More Embryonic Stem Cell Hype, Less Reality and Ethics

October 9, 2014

Excitement over a newly-released paper on stem cells making insulin is a tribute to the Harvard stem cell Press Office.


The actual report is quite a bit less earth-shaking than you might be led to believe by the Harvard press office.  The science itself, in a paper from the lab of Dr. Doug Melton published in the journal Cell, provides an incremental improvement in the derivation of functional (insulin-secreting) beta cells.  Melton’s lab developed an improved method to generate millions of insulin-secreting cells from human embryonic stem cells (hESC, which require the destruction of a young human being) and from human induced pluripotent stem cells (hiPSC, the stem cells created from normal skin cells, without using embryos.)  The multistep protocol, which took 4-5 weeks and treatment with eleven different factors, produced insulin-secreting cells which the paper termed “SC-β” cells, that secreted about half the amount of insulin as normal adult beta cells from the pancreas.  Previous attempts resulted in insulin-secreting cells that were immature and more like fetal than adult cells.  In this new report, the authors note that global gene expression analysis showed “SC-β cells made ex vivo are most similar, but not completely identical, to cadaveric beta cells.”  The SC-β cells secreted insulin in response to different glucose levels in the lab dish and when injected into immunocompromised mice.  When the new SC-β cells were tested in a diabetic mouse model, 5 out of 6 mice survived up to 4 months, compared to 1 out of 6 control mice.


Embryonic Stem Cells Unnecessary

The paper itself makes the case that embryonic stem cells are not needed for even this incremental advance or any subsequent work.  The authors tested batches of SC-β cells made from hESC as well as from hiPSC.  The results were equivalent no matter the starting cell type.  So for any future production of SC-β cells, the authors have shown that no embryonic stem cells are necessary.


Unanswered Questions—Transplant Rejection and Safety

The paper and its results do not address some significant questions related to these new SC-β cells—immune rejection and safety (tumor formation).  The cells were tested in immunocompromised mice, so they were free from immune attack.  This will be an issue in any potential treatment if the SC-β cells are derived from hESC.  Use of hiPSC made from a diabetic patient might provide a way around immune attack on the SC-β.


Safety, especially from aberrant cell growth including tumor formation, is always an issue with pluripotent stem cells, especially hESC.  In the mouse experiment, the authors note that large masses of tumors were not seen, but also point out: “A much larger number of transplants and more extensive histological examination will be needed to assess the possibility of undesired cell growth in the grafts.” 


While the Harvard press release discusses testing of an implantation device to protect SC-β cells implanted into mice, this simply makes the point that the issues of immune rejection, as well as keeping the implanted cells from running free in the patient, have not been tackled.  In the end, this combination device is simply a potential cell-based insulin pump, not a cure for diabetes.


Embryonic Stem Cells Questionable

In the past, the obsession with ESC has led to some questionable claims about their abilities to treat diabetes.  Their ability to make authentic insulin, in quantities that would be useful, were first trumpeted and then shown to be incorrect and even artifactual (see, e.g., here and here).  In fact, teratoma formation was often the result or even the inducer of insulin secretion from ESC.


In fact, the high-efficiency production of insulin-secreting cells from hESC and hiPSC has been done before today’s announcement—similar results were published in September 2014 by Rezania et al.  That report also failed to address the questions that the current paper did not address, such as transplant rejection.


Other Ways to Make Insulin-Secreting Cells—No Embryonic Stem Cells Needed

The obsession with ESC continues to make headlines, but not help patients.  Even Melton’s lab has shown various other ways to make insulin-secreting cells, including: stimulating growth of pancreatic beta cells (which improves glucose tolerance) by expression of betatrophin growth factor; direct reprogramming to turn other pancreatic cells into new insulin-secreting cells within the body; and regeneration of insulin-secreting beta cells by the normal pancreas, achieved by stopping the autoimmune attack typical of Type 1 diabetes.


This latter result is important, because it addresses the underlying cause of Type 1 diabetes:  the autoimmune attack on the insulin-secreting cells.  Stopping the autoimmune destruction of beta cells allows the body to regenerate normal, insulin-secreting cells from the body’s own adult stem cells and progenitors.


Other scientists have shown the real promise of this approach. 

Faustman et al. used a simple treatment with BCG to achieve a transient improvement in patients, providing proof of principle for the concept.


Zhao et al. used cord blood-derived adult stem cells to “re-educate” the immune cells of diabetic patients, providing lasting improvement in metabolic control.


The best results thus far for Type 1 diabetic patients has resulted from the collaboration of Voltarelli and Burt, using immunosuppression to remove rogue immune cells followed by transplantation of the patient’s own adult stem cells.  Their success was reported in 2007 and in 2009 in JAMA.  This was able to induce complete remission (insulin independence) in most patients with early onset type 1 diabetes mellitus.  As they noted after publication of their second paper in 2009: “It's the first therapy for patients that leaves them treatment-free — no insulin, no immune suppression for almost five years.”  Sadly, Dr. Voltarelli died in 2012, but his team continues to work on effective patient treatments.


Adult stem cells remain the gold standard for real patient treatments.