A Mechanical Artificial Red Cell:
Saturation @ PO2
|Atmosphere||159 mm||100%||0.3 mm||0 mm||600.7 mm||760 mm|
|Inspired||158 mm||100%||0.3 mm||5.7 mm||596 mm||760 mm|
|Tracheal||150 mm||100%||0.3 mm||47 mm||562.7 mm||760 mm|
|Expired||116 mm||99%||32 mm||47 mm||565 mm||760 mm|
|Alveolar (lung)||100 mm||98%||40 mm||47 mm||573 mm||760 mm|
|Arterial||95 mm||95%||40 mm||47 mm||573 mm||755 mm|
|Venous||40 mm||70%||46 mm||47 mm||573 mm||706 mm|
|Tissues (resting)||30 mm||55%||50 mm||47 mm||573 mm||700 mm|
|Tissues (working)||20 mm||33%||54 mm||47 mm||573 mm||694 mm|
Possible artificial oxygen carriers have been investigated for eight decades, starting with the first documented use of hemoglobin solutions in humans in 1916 . The commercial potential for successful blood substitutes has been estimated at between $5-10 billion/year [18-19], so the field is quite active [20-23]. Current blood substitutes are either hemoglobin formulations or fluorocarbon emulsions.
When tetrameric hemoglobin is freed from the red cell it loses effectiveness in three ways. First, it dissociates to dimers that are rapidly cleared from circulation by the mononuclear phagocytic system (10-30 minute half life) and by the kidneys (1 hour half life). Second, it binds O2 more tightly, reducing deliverability of O2 during tissue hypoxia. Third, during storage, hemoglobin may be oxidized to useless methemoglobin due to the absence of the protective enzyme methemoglobin reductase normally present in red cells.
Efforts to modify hemoglobin to increase intravascular dwell time have followed many pathways. Hemoglobin (in solution) has been cross-linked (either internally or with a macromolecule), polymerized, modified by recombinant DNA techniques, or microencapsulated. Encapsulation is most promising, given that all vertebrate hemoglobin is contained in cells to maintain its stability, preserve function, and protect the host from toxicity. Chang  reported making synthetic lecithin-cholesterol microcapsules containing hemoglobin as early as 1957. Hunt  reported liposome encapsulation of hemoglobin in 1983; his 0.1-1.5 micron "neohemocytes" proved nontoxic to rat kidneys of rats but were cleared from circulation with a half-life of 5.8 hours. Liposome-encapsulated hemoglobin (0.4-micron microspheres) can be stored dry up to 3 months without loss of function , and can sustain life in rats transfused isovolemically down to red cell hematocrits of only 4% . Taking a different approach, Suslick used ultrasound to weld roughly 1 million hemoglobin molecules into 2-micron hollow-core microspheres that carry 50% more oxygen per unit volume than natural red cells; the product, which remains 80% effective after 6 months' storage, is in preliminary animal and human trials .
Fluorocarbons offer a simpler approach to oxygen transport and delivery that relies on physical solubilization rather than binding of the oxygen molecules. Liquid fluorocarbons selected for the preparation of injectable oxygen carriers are typically molecules of 8-10 carbon atoms with molecular weights in the 450-500 range, dissolving 20-25 times as much O2 as water and delivering about the volume of oxygen to the tissues as an equal weight of hemoglobin. Mice survive immersion in fluorocarbon through which oxygen is bubbled , and rats breathing 95% oxygen have survived total blood replacement .
Because they are insoluble in water, fluorocarbons are administered in the form of emulsions of 0.1-0.2 micron droplets dispersed in a physiologic solution similar to fat emulsions routinely used for parenteral nutrition. After opsonization and phagocytosis of the emulsion droplets, the fluorocarbon is transferred to lipid carriers in the blood and released during passage through the pulmonary capillary bed. Thus fluorocarbons are not metabolized but are excreted unchanged by exhalation as a vapor through the lungs, typically in 4-12 hours for the present emulsions. Commercial preparations have been available for more than 20 years and have been administered to at least 2000 coronary perfusion patients without significant side effects.
At least four hemoglobin formulations and one fluorocarbon are in Phase I safety trials, and one company has filed an application to conduct an efficacy trial . Most of the red cell substitutes under trial at present have far too short a survival time in the circulation to be useful in the treatment of chronic anemia, and are not specifically designed to regulate carbon dioxide or to participate in acid/base buffering. Several cell-free hemoglobin preparations evidently cause vasoconstriction, decreasing tissue oxygenation , and there are reports of increased susceptibility to bacterial infection  due to blockade of the monocyte-macrophage system [33-35], complement activation , free-radical induction [37-38], and nephrotoxicity [39-41].
Fluosol, a widely available fluorocarbon perfusant, is approved by the FDA for use in patients at high risk of ischemic complications during percutaneous transluminal coronary angioplasty, but its experimental use in the treatment of severe anemia has been disappointing and cannot yet be recommended . The greatest physiological limitation is that oxygen solvates linearly with partial pressure, so clinically significant amounts of oxygen can only be carried by prolonged breathing of potentially toxic high oxygen concentrations.
Nanotechnology was first proposed by Nobel physicist Richard P. Feynman in December 1959, in a talk in which he also issued a seemingly "impossible" challenge to build a working electric motor no larger than a 1/64th-inch cube, backed by a $1000 prize to spur interest in the new field . Just 11 months later, engineer William McLellan had constructed a 250-microgram 2000-rpm motor out of 13 separate parts and collected his reward . Today a new $250,000 Feynman Prize, sponsored by the Foresight Institute, is available to any engineer who builds the first programmable nanometer-scale robotic arm .
Development successes leading towards molecular nanotechnology have been achieved along so many independent pathways that only a few illustrative milestones can be mentioned here. A Scanning Tunneling Microscope was used in 1989 to spell out "IBM" using 35 individual xenon atoms on a nickel surface . Atomic Force Microscopes (AFMs) have performed nanomachining operations on planar MoO3 crystals: applying 100 nanonewtons at the tip, two rectangular slots and a 50-nm rectangular sliding member were milled from a crystal, then the member was slid repeatedly from one slot to the other, making a reversible mechanical latch ; 10-nm features can be milled with AFMs. Individual nucleotides have been distinguished and manipulated on stretched DNA strands using AFMs [48-49].
Chemists have built a large number of potentially useful rigid nanoparts including molecular-scale rods, rings, springs, cubes, spheres, tetrahedrons, hollow tubes, propellors, and tongs , and wire-frame nanostructures of many shapes made of polymerized DNA [51-52]. They have begun experimenting with molecular "construction kits" . Chemists have also manufactured self-assembling multi-nanopart assemblies such as rotaxane "molecular shuttles" which move back and forth ~500 times/sec like a molecular abacus , and N-catenanes , a series of up to five interlocked rings (five nanoparts) averaging 75 atoms per ring arranged in the shape of the Olympic logo, the largest mechanically-interlocked molecule synthesized to date. Nobel chemist Jean-Marie Lehn has reviewed recent progress in supramolecular devices, supramolecular self-organization, and programmed chemical systems .
Among the top-down pathways, research in Micro Electro-Mechanical Systems (MEMS) has produced 100-micron-scale accelerometers, flow valves, pistons, gear trains, and piezo-driven motors that are available today off-the-shelf in mass quantities; microgrippers used to manipulate individual 2.7-micron polystyrene spheres, dried red blood cells, and various protozoa ; and a 1/1000th-scale working electric car just 2.8 millimeters wide, complete with motor, wheels, body, spare tire, bumpers, and even a 10-micron thick license plate .
Given a future ability to precisely engineer complex, micron-scale machines, it is possible to imagine a complete microscopic chemical factory  that avoids the shortcomings of current artificial blood technologies and simulates most major biochemical functions of the natural erythrocyte. Proper de novo design demands the simplest system capable of performing the desired task.
Given the goal of oxygen transport from the lungs to other body tissues, the simplest possible design for an artificial respirocyte is a microscopic pressure vessel, spherical in shape for maximum compactness.
Most proposals for durable nanostructures employ the strongest materials, such as flawless diamond or sapphire constructed atom by atom, with Young's modulus 1012 N/m2 (107 atm) and conservative working stress (~0.2 times tensile strength) of 1010 N/m2 (100,000 atm) . Tank storage capacity is given by Van der Waals equation which takes account of the finite size of tightly packed molecules and the intermolecular forces at higher packing densities: P = [nRT/(V-nB)] - [An2/V2], with P in atm, n in moles of gas, R = 8.206 x 10-5 m3-atm/mole-°K, T = 310 °K (human body temperature), V in m3, and constants A and B determined experimentally for each gas. Table 2 shows significant gains in gas molecule packing density up to 1000 atm, smaller gains at 10,000 atm, and no significant gain at >100,000 atm. Rupture risk and explosive energy rise with pressure, so a standard 1000 atm peak operating pressure appears optimum, providing high packing density with an extremely conservative 100-fold structural safety margin. (By comparison, natural red blood cells store oxygen at an equivalent 0.51 atm pressure, of which only 0.13 atm is deliverable to tissues.)
|Applied Pressure||1026 Molecules
|A (m6-atm/mole2)||1.36 x 10-6||3.59 x 10-6||1.39 x 10-6|
|B (m3/mole)||3.18 x 10-5||4.27 x 10-5||3.91 x 10-5|
In the simplest case, oxygen release could be continuous throughout the body . Slightly more sophisticated is a system responsive to local O2 partial pressure, with gas released either through a needle valve  (as in aqualung regulators) controlled by a heme protein that changes conformation in response to hypoxia , or by diffusion via low pressure chamber into a densely packed aggregation of heme-like molecules trapped in an external fullerene cage porous to environmental gas and water molecules, or by molecular sorting rotors (Section 2.2.2).
These simple proposals have two principal failings. First, once discharged the devices become useless. As with current blood substitutes, discharge time is too short. In the absence of functioning red cells the O2 contained in a 1 cm3 injection of 1000 atm microtanks would be exhausted in 2 minutes. Second, the proposals involve placement of numerous point source O2 emitters throughout the capillary bed in conjunction with the existing erythrocyte population. These extra emitters are functionally equivalent to red blood cells whose CO2 transport and acid-buffering capabilities have been selectively disabled. Their addition to the blood pushes respiratory gas equilibrium toward higher CO2 tension and elevated hydrogen ion concentration, which could lead to carbon dioxide toxicity and acidosis (hypercapnia), especially in anemic, nonrespiratory, or ischemic patients, and to hyperoxic hemolysis and other complications.
Neither problem may be overcome using passive systems alone. The easiest way to extend duration is to provide for recharging the microvessels with oxygen gas, preferably via the lungs since direct regeneration of O2 from CO2 is energetically prohibitive. The easiest way to prevent carbon dioxide toxicity is to provide additional tankage for CO2 transport and some active means for gas loading at the tissues and unloading at the lungs. Note that physically stored CO2 makes no net addition to blood acidity. Respirocytes operating in the absence of red cells would generate little CO2-related acidity. Proper blood pH could probably be maintained by the kidneys alone.
The key to successful respirocyte function is an active means of conveying gas molecules into, and out of, pressurized microvessels. Molecular sorting rotors have been proposed  that would be ideal for this task (Figure 1). Each rotor has binding site "pockets" along the rim exposed alternately to the blood plasma and interior chamber by the rotation of the disk. Each pocket selectively binds a specific molecule when exposed to the plasma. Once the binding site rotates to expose it to the interior chamber, the bound molecules are forcibly ejected by rods thrust outward by the cam surface.
Figure 1. Molecular Sorting Rotor (modified from Drexler )
larger image (640x577 pixels, 128 K)
Molecular sorting rotors can be designed from about 105 atoms (including the housing), measuring roughly 7 nm x 14 nm x 14 nm in size with a mass of 2 x 10-21 kg. These devices could sort small molecules of 20 or fewer atoms at a rate of 106 molecules/sec with laminar flow for an energy cost of 10-22 joule/molecule sorted, and pump against head pressures up to 30,000 atm at an additional energy cost up to 10-19 joule/molecule . Rotors are fully reversible, so they can be used to load or unload gas storage tanks, depending upon the direction of rotor rotation. (It should be possible to recover most of the sorting energy by adding a generator subsystem, or by compressing one gas using energy derived largely from the decompression of the other  using differential gearing. Neither alternative, which might reduce power consumption by a factor of ~10-100, was pursued in the present design because the energy resource -- serum glucose -- appears plentiful.)
Typical molecular concentrations in the blood for target molecules of interest (O2, CO2, N2 and glucose) are ~10-4, which should be sufficient to ensure at least 90% occupancy of rotor binding sites at the stated rotor speed . Each stage can conservatively provide a concentration factor of 1000, so a multi-stage cascade (Figure 2) should ensure storage of virtually pure gases. Since each 12-arm outbound rotor can contain binding sites for 12 different impurity molecules, the number of outbound rotors in the entire system can probably be reduced to a small fraction of the number of inbound rotors.
Figure 2. Sorting Rotor Cascade (modified from Drexler )
larger image (640x427 pixels, 91 K)
Receptors with binding sites for specific molecules must be extremely reliable (high affinity and specificity) and survive long exposures to the aqueous media of the blood. The following are examples of well-known receptors that could be studied for useful features in this application.
Oxygen transport pigments are conjugated proteins, that is, proteins complexed with another organic molecule or with one or more metal atoms. Transport pigments contain metal atoms such as Cu2+ or Fe3+ making binding sites to which oxygen can reversibly attach. Besides hemoglobin (MW 68,000) and myoglobin (MW 17,000) in man, other natural respiratory pigments include hemocyanin [62-63], a blue copper-based pigment found in molluscs and crustaceans (MW 1-7 million) and chlorocruorin , a green iron-based pigment found in marine polychaete worms (MW 3 million), both of which are only about 25% as efficient as oxygen carriers as hemoglobin ; hemerythrin , a purple iron-based pigment found in some molluscs and worms (MW 100,000), about 10% as efficient as hemoglobin; vanadium chromagen , a pigment found in the blood of sea squirts, ascidians and tunicates, in apple-green, blue, and orange varieties, due to the presence of different oxides of vanadium .
Artificial reversible oxygen-binding molecules have also been studied, including cobalt-based porphyrins such as coboglobin (a cobalt-based analog to hemoglobin) and cobaltodihistidine [67-71], other metallic porphyrins [67,72-73], simple iron-indigo compounds , iridium complexes such as chloro-carbonyl-bis(triphenylphosphine)-iridium [75-76], a simple cobalt/ammonia complex , zeolite-bound divalent chromium , nonporphyrin lacunar iron complexes , and heme-linked NADPH oxidase . Implantable blood oxygen sensors such as Medtronic's hemodynamic monitor are already in clinical trials. Unlike hemoglobin, hemocyanin, hemerythrin and coboglobin are not poisoned by carbon monoxide; neither will respirocytes.
Many proteins and enzymes have binding sites for carbon dioxide. For example, hemoglobin reversibly binds CO2, forming carbamino hemoglobin. A zinc enzyme present in red blood cells, carbonic anhydrase , catalyzes the hydration of dissolved carbon dioxide to bicarbonate ion, so this enzyme has receptors for both CO2 and H2O. The first step in chlorophyllic photosynthesis, in which CO2 is added to a 5-carbon sugar, is catalyzed by ribulose biphosphate carboxylase, probably the world's most abundant enzyme because it accounts for more than half the soluble protein in every green leaf on Earth . (This enzyme actually has two separate CO2 binding sites, and a third oxygenase site that binds O2.)
Many molecules bind water reversibly, including a wide variety of deliquescent crystals, efflorescent minerals, hydrophilic and polar amino acids, and numerous enzymes such as carbonic anhydrase, hydrolases and dehydratases .
Binding sites for glucose are common in nature. For example, cellular energy metabolism starts with the conversion of the 6-carbon glucose to two 3-carbon fragments (pyruvate or lactate), the first step in glycolysis. This is catalyzed by the enzyme hexokinase, which has binding sites for both glucose and ADP. Another common cellular mechanism is the glucose transporter molecule, which carries glucose across cell membranes and contains several binding sites [81-82]. Other glucose-binding proteins are found in the intestines , liver , kidney, adipose tissue, and elsewhere. Implantable glucose sensors have been developed by Becton, Dickinson Inc., by Japan's University of Osaka, and others.
Finally, certain microorganisms, including the Rhizobium genera of bacteria found on leguminous plant roots, some free-living soil bacteria such as the Azotobacter, and a few species of blue-green algae such as Anabaena cylindrica, achieve biological fixation of atmospheric N2 using an enzyme complex called nitrogenase . Nitrogenase is extremely labile in the presence of oxygen, but its nitrogen binding sites might reward further study.
Once they are structurally and functionally known, receptors (in all required conformations) for each of the different gases to be transported may be incorporated into the rotors as precisely shaped and charged diamondoid surfaces and cavities using the manufacturing techniques suggested by Drexler  for atom-by-atom assembly, including mill-style and manipulator-style hierarchical mechanosynthesis, shape description languages, parts arrays and design compilers.
The upper limit of physical device size is easy to specify because respirocytes must have ready access to all tissues via blood vessels. They cannot be larger than human capillaries, which average 8 microns in diameter but may be as small as 3.7 microns  -- so narrow that natural red blood cells (7.82 micron x 2.58 micron biconcave disks ) must fold in half to pass, single-file.
As respirocyte design radius shrinks, surface area per unit enclosed volume rises rapidly. Smaller cells may have more difficulty defending against environmental insults and face more numerous potential filtration sites throughout the body. (For example, liver endothelial cells, Kupffer cells, and the fenestrated endothelium of the glomerular membrane in the kidney filter particles <100 nm from the blood, but Kupffer cells of the liver have exclusive phagocytosis of foreign particles >100 nm .) These factors argue for larger device sizes.
The minimum possible respirocyte diameter is driven by operational requirements and by minimum component size. For instance, the smallest reasonable computer requires 105 nm3, a 58-nm diameter sphere (Section 3.4). Additionally, gas is loaded using molecular sorting rotors mounted on the surface of a spherical tank. In the baseline design (diameter D = 1 micron), 37.28% of tank surface consists of sorting rotors and related subsystems (Section 3.5). The number of sorting rotors scales with tank volume or R3; tank surface scales as R2; so the percentage of tank surface in rotors (Rotor/Surface Ratio or RSR) scales linearly with R (RSR ~ qD, q = 0.3728 fractional surface coverage). RSR ~ 1.00 (100%) coverage occurs at D = 2.68 microns, the upper limit. Careful review of the baseline design suggests that the minimum rotor area necessary to achieve all performance specifications while maintaining 10:1 subsystem redundancy (Section 3.5) is about 17,000 nm2, which implies D = 0.245 micron at RSR = 0.0902 (9.02%). Eliminating all redundancy reduces rotor requirements to 1700 nm2, which implies D = 0.114 micron at RSR = 0.0412 (4.12%). The above considerations suggest a reasonable range for respirocyte diameter of 0.2-2 microns; the present study assumes a spherical respirocyte diameter of ~1 micron.
Another design issue that arises when operating in an aqueous medium is buoyancy, which can readily be controlled by loading or unloading water ballast. The smaller the respirocyte, the slower it settles out of suspension, as given by Stokes Law: Vt=2gR2(r1-r2)/9h, where Vt is terminal velocity (cm/sec), g is the acceleration of gravity (981 cm/sec2), r1 and r2 are device and fluid densities (gm/cm3), and h is coefficient of viscosity (1.1 centipoise for plasma, 4 centipoise for whole blood at 310 K).
A 1-micron respirocyte with density ranging from 679 kg/m3 for tanks empty (vacuum) to 1370 kg/m3 for all tanks full would rise or fall at a maximum rate of 0.1-0.6 mm/hour; the settling rate for 0.1-micron particles is 100 times slower. By comparison, even the small difference in density between individual red cells (1100 kg/m3) and blood plasma (1025 kg/m3 at 310 K) causes red cells to settle out of suspension at a much faster 4-10 mm/hour depending on hematocrit (the volumetric % of blood occupied by red cells, typically 45% in humans) and degree of RBC aggregation. Natural erythrocytes appear unhandicapped by their faster settling rate, so active ballast management for artificial respirocytes (which settle slower) is probably unnecessary in normal operations.
Once a therapeutic purpose is completed, it may be desirable to extract artificial devices from circulation. Onboard water ballast control is extremely useful during respirocyte exfusion from the blood. During nanapheresis, blood to be cleared may be passed from the patient to a specialized centrifugation apparatus where acoustic transmitters (Section 3.2) command respirocytes to establish neutral buoyancy. No other solid blood component can maintain exact neutral buoyancy, hence those other components precipitate outward during gentle centrifugation and are drawn off and added back to filtered plasma on the other side of the apparatus. Meanwhile, after a period of centrifugation, the plasma, containing mostly suspended respirocytes but few other solids, is drawn off through a 1-micron filter, removing the respirocytes. Filtered plasma is recombined with centrifuged solid components and returned undamaged to the patient's body. The rate of separation is further enhanced either by commanding respirocytes to empty all tanks, lowering net density to 66% of blood plasma density, or by commanding respirocytes to blow a 5-micron O2 gas bubble to which the device may adhere via surface tension, allowing it to rise at 45 mm/hour under normal gravitational acceleration.
|© Copyright 1996-1999, Robert A. Freitas Jr. All rights reserved.|
Foresight materials on the Web are ©1986–2019 Foresight Institute. All rights reserved. Legal Notices.