. S. ~OIIII~. and 4 . N .SI/IIV/IICT.
_I. Biol. Cheni. 7 5 / .
I290 1 1976)
[34] R. \Id/11and J . Sr~,i11/111ri/r.
Biochem. Biophys. Res. Comiiiuii. 51). X87
[ 3 5 ] J M. H<II.I.I\ and H 5. BI~II,\IIUIII. J . Lab. C'llti. Med. 86. 564 11975).
[36] J. lfo/rld!rcv. P. 11. R<J\*.nnd H: 1. EUI,JII. Proc. Natl. Acad. Sci.
LISA 7 1 . 4x64 11974)
f Q. I f , (;I/IWII. Biochcm. Biophys. Rcs. C-ommun. Ol.
[37] K . . \ l o / / ~ ~;itid
237 119741.
I! I EUIOII. J . t f o / r i < / i i ~ vP.
. D. Rm\. R. (;. Tw/zrrdirr. and E D.
H c r k , ~ ~ .Btochem
Biophys. Rcs. C o m m u n OY. 53X 11976).
1.793 \.I R. Whrcvrrrrrri and C . L Corrirrii. Biochem. Biophys. Res. C o m m u n .
in press.
[40] K . It! Hrrdrl and S. Eivrr.1. _1 Mol. B1ol. KO. 445 I 1973).
. Proc. Natl. Acad. Sci. USA 71). 1506 11973).
[41] R C W r //i~ ir ii~
[42] i t : I4 bi'il\orr. 11. R. 1.1r::oml. J . 7. /'~~IIII~~~~III.
and C. S. J~I/IIIVIII.
Jr.. Proc. Natl. Acad. Sol. USA 71. 1160 119741
14.71 7. R ~ ~ i l i d ~ 1 l ' and
~ J J i SJ . H . K,wiiiq. J. M a p Reaoti. 15. 344 119741.
14.11 R. W Bridil Proceedingc of the Symposium o n the Mo1ecul;ir and
Cellular Aspects of Sickle Cell Disease. D H E W Pub. No. 76-1007.
Dallas 1976. p. 145.
[45] 7: R 1.iird~Iriiiir. S . H . Kiiciiicg. 7: 5~111\~io.~.
and J I - . B e r i l ~ ,pri\ate
communication ( 1976).
1461 (;. 1.. C 01111111. h'. 21. ~ > I / u I ~ ~ I JK.
c ~ . k ~ i J i l l l ~ l / ~and
~ l . .M.R. W~rl~~rrrii~ri.
Arch Biochem Biophys. / 6 2 . 4x7 119741.
J I I .R . W i r ~ ~ ~ r iaiid
i i ~ ~G.
~ iL.
Arch. Biochcm.
1471 5. C. ~ ' / I ~ ~ I I / ~ ~ ~.\I.
Hiophys. / 6 6 . I93 11975).
~. .Wir/wiiiuii. K ~ i i r ~ r i i ~ and
~ l ~ ~G.
r . 1.. Corirmi. Arch.
14x1 4 I f . ~ ~ / I I I ~M.
Biochem. Biophys. 167. 145 11975).
[49] G. L. Cortir~iiand M. R . W~II~~IIIIIII. Arch. Biochem. Biophys 177.
293 119761.
.A IfC/iiiirriq. and ( i . 1. ( . I , I I C I J I I . Hiochem.
[ 5 0 ] 41. R W~I~WIII~III. K . ~IIIII~I~I~~I.
Biophys. Res. Cornmiin. 63. 5x0 119751.
1511 4 . P. ~\!~III~III. _I. Mol. B i d XI. 4x3 (19741.
[52] 4 1'. 1fiii1011.J. Mol. B i d Y5. 2x9 (lY75).
[53] . A . 1'. Lliririrrr. Hiophys. Chem. 1. 3x7 (19741.
the S!mposium on the Molccuiar and
[54] 4. 1'. ~ ~ I I I ~ O I I I'roceedinp.;
Cellular Aspects of Sickle Cell Disease. D H E W P u b No. 76-1007,
Dallas 1976. p. 257.
[55] C. .4. S f m u i i . J Esp. Med. 122. 341 (19661.
C 0 M M U N I CAT1 0 N S
Conformational Studies on Twelve-Membered Heterocycles. Crystal Structure of 5,18-Dimethyl-5,18-diazatribenzo[ a,e,i]cyclododecene-6,17( 5H,18H)-dione
By M! David Ollis, Julia Stephuriidou Stephtrriutou, J . Fraser
Stoddart, Andrew Quick, Doriald Rogers, and David J . WilIiams['l
The synthesis and conformational behavior of the N,N',N"trimethy1"'and N,N',N"-tribenzyl['' derivatives of trianthranilide have been described recently. We now discuss the confor-
Prof. Dr. W. I).
0111s ['I.
Dr. J. F. Stoddart. and Mrs. J. S. Siephanatou
Department of Chemistry. T h e University
Shefield S 3 7 H F (England)
Prof. Dr. D. Rogers. Dr. A . Quick. a n d D. J. Williams ['I
Chemical Crystallography Laboratory, Imperial College
L o n d o n SW7 ZAY (England1
['IAuthors to w h o m correspondence should be addressed.
Без категории
Размер файла
1 050 Кб
sickle, molecular, aspects, disease, cells
Пожаловаться на содержимое документа