Researchers may finally have a way to study the effects of progranulin loss of function on TDP-43 pathology in cell culture. Scientists led by Martina de Majo and Erik Ullian at the University of California, San Francisco, knocked out the progranulin gene in organoids grown from induced human neurons and astrocytes. In the February 23 Stem Cell Reports, they reported that the astrocyte GRN knockout spurs TDP-43 into aggregating in the organoids. It also messes up gene splicing in the neurons, even if they carry normal copies of the PGRN gene. This is the first in vitro model to spontaneously develop the TDP-43 pathology seen in frontotemporal dementia and amyotrophic lateral sclerosis and suggests that astrocytes may drive some of that pathology in disease.

  • Knocking out progranulin in cultured neurons barely affects TDP-43.
  • In three-dimensional neuron/astrocyte organoids, it causes TDP-43 pathology.

“It is exciting to see multiple features of PGRN loss-of-function pathology, including TDP-43 loss of function, reproduced in the organoid system,” wrote Thomas Kukar of Emory University, Atlanta. “These results provide convincing evidence that neurons, astrocytes, and likely many cell types are impaired by loss of PGRN function.”

Progranulin supports the function of lysosome function and neuronal survival. In people with loss-of-function mutations in the GRN gene, aggregated and hyperphosphorylated TDP-43 accumulate in the cytosol of neurons. Normally found in the nucleus, this RNA/DNA binding protein regulates gene splicing, in particular the removal of “cryptic” exons (Aug 2015 news). Without TDP-43, mis-spliced proteins interfere with brain and nerve cell function (Feb 2023 news; Apr 2021 news).

Alas, capturing this PGRN/TDP-43 nexus in cells has been difficult. Loss of progranulin function in neural progenitor cells, or induced human neurons, made them more prone to apoptosis but provoked no TDP-43 aggregation (Sep 2011 news; Almeida et al., 2012). Likewise, when de Majo and colleagues knocked out GRN in human induced neurons and astrocytes in co-culture, TDP-43 did not aggregate.

Enter organoids. To better model how GRN loss of function relates to TDP-43 pathology, co-first authors de Majo and Mark Koontz turned to the human induced astrocyte/neuron mini organoid system they had previously developed (Krencik et al., 2017). They knocked out GRN in the neurons, the astrocytes, or both, and then looked for signs of TDP-43 shenanigans four weeks later.

The GRN knockout organoids grew into uniform spheres as quickly as their wild-type counterparts. However, the knockouts made twice as much phosphorylated TDP-43, accumulated TDP-43 in the cytosol, and contained four times as much mis-spliced stathmin 2 as did wild-type organoids (see image below). All these are signs of TDP-43 pathology (Jan 2019 news).

“The extent of the TDP-43 phenotype, and how reproducible the findings were, shocked me” Ullian told Alzforum.

In ALS/FTD, neurons typically harbor TDP-43 aggregates, but the authors did not pinpoint the inclusions to either cell type. Ullian said that only neurons express stathmin 2, therefore he believes the mis-spliced STMN2 indicates faulty TDP-43 in those cells.

Roving TDP-43. While TDP-43 stayed within the nucleus (blue) in organoids of wild-type neurons and astrocytes (left), the protein entered the cytosol if GRN was knocked out in neurons (second from left), astrocytes (third from left), or both cell types (right). [Courtesy of de Majo et al., Stem Cell Reports, 2023.]

Which cells drove the TDP-43 changes? Organoids of progranulin-null neurons and wild-type astrocytes did not accumulate TDP-43 or mis-splice STMN2. However, in organoids of wild-type neurons and GRN knockout astrocytes, TFP-43 aggregated and STMN2 mis-spliced. This phenotype was worse when both cell types lacked progranulin. “Diseased human astrocytes can drive neurodegenerative phenotypes in healthy human neurons,” the authors deduced.

How could that be? RNA-sequencing revealed that these astrocytes expressed abnormally low amounts of the phagocytosis receptor genes MERTK, MEGF10, and AXL, which are crucial for these glia to mop up debris and prune synapses (Jun 2020 news; Apr 2021 news; Dec 2013 conference news). Indeed, GRN knockout astrocytes engulfed fewer fluorescently labeled synaptosomes than did wild-type astrocytes. How this induces TDP-43 pathology in the neurons is unclear, but the authors found that when astrocytes lacked progranulin, synapses in the organoids proliferated, suggesting that the astrocytes were failing to adequately support the neurons.

Next, Ullian plans to create organoids with human iPSCs engineered to have heterozygous GRN loss-of-function mutations, while De Majo will make organoids using stem cells from people who have ALS or FTD. In people, FTD arises from GRN haploinsufficiency. Researchers in Ullian’s lab are also creating the next generation of organoids with neurons, astrocytes, and microglia.—Chelsea Weidman Burke


  1. In this study, the authors differentiated progranulin-deficient iPSCs into 3D organoids that contain both neurons and astrocytes, and observed TDP-43 pathology. Interestingly, progranulin-deficient astrocytes drive TDP-43 pathology. These observations are important, because this in-vitro experimental system recapitulates TDP-43 pathology. Hence it offers an opportunity to investigate the underlying mechanisms, in particular, the contributions of astrocytes to TDP-43 pathology found in various neurodegenerative diseases.

Make a Comment

To make a comment you must login or register.


News Citations

  1. Does New Role for ALS-Linked Protein Help Explain Neurodegeneration?
  2. Can ‘Cryptic Peptides’ Peg People with TDP-43 Pathology?
  3. Sans Nuclear TDP-43, Splicing of An ALS/FTD Gene Goes Awry
  4. Systems Biology Approaches Get Wnt of Progranulin’s Role in FTD
  5. Microtubule Regulator Connects TDP-43 to Axonal Dysfunction
  6. Too Phatal: How Microglia, Astrocytes Snuff Out Dying Neurons
  7. Microglia Build Plaques to Protect the Brain
  8. Glial Cells Refine Neural Circuits

Paper Citations

  1. . Induced pluripotent stem cell models of progranulin-deficient frontotemporal dementia uncover specific reversible neuronal defects. Cell Rep. 2012 Oct 25;2(4):789-98. PubMed.
  2. . Systematic Three-Dimensional Coculture Rapidly Recapitulates Interactions between Human Neurons and Astrocytes. Stem Cell Reports. 2017 Dec 12;9(6):1745-1753. Epub 2017 Nov 30 PubMed.

Further Reading

Primary Papers

  1. . Granulin loss of function in human mature brain organoids implicates astrocytes in TDP-43 pathology. Stem Cell Reports. 2023 Mar 14;18(3):706-719. Epub 2023 Feb 23 PubMed.