pixabay.com
Introduction
In April 2020, Yang Hui's research group, a researcher at the Institute of Neuroscience of the Chinese Academy of Sciences, published a research paper at Cell. In July 2020, Professor Fu Xiangdong of the University of California reported Yang Hui for "academic ethics misconduct such as plagiarism and suspected fraud." It has caused an uproar in China's domestic academic circles. On February 1, 2021, professors at Harvard University, Hopkins University and other universities (including academicians of the National Academy of Sciences) published articles and raised a variety of questions. Neuroscientists at home and abroad have published public opinions in this article.
Related articles
Exclusive controversy | Shanghai Academic Rising Star was reported, and the reviewer appeared to say that he had rejected the manuscript
Yang Hui issued another statement: He apologized for not communicating the progress of the work in a timely manner
Fu Xiangdong's first response: the controversy with yang hui papers of the Institute of Neurology
Written by | Di Li Hui
● ● ●
In July 2020, a whistleblower letter went viral. Fu Xiangdong, a professor in the Department of Cellular and Molecular Medicine at the University of California, San Diego, reported Yang Hui, a researcher at the Institute of Neurology of the Chinese Academy of Sciences, to the Chinese Academy of Sciences, the Ministry of Science and Technology, and the Foundation Committee, saying that he had "plagiarized and suspected fraud and other academic ethics misconduct." The questioned Yang Hui's paper was published in Cell on April 30, 2020.
After that, Fu Xiangdong confirmed to the Intellectuals that after communicating with the Nerve Institute to provide evidence of DNA primer orders without success, he did deliver the report letter. Although Yang Hui also responded through other media, Fu Xiangdong believed that "obviously avoiding the important is light, and the text is too exaggerated." Implicitly, there was no positive response to his question.
Months later, when the controversy between the two men calmed down, several experts in the field of nerve regeneration still published articles, suspecting that what Yang Hui saw in the Cell paper may be an illusion, and some of the conclusions need to be independently repeated.
So, where is the doubt?
<h1 class="pgc-h-arrow-right" data-track="25" >01 Direct "contradiction"</h1>
The hierarchical structure of the retina and different cell types, the red one is the cone cell, the blue one is the rod cell, the brown one is the retinal ganglion cell, and the purple one is one of the main glial cells of the retina, Müller cell.
Image source Wikipedia https://commons.wikimedia.org/wiki/File:Retina_layers.svg
Of the three most common retinal diseases, older macular degeneration and diabetic retinopathy lead to loss of rods and cone cells as photoreceptors; glaucoma leads to loss of retinal ganglion cells. The loss of these three retinal neurons has caused most of the irreversible blindness. In the United States alone, these three diseases afflict nearly 15 million people, and their prevalence is increasing as the population ages.
Then, if there is a therapy that can regenerate these lost retinal neurons, it has important clinical significance.
In April 2020, Macau University of Science and Technology professors Zhang Kang and Fu Xiangdong co-published an article in which the expression of a certain protein PPTBP1 in Müller cells was knocked down, and it was found that Mueller cells transformed into cone cells, and the light response involved in the cone was significantly restored.
Immediately after the Yang Hui Cell paper published on April 30, different technical methods were used to knock down the expression of the same PPTBP1 in Mueller cells, but mueller cells were converted into retinal ganglion cells.
In the February 1, 2021 issue of The Journal of Clinical Investigation, Johns Hopkins Professor Seth Blackshaw and Harvard University Professor Josh Sanes (hereinafter referred to as JCI Articles) were published, and after examining some of the recent seemingly "powerful" studies, they found many contradictions, including the above two studies, the authors pointed out that -
All of them knocked down the expression of PPTBP1 in retinal Mueller cells, but Mueller cells transformed into different types of neurons — one is a cone cell, the other is a retinal ganglion cell, and this difference needs to be explained.
Both Blackshaw and Sanes are veteran experts on visual phylogenetics, damage, and regeneration, and the latter are members of the National Academy of Sciences.
In fact, in addition to the difference pointed out in the JCI article, some of the conclusions of Fu Xiangdong's Nature paper published on June 24, 2020, and Yang Hui's Cell paper have a more direct contradiction.
For example, Fu Xiangdong reduced the expression of the PBP1 gene through RNA interference technology, and converted astrocytes into dopaminergic neurons in the dense substantia nigra of the midbrain of Parkinson's mice, and further projected onto the striatum.
Yang Hui, through gene editing technology, knocked down the expression of the PBP1 gene, directly in the striatum region of the brain of Parkinson's mice, and converted astrocytes into dopaminergic neurons.
However, in addition to the midbrain, Fu Xiangdong also did more verification in other brain regions, and found that knocking down PBP1 in mouse striatal astrocytes did not regenerate dopaminergic neurons.
This directly conflicts with Yang Hui's conclusions.
"95% of the striatum are GABAergic neurons, all the articles with transdifferentiation get such neurons, and only by adding factors specific to dopaminergic neurons can we get a certain amount of dopaminergic neurons." Experienced people can see at a glance that Yang Hui's cells are all dead cells. Everyone knows that antibodies easily attach to dead cells, so they draw irrational conclusions. Fu Xiangdong told The Intellectual.
Yang's results have also surprised other neuroscientists.
"Neurons in the striatum are mainly GABAergic (neurons), and in such a microenvironment, the glial cells of the striatum are theoretically more likely to become GABAergic neurons. If you want to become a dopamine neuron, I think you should need some specific factors to push. Chen Gong, a professor at the Guangdong-Hong Kong-Macao Institute of Central Nervous System Regeneration at Jinan University, told Intellectuals.
He further stated that if dopamine neurons are regenerated in the striatum, it should be better understood whether these non-existent, sudden appearance of dopamine neurons may have an adverse effect on the local neural circuits of the striatum, and even the broader brain neural circuits.
<h1 class="pgc-h-arrow-right" data-track="25" >02 "crazy long" axon? </h1>
In addition to these direct contradictions, Yang Hui's paper has other surprising places.
In his paper, not only were Mueller cells converted into retinal ganglion cells, but the nascent cellular axons were also able to rapidly project through the optic nerve to the lateral knee nucleus (LGN) and superior thalamus (SC) and repair the visual function of the damaged model animals.
The authors of the JCI paper remind that the regeneration efficiency of axonated retinal ganglion cells in the adult mammalian retina is relatively low, and the number of regenerated axons is also very limited, and it rarely reaches the central targets.
Screenshot of Yang Hui's Cell paper, the axons of retinal ganglion cells pass through the optic nerve and transmit visual signals to the outside of the retina, reaching the lateral knee nucleus (LGN) and superior thalamus (SC) areas of the brain.
"Axon growth takes longer than other intermediate states and should be easier to observe and document. For long-distance axonal projection neurons, the production of new neurons is only the first step in functional recovery, and each step of later axon growth, orientation, synapse formation, and myelin sheath formation is essential. Zhou Fengquan, a professor at the Johns Hopkins University School of Medicine, told Intellectuals that in the field of axon regeneration, how to correctly guide the regeneration of axons back to their original targets in the nervous system of adult animals (lacking many guiding factors in the developmental process and supporting the external environment) is still a difficult problem.
In a review article published in The FEBS Journal in December 2020, Zhou Fengquan believed that the axons of the newly generated retinal ganglion cells in Yang Hui's paper grew too fast and were too numerous, and he specifically elaborated in the paper -
"Within two weeks, Yang Hui studied the axons of newly generated retinal ganglion cells from the outer nuclear layer (ONL) all the way to the superior mound of the midbrain. 7.5 mm per week was grown, and most of the axon regeneration studies reached the optic cross after four weeks, only 1 mm per week, and most of the regenerative axons to the optic cross could not find the path of the target brain nucleus; in addition, in another study that followed the cell transplantation strategy, of all the transplanted P0 mouse retinal ganglion cells with high regeneration ability, only a very small number of transplanted P0 mouse retinal ganglia cells were able to achieve axon regeneration, and those that did extend the axons, after 3 weeks of retinal ganglion cell layer transplantation, only a very limited number of retinal ganglion cells extended the axon to the outlet of the retinal optic nerve, far from the opticus and the metathalamus. ”
Fu Xiangdong believes that Yang Hui's work believes that the optic nerve can grow at a rate of ten times normal development, "completely contrary to scientific common sense." ”
<h1 class="pgc-h-arrow-right" data-track="25" >03 Is it an illusion? </h1>
How to explain these contradictory and even counterintuitive phenomena? Are glial cells really converted into neurons?
After examining many of the current emerging studies of neural in situ regeneration, the two authors of the JCI article, Zhou Fengquan and many other experts in the industry realized that to exclude illusions, more tests need to be done to improve the evidence chain.
In the JCI paper, the authors list three possible "illusions", and Figure B suggests that glial cells may not have transformed into new neurons. Image from reference 1 screenshot of the paper.
The first problem is "leakage".
Our brains, such as the retina, have two main types of cells, neurons and glial cells. The overall goal of these experiments is to turn some of the glial cells into neurons, and the popular method is to use the virus to specifically deliver a certain toolkit (including promoters) to the glial cells for intervention (such as overexpression or knocking out a gene), and in order to track these glial cells, the toolkit also carries fluorescent protein genes. In this way, in principle, it is possible to see which glial cells really become neurons.
But there is a gap between ideals and reality, and in practice, the design of some kits does not guarantee that only glial cells will be manipulated or labeled.
"What you mean is that you think you're designing a reaction that should be 100 percent in the glial cells, but it's not exactly, and the higher the virus titer you use, the more likely the response is to take place inside the neurons, so if you use an excess of virus, it's possible to see the fluorescent proteins that were supposed to be expressed in the glial cells directly inside the neurons, so you think you've seen the new neurons after the transformation, but maybe just the original neurons." Chen Gong explained.
As far as the leakage is concerned, Zhou Fengquan believes that the specificity of the AAV virus and GFAP promoter used in the experiments of Yang Hui et al. is a problem.
"The AAV virus used first infects different kinds of cells within the retina, including ganglion cells (RGCs) and glial cells (such as Mueller cells). At the same time, gene editing specifically for Muller glial cells is mainly through the use of GFAP glial-specific promoters, but GFAP promoter glial cell specificity is not 100% reliable. Genes containing GFAP promoters may also be expressed in neuronal cells. He said.
However, Chen Gong also reminded that the existence of leakage should not completely negate the possibility of glial cells transdifferentiation into neurons.
In practice, his team and many others internationally have used retroviruses to demonstrate the transdifferentiation of glial cells because retroviruses only target dividing glial cells and do not leak into non-dividing neurons. His team also used glial cell lineage tracer mice to demonstrate that glial cells that were previously labeled for tracing could indeed be converted into neurons.
"Therefore, we must comprehensively and objectively evaluate the new technology of in situ nerve regeneration, and we cannot beat the entire field to death because someone has done the wrong experiment." He said.
As a pioneer in this field, Chen Gong has also published articles clarifying some comments. In his view, although AAV has leakage phenomena, as long as the titer is lowered, the leakage ratio can be reduced to a very low level.
"We generally recommend reducing the false positive ratio to 5% or less with a titer of 1x108-10GC/ml. Some laboratories have used excessive amounts of AAV, such as titers in excess of 1x1013GC/ml, and the false positives caused by this high titer are of concern. Some experimental results require third-party verification. ”
Curiously, however, although the different AAVs used in Yang Hui's Cell paper were all "high" titers of 1013, no leaks were observed.
<h1 class="pgc-h-arrow-right" data-track="25" >04 Missing intermediate state? </h1>
Another way to show that transformation does occur is to prove the existence of intermediate states.
Imagine that the transformation of glial cells into nerve cells cannot happen instantaneously, but must have undergone some intermediate process and state - the morphology of the cell may change, the position of the cell is gradually migrating, and the molecules and functions inside the cell are gradually changing.
In the previous whistleblower letter, Fu Xiangdong pointed out that Yang Hui Cell's paper lacked the intermediate state of transformation -
After Yang Hui's "Cell" paper was published online in April, many experts in the field of neurobiology immediately raised a series of questions about the quality and reliability of its data, pointing out that the paper did not have conclusive evidence to prove the cell migration, asymptotic cell morphology and gene expression transformation, and newly acquired neuroelecological properties that must induce newly produced neuronal cells. In layman's terms, since there is no support from a series of experimental data on the process of cell transdifference, it is difficult to say that their conclusions are true and reliable, and it cannot be ruled out that the positive results they see are due to common experimental illusions in the laboratory. ”
Fu Xiangdong believes that most of the evidence supporting Yang Hui's conclusion is likely due to the leaky expression of GFAP-CREs in endogenous neurons.
The process of glial cell transition to neurons is accompanied by migration changes in morphology, transcriptome, and location. (Image from reference 2 paper screenshot)
So, what can be done to be convincing?
"It's like a person taking the elevator from the 10th floor to the 1st floor, always going down the layers, not jumping to the first floor all at once." Our team observed the intermediate state of astrocyte-to-neuronal transformation in both mouse and macaque cerebral cortex, that is, the early transformation of cells with both partial glial and partial properties of neurons, and the existence of this intermediate state is strong evidence of transdifferentiation between cells," Chen explains.
Zhou Fengquan also suggested that the intermediate state of the transition should be captured as intensively as possible.
For example, Mueller glial cells and retinal ganglion cells in the retina are distributed in different layers, and the transformation means that muller glial cells have to "climb" to their destination step by step, and the cell morphology and transcriptome undergo gradual changes (single-cell sequencing can be used to do such an analysis); eventually, the newly transformed retinal ganglion cells need to further track their axon growth, accurate projection to the corresponding brain areas and perform normal functions (current brain tissue cleaning and deep 3D imaging technology can do such observations).
<h1 class="pgc-h-arrow-right" data-track="25" >05 needs to be repeated</h1>
Questioning and criticism are the norm of science and a source of progress, especially in the face of a multiplicity of emerging fields. Many times, it is not easy to distinguish between facts and false appearances, but "truthful reporting" is the basic norm and bottom line that should be.
Despite all the criticisms available, many studies claim to alleviate or treat a neurological disorder in animal models. Obviously, the more reliable the results of basic research, the more favorable it is for clinical translation at a later stage. The "tantalizing" results reported by Yang Hui and others, whether it is the treatment of neurotic blindness or Parkinson's, are not surprising to attract a wider range of attention, and if it is an unimportant result, there will be few people competing.
Zhang Kang, who is also facing JCI questions, told The Intellectual, "Selecting four-month-old or even older rd10 mice has been unable to detect photosensitive cells and retinal electrograph signals, combined with the GFAP-CRE tracking system, we believe that the photoreceptor cells and functions formed later are transformed from Mueller glial cells."
He also admitted that in the later stage, it is necessary to improve the in vivo tracking system, including the use of longer or full-length GFAP promoters, double cres, set up a strict control group, sufficiently intensive sampling time points and comprehensive analysis methods, step by step tracking the transformation process of Muller glial cells into nerve cells, and conduct a comprehensive analysis of the formed cells to prove that the nerve cells seen are really from Mueller glial cells.
Fu Xiangdong also told The Intellectuals that they are designing experiments to comprehensively solve the problem of AAV leakage in response to the JCI article.
At press time, Yang Hui did not respond to intellectual email inquiries.
"I believe that such a windy study will be repeated by someone, and only repetition can make a final judgment." Chen Gong said.
Resources:
1. Fu Xiangdong's first response: The controversy with Yang Hui's paper at the Institute of Neurology began and ended, and http://zhishifenzi.com/column/depthview/9770?category=depth
2. Seth Blackshaw, Joshua R. Sanes. Turning lead into gold: reprogramming retinal cells to cure blindness. J Clin Invest. 2021;131(3):e146134. https://doi.org/10.1172/JCI146134.
3. Qian C, Dong B, Wang XY, Zhou FQ. In vivo glial trans-differentiation for neuronal replacement and functional recovery in central nervous system. FEBS J. 2020 Dec 22. doi: 10.1111/febs.15681. Epub ahead of print. PMID: 33351267.
4. Fu X, et al. Visual function restoration in genetically blind mice via endogenous cel- lular reprogramming [preprint]. https://doi. org/10.1101/2020.04.08.030981. Posted on bioRxiv April 8, 2020.
5. Zhou H, et al. Glia-to-neuron conversion by CRISPR-CasRx alleviates symptoms of neurological disease in mice. Cell. 2020;181(3):590–603.e16.
6. Qian, H., Kang, X., Hu, J. et al. Reversing a model of Parkinson’s disease with in situ converted nigral neurons. Nature 582, 550–556 (2020). https://doi.org/10.1038/s41586-020-2388-4
7. Ge, L.J., Yang, et al. (2020). In vivo Neuroregeneration to Treat Ischemic Stroke Through NeuroD1 AAV-Based Gene Therapy in Adult Non-human Primates. Frontiers in cell and developmental biology 8, 590008.
8. Xiang, Z., Xu, L., Liu, M., Wang, Q., Li, W., Lei, W., and Chen, G. (2021). Lineage tracing of direct astrocyte-to-neuron conversion in the mouse cortex. Neural regeneration research 16, 750-756.
9. Guo, Z., Zhang, L., Wu, Z., Chen, Y., Wang, F., and Chen, G. (2014). In vivo direct reprogramming of reactive glial cells into functional neurons after brain injury and in an Alzheimer's disease model. Cell stem cell 14, 188-202.