Scientific RepoRts | 7: 190 | DOI:10.1038/s41598-017-00250-4
misexpression is associated with
Alzheimer’s disease progression
& Thomas P. Sakmar
Recent reports show transcription preference for long genes in neuronal tissues compared with non-
neuronal tissues, and a gene-length dependent change in expression in the neurodevelopmental
disease Rett syndrome (RTT). Whether the gene-length dependent changes in expression seen in
RTT might also be seen in neurodegenerative diseases is not yet known. However, a reasonable
hypothesis is that similar eects might be seen in neurodegenerative diseases as well as in RTT since
a common general feature of both illnesses involves progressive dysfunction of synapses. Here, we
demonstrate a clear length-dependent gene misexpression in the most prevalent neurodegenerative
disease, Alzheimer’s disease. We show that the eect is associated with disease progression and can be
attributed specically to neurons. In particular, we observed gene length-dependent down regulation
on the level of the whole tissue and gene length-dependent up regulation on the level of single cells.
Our analysis shows that a gene-length eect on expression can be found in degenerative neurological
illnesses, such as Alzheimer’s disease. Additional investigation to elucidate the precise mechanism
underlying gene-length dependent changes in expression is warranted.
A recent study by Gabel et al.
showed misexpression of long genes in the Rett syndrome (RTT), a devastating
neurodevelopmental disease caused mainly by mutations in the gene for MECP2 (methyl CpG binding protein 2)
Evidence was presented for the involvement of DNA-methylation in this process. To the best of our knowledge,
this is the only study as of today that shows misexpression of long genes in a human brain condition. Given that
the literature on length-dependent gene expression points at increased expression of long genes in neurons
and that neurons die in neurodegenerative diseases
, we hypothesized that this eect should also be evident in
neurodegeneration in addition to neurodevelopmental conditions. Further support for this hypothesis also comes
from ndings on a class of enzymes that relieve supercoiling during DNA transcription, for example topoisomer-
ases. ese enzymes are associated with synaptic function
and are implicated in a variety of brain disorders
including neurodegenerative disorders
. To investigate the involvement of this eect in neurodegeneration, we
studied available datasets of the most common neurodegenerative disease, Alzheimer’s disease (AD).
In principle, length-dependent gene expression eects might be measured at the level of whole tissues, or
at the level of specic cell populations within a tissue. e former usually, though not necessarily, points at a
dierence in cell composition while the latter points at dierent transcriptional characteristics of one cell type
compared to another.
Still no clear statistical framework exists to measure gene-length dependent eects in samples. We therefore
implemented two statistical approaches in analyzing our observations. We compared the 95% condence inter-
vals of Bootstrap resampling with replacement for the measured data and performed a threshold-dependent
analysis that calls for dierentially expressed genes. Our results show clear gene-length dependent misexpression
in AD at both the whole tissue level and the cell specic levels and clear association with disease progression.
Results and Discussion
While previous studies reported a gene-length dependent expression based on analysis of whole tissues from
mouse and human samples, and separately based on single-cell analysis from cell lines
, we searched for this
eect in specic cell populations from adult human brain dataset that includes 466 cells (GSE67835
), to examine
Laboratory of Chemical Biology and Signal Transduction, The Rockefeller University, 1230 York Ave., New York,
NY, 10065, USA.
Department of Neurobiology, Care Sciences and Society, Division for Neurogeriatrics, Center for
Alzheimer Research, Karolinska Institutet, 141 57, Huddinge, Sweden. Correspondence and requests for materials
should be addressed to S.B. (email: email@example.com) or T.P.S. (email: firstname.lastname@example.org)
Received: 1 December 2016
Accepted: 15 February 2017
Published: xx xx xxxx