Structure and Cytoskeleton of Circulating Platelets
Megakaryocyte development culminates with the release of mature discoid platelets
with average dimensions of 3.0 × 5.0 μm. The plasma membrane of the resting platelet
is replete with transmembrane receptors. As much as 20% of the surface area is occupied
with receptors, which include ligands for thrombin, IgG (FcγRIIA), and the major surface
glycoproteins GPαIIbβ3 and GPIb/IX/V. The cytoplasm of platelets contains two types of secretory granules:
dense- and α- granules. The disc shape of the platelet is maintained by a very unique
and specialized cytoskeleton, which functions as the molecular struts and girders
of the cell.[1] The main components of the platelet cytoskeleton are a marginal microtubule coiled
repeatedly into a band, a spectrin-based membrane skeleton, and a three-dimensional
network of cross-linked actin filaments.
The marginal band of resting platelets is located at the cell periphery and consists
of multiple microtubules, coiled up 8 to 12 times.[2] Microtubules are long, hollow polymers formed from tubulin dimers that provide the
force for movement in many cellular functions, including mitotic chromosome segregation
as well as organelle translocation. Platelets and megakaryocytes contain primarily
the β-1 isoform of tubulin. β-1 Tubulin is the most highly divergent β tubulin isoform
and is hematopoietic specific. Analysis of β-1 tubulin knockout mice points to a critical
cellular function for β-1 tubulin in platelet biogenesis. These mice exhibit a bleeding
phenotype, have reduced platelets, and contain defective microtubule coils.[3] Consistent with these observations, a mutation in the human β-1 tubulin gene has
been identified, and analysis has revealed large spherical platelets with a microtubule
coil that is highly disorganized. Platelets also contain the motor proteins cytoplasmic
dynein and kinesin, which can bind to microtubules and exert force. The generation
of the microtubule coil is a critical event in the final stages of platelet production,
and the mechanisms of its construction are discussed in more detail later in this
review.
An elaborate network of spectrin and associated proteins laminates the underside and
provides structural support for the plasma membrane of the resting platelet.[1] The spectrin strands interact to form a hexagonal lattice and are densely decorated
with attached membrane glycoprotein. The platelet spectrin network is structurally
similar to the membrane skeleton of the erythrocyte, except its spectrin molecules
interconnect to actin using the ends of long filaments instead of short oligomers
of actin. Individual molecules of actin come together to form short oligomers that
subsequently undergo assembly into long filaments. Actin is the most abundant of all
proteins in platelets, making up about 20% of the total protein. The filamentous actin
element of resting platelets is arranged into approximately 2,000 actin filaments
that are interconnected, at various points, into a rigid network by the actin crosslinking
protein filamin A (FLNa).[1] FLNa subunits self-associate in solution into long homodimers.[4] FLNb is expressed in platelets at one-tenth the level of FLNa (12,000 copies/platelet).
Andrews and Fox[5] were the first to discover that FLNa links GPIb/IX/V (which forms the von Willebrand
factor receptor on the platelet membrane) to actin filaments in the cytoplasm. The
second rod domain of FLNa (repeat 17) has a binding site for the cytoplasmic tail
of GPIb-α component of GPIb/IX/V, and cell biologic experiments have shown that most
platelet FLNa ( ≥ 90%) is in complex with GPIb-α.[4] Because a large percentage of FLNa is attached to actin, it positions the von Willebrand
factor receptor on the surface of the platelet over the underlying linear filaments
in the platelet cytoskeleton. Platelets from patients with Bernard-Soulier syndrome
(who lack the von Willebrand factor receptor) also lack this connection, and their
platelets are unusually large and fragile. As Bernard–Soulier syndrome patients have
a severe bleeding diathesis and thrombocytopenia, this suggests that GPIb/IX/V is
important for normal platelet generation.[6] This possibility is further suggested by the low platelet count and unusual morphology
of platelets observed in mice lacking the GPIb-α receptor.[7] Although the FLNa–von Willebrand factor receptor connection has been characterized
at length in platelets, the role of this molecular linkage in platelet production
is not fully understood.
Mechanisms of Platelet Production
Proplatelet Model of Platelet Production
Megakaryocytes are highly specialized cells that expand and become polyploid through
a process called endomitosis. As megakaryocytes develop, their cytoplasm increases
in size and becomes full of platelet-specific granules; at the same time the megakaryocyte
develops a highly invaginated demarcation membrane system. Megakaryocyte maturation
comes to completion with the release of platelets into the bloodstream. Past studies
focusing on the mechanics of this process have been hindered by the rarity of megakaryocytes
in the bone marrow (< 0.1% of the total cells) and the lack of in vitro systems that
faithfully reconstitute platelet biogenesis. However, the discovery of thrombopoietin,
the major cytokine that regulates the growth and development of megakaryocytes, and
the emergence of cell culture systems that reconstitute bona fide platelet generation
have resulted in substantial progress toward understanding the maturation of megakaryocytes.[8]
The currently favored model of platelet production recognizes that differentiated
megakaryocytes extend long cytoplasmic processes, designated proplatelets.[9] Proplatelets function as the assembly lines of platelet production and comprise
platelet-sized swellings in tandem arrays that are linked by thin cytoplasmic bridges
([Fig. 1]).[10] Although extensive characterization of proplatelets remains incomplete, these processes
have been recognized both in vitro and in vivo, and proplatelet-producing megakaryocytes
generate platelets that are structurally and functionally similar to blood platelets.[11] Proplatelets have been observed extending from megakaryocytes in the bone marrow
through junctions in the lining of blood sinuses where they have been hypothesized
to be released into the circulation and undergo additional fragmentation into platelets.[9] Furthermore, mice lacking distinct hematopoietic transcription factors, such as
nuclear factor erythroid-derived 2, have severe thrombocytopenia and fail to produce
proplatelets in vitro, highlighting the parallel to platelet genesis in vivo.[12] Before proplatelet-based models, platelets were thought to form in the cytoplasm
of the megakaryocyte and undergo fragmentation along the boundaries of the demarcation
membrane system.[13] However, evidence now indicates the primary role of the demarcation membrane system
is to serve as an extensive membrane reservoir that is required for proplatelet elaboration.[14]
[15]
[16] The exact mechanism and the locations of platelet production are, however, still
controversial. Production of platelets from megakaryocytes within capillary beds has
been proposed as an alternative mechanism, but it is difficult to demonstrate in the
mouse or other species. Proplatelet formation, originally identified within bone marrow
sinusoids, may also take place partially in the bloodstream.
Fig. 1 Generation of proplatelets by a mouse megakaryocyte. Time-lapse sequence of a megakaryocyte,
showing the essential events that lead to elaboration of proplatelets in vitro. (A)
Platelet production starts when the megakaryocyte cytoplasm starts to erode at one
pole (arrow). (B) The majority of the megakaryocyte cytoplasm has been converted into
multiple proplatelet processes that continue to lengthen and form swellings along
their length. These processes are highly dynamic and undergo bending and branching.
(C) Once the bulk of the cytoplasm has been converted into proplatelets, the entire
process ends in a rapid retraction that separates the released proplatelets from the
residual cell body. (Adapted from: Italiano JE Jr., Lecine P, Shivdasani RA, Hartwig
JH. Blood platelets are assembled principally at the ends of proplatelet processes
produced by differentiated megakaryocytes. J Cell Biol 1999;147(6):1299–1312.)
Stages of Platelet Production
The development of megakaryocyte cultures that faithfully reconstitute platelet formation
has provided model systems to study megakaryocytes in the act of forming proplatelets.[11]
[17] Live cell microscopy reveals both spatial and temporal changes leading to the generation
of proplatelets ([Fig. 1]).[18] Remodeling of the megakaryocyte cytoplasm concentrates almost all the intracellular
contents into proplatelet processes and their platelet-size particles, which, in the
final stages, appear as beads linked by thin cytoplasmic bridges. The transformation
unfolds over 3 to 10 hours and begins in a polarized fashion with the erosion of one
side of the megakaryocyte cytoplasm. Thick pseudopodia-like processes initially form
and then extend into thin tubes of uniform diameter of 2 to 4 μm. The thin tubules
subsequently undergo a dynamic bending and branching process and develop periodic
platelet-sized swellings that span their length ([Fig. 2]). Ultimately, the megakaryocyte is converted into a residual naked nucleus surrounded
by an elaborate network of proplatelets ([Fig. 1C]). Megakaryocyte maturation ends when a fast retraction that separates the proplatelet
fragments from the cell body releases the processes into culture. The subsequent rupture
of the cytoplasmic bridges between platelet-size segments is thought to release individual
platelets into the circulation.[18]
Fig. 2 Architecture of proplatelets. Image of a differential interference contrast micrograph
of proplatelets extending from a mouse megakaryocyte culture. The hallmark features
of proplatelets, including the tip, shaft, branch points, and platelet-sized swellings
that decorate the length, are observed. Scale bar, 5 μm.
Microtubules Power the Elongation of Proplatelets
Some of the first insights into the cytoskeletal mechanisms that power platelet formation
dates from the work of Tablin and Leven, who used microtubule inhibitors to establish
that proplatelet extension is dependent on microtubules.[19] The megakaryocyte cytoskeleton, which is composed of microtubules, actin, and spectrin,
functions as the engine that powers the production of platelets ([Table 1], [Fig. 3]). Proplatelet formation and elongation is dependent on microtubule function, because
treatment of megakaryocytes with drugs that disassemble microtubules, such as vincristine
or nocodazole, blocks proplatelet formation.[19] Examination of the microtubule cytoskeletons of proplatelet-producing megakaryocytes
provides insights into how microtubules contribute to platelet production ([Fig. 4]). The microtubule cytoskeleton in megakaryocytes undergoes a striking remodeling
during proplatelet production.[18] In immature, round megakaryocytes without proplatelets, microtubules extend out
from the cell center (centrosome) to the cortex. As thick pseudopodial processes form
during the early stage of proplatelet production, cortical microtubules combine into
thick bundles located under the plasma membrane of these structures. When pseudopodia
start to extend (at an average rate of about 0.85 μm/min), microtubules form thick
linear arrays that core the length of the proplatelets. The bundles of microtubules
are thickest in the region of the proplatelet near the body of the megakaryocyte but
then thin to bundles of 5 to 10 microtubules near the tips of the proplatelets. The
distal end of each proplatelet always has a platelet-size enlargement containing a
microtubule bundle that makes a U-turn and loops just beneath the plasma membrane
and reenters the shaft to form a teardrop-shaped or tennis racket–shaped structure.
Because microtubule coils similar to those observed in blood platelets are detected
only at the tips of proplatelets and not within the platelet-size swellings found
along the length of proplatelets, putative platelets are generated only at the ends
of proplatelets.[18]
Fig. 3 Model of platelet production. (A) The formation of proplatelets begins with the extension
of thick pseudopodia that use cortical bundles of microtubules to extend and form
thin proplatelets with bulbous ends. Proplatelet membranes are laminated with an undercoat
of spectrin. The ends of proplatelets contain a bundle of microtubules that loop on
themselves. Proplatelet elongation involves the sliding of microtubules past one another,
driven by the molecular motor cytoplasmic dynein. As proplatelets extend, development
of the membrane surface area necessitates the outflow of the invaginated membrane
reservoir, a process that requires reorganization of the membrane skeleton. Mitochondria
and granules traffic (as indicated by the orange and yellow spheres) to the tips of
proplatelets along microtubules, which function as the highways of the cell. Actin
promotes the branching and amplification of proplatelet tips, representing a mechanism
to augment the numbers of proplatelet tips and ultimately, platelets. (B) The entire
megakaryocyte cytoplasm is converted into a mass of proplatelets. (C) Proplatelets
continue to morph into preplatelets (anucleate discoid particles 2–10 μm across),
which are released from the cell. Preplatelets reversibly convert into barbell proplatelets
(inset, top right), a process that is driven by microtubule-based forces. The membrane
skeleton stabilizes this barbell form. Platelets release from proplatelet ends after
the final fission event. The nucleus is eventually extruded from the proplatelets,
ending the role of the megakaryocyte in this process.
Fig. 4 Organization of microtubules within proplatelets. Immunofluorescence micrograph of
murine megakaryocytes grown in culture and labeled with β-1 tubulin antibodies indicate
that microtubules extend the entire length of proplatelets, including the tips and
shaft. Immunofluorescence studies further demonstrate that coils (arrows) of microtubule
similar to those seen in mature platelets occur in both proplatelets and released
platelet-sized particles. Scale bar, 5 μm.
Table 1
Cytoskeletal machinery of platelet production
|
Cytoskeletal component
|
Major function(s) in platelet production
|
|
Microtubules
|
Proplatelet elongation
|
|
Actin
|
Proplatelet bending and branching
|
|
Spectrin
|
Formation of the invaginated membrane system and stabilization of proplatelet architecture.
|
By directly visualizing microtubule dynamics in living megakaryocytes using green
fluorescent protein technology, we have obtained insights into how microtubules provide
the force to power proplatelet elongation.[20] End-binding protein three (EB3), a microtubule plus end-binding protein associated
only with growing microtubules, attached to green fluorescent protein (GFP) was retrovirally
expressed in mouse megakaryocytes and used as a marker to track the dynamics of microtubule
plus ends. Round, immature megakaryocytes without proplatelets use a centrosomal-coupled
microtubule nucleation/assembly reaction, which appears as a prominent starburst pattern
when visualized with EB3-GFP. Microtubules assemble only from the centrosomes and
grow outward into the cell cortex, where they turn and run in parallel with the cell
edges. However, just before proplatelet production commences, centrosomal assembly
ceases and microtubules begin to condense into the cortex. Fluorescence time-lapse
microscopy of proplatelet-producing megakaryocytes expressing EB3-GFP demonstrates
that as proplatelets extend, microtubule polymerization takes place constantly throughout
the entire length of the proplatelet, including the tip, swellings, and shaft. The
rates of microtubule assembly (average, 10.2 μm/min) are about 10-fold faster than
the rate at which proplatelets grow, suggesting assembly and proplatelet elongation
are not tightly coupled.[20] The EB3-GFP studies also demonstrated that microtubules assemble in both directions
in proplatelets. This reveals that the microtubules within the bundles have a mixed
polarity ([Fig. 5]).
Fig. 5 Direct visualization of microtubule assembly in living megakaryocytes expressing
end-binding protein three (EB3)–green fluorescent protein (GFP). (A) The first frame
from a time-lapse movie of a living megakaryocyte that was directed to express EB3-GFP.
The cell body (CB) is at the right of the micrograph and proplatelets (PP) extend
to the left. EB3-GFP labels growing microtubule plus-ends in a characteristic “comet”
staining pattern (arrowheads) that has a bright front and dim tail. The moving comets
are found along the proplatelets as well as in the body of the megakaryocyte. Scale
bar, 5 mm. (B) Kymograph (movement over time of the boxed region in A). Images are
of every second. EB3-GFP comets undergo bidirectional movements in proplatelets demonstrating
that microtubules are organized as bipolar arrays. Some EB3-GFP comets that move toward
the tip are highlighted in green; others that move toward the cell body are highlighted
in red.
Although microtubules are constantly assembling in proplatelets, polymerization by
itself does not supply the force for proplatelet elongation. Proplatelets continue
to extend at normal rates even when microtubule assembly is momentarily blocked by
drugs that inhibit net microtubule polymerization, suggesting an alternative mechanism
for proplatelet extension. In line with this idea, proplatelets have an internal microtubule
sliding mechanism.[20] The minus-end directed microtubule molecular motor protein cytoplasmic dynein localizes
along the length of microtubules of the proplatelet. Cytoplasmic dynein appears to
participate directly in microtubule sliding, because inhibition of cytoplasmic dynein,
through dissociation of the dynactin complex, blocks proplatelet formation. Microtubule
sliding can also be reactivated in permeabilized proplatelets. Addition of ATP, known
to support the enzymatic activity of microtubule-based molecular motors, activates
proplatelet elongation in the permeabilized proplatelets that contain both cytoplasmic
dynein and its regulatory complex called dynactin. Thus, dynein-driven microtubule
sliding appears to be a crucial event in powering proplatelet elongation.
Amplification of Proplatelet Ends via Bending and Branching
At the same time as the microtubule system is used to propel the extension of proplatelets,
an actin-powered process is used to branch off the shaft of the proplatelet, in so
doing increasing the number of proplatelet tips available to participate in platelet
production ([Fig. 6]). This unique process starts when a region of the proplatelet shaft is bent into
a U-shape. A new daughter process next juts out from the middle of this bend and extends.
Some of the microtubules within the bent segment of the loop separate from the bundle
to form a bulge. Sliding and polymerization of these microtubules creates a new daughter
proplatelet process.[17] This branching mechanism is driven by actin-based forces and is inhibited when megakaryocytes
are treated with drugs such as the cytochalasins, a family of toxins that block actin
monomers from polymerizing. Although the precise details on how the actin cytoskeleton
powers this event are lacking, actin filament assemblies occur periodically along
the length of the proplatelet shaft and are used as muscles to bend the rigid microtubules.
At these locations, actin filaments polymerize; although much is known about the regulation
of actin polymerization in platelets, there is very little information on how actin
polymerization is regulated or stimulated at these locations along proplatelets. Strong
connections between the microtubules and the actin filaments must be established to
transmit the actin-powered bending forces to the microtubule bundle. One possibility
is that actin filament–associated myosin motors may provide the force for bending.
Myosins comprise a family of molecular motor proteins most known for their role in
the contraction of muscle and a wide range of other motilities. Certain diseases suggest
an important role for myosin II in platelet generation. Mutations of MYH9 (the gene that encodes myosin IIA) that modify myosin IIA at the site involved in
myosin thick filament formation are implicated in MYH9-related disorders (e.g., May–Hegglin anomaly, Sebastian syndrome), a disorder where
platelet numbers are reduced in numbers and are abnormally large.
Fig. 6 Bending and branching of proplatelets. The bifurcation of proplatelets is observed
in these phase-contrast images taken 10 minutes apart showing the bending and branching
of a proplatelet extension. The bends are converted into loops that become compressed
and extend, resulting in a bifurcation of the original tube. White arrowheads indicate
the branch points. (Adapted from: Italiano JE Jr., Lecine P, Shivdasani RA, Hartwig
JH. Blood platelets are assembled principally at the ends of proplatelet processes
produced by differentiated megakaryocytes. J Cell Biol 1999;147(6):1299–1312.)
Mechanisms of Organelle Transport and Capture during Platelet Production
Essential in the process of platelet generation is the distribution of granules and
organelles into nascent platelets. In addition to functioning as the primary machinery
to elongate proplatelets, the microtubules that line the length of proplatelets provide
a secondary role—highways for the transport and delivery of membrane, organelles,
and granules into proplatelets and assembling platelets at the ends of proplatelets.
By monitoring the distribution and dynamics of organelles/granules in living cells,
it was established that individual organelles are sent one by one from the megakaryocyte
body into the proplatelets, where they move in both directions until they are captured
at the ends of proplatelets.[21] Thin-section electron microscopy and fluorescence microscopy studies indicate that
organelles are in direct contact with microtubules, and actin inhibitors do not block
the motion of organelles. Thus, motility appears to involve microtubule-based forces.
Microtubules are polar structures with a clearly defined directionality, as indicated
by a plus and minus end. The bidirectional movement of organelles is imparted by the
bipolar organization of microtubules within the proplatelet. Beads coated with the
molecular motor kinesin, which only moves in the plus-end direction, translocate in
both directions over the microtubules within permeabilized proplatelets. Of the two
major microtubule-based motors, only the plus-end–directed kinesin is located in a
pattern similar to organelles and is most likely responsible for translocating these
cargo along microtubules.[21] Taken together, the data suggest that a twofold mechanism of organelle movement
occurs during platelet production. First, organelles and granules move along microtubules,
and, second, the microtubules themselves can slide in relation to other motile filaments
to move organelles along proplatelets.
Spectrin-Based Membrane Skeleton in Platelet Production
Although the role of microtubules and actin filaments in platelet production have
been analyzed at length, the role of the membrane skeleton in this process has only
recently emerged. Rapid-freeze high-resolution electron microscopy reveals that proplatelets
contain a dense membrane skeleton similar in structure to that observed in the platelets
in blood.[22] The main fibrous component of this skeleton is spectrin. The nonerythroid spectrin
subunits, β-II and α-II spectrin, are predominately expressed in megakaryocytes, proplatelets,
and platelets. However, erythroid β-I and α-I spectrin isoforms are also present.
Assembly of spectrin tetramers is required for generation of the invaginated demarcation
membrane system and ultimately proplatelet production because expression of a spectrin
tetramer-disrupting peptide in megakaryocytes inhibits the progression of both. In
addition, introduction of this spectrin-disrupting construct into a detergent-permeabilized
model system rapidly destabilizes proplatelet morphology, resulting in enormous swelling
and blebbing. Spectrin tetramers also stabilize the barbell shapes of the penultimate
stage in platelet generation from proplatelets.[22] Taken together, these observations suggest a role for spectrin in distinct events
of megakaryocyte development through its participation in the generation of demarcation
membranes and in the maintenance of proplatelet structure.
Release of Individual Platelets
In vivo, proplatelets extend into bone marrow vascular sinusoids, where they are released
and enter the bloodstream. Previously, our understanding of platelet release was based
on static images of scanning electron micrographs of the bone marrow sinusoids.[9] More recently, Junt and colleagues have used intravital fluorescence microscopy
to directly visualize proplatelet production in the opened cranial marrow cavity of
living mice.[23] Fluorescently labeled megakaryocytes could be seen to protrude proplatelets and
release megakaryocyte fragments into the marrow sinusoids of living mice. Notably,
these anucleate mekaryocyte fragments typically exceed platelet dimensions, suggesting
that platelet morphogenesis continues in the circulation. In line with these observations,
we identified a previously unrecognized intermediate stage in platelet formation and
release, which we termed the preplatelet.[24] Preplatelets, which appear as “giant discoid platelets,” 3 to 10 μm in diameter,
retain the capacity to convert into barbell-shaped proplatelets and undergo fission
into individual platelets. Inhibitors of microtubule assembly block the transition
of preplatelets to barbells. Furthermore, taxol, which stabilizes microtubules and
stimulates microtubule polymerization, promotes the conversion of preplatelets into
platelets. Thus, the conversion of preplatelets to barbell proplatelets is powered
by microtubule-based forces.[24] It is tempting to speculate that the preplatelet fission mechanism is a major determinant
of platelet size and that some macrothrombocytopenias (Bernard-Soulier syndrome, MYH9-related disorders, etc.) represent a malfunction in converting preplatelets into
barbell-shaped proplatelets. It is likely that the microtubule motors that drive proplatelet
extension are involved in aspects of platelet release as well as in the process of
microtubule coiling. Force constraints deriving from cortical microtubule band diameter
and thickness play a major role in determining barbell conversion, and mathematical
modeling suggests that platelet size is limited by microtubule bundling, elastic bending,
and actin-myosin-spectrin cortical forces. In support of this concept, laser scanning
cytometry has provided high-resolution images of both preplatelets and barbell-shapes
in blood.[25] It was demonstrated that individual human platelets have the innate capacity to
duplicate and form new cell bodies that undergo fission into platelets.[26] The morphologic similarities between platelets that form new cell bodies and preplatelets
are striking. Whether or not newly released platelets exhibit a preplatelet phenotype,
which may allow them to form barbell shapes and divide again, is not clear.
Translating Thrombopoiesis Biology into Medicine
It has been 17 years since the discovery of thrombopoietin,[26] and drugs that are thrombopoietin mimetics are beginning to make an impact on the
treatment of thrombocytopenia. Despite this progress, we need to address the next
major advances in this field. In the United States alone more than 2 million platelet
transfusions occur each year, all with platelets from volunteer donors. Because of
their essential role in hemostasis, platelets are used for patients who have experienced
traumatic injury or are undergoing chemotherapy. The supply of platelets has long
been a problem for hospitals and blood banks. The short storage time (5 days) of platelet
products, which must be kept at room temperature, is also a challenge. In addition,
the risk of transfusion of transmitted diseases and shortages in supply provide additional
problems associated with donor platelets. Clearly, the ability to continuously generate
platelets ex vivo in a bioreactor would provide a more advanced way to generate a
product to treat thrombocytopenia. Since thrombopoietin was identified as the major
regulator of platelet production, it has been used to make enriched populations of
megakaryocytes. In 1995 Choi and colleagues demonstrated that platelets generated
in vitro from proplatelet-displaying human megakaryocytes were functional.[11] Since then, both megakaryocytes and platelets have been differentiated from multiple
sources, including embryonic stem cells and induced pluripotent stem cells.[28]
[29] However, despite all these various sources of megakaryocytes and platelets, the
yields of in vivo generated platelets have not come close to what is necessary for
clinical application. In the bone marrow specialized microenvironments, called niches,
regulate megakaryocyte development and platelet production through a complex crosstalk
between many cell types. The establishment of an in vitro model that faithfully recapitulates
the fundamental interactions of the niche components in a controlled setting could
advance the development of in vitro platelets for transfusion. For example, Lasky
and colleagues have constructed a purpose-built three-dimensional hydrogel scaffold
that functions as a bioreactor for platelet production. In this system, the authors
used CD34 positively selected human cord blood cells in a three-dimensional hydrogel
scaffold coated with thrombopoietin and/or fibronectin to increase platelet output.[30] In this three-dimensional model, the manipulation of oxygenation and flow-induced
shear stress appears to increase the yield of in vitro platelets derived from cord
blood.[31] Building on these models, Pallotta and colleagues have also recently developed a
silk-based three-dimensional system that partially recapitulates the spatial reconstruction
of the bone marrow environment and produces platelets in vitro.[32] Overall, these advances suggest that it will be important to mimic physiology and
use biologically inspired engineering to advance this technology to the clinic.
Platelets: Not Just the Band-Aids of the Blood
Although their primary function is to prevent bleeding, recent data suggest that platelets
contribute to a diverse array of processes that go way beyond thrombosis and hemostasis.
Platelets have been implicated in many different processes, including the development
of the lymphatic system, liver regeneration, inflammation, and cancer.[33] From the perspective of platelet production, it will be important to understand
how platelet generation is altered or “reprogrammed” to affect these other functions.
We have recently investigated how platelets regulate angiogenesis and innate immunity.
Platelets and New Blood Vessel Growth
A body of clinical and experimental data suggest that platelets influence tumor development
by transporting and delivering angiogenesis regulatory proteins.[34]
[35] The ability of platelets to interact with the endothelium is a key factor that allows
them to regulate angiogenesis. Some of the first data suggesting that platelets can
modulate angiogenesis were reported by Gimbrone et al, who demonstrated that perfusion
of plasma depleted of blood platelets resulted in instability of the endothelial layer
and hemorrhages.[36] Other studies in animals demonstrated that absence of platelets or low platelet
count led to increased permeability of the vasculature, the likely result of excessive
space between endothelial cells. It was later demonstrated that platelets modulate
angiogenesis by showing that platelets could stimulate the formation of capillary-like
and tubelike structures when added to human umbilical vein endothelial cells in culture.
Although platelets have been presumed to contribute to tumor development by providing
numerous stimulators and inhibitors of angiogenesis, the regulatory role of platelets
in this process is not fully understood. Platelets contain numerous regulators of
new blood vessel growth, which can be delivered to the endothelium when platelets
activate.[37] The proangiogenic regulatory proteins vascular endothelial growth factor, platelet-derived
growth factor, epithelial growth factor, basic fibroblast growth factor, metalloproteinases,
and sphingosine 1-phosphate have all been identified in platelets.[38] The stimulators in platelets are counterbalanced by the platelet angiogenesis inhibitors,
including platelet factor 4, thrombospondin-1, endostatin, tissue inhibitor of matrix
metalloproteinases, and angiostatin.[38] Most angiogenesis regulatory proteins identified in platelets have been localized
to α-granules, the major storage granule of platelets. Each human platelet contains
about 40 to 80 α-granules. By attaching to the endothelium of injured vessels and
then releasing their contents, platelets deliver high concentrations of angiogenesis
regulatory proteins in a precise manner.
The levels of these angiogenic regulatory proteins in platelets appear to play a role
in tumor angiogenesis. It has been reported that when a microscopically sized human
tumor is present in a mouse, circulating platelets take up and sequester specific
angiogenesis regulatory proteins, such as platelet factor 4, vascular endothelial
growth factor, and basic fibroblast growth factor.[39] The angiogenic regulatory proteins are sequestered in the platelets at a significantly
higher concentration than is observed in the plasma. This new platelet property may
lead to the development of a biomarker for very early detection of tumor recurrence.[40] Whereas an association between new blood vessel growth and platelets has long been
recognized, the cause and effect relationship linking the two has been unclear. Given
that platelets contain both stimulators and inhibitors of angiogenesis, packaged into
a homogeneous population of α-granules, the question becomes how can platelets have
either a proangiogenic or antiangiogenic effect? The release of a mixture of both
pro- and antiangiogenic regulatory proteins from platelets should cancel the effect
of each other. Several groups have demonstrated that platelets can preferentially
secrete a platelet releasate that has either a pro- or antiangiogenic effect.[41]
[42] The treatment of human platelets with the selective protease activated receptor
(PAR)-4 agonist resulted in release of endostatin-containing granules but not vascular
endothelial growth factor–containing granules, whereas the selective PAR-1 agonist
liberated vascular endothelial growth factor but not endostatin-containing granules.
This differential release is also observed with physiologic agonists.[43] Activation of human platelets with adenosine diphosphate stimulates the release
of vascular endothelial growth factor but not endostatin, whereas thromboxane A2 releases
endostatin but not vascular endothelial growth factor. Activation with adenosine diphosphate
also promotes the formation of capillary structures by human umbilical vein endothelial
cells in vitro. Conversely, thromboxane A2-activated releasate inhibits formation
of capillary structures. We have also tested the hypothesis that cancer cells preferentially
stimulate platelets to secrete their proangiogenic payload, providing a mechanism
for how tumors may hijack platelets to promote new blood vessel growth. In support
of this idea, the MCF-7 breast cancer cell line stimulates secretion of vascular endothelial
growth factor and a proangiogenic releasate from platelets. Interestingly, the antiplatelet
agent aspirin blocked platelet-mediated angiogenesis after introduction to the MCF-7
cells, pointing to a potential mechanism for how aspirin may influence malignancy.[43] Taken together, these data suggest that manipulation of differentially mediated
release of angiogenesis regulatory proteins from platelets may provide a new modality
for treatment of cancer. A better understanding of the mechanisms by which platelets
release angiogenic regulatory proteins should yield strategies for therapeutic benefit.
Platelets and Immunity
Platelets play a role in immunity by expressing members of the toll-like receptor
(TLR) family.[44] TLRs are a class of proteins that play an essential role in the innate immune system
by recognizing molecules that are broadly shared by pathogens. TLRs have been extensively
characterized in macrophages, dendritic cells, and neutrophils and support immune
activation in response to conserved molecular motifs on pathogens. Although platelets
express TLRs 1 through 9, most work has focused on TLRs 1 to 6, which are expressed
on the surface of the platelet and believed to trap bacteria for elimination by phagocytes.[45]
[46]
[47] We recently demonstrated that the TLR9 transcript is specifically up-regulated during
platelet production and is localized to a novel electron-dense tubular system named
the T-granule.[48] TLR9 colocalizes with protein disulfide isomerase and is associated with either
vesicle-associated membrane protein 8 or vesicle-associated membrane protein 7, which
are molecules involved in vesicle fusion that regulate TLR9 distribution in platelets
during activation. Type IV collagen specifically increases P-selectin, a cell adhesion
molecule, and TLR9 surface expression and augments oligodeoxynucleotide sequestration
and platelet aggregation upon the addition of viral and bacterial oligodeoxyribonucleotides.[48] Increased surface expression of TLR9 in platelets and type C CpG sequestration may
aid in the regulation and sequestration of circulating levels of bacterial DNA in
the blood and consequently help manage the inflammatory response after lysis of bacterial
cells. A more thorough understanding of the function that TLRs play in regulating
the activation of platelets will likely yield new therapies for the treatment of cardiovascular
infections.
