Chemical-Only Cell Reprogramming Cocktails Direct Converts Skin Cells into Neurons


Two Chinese laboratories have independently transformed skin cells into neurons using only a cocktail of chemicals. One laboratory used skins cells from Alzheimer’s patients and the other used healthy laboratory mice, and therefore, the protocols developed by each laboratory differ. However, the success of these protocols suggests that it might be economically possible to use neurons made a patient’s own cells to test drug regiments for clinical purposes.

These two studies reinforce the idea that a purely chemical approach represents a promising way to scale up cell reprogramming research that might avoid the technical challenges and safety concerns associated with the more popular method of using transcription factors.

One of the challenges of reprogramming cells to change their identity is that you may end up with cells that look normal on the outside, but inside, many of their internal workings are quite different from the cell type you want to make. In these two papers, neurons made from chemically reprogrammed cells showed neuron-specific gene expression, action potentials, and synapse formation, which is strong evidence that these protocols produce fully operational neurons.

In both cases, the protocols employed decreased the expression of skin-specific genes and increased the expression of neuron-specific genes. These chemicals promoted neuronal cell fates by coordinating multiple signaling pathways that worked together to commit the cells to a neuronal fate. This direct reprogramming procedure bypasses the so-called “proliferative intermediate stage” that put cells under stress and increases the mutation rates. Therefore direct conversion protocols are inherently safer than other reprogramming protocols.

The paper from the laboratory of Jian Zhao (Cell Stem Cell 2015;17(2):204) designed a purely chemical protocol to convert skin cells from human Alzheimer’s disease patients into neurons. Direct reprogramming protocols are available for converting human skin cells into neurons, but these protocols require that cells be transfected with genes that encode transcription factors. Such manipulation requires that cells be treated with viruses or subjected to potentially stressful transfection conditions. This purely chemical protocol is a potentially welcome alternative that would be both safer and easier. The chemicals used in these procedures are easy to synthesize, stable, and standardization of the procedures would also be much easier.

The paper that uses a purely chemical protocol to directly reprogram mouse skin cells comes from the laboratory of Hongkui Deng (Cell Stem Cell 2015;17(2):195) is the culmination of four years of work. The main hurdle was suppressing skin-specific gene expression. Then Dong identified a compound called I-BET151 that suppressed skin cell-specific gene expression. This allowed Deng and his colleagues to successfully reprogram mouse cells with a purely chemical protocol.

The next step for both of these laboratories is to show that, in principle, these chemically reprogrammed cells can be used for therapeutic purposes. Such a proof-of-principle experiment will put direct reprogramming on the map for regenerative medicine in a powerful way.

Direct Reprogramming Cells with Recombinant Proteins


In my opinion, for what it’s worth, we will probably see direct reprogramming take a prominent place in regenerative medicine in the future. It will not be in the near future, but as direct reprogramming becomes better understood and more feasible, it will probably become a central part of the discussion of regenerative medical strategies.

Direct reprogramming, which is also known as lineage conversion, uses cell type-specific transcription factors to convert a mature, adult cell into a different type of mature, adult cell. The cell does not pass through a pluripotent intermediate, and becomes a wholly different type of cell.

Of course, forcing the expression of lineage-specific transcription factors in cells requires that they be treated with recombinant viruses or other such tools. These genetic manipulations present problems for regenerative medicine, since such viruses can cause mutations or cause the introduced genes to be constantly activated, both of which can cause cells to die to grow uncontrollably. Genetically engineering cells needs to be done in a “kinder and gentler” way (to quote George HW Bush).

To that end Dennis Clegg and his colleagues from the Center for Stem Cell Biology and Engineering at UC Santa Barbara have used specially designed proteins to directly cultured retinal pigmented epithelial cells to neurons.

Newly discovered C-end rule (CendR) cell- and tissue-penetrating peptides have a arginine-rich sequence at the end of proteins that allows them to bind particular cell receptors and be internalized into the cell. These CendR peptides bind to the NRP-1 protein and are internalized. Several laboratories have used CendR peptides to increase the efficacy of anti-cancer drugs in experimental cases (see Alberici L, et al (2013) Cancer Res 73:804–812; Sugahara KN, et al. (2010) Science 328:1031–1035; Sugahara KN, et al. (2009) Cancer Cell 16:510–520; and Roth L,, et al. (2012) Oncogene 31:3754–3763).

By tacking a CendR peptide to the end of the Sox2 protein, Clegg and others were able to convert retinal pigmented epithelial (RPEs) cells to neurons. The Sox2 protein is highly expressed in neural progenitor cells. Other studies have shown that Sox2 can reprogram mouse and human fibroblasts to neural stem cells (Ring KL, et al. (2012) Cell Stem Cell 11:100–109). Thus, Sox2 should do the trick.

Making cultured RPE cells from embryonic stem cells is relatively easy to do. Therefore, Clegg and his coworkers made cultured RPEs and then treated them with viruses that expressed Sox2. The cultured RPEs showed conversion to neurons and the expression of neuron-specific genes.

Since they had established that Sox2 could convert RPEs to neurons, they tried recombinant Sox2 protein with the CendR peptide RPARPAR at the end of the protein. After 60 days in culture, the cells expressed a host of neuron-specific genes, and were capable of taking up a dye that only active neurons can take up (FM1-43).

Reprogramming human fetal RPE (hfRPE) cells to neurons using recombinant SOX2 proteins. (A): Efficiency of hfRPE cells to be reprogrammed to neuron-like cells after recombinant proteins was added to the media every 24 hours for 30 days. (B): Efficiency of hfRPE cells to be reprogrammed by adding SOX2-RPARPAR recombinant protein every 48 hours for different time courses. (C): Representative images of hfRPE (fRPE1914) cells during reprogramming to neuron-like cells after 30, 40, and 50 days in culture with SOX2-RPARPAR protein. Scale bars = 100 μm. (D): Representative images of hfRPE (fRPE1914) cells reprogrammed to neuron-like cells expressing neuronal markers, but not an RPE marker (PAX6), using SOX2-RPARPAR protein. Scale bars = 50 μm. Abbreviations: D, days; RPE, retinal pigmented epithelial cells.
Reprogramming human fetal RPE (hfRPE) cells to neurons using recombinant SOX2 proteins. (A): Efficiency of hfRPE cells to be reprogrammed to neuron-like cells after recombinant proteins was added to the media every 24 hours for 30 days. (B): Efficiency of hfRPE cells to be reprogrammed by adding SOX2-RPARPAR recombinant protein every 48 hours for different time courses. (C): Representative images of hfRPE (fRPE1914) cells during reprogramming to neuron-like cells after 30, 40, and 50 days in culture with SOX2-RPARPAR protein. Scale bars = 100 μm. (D): Representative images of hfRPE (fRPE1914) cells reprogrammed to neuron-like cells expressing neuronal markers, but not an RPE marker (PAX6), using SOX2-RPARPAR protein. Scale bars = 50 μm. Abbreviations: D, days; RPE, retinal pigmented epithelial cells.

The efficiency for this experiment was lousy (0.3%) as opposed to the efficiency for the use of recombinant viruses (11%). Nevertheless, this experiment shows that it is possible to directly reprogram cells without using recombinant viruses.

Putting Peps in Your Heps


The liver is a special organ that performs a whole host of essential functions. The liver stores iron, vitamins and minerals; it detoxifies alcohol, drugs, and other chemicals that accumulate in our bloodstreams, and it produces bile (used to dissolve fats so that they can be degraded), and blood-based proteins like clotting factors and albumin. The liver also stores sugar in the form of glycogen. All of these tasks are undertaken by a single cell type, the hepatocyte (otherwise known as a liver cell).

human-liver-diagram

When your liver fails, you get really sick. This was greatly illustrated to me by one of my colleagues where I teach whose wife suffered extensive liver damage as a result of her battle with lupus (short for systemic lupus erythematosus, an autoimmune disease). Now that this dear lady has had a liver transplant, she is a new person. What a difference a healthy liver makes.

What can regenerative medicine do for patients with failing livers? Human pluripotent stem cells, either embryonic stem cells or induced pluripotent stem cells, can be directed to differentiate into liver cells in culture, but the liver cells made by these cells are very immature. They express proteins commonly found in fetal liver cells (for example, alpha-fetoprotein) and they also lack key enzymes associated with adult cells (such as cytochrome P450s). Rashid and others in the Journal of Clinical Investigation (2010; 120: 3127-3136) showed this. The development of three-dimensional culture systems have increased the maturity of such cells, but there is still a long way to go (see T Takebe and others, Nature 2013; 499:481-484 and J Shan and others, Nature Chemical Biology 2013; 9: 514-520).

Two papers from the journal Cell Stem Cell might show a way forward to making mature liver cells for regenerative liver treatments without destroying embryos or even using and pluripotent stem cell lines. These papers utilize the procedure known as “direct reprogramming,” otherwise known as “direct lineage conversion.” Direct reprogramming requires the forced overexpression of particular genes that causes the cells to switch their cell types.

In the first of these papers, Pengyu Huang and his colleagues from the Chinese Academy of Sciences in Shanghai, China overexpressed a three-gene combination in mouse embryonic fibroblasts that converted the cells into hepatocytes at an efficiency of 20% after 14 days in culture. This gene combination, known as 3TF (HNF4/HNF1A/FOXA3), converted the mouse embryonic skin cells into mature liver cells that made blood proteins and drug-processing enzymes. The only problem was that these mature cells could not grow in culture because they were mature. Therefore, Huang and others infected these cells with a virus called SV40, which drove the cells to divide. Now these cells could be grow in culture and expanded for further experiments.

When transplanted into the livers of mice with failing livers, the induced liver cells made by Huang and others restored proper liver function and allowed the mice to survive.

A second paper by Yuanyuan Du and others from the Peking-Tsinghua Center for Life Sciences at Peking University in Beijing, China, used a large gene combination to make mature liver cells from human skin fibroblasts. This gene combination included eight genes (HNF1A/HNF4A/HNF6/ATF5/PROX1/CEBPA/p53 ShRNA/C-MYC) that converted the human skin cells into liver cells after 30 days in culture at an efficiency of nearly 80%. Again, these cells metabolized drugs as they should, made blood proteins, took up cholesterol, and stored glycogen. Du and others compared the gene expression profile of these human induced hepatocytes or “hiHeps” to the gene expression profile of liver cells taken from liver biopsies. While there were differences in gene expression, there was also significant overlap and a large overall similarity. In fact the authors state, “these results indicate that hiHeps show a similar expression profile to primary human hepatocytes.”

Next, Du and others used three different mouse models of liver failure in all three cases, the hiHeps were capable of colonizing the damaged liver of the mouse and regenerating it. Mind you, the hiHeps did not do as good a job as human primary hepatocytes, but they still worked pretty well. This shows that this direct reprogramming protocol, as good as it is, can still be optimized and improved.

These studies show that the production of highly functional human hepatocyte-like cells using direct reprogramming is feasible and represents an exciting step towards the production of a supply source of cells for drug development, and therapies for liver disease.