The FTLD risk factor TMEM106B controls lysosomal trafficking and dendrite outgrowth

Frontotemporal dementia is the second most common neurodegenerative
disease in people younger than 65 years. Patients suffer from
behavioral changes, language deficits and speech impairment.
Unfortunately, there is no effective treatment available at the
moment. Cytoplasmic inclusions of the DNA/RNA-binding protein
TDP-43 are the pathological hallmark in the majority of FTLD cases,
which are accordingly classified as FTLD-TDP. Mutations in GRN, the
gene coding for the trophic factor progranulin, are responsible for
the majority of familiar FTLD-TDP cases. The first genome-wide
association study performed for FTLD-TDP led to the identification
of risk variants in the so far uncharacterized gene TMEM106B.
Initial cell culture studies revealed intracellular localization of
TMEM106B protein in lysosomes but its neuronal function remained
elusive. Based on these initial findings, I investigated the
physiological function of TMEM106B in primary rat neurons during
this thesis. I demonstrated that endogenous TMEM106B is localized
to late endosomes and lysosomes in primary neurons, too. Notably,
knockdown of the protein does neither impair general neuronal
viability nor the protein level of FTLD associated proteins, such
as GRN or TDP-43. However, shRNA-mediated knockdown of TMEM106B led
to a pronounced withering of the dendritic arbor in developing and
mature neurons. Moreover, the strong impairment of dendrite
outgrowth and maintenance was accompanied by morphological changes
and loss of dendritic spines. To gain mechanistic insight into the
loss-of-function phenotypes, I searched for coimmunoprecipitating
proteins by LC-MS/MS. I specifically identified the
microtubule-binding protein MAP6 as interaction partner and was
able to validate binding. Strikingly, overexpression of MAP6 in
primary neurons phenocopied the TMEM106B knockdown effect on
dendrites and loss of MAP6 restored dendritic branching in TMEM106B
knockdown neurons, indicating functional interaction of the two
proteins. The link between a lysosomal and a microtubule-binding
protein made me study the microtubule dependent transport of
dendritic lysosomes. Remarkably, live cell imaging studies revealed
enhanced movement of dendritic lysosomes towards the soma in
neurons devoid of TMEM106B. Again, MAP6 overexpression phenocopied
and MAP6 knockdown rescued this effect, strengthening the
functional link. The MAP6-independent rescue of dendrite outgrowth
by enhancing anterograde lysosomal movement provided additional
evidence that dendritic arborization is directly controlled by
lysosomal trafficking. From these findings I suggest the following
model: TMEM106B and MAP6 together act as a molecular brake for the
retrograde transport of dendritic lysosomes. Knockdown of TMEM106B
and (the presumably dominant negative) overexpression of MAP6
release this brake and enhance the retrograde movement of
lysosomes. Subsequently, the higher protein turnover and the net
loss of membranes in distal dendrites may cause the defect in
dendrite outgrowth. The findings of this study suggest that
lysosomal misrouting in TMEM106B risk allele carrier might further
aggravate lysosomal dysfunction seen in patients harboring GRN
mutations and thereby contribute to disease progression. Taken
together, I discovered the first neuronal function for the FTLD-TDP
risk factor TMEM106B: This lysosomal protein acts together with its
novel, microtubule-associated binding partner MAP6 as molecular
brake for the dendritic transport of lysosomes and thereby controls
dendrite growth and maintenance.