True blood: Red blood cells manufacturing from stem cells – Interview with Eun Jun Baek

by Alexey Bersenev on March 19, 2012 · 0 comments

in interviews

Post to Twitter Send Gmail Post to LinkedIn

I’m pleased to introduce a researcher from South Korea – Eun Jun Baek. She is a Professor at Hanyang University (Seoul, S. Korea). For the last few years she dedicated her research to red blood cells (RBC) production from stem cells. You can look at her recent work here, here and here. She kindly agreed to answer my questions.


1. Dear Dr. Baek, let me begin from the discussion about the cell source for blood production. According to our recent poll, most professionals are in favor of adult stem cells as the best source for blood manufacturing. What do you think is the best starting material for RBC manufacturing? What are advantages and disadvantages each of the source?

I think the cord blood (CB) is the best source for blood manufacturing in the near future because it is easy to get with various choices of blood types, and the final RBC product is safe to be transfused without worrying about tumor formation. Even though the current technology is unable to fit the huge amount of blood demand, the CB derived RBC products have shown the best results in efficiency of cell generation process, the ease with maintaining stem cell sources, enucleation rate, practically the highest expansion fold to RBCs from one hematopoietic stem cell, and cost effectiveness through the whole process. Also, it would be possible to supply blood to the recipients with the rare phenotypes or irregular antibodies in their serum due to allo-immunization.
If induced pluripotent stem cell (iPSC) or embryonic stem cell (ESC) lines could be established from group O RhD negative Kell negative donors, theoretically it is possible to generate limitless cell sources. However, there are many hurdles to be solved and I have no idea when it will come true.
Recently several papers reported that the expansion fold increased dramatically starting from one CB unit and they could produce 500 units of PRBCs by calculation (Timmins NE, et al. Tissue Engineering Part C 2011). For human clinical trials using the manufactured RBCs, more than 7~10 years might be needed for development of cell expansion and enucleation process, cell storage study, the safety and function examination, and finally, clinical trials. I also hope that researchers will develop better ways to make RBCs from human embryonic stem cells or iPSC which can proliferate potentially limitless.

In case of cord blood, what would be the best starting population – total nucleated cells or CD34+ total or erythroid progenitors?

For the mass production system, starting cell sources and their banking could be managed in two ways. Master cell bank would be CD34+ cells and working cell bank would be basophilic erythroblast which usually takes 8 days of culture starting from CB CD34+ cells. When I tried total nucleated cells as a starting source, the cells in a culture plate are too heterogenous consuming media and cytokines while accumulating metabolic wastes. However, as the process of isolating CD34+ cells needs time (2~4 hrs), workload and costs, developing new methods of generating RBCs starting from total nucleated cells is a still attractive topic.
Once I get CB sample, the most isolated CD34+ cells are aliquoted and frozen in a liquid nitrogen tank and some fresh portions are cultivated. At days 8 to 12, the cell number is too high to handle, some fractions of cells are frozen again. Almost my work was done with frozen CD34+ cells or frozen basophilic erythroblasts and they have showed good viability with full maturation to RBCs.

2. In your recent study, you were able to produce RBCs in clinical-grade conditions and in large quantities. Can you please tell us what those clinical-grade conditions were and what was the scale of production? What are the critical limiting factors in RBC manufacturing?

The reagents used were applicable for human clinical trial. For example, we did not use serum/plasma, or feeder cells, which are usually grown with serum or plasma. During the last several days of culture, we didn’t use any cytokines. Also, animal derived reagents such as bovine albumin or transferrin was not added. Also other added reagent of vitamin C and poloxamer was applicable to human.
In my 2011 paper, CD34+ cells were grown in 6 well plates and from day 8 the cells were moved to the 75T plate. For some assays using cultured RBCs, the number of plates were needed up to 15. The expansion scale in my protocol was not the highest compared the data mentioned in other articles. The final goal was not to maximize the expansion and proliferation of erythroblasts. As the cell viability in terminally maturated phase was very poor especially in the absence of serum/plasma and feeders, we focused to increase viability and enucleation when the cell proliferation was minimal. To maximize the proliferation of immature cells not only for the final enucleated RBCs, we can change the protocols by elevating the cytokine levels for longer culture days and add plasma/serum.
For human trials, the most critical hurdle is limitations in stem cell sources, inefficient proliferation, low enucleation rate, and the high cost through the whole process. However, I believe the researchers would overcome all the problems.

3. What was a result in testing of “shelf storage” of the final product?

Previously, there was no report for shelf storability of manufactured blood, probably because the generated cell condition was not that good. Cell morphology in previous papers demonstrated irregular cytoplasmic membrane with many vacuoles which mean they are almost dead. Also there are few papers with functional assay data. Those weak cells cannot survive longer and could not be transfused. If the cultured cells are strong comparable to donated RBCs, then they could be stored up to 1 month like procured blood. Even though some damaged reticulocytes during culture made the stored cell count suddenly dropped, the generated RBCs could be stored up to 4 weeks.

4. The cost-effectiveness is a big issue in clinical cell manufacturing. If you able to produce 1012 RBC from 1 unit of cord blood (as you mentioned in the study), how much would it cost? I think, the main problem is a large volume of culture medium. How much medium will take to produce 1 unit of RBC for transfusion from 1 unit of cord blood?

If we can make only one unit of packed RBC from one CB unit, then it would be better to transfuse RBCs in the cord blood. To fit the cost effectiveness more than 10 units should be produced from one CB unit. In my study, however, my focus was not expansion of cells but establishing culture conditions at terminal erythropoiesis when the cells are not well proliferating and easily die.
On the contrary, Timmins NE, et al. recently reported that they could produce 500 packed RBC units from one CB unit. Of course, the large volume of culture medium is one of the problems to be solved. Till now, the cost is estimated up to US$8330 (Timmins et al, 2011) due to the high cost of cytokines and medium. Once the blood manufacturing-company could produce those inserted reagents in a large scale, the price of one unit of cultured RBCs might be reduced to US$1000. At this prices, many patients would want to receive the manufactured blood than donated one. The donated blood always has the risk of transfusion transmissible infection and contamination even after the various examinations. Also, transfusion effects could be greater than donated one. (Contrary to the donated blood containing old RBCs, culture blood is homogenously young cells and could survive longer in the recipient, which is very helpful in chronic anemic patient.) Also, if we could produce many batches of RBC units from an embryonic cell line, the patient will minimally exposed to the transfused cells. These advantages outscore the expensive costs of artificial blood.

5. What was the purity of RBC in the final product according your protocol? Did you see any cells- contaminants?

In the final day of culture, there exist enucleated RBCs, extruded nucleus, still proliferating erythroblasts, and a few myeloid cells. To isolate pure RBCs from the mixture, we used leukocyte removal filter which is commonly used in a hospital blood bank. The filter could remove almost other cells and extruded nuclei. Finally, the purity of RBC in the final product could reach up to 100 %. Also, in case using CB as a stem cell source, the very few extruded nuclei and other lineage cells would not harm the blood recipient even without irradiating the product. (In our published article we used just a little fraction of the leukocyte removal filter to minimize the loss of the RBC. Therefore, you can observe some debris of extruded nuclei.)

What would be the criteria for the quality of GMP-manufactured RBC product?

GMP quality for clinical-grade cells should offer sustainable defined quality and safety of cell products, not just producing facilities with clean rooms. To avoid contamination and infections, animal substance-free culture media, feeder-free culturing methods are essential through every process of stem cell isolation, passaging, cell expansion and cryopreservation procedures.
The best criteria is a cell product using a chemically defined culture medium with GMP quality having only human derived substances. To avoid xeno-components such as bovine albumin, I used human transferrin, and human albumin, and serum-free/feeder cell free protocols. Also the storage media contained only chemically defined media and CPDA anticoagulant which is routinely used in donated RBCs.
But regarding ESC or iPSC sources, the GMP criteria is very complicated. Even though the current technology enables to maintain those cell lines in serum free system, the efficiency is problematic in a mass production system. Moreover, there is no FDA approved ESC/iPSC lines. Also the feeder, enzyme, and all reagents should be GMP quality. Even the ideal cell lines are established, the cost will be huge to fit those GMP criteria in those cell lines.

6. The case of the first-in-human RBC transfusion made from cord blood stem cells was published last year. Although exciting, some researchers noted that the therapeutic transfusion based on this protocol still unrealistic. Can you comment on this study and Douay’s group protocol? How close we’re getting to clinical trials now?

They used very small amount of isotope-tagged RBCs. Also, the growing system is very unrealistic in terms of the relatively low expansion fold and unrealistic the whole amount of media needed. But, they are pioneers in this field and have achieved the goal of first human clinical trial. Now, the mass production system should be extensively studied using bioreactors and I’m sure several teams are challenging on this subject. If the expansion system is established, then the cost downsizing and getting permission from FDA would be necessary. Then, phase I, II, and III clinical trials should go on, which also need a huge cost and time. But, I’m sure finally some researchers and companies will fulfill this. I hope we can be thankful to the new pioneers in near future.

7. What role, do you think, artificially produced RBC will play in the future? Will it take a particular niche in transfusion medicine and hematology or potentially could replace blood donations?

Yes, I hope the artificial blood could finally replace some portions of blood donation. If we know we can buy blood from manufacturer, many people require the reward for their donation, and the voluntary sacrifice would markedly drop. The power to motivate blood donations has been that ‘there is no alternative method other than your blood. Your donation saves one life.” If this concept of noble and lofty donation is broken, insufficiency of supply would be worse. And, once the artificial blood could replace the donation, the production capacity would enlarge for routine clinical use and the cost would be lower and lower.

Thank you very much for the interview! I wish you good luck in your research!

All Stem Cell Assays interviews
Suggest us an interview!

{ 0 comments… add one now }

Leave a Comment

Previous post:

Next post: