25 Mar 2019

Does Alzheimer’s disease retard the birth of new neurons in the hippocampus? Data from mouse models suggests it might, but evidence from human brain has been lacking. In today’s Nature Medicine, researchers led by María Llorens-Martín at the Universidad Autónoma de Madrid describe refinements to postmortem tissue processing that allowed them to reliably detect thousands of newborn neurons in the human hippocampus. Armed with this technique, they found compelling evidence that neurogenesis persists in cognitively healthy people until the end of life, but drops off dramatically as soon as Alzheimer’s pathology takes hold. “This is the first demonstration of this phenomenon in humans,” Llorens-Martín told Alzforum.

Other researchers praised the technical advance. “I was impressed by the rigor they brought to bear on these studies, and I hope other groups adopt the neuropathology standard they have laid out,” Howard Federoff at the University of California, Irvine, told Alzforum.

Orly Lazarov at the University of Illinois at Chicago called the paper important. “This confirms many of the observations we have seen in mouse models, and has implications for therapy,” she said. At the same time, she and others noted that many key questions remain, including what causes neurogenesis to ebb in the Alzheimer’s brain, whether that harms cognition, and whether stimulating the birth of new neurons would ameliorate cognitive symptoms.

The first report of adult neurogenesis in human brain upended long-held dogma that the process simply didn’t happen (Oct 1998 news). Several subsequent studies confirmed the finding, but controversy remained, with other recent work reporting negligible numbers of new neurons in adult brain (Apr 2018 news; Cipriani et al., 2018). 

Llorens-Martín and colleagues hypothesized that variations in how brain tissue is processed and preserved might account for these discrepancies. They undertook an extensive study to find the conditions that gave the strongest antibody staining for doublecortin, the gold-standard marker for immature neurons. First author Elena Moreno-Jiménez determined that the most crucial factor was fixation time; soaking samples for more than 12 hours in paraformaldehyde nearly abolished the doublecortin signal. This creates a problem for detecting neurogenesis, as nearly all brain tissue is fixed for at least 24 to 48 hours, and in some cases may steep for weeks or even months in paraformaldehyde, Llorens-Martín told Alzforum.

With further research, however, the authors discovered they could recover the doublecortin signal from tissue that had been fixed for one or two days. To remove excess fixative, they treated samples with sodium borohydride, which reduces aldehydes to alcohols. Next, they immersed samples in heated citrate buffer, which breaks cross-links between formaldehyde and other proteins, allowing antibodies to bind the protein of interest. Finally, they used a Sudan Black solution to mop up the autofluorescence caused by fixation. This procedure brought back robust doublecortin staining in dentate gyrus, where neurogenesis occurs, while showing little background in non-neurogenic areas of the hippocampus, such as CA1. Based on these findings, the authors developed a standard protocol, consisting of processing brain tissue within 10 hours of death, fixing in 4 percent paraformaldehyde for 24 hours, and treating as described above to recover antigen binding and remove autofluorescence.

Using this procedure, the authors screened dentate gyrus samples from 13 people who died between the ages of 43 and 87. All had been free of any neurological conditions or cognitive decline at the time of death, and their brains were judged to be Braak stage zero. The youngest samples contained about 40,000 doublecortin-positive cells per square millimeter, while the oldest averaged around 25,000. The rate of decline in neurogenesis with age is similar to that reported in other human studies, Llorens-Martín told Alzforum (Knoth et al., 2010; Jun 2013 news). 

Aging vs. AD. Neurogenesis persists into old age in healthy elderly (white circles), but drops sharply when Alzheimer’s pathology is present (red circles). [Courtesy of Moreno-Jiménez et al., Nature Medicine.]

In Alzheimer’s disease brains, it was a different matter. The authors examined dentate gyrus samples from 45 people with AD who had died between age 52 and 97. Their brains spanned Braak stages 1 to 6. At stage 1, these tissues mustered about 20,000 newborn neurons per square millimeter, fewer than the average 25,000/mm² seen in the oldest samples from healthy people. At later Braak stages, the numbers were even lower, averaging around 10,000/mm² (see image above). The data match findings from mice, where amyloidosis correlates with reduced neurogenesis (Mar 2010 news). 

What might cause this? Llorens-Martín noted that neurogenesis falls off even at the earliest Braak stage, before plaques or tangles have taken hold in dentate gyrus. This suggests that neurogenesis defects precede pathology, but more research will be needed to determine if impaired neurogenesis is a cause, a consequence, or a parallel process to proteinopathy, she said. Dennis Selkoe at Brigham and Women’s Hospital, Boston, noted in an email to Alzforum that Aβ oligomers accumulate in hippocampus and are known to harm neurogenesis, which suggests this form of Aβ could be the culprit. “These data once more support starting therapeutics as early as possible,” Selkoe wrote.

Llorens-Martín would also like to know in what way neurogenesis falters. Do progenitor cells slow their proliferation, or do newborn neurons fail to mature or survive? Her group found some evidence for a block in maturation. In AD samples at higher Braak stages, they detected lower levels of markers of more mature neurons, suggesting the cells may become stuck at immature stages. Intriguingly, recent work suggests that ApoE4 can stunt neuron maturation (Aug 2018 news). 

Proliferation changes could also bear some of the blame, however. A recent study from Bruce Yankner and colleagues at Harvard Medical School found that neural progenitor cells generated from people with sporadic AD differentiate more readily than do progenitor cells from healthy controls. This would deplete the progenitor pool, eventually crippling neurogenesis (Feb 2019 news). “The results [from Moreno-Jiménez et al.] are consistent with our recently reported observation,” Yankner wrote to Alzforum (full comment below).

Would stimulating neurogenesis help those with AD? Some evidence from mice suggests so (Nov 2011 news; Sep 2018 news). However, researchers agreed they need more data about what goes wrong in human brain before they can answer this question. Llorens-Martín plans to analyze single-cell proteomic and genomic data from human samples to gather more clues.—Madolyn Bowman Rogers

Comments

1. • Bruce Yankner
     Harvard Medical School
   • Posted: 25 Mar 2019

   This is a rigorous and well-controlled study addressing the controversial issue of adult hippocampal neurogenesis in aging humans. The data demonstrating a relationship between parameters of tissue processing, particularly postmortem interval and duration of fixation, and detection of doublecortin (DCX)-positive immature neurons in the dentate gyrus is clear and may underlie some of the disparate findings on adult human neurogenesis from different groups.

   The results are consistent with our recently reported observation of accelerated differentiation and depletion of neural progenitors at an early stage in induced pluripotent stem cell models of sporadic AD and APOE4 (Meyer et al., 2019). Our findings predict that the neural progenitor pool would be depleted in individuals predisposed to AD, and during aging might reach a tipping point after which neurogenesis declines. The Moreno-Jimenez report did not analyze the neural progenitor population in aging control or AD brains, making it difficult to know if loss of neural progenitors is a cause of the reduced neurogenesis they observe in AD dentate gyrus. They did, however, investigate if altered maturation of immature neurons might be involved and concluded that there was some effect on maturation based on an association between maturation-related markers and the progression of pathology as determined by Braak scores. However, examination of the data (Fig. 4 c-i) suggests that for many of the maturation markers, there is a decline between control and Braak 1, with little difference between Braak 1 and Braak 6 (PSA-NCAM, Prox1, NeuN, and calbindin). This raises the possibility that the major change in neurogenesis occurs very early in the disease. To resolve this issue, it will be important for subsequent studies to analyze the state of the neural progenitor population at different stages of AD, beginning with cases of mild cognitive impairment.

   References:

   Meyer K, Feldman HM, Lu T, Drake D, Lim ET, Ling KH, Bishop NA, Pan Y, Seo J, Lin YT, Su SC, Church GM, Tsai LH, Yankner BA. REST and Neural Gene Network Dysregulation in iPSC Models of Alzheimer's Disease. Cell Rep. 2019 Jan 29;26(5):1112-1127.e9. PubMed.

   View all comments by Bruce Yankner
2. • David Allsop
     Lancaster University, UK
   • Posted: 04 Apr 2019

   Reading this paper reminded me of some data that we published a few years ago on a brain-penetrant retro-inverso peptide inhibitor that blocks Aβ oligomer and fibril formation (see Parthsarathy et al., 2013). One of the most striking findings arising from the repeated injection of APP/PS1 transgenic mice with this peptide inhibitor was the dramatic increase (by 210 percent, p=0.0001) in the mean number of young doublecortin-expressing neurons in the dentate region of the hippocampus, compared to control/saline treated animals. This suggested to us at the time that Aβ oligomers could have a major negative impact on brain neurogenesis. Perhaps this is an aspect of amyloid therapy that has been largely overlooked and deserves further investigation.

   References:

   Parthsarathy V, McClean PL, Hölscher C, Taylor M, Tinker C, Jones G, Kolosov O, Salvati E, Gregori M, Masserini M, Allsop D. A novel retro-inverso peptide inhibitor reduces amyloid deposition, oxidation and inflammation and stimulates neurogenesis in the APPswe/PS1ΔE9 mouse model of Alzheimer's disease. PLoS One. 2013;8(1):e54769. PubMed.

   View all comments by David Allsop

Make a Comment

To make a comment you must login or register.

References

News Citations

1. Humans Sprout New Neurons 22 Nov 2002
2. Newborn Neurons in the Adult Brain: Real Deal, or Glial Imposters? 12 Apr 2018
3. Newborn Neurons Abundant in Adult Human Hippocampus 7 Jun 2013
4. Early Casualty, Neurogenesis Cripples Cognition in AD Mice 19 Mar 2010
5. Alzheimer’s Disease-Related Proteins Needed for Neurogenesis 3 Aug 2018
6. In Alzheimer’s, Too Little REST Spurs Too Much Neurogenesis 4 Feb 2019
7. With a Growth Factor, Neurogenesis and Hippocampus Make a Comeback 4 Nov 2011
8. Exercise Pill? Pharmacological Mimics Boost Cognition in Lazy Mice 9 Sep 2018

Paper Citations

1. Cipriani S, Ferrer I, Aronica E, Kovacs GG, Verney C, Nardelli J, Khung S, Delezoide AL, Milenkovic I, Rasika S, Manivet P, Benifla JL, Deriot N, Gressens P, Adle-Biassette H. Hippocampal Radial Glial Subtypes and Their Neurogenic Potential in Human Fetuses and Healthy and Alzheimer's Disease Adults. Cereb Cortex. 2018 Jul 1;28(7):2458-2478. PubMed.
2. Knoth R, Singec I, Ditter M, Pantazis G, Capetian P, Meyer RP, Horvat V, Volk B, Kempermann G. Murine features of neurogenesis in the human hippocampus across the lifespan from 0 to 100 years. PLoS One. 2010;5(1):e8809. PubMed.

Further Reading

No Available Further Reading