Matthew Neal, a trauma surgeon at the renowned University of Pittsburgh Medical Center (UPMC), regularly receives calls from smaller facilities reporting that a patient has suffered a traumatic brain injury and needs a blood platelet transfusion to stop the bleeding, but the facility doesn’t have any. The demand for platelets is infrequent and platelets expire after only five days, so stocking up would lead to waste. The patient transfers to UMPC, squandering precious time. 

While grappling with a blood donation pipeline mangled by the COVID-19 pandemic, even UPMC sometimes ran out of platelets. “We saw for one of the first times in our modern history shortages even at our largest, most advanced academic medical centers,” Neal said. “That elevated the need for alternatives to our current blood management program to a new level.” 

One alternative with lifesaving potential in these emergency situations is artificial blood.   Scientists have long pursued this enigmatic fluid, with experimental transfusionists Christopher Wren and Robert Boyle at the University of Oxford attempting to replace blood with milk and liquor following the discovery of circulation in 1616 (1,2,3). Armed with today’s knowledge of the complexity of blood’s composition and function, researchers tackle artificial blood one component at a time. They adapt hemoglobin to perform the critical oxygen delivery function of red blood cells. They design synthetic nanoparticle based platelets for blood clotting. They freeze 

dry the rich mixture of blood proteins found in plasma and explore recombinant forms grown in the lab. Finally, they combine all of these synthetic components to achieve a goal 400 years in the making: artificial whole blood. 

Sweet hemes

The red blood cell owes its role as the blood’s oxygen transporter to its hemoglobin protein; each of its four subunits contains an iron centered heme group. When the iron in one heme group binds oxygen, the others follow suit, allowing hemoglobin to efficiently carry large quantities of oxygen throughout the body. 

Researchers are developing hemoglobin based products to improve oxygen delivery during traumatic blood loss and for red blood cell disorders such as sickle cell anemia. These synthetic products do not need to be blood type matched, and some may be stored for years, whereas red blood cells last for only about 42 days in the refrigerator. While synthetic substitutes circulate with a half-life on the order of hours as opposed to days, they provide a life sustaining bridge to a red blood cell transfusion. “We don't need to have a red blood cell that has the normal lifespan of a red cell or that takes the place completely of a red cell,” Neal said. “We need something that provides oxygen carrying capacity until the patient can get to definitive care.” 

While administering the red blood cell’s oxygen carrying powerhouse on its own might seem an obvious approach, hemoglobin turns from life giving to toxic outside the protective interior of the cell. While the red blood cell contains an enzymatic reduction system to keep the iron in its active oxidation state, the iron in cell free hemoglobin can oxidize to forms that cannot bind oxygen and may damage DNA, proteins, and lipids. As the fragile interactions between its subunits are disrupted in the tumultuous environment of the bloodstream, hemoglobin can break apart. The resulting fragments and any intact protein can penetrate the blood vessel walls, where oxygen bound to hemoglobin can react with nitric oxide released from the vessel’s endothelial lining. Nitric oxide is a key signaling molecule that relaxes blood vessels, and depleting it through this reaction can cause dangerous vascular constriction. Hemoglobin that escapes the blood vessel can travel, causing oxidative damage to tissues and organs throughout the body. 

To overcome these challenges, researchers introduced crosslinking reagents that formed stable bonds within and between multiple hemoglobin proteins, forming large protein clumps that could neither breach the blood vessel wall nor fall apart into fragments that could. These polymerized hemoglobin proteins yielded positive outcomes in “compassionate use” medical scenarios in which no other treatment options are available. Physicians have used them to sustain people with rare blood types while searching for type matched red blood cells and patients who cannot receive a blood transfusion due to their religious practices (in which case, they repeatedly administered the product until the patient’s red blood cells regenerated) (4,5). However, full approval of these polymerized hemoglobin oxygen carriers has been hindered by adverse cardiovascular effects observed during phase III clinical trials (6).

Still, Andre Palmer, a chemical and biomolecular engineer at The Ohio State University, isn’t ready to give up on polymerized hemoglobin. “As engineers, we are always interested in scaling up the production of materials and making sure we have a very cost effective process,” he said. “Polymerizing hemoglobin is the most cost effective way of increasing the molecular radius of the hemoglobin to prevent side effects.” 

According to Palmer, previous generations of polymerized hemoglobin based oxygen carriers might have been small enough to squeeze through blood vessel walls, resulting in side effects. By synthesizing polymerized bovine hemoglobin of various molecular weights and administering them in rodent models, Palmer’s team identified a 500 kDa threshold that gives lower measures of vascular constriction, indicating better retention of the protein in the circulatory system (7,8). The team developed an inexpensive, high throughput flow filtration system to efficiently purify their supersized polymerized hemoglobin, eliminating smaller side products (9).

Slimy solutions

Even if free hemoglobin is large enough to remain confined to the blood vessel, it can oxidize, impairing oxygen transport and causing oxidative damage to the vessel tissues. It can also scavenge nitric oxide from within the blood vessel, contributing to vascular constriction. Researchers have identified mutations in human hemoglobin that facilitate iron reduction and hinder nitric oxide binding (10,11). 

Jacob Elmer, a chemical and biological engineer at Villanova University, set out to genetically engineer a superior human hemoglobin and ended up purchasing 1,000 earthworms. “I was looking around for inspiration from nature — what kind of mutations can I make in this protein? And I was just won over by this naturally occurring earthworm hemoglobin,” he said. “I started off trying to make the mutations that I saw in earthworm hemoglobin in human hemoglobin, but when it came down to it, it was just easier to make the earthworm hemoglobin.” 

Earthworms are among a small group of organisms that lack red blood cells, which forced their hemoglobin to evolve to function safely outside of the cell. Elmer has explored using hemoglobin from many of them. “We've extracted hemoglobins from lots of other organisms: snails, leeches, a lot of marine worms, a lot of terrestrial worms, and earthworm hemoglobin somehow has still emerged to be the best,” he said. “But I really liked when we were extracting the hemoglobin from leeches because there was a poetic sense of justice there. Leeches normally steal blood from other organisms; we were stealing their blood. But unfortunately, that didn't go anywhere.” 

Earthworm hemoglobin has several tantalizing features that solidify its status as reigning champ of alternative hemoglobins. With 180 subunits and a molecular weight more than 50 times that of human hemoglobin, earthworm hemoglobin can’t slip out of the blood vessel. The protein is held together in the bloodstream by strong chemical bonds and subunits dedicated solely to crosslinking. “I teach a class on protein engineering, and every one of the things that I teach about protein stability, every one of the tricks and techniques that exist, earthworm hemoglobin has applied here to maintain that structural stability,” Elmer said. 

The electrical environment of the earthworm heme group causes the iron to oxidize more slowly than in human hemoglobin. It is also easily re-reduced by chemicals in the blood such as ascorbic acid, helping earthworm hemoglobin maintain its active oxidation state for 48 hours in a rodent model (12). The heme pocket is also smaller and cannot fit both oxygen and nitric oxide at the same time, which prevents nitric oxide depletion. 

Elmer and his collaborators used earthworm hemoglobin to replace lost blood in mice and hamsters, resulting in successful resuscitation without side effects (13,14). He hasn’t observed an immune response in these rodents even after repeated administration, although he plans to further evaluate safety and immunogenicity in more complex animal models. 

It really is a product that a medic could have in his backpack for weeks at a time, just sitting there in the deserts of Afghanistan or Iraq if necessary. 
- Jacob Elmer, Villanova University

Elmer also discovered that earthworm hemoglobin remains functional for over a week when stored at an elevated temperature of 37 °C (12). This temperature stability may allow it to be used in environments lacking cold storage, such as in military settings, where bleeding is the number one cause of potentially survivable, traumatic prehospital death (15). “It really is a product that a medic could have in his backpack for weeks at a time, just sitting there in the deserts of Afghanistan or Iraq if necessary,” Elmer said. 

Why reinvent the cell?

Rather than exploring how to tame cell free hemoglobin by introducing modifications found in crawling creatures, other researchers are enclosing the protein in a red blood cell mimic. “We have to pay attention to what nature has already done because evolution has produced the current state of affairs after hundreds of millions of years of selective pressure,” said Allan Doctor, a pediatric critical care specialist at the University of Maryland School of Medicine. “You make almost 500 kilos of red cells in your lifetime in a situation where the body's very parsimonious about where energy goes and nothing is wasted. If this significant portion of our energy and nutritional budget is devoted to doing things a certain way, I think I should try to imitate that.” 

Doctor cofounded the company KaloCyte, where he and his colleagues developed an artificial red blood cell called ErythroMer for commercial use. Their artificial cell currently uses human hemoglobin obtained from expired donor red blood cells, although the team is developing a system to brew the recombinant protein in yeast. The team encapsulated hemoglobin in a cell like vesicle made from synthetic versions of lipids in the cell membrane and stabilized it with a polyethylene glycol (PEG) polymer. Snuggled safely inside this synthetic cell, the hemoglobin cannot break apart or escape the blood vessel. And just like in a red blood cell, the membrane provides a barrier to nitric oxide diffusion, limiting its interaction with hemoglobin. 

The red blood cell’s enzymatic iron reduction system, which is fueled by continuous glucose burning, is more challenging to mimic. To recapitulate this dynamic process in a stagnant cell, the team packages the hemoglobin with leucomethylene blue, a chemical similar to ascorbic acid, that reduces any oxidized iron to its active oxidation state. 

ErythroMer also tunes hemoglobin-oxygen binding according to the body’s oxygen demand. The membrane features a short peptide that dangles inside the cell and binds to a small molecule. In the lungs, where the pH is greater than 7.4, the molecule remains bound to the peptide, confined to the vesicle’s inner wall. When the cell travels throughout the body and encounters oxygen-deprived tissue where the pH is less than 7.4, the peptide gains protons, breaking its bond with the small molecule. The molecule then binds hemoglobin, displacing and releasing oxygen. When the pH increases as the cell returns to the lungs, the peptide loses the protons and rebinds to the small molecule, allowing hemoglobin to take up oxygen again. This system mimics the red blood cell’s acute regulation of hemoglobin’s affinity for oxygen throughout the circulatory system using a complex network of transmembrane enzymes and ion channels. 

Other hemoglobin based products with a fixed oxygen affinity bind and release oxygen in response to oxygen gradients throughout the body, but the affinity must strike a balance between efficient uptake in the lungs and delivery to the tissues. The context dependent affinity of ErythroMer optimizes both of these opposing functions, allowing less product to be used to transport the same amount of oxygen, Doctor said. The researchers replaced 70 percent of a mouse’s blood volume with ErythroMer and found that it delivered oxygen as effectively as real blood (16). When they removed half the blood from rabbits, left them in shock for one hour, then administered ErythroMer, it resuscitated the rabbits with similar efficacy to real blood (16). While the team hasn’t observed any organ toxicity, vascular constriction, or other side effects so far, they will continue to evaluate the product in ferret and primate models. 

Hiromi Sakai, a chemist at Nara Medical University in Japan, developed a similar PEG stabilized, lipid vesicle encapsulated hemoglobin product with survival rates comparable to red blood cells in rat and rabbit models of hemorrhage (17,18). Following these promising results, the team evaluated the product for safety in a first-in-human phase I clinical trial (19). Some participants experienced adverse effects including rash and fever, but they were tolerable and temporary. “These side effects appeared only in humans so far,” Sakai said. “So, we are a bit discouraged, but there are some remedies for such side reactions.” For example, pretreatment with histamine receptor blocker or immunosuppressant drugs may prevent these responses. Sakai isn’t sure what part of the vesicle is to blame, but he is interested in evaluating the effect of tweaking the lipid formulation in the future. 

While immunogenicity testing with ErythroMer is ongoing, the researchers haven’t observed an increased immune response relative to stored blood so far. Still, Doctor is not surprised by the results of Sakai’s trial. “In the end, it is unlikely that it will be completely immune silent. If you put something artificial like this in the bloodstream, it would be surprising if absolutely nothing happened. The question is does the risk-benefit [analysis] favor use in the setting of bleeding to death?” Doctor said. “There might be some unavoidable mild immune reaction. And in that case, as long as there's no significant long-term consequence, it'll still be worth doing.”

As events that demand large quantities of blood become more common in civilian life, developing these products may be worth doing now more than ever. “There were not a lot of mass casualty situations when I started working on this. And sadly, since that time… there’s almost combat like problems in cities that really risk running out of blood,” Doctor said. “So, there’s a need to have mass casualty depots stored around the country.” 

To create these stockpiles, the researchers have freeze dried ErythroMer, forming a powder that can be redissolved in sterile water or salt solutions and administered at the point of injury. The powder lasts one to two years under ambient conditions, providing a shelf stable inventory that can be accessed in case of an emergency. 

Stick to it

When blood vessels are damaged, platelet cell fragments coasting in the blood’s lazy river spring into action to plug them up and stop the bleeding. Like the first row of bricks laid on the ground to build a wall, platelets adhere to the interior of the blood vessel. The injury disrupts the vessel’s endothelial lining, exposing the collagen protein underneath. Platelets can bind directly to collagen or use von Willebrand factor, a molecule secreted by damaged endothelial cells, as a double-sided tape like intermediate. To add additional layers of bricks and continue constructing the wall, the platelets must also stick to each other. Injury activates a platelet surface receptor that binds the blood protein fibrinogen. Fibrinogen binds platelets at both ends, linking them together. Injury activated platelets also undergo a change in their membranes in which a negatively charged phospholipid on the inner surface flips to the outer surface, where it causes coagulation factors to assemble. This assembly amplifies a series of coagulation reactions that produces thrombin, an enzyme that converts fibrinogen to fibrin. Fibrin proteins bind to injury activated platelets and to one another, forming a strong network around the platelet aggregate that stabilizes it like mortar on a brick wall.  

Researchers have developed synthetic platelets that mimic one or more of these functions to staunch persistent bleeding in people with severe injuries or platelet deficiencies. These products work together with the body’s natural coagulation mechanisms to augment clotting activity during uncontrolled blood loss. “The clot promotion is an important part because unless you seal the bleeding wound, no matter what you give that person, they’re just going to bleed out,” said Anirban Sen Gupta, a biomedical engineer at Case Western Reserve University. Synthetic platelets must support clot formation only at the injury site to prevent clots from obstructing blood flow elsewhere in the body. It’s no small feat, but given the exceptionally short shelf life of natural platelets, these synthetic versions are particularly valuable. 

Shinji Takeoka, a biomedical scientist at Waseda University in Japan, developed an artificial platelet that amplifies the platelet aggregation process. He decorated a lipid particle with a peptide that binds to the fibrinogen receptor on injury activated platelets, helping them clump together. The lipid particle also encapsulates adenosine diphosphate (ADP) and releases it at the injury site, likely due to mechanical stress. The ADP signaling molecule activates the fibrinogen receptor on additional platelets, allowing them to bind to the lipid particles in a positive feedback loop. “This simple strategy is beneficial for emergency use,” Takeoka said. 

Takeoka and his collaborators demonstrated a clear benefit by administering the synthetic platelets in a low platelet count rabbit model of traumatic hemorrhage, yielding a 100 percent survival rate during the three day postinjury monitoring period (20). The platelets also increased survival relative to particles lacking the fibrinogen receptor binding peptide or the ADP payload. 

Ashley Brown, a biomedical engineer at North Carolina State University and the University of North Carolina at Chapel Hill, developed a synthetic platelet that helps build the fibrin network. She designed a hydrogel microparticle presenting a peptide that binds to fibrin proteins, recruiting them to form a fibrin stabilized clot. “It basically serves as a scaffold where more fibrin can be deposited and grow to make a better clot,” Brown said. “It gives it a better chance of forming a robust enough clot to stop the bleeding.” 

To ensure injury specific clotting, the peptide binds selectively to fibrin generated at the site of injury over its circulating fibrinogen precursor. The malleable particles can also recapitulate a natural process in which the fibrin network contracts, increasing clot density and stability (21). Brown’s synthetic platelets form clots as dense as those made up of real platelets in vitro and they reduce blood loss in a rodent model of traumatic bleeding (22).    

Sen Gupta and Neal collaborated to develop a synthetic platelet that helps build the platelet wall at every step. Their design features a lipid nanoparticle with three distinct peptides attached. The first peptide binds to collagen and the second binds to von Willebrand factor, mimicking the adhesion of platelets to the injured blood vessel. The third peptide binds to the fibrinogen receptor on injury activated platelets, facilitating platelet aggregation in the same manner as Takeoka’s technology. 

To enable their synthetic platelet to promote fibrin based clot stabilization, the researchers explored two different strategies to increase the amount of thrombin (and therefore the conversion of fibrinogen to fibrin) with spatial control. One approach is to decorate the nanoparticle with the negatively charged phospholipid that increases thrombin production when presented on the platelet surface (23). The phospholipid is cloaked by a polymer while the particle circulates the body, shielding the negative charge. When it encounters an enzyme found in high concentrations at the site of the injury, the polymer gets cleaved off, exposing the negative charge and initiating the pathway that amplifies thrombin and fibrin production. The second strategy is to simply encapsulate thrombin itself inside the nanoparticle (24). When the particle is degraded by another enzyme upregulated at the injury site, the thrombin is released and begins forming fibrin. Directly administering thrombin in this way could be effective for people whose thrombin production cascade is impaired. 

The most important result that we have gotten so far in a variety of animal models is that it produces clots and reduces bleeding in a very injury site specific way. 
- Anirban Sen Gupta, Case Western Reserve University

These synthetic platelets have improved survival in rodent and pig models of traumatic hemorrhage without any signs of off target clotting (23,25). “The most important result that we have gotten so far in a variety of animal models is that it produces clots and reduces bleeding in a very injury site specific way,” Sen Gupta said. The team is now conducting dose escalation studies to more rigorously evaluate systemic clotting risks, immunogenicity, and any other side effects to establish a therapeutic window. They’re also developing a freeze-dried version of the synthetic platelet that is shelf stable, portable, and dissolvable on demand. While storage evaluation is ongoing, the synthetic platelets tolerate temperatures from -4 to 50 °C and are stable under ambient conditions for nine months — and counting.   

Plasma pow(d)er

The synthetic platelets can’t act alone, however. They must be administered in plasma, the liquid component of blood that houses a rich mixture of proteins including fibrinogen and other coagulation factors. Plasma proteins also maintain the osmotic pressure that keeps blood inside the vessels and helps protect organs against inflammatory damage. A landmark trial found that giving two units of thawed plasma in the prehospital setting reduced mortality in trauma patients by 10 percent (26). “There's nothing else that we've done in decades of research that has reduced mortality by 10 percent,” Neal said. 

Plasma needs to be thawed because it’s kept frozen, which extends its shelf life from five days as a refrigerated liquid to one year. “You cannot [thaw it] in the middle of the street or in the Rocky Mountains or in the Outback,” said Jose Cancelas, a hematologist at the University of Cincinnati College of Medicine. “Having an accessible plasma product that is shelf stable and portable is, I think, unequivocally on the list of highest priorities for innovation in the trauma resuscitation space,” Neal said. 

Cancelas led a dose escalation clinical trial to evaluate the safety of a freeze-dried plasma product that can be stored at room temperature for one year (27). Researchers extracted plasma from donors, freeze dried it, dissolved it in sterile water, and readministered it to the same donors. Even at the highest dose, they saw no serious adverse events and no increase in mild side effects compared to participants given thawed plasma. The researchers also measured changes in the levels of key clotting proteins in the plasma after collecting, freeze drying, resuspending, and readministering it. While the process degraded small amounts of fragile proteins, the levels of active proteins in vivo were high enough for sufficient clotting activity. It took only about a minute to reconstitute the freeze dried plasma as a liquid.  

Despite these promising results, mass production of freeze-dried plasma is limited by the fact that it is donor dependent. People with type O negative blood have neither A nor B antigens on their red blood cells, making them universal red blood cell donors, but they have both A and B antibodies in their plasma. Those with type AB blood do not have these antibodies against other blood types in their plasma and are therefore universal plasma donors. “But only five percent of the population is AB, so that’s a problem to access significant amounts of that plasma,” Cancelas said. While blood from multiple donors with different blood types can be pooled to approximate a universal plasma, each donor increases the risk of contamination with a virus. Heat or chemicals can be used to kill pathogens in the plasma, but these treatments may destroy delicate proteins.  

In place of plasma derived proteins, several lab grown recombinant clotting factors have been developed to treat hemophilia (28,29). In theory, a synthetic version of plasma could be made entirely from recombinant proteins, but “it's not a trivial matter,” said Maureane Hoffman, a pathologist at Duke University. “Just because you can express one [protein] doesn't mean that you can express another. They’re structurally related, but they have their quirks.” Cost and effort scales with each of plasma’s many proteins. “If it was just a few clotting factors, we could make it,” Cancelas said. “But the problem with plasma is that it’s not five or 10 proteins; it’s tons of proteins.” One reference set catalogs 697 of them (30).

Synthetic protein producing cells hold promise in the development of recombinant plasma. Avi Schroeder, a chemical engineer at the Technion – Israel Institute of Technology, developed a lipid vesicle encapsulating a bacteria based protein expression system. The system releases synthesized proteins through tiny ruptures in the lipid membrane, leaving the vesicle intact to continue producing protein. Schroeder’s team recently developed a synthetic cell that makes recombinant human basic fibroblast growth factor (bFGF) to promote blood vessel formation (31). The cells produced physiologically relevant quantities of bFGF in three hours, triggering the growth of new blood vessels when injected into mice.  

By honing their singular protein manufacturing function, synthetic cells can produce some proteins more efficiently than natural cells. While the researchers didn’t evaluate the lifespan of the protein machinery past six hours, dynamic production of plasma proteins in synthetic cells might extend their circulation time relative to that of a single administration of recombinant plasma proteins. This feature would be especially advantageous as some recombinant clotting factors have shown shorter half-lives than their plasma derived counterparts (29).  

While they haven’t yet used synthetic cells to produce plasma proteins, Schroeder’s team has adapted the platform to express several unique proteins. “Instead of a production plan that can produce one protein or two proteins, you could technically produce almost any protein using the synthetic cell by only changing the DNA code,” Schroeder said. This versatility could also enable the development of plasma tailored for different applications. “You could produce possibly different proteins for different patients only by swapping the DNA,” Schroeder said. For example, Doctor is interested in plasma designed with extra clotting activity for people with uncontrolled bleeding. 

It’s unlikely that any recombinant expression system will be able to churn out 700 proteins, so synthetic plasma may end up semisynthetic. Proteins that don’t fare well during the decontamination treatments could be incorporated recombinantly. Although it’s difficult to predict which manipulations may lead to antibody formation for any given protein, a hybrid of donor derived and recombinant plasma proteins could theoretically be finetuned to reduce the levels of natural antibodies and improve safety.

All together now

While each synthetic blood component could be administered independently, researchers are interested in combining them to form artificial whole blood. “That's the Holy Grail,” Neal said. “When we resuscitate bleeding trauma patients, our standard of care is to initiate by giving whole blood because that's what the patient is bleeding.”

“We know that it is an advantage to combine them because each of them brings a specific function,” Sen Gupta said. “Part of our current ongoing research is to figure out the optimized processes by which they can be combined.”

While Sen Gupta and other researchers are interested in developing a freeze-dried mixture of artificial red blood cells, synthetic platelets, and plasma for point of care whole blood, Takeoka and Sakai have already demonstrated that their products can be administered in tandem for both clotting and oxygen delivery function (32). The researchers induced a lethal liver laceration in rabbits and administered Takeoka’s synthetic platelets (suspended in plasma extracted from the rabbits) after a period of blood loss. Once bleeding had ceased, they treated the rabbits with Sakai’s hemoglobin vesicles, which did not disrupt the clotting activity. Similarly, administering the synthetic platelets before the hemoglobin vesicles did not affect the vesicles’ function, indicating that the two artificial blood components interact favorably in vivo. The team monitored the rabbits for 24 hours and observed that their intervention produced a survival rate similar to the same course of treatment with real platelets in plasma and red blood cells. 

“It’s an estimate of actual blood,” Takeoka said. “It’s a dream for us.”

References

1. Sarkar, S. Artificial blood. Indian J Crit Care Med  12, 140-144 (2008).
2. Learoyd, P. The history of blood transfusion prior to the 20th century− part 1. Transfus Med  22, 308-314 (2012).
3. Learoyd, P. The history of blood transfusion prior to the 20th century− part 2. Transfus Med  22, 372-376 (2012). 
4. Cothren, C., Moore, E.E., Offner, P.J., Haenel, J.B., & Johnson, J.L. Blood substitute and erythropoietin therapy in a severely injured Jehova’s Witness. N Engl J Med  346, 1097-1098 (2002). 
5. Lundy, J.B., Lewis, C.J., Cancio, L.C., & Cap, A.P. Experience with the use of Hemopure in the care of a massively burned adult. Int J Burns Trauma  4, 45-48 (2014). 
6. Levy, J.H. et al. Polymerized bovine hemoglobin solution as a replacement for allogeneic red blood cell transfusion after cardiac surgery: results of a randomized, double-blind trial. J Thorac Cardiovasc Surg  124, 35−42 (2002). 
7. Williams, A.T. et al. Polymerized hemoglobin with increased molecular size reduces toxicity in healthy guinea pigs. ACS Appl Bio Mater  3, 2976-2985 (2020). 
8. Cabrales, P. et al. Effects of the molecular mass of tense-state polymerized bovine hemoglobin on blood pressure and vasoconstriction. J Appl Physiol  107 1548−1558 (2009). 
9. Cuddington, C.T., Wolfe, S.R., & Palmer, A.F.. Biophysical properties of tense quaternary state polymerized human hemoglobins bracketed between 500 kDa and 0.2 μM in size. Biotechnol Progress  38, e3219 (2022).
10. Cooper, C.E. et al. Engineering tyrosine residues into hemoglobin enhances heme reduction, decreases oxidative stress and increases vascular retention of a hemoglobin based blood substitute. Free Radic Biol Med  134, 106-118 (2019). 
11. Varnado, C.L. et al. Development of recombinant hemoglobin-based oxygen carriers. Antioxid Redox Signal  18, 2314-2328 (2013). 
12. Muzzelo, C., Neely, C., Shah, P., Abdulmalik, O., & Elmer, J. Prolonging the shelf life of Lumbricus terrestris erythrocruorin for use as a novel blood substitute. Artif Cells Nanomed Biotechnol  46, 39-46 (2018).
13. Palmer, A.F. & Elmer, J. Biophysical properties of Lumbricus terrestris erythrocruorin and its potential use as a red blood cell substitute. J Funct Biomater  3, 49-60 (2012).
14. Dowd, S & Elmer, J. Lumbricus terrestris erythrocruorin: a novel blood substitute from a terrestrial earthworm. In Blood Substitutes and Oxygen Biotherapeutics (eds Liu, H, Kaye, AD & Jahr, JS) 277-284 (Springer Nature Switzerland, 2022).
15. Defense Advanced Research Projects Agency. Developing Field-Deployable, Shelf-Stable Whole Blood Substitutes to Treat Hemorrhage. At <https://www.darpa.mil/news-events/2021-05-11>. 
16. Mittal, N. et al. ErythroMer (EM), a nanoscale bio-synthetic artificial red cell. In Blood Substitutes and Oxygen Biotherapeutics (eds Liu, H, Kaye, AD & Jahr, JS) 253-265 (Springer Nature Switzerland, 2022).
17. Yuki, Y et al. Efficacy of resuscitative infusion with hemoglobin vesicles in rabbits with massive obstetric hemorrhage. Am J Obstet Gynecol  224, 398.e1-11 (2021).
18. Takase, B., Higashimura, Y., Asahina, H., Ishihara, M., & Sakai, H. Liposome-encapsulated hemoglobin (HbV) transfusion rescues rats undergoing progressive lethal 85% hemorrhage as a result of an anti-arrhythmogenic effect on the myocardium. Artif Organs  45, 1391-1404 (2021).
19. Azuma, H. et al. First-in-human phase 1 trial of hemoglobin vesicles as artificial red blood cells developed for use as a transfusion alternative. Blood Adv  6, 5711-5715 (2022). 
20. Nishikawa, K. et al. Fibrinogen c-chain peptide-coated, ADP-encapsulated liposomes rescue thrombocytopenic rabbits from non-compressible liver hemorrhage. J Thromb Haemost  10, 2137-2148 (2012). 
21. Brown, A.C. et al. Ultrasoft microgels displaying emergent platelet-like behaviors. Nat Mater13, 1108-1114 (2014).
22. Nandi, S. et al. Synthetic Platelet Microgels Containing Fibrin Knob B Mimetic Motifs Enhance Clotting Responses. Adv Therap  4, 2100010 (2021).  
23. Sekhon, U.D.S. et al. Platelet-mimicking procoagulant nanoparticles augment hemostasis in animal models of bleeding. Sci Transl Med  14, eabb8975 (2022). 
24. Girish, A. et al. Platelet-Inspired Intravenous Nanomedicine for Injury-Targeted Direct Delivery of Thrombin to Augment Hemostasis in Coagulopathies. ACS Nano  16, 16292-16313 (2022). 
25. Hickman, D.A. et al. Intravenous synthetic platelet (SynthoPlate) nanoconstructs reduce bleeding and improve ‘golden hour’ survival in a porcine model of traumatic arterial hemorrhage. Sci Rep  8, 3118 (2018). 
26. Sperry, J.L. et al. Prehospital plasma during air medical transport in trauma patients at risk for hemorrhagic shock. N Engl J Med  379, 315-326 (2018).
27. Cancelas, J.A., et al. Characterization and first-in-human clinical dose-escalation safety evaluation of a next-gen human freeze-dried plasma. Transfusion  62, 406-417 (2022).
28. Peyvandi, F., Miri, S., & Garagiola, I. Immune responses to plasma-derived versus recombinant FVIII products. Front Immunol  11, 591878 (2021). 
29. Hoffman, M. & Konkle, B.A. Recombinant factor VIIa: Administration and adverse effects. At <https://www.uptodate.com/contents/recombinant-factor-viia-administration-and-adverse-effects>. 
30. Schenk, S., Schoenhals, G.J., de Souza, G., & Mann, M. A high confidence, manually validated human blood plasma protein reference set. BMC Med Genomics  1, 41 (2008).
31. Chen, G. et al. Implanted synthetic cells trigger tissue angiogenesis through de novo production of recombinant growth factors. Proc. Natl. Acad. Sci. U.S.A.  119, e2207525119 (2022). 
32. Hagisawa, K. et al. Combination therapy using fibrinogen γ-chain peptide-coated, ADP-encapsulated liposomes and hemoglobin vesicles for trauma-induced massive hemorrhage in thrombocytopenic rabbits. Transfusion  59, 3186-3196 (2019).