NHE6 Paper

Ilie et al.
Molecular Neurodegeneration (2016) 11:63

DOI 10.1186/s13024-016-0129-9
RESEARCH ARTICLE Open Access
A Christianson syndrome-linked deletion

mutation (Δ287ES288) in SLC9A6 disrupts
recycling endosomal function and elicits
neurodegeneration and cell death
Alina Ilie1, Andy Y. L. Gao2, Jonathan Reid1, Annie Boucher1, Cassandra McEwan1, Hervé Barrière1,
Gergely L. Lukacs1, R. Anne McKinney2 and John Orlowski1*
Abstract
Background: Christianson Syndrome, a recently identified X-linked neurodevelopmental disorder, is caused by
mutations in the human gene SLC9A6 encoding the recycling endosomal alkali cation/proton exchanger NHE6.
The patients have pronounced limitations in cognitive ability, motor skills and adaptive behaviour. However, the
mechanistic basis for this disorder is poorly understood as few of the more than 20 mutations identified thus far
have been studied in detail.
Methods: Here, we examined the molecular and cellular consequences of a 6 base-pair deletion of amino acids
Glu287 and Ser288 (ΔES) in the predicted seventh transmembrane helix of human NHE6 expressed in established cell
lines (CHO/AP-1, HeLa and neuroblastoma SH-SY5Y) and primary cultures of mouse hippocampal neurons by
measuring levels of protein expression, stability, membrane trafficking, endosomal function and cell viability.
Results: In the cell lines, immunoblot analyses showed that the nascent mutant protein was properly synthesized
and assembled as a homodimer, but its oligosaccharide maturation and half-life were markedly reduced compared
to wild-type (WT) and correlated with enhanced ubiquitination leading to both proteasomal and lysosomal
degradation. Despite this instability, a measurable fraction of the transporter was correctly sorted to the plasma
membrane. However, the rates of clathrin-mediated endocytosis of the ΔES mutant as well as uptake of companion
vesicular cargo, such as the ligand-bound transferrin receptor, were significantly reduced and correlated with
excessive endosomal acidification. Notably, ectopic expression of ΔES but not WT induced apoptosis when
examined in AP-1 cells. Similarly, in transfected primary cultures of mouse hippocampal neurons, membrane
trafficking of the ΔES mutant was impaired and elicited marked reductions in total dendritic length, area and
arborization, and triggered apoptotic cell death.
Conclusions: These results suggest that loss-of-function mutations in NHE6 disrupt recycling endosomal function
and trafficking of cargo which ultimately leads to neuronal degeneration and cell death in Christianson Syndrome.
Keywords: NHE6/SLC9A6, Christianson syndrome, X-linked intellectual disability, Protein misfolding, Ubiquitination,
Endosomal pH homeostasis, Membrane trafficking, Apoptosis
(Continued on next page)
* Correspondence: [email protected]
1
Department of Physiology, McGill University, Bellini Life Sciences Bldg., Rm,
166, 3649 Promenade Sir-William-Osler, Montreal, QC H3G 0B1, Canada
Full list of author information is available at the end of the article
© 2016 The Author(s). Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0
International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and
reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to
the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver
(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
Ilie et al. Molecular Neurodegeneration (2016) 11:63 Page 2 of 28
(Continued from previous page)

Abbreviations: 7-AAD, 7-amino actinomycin D; AMPAR, α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid
receptor; AP-1, Chinese hamster ovary (CHO) cells deficient in the Na+/H+ antiporter 1 isoform; AP2, Adaptor protein
2; Arf6, ADP-ribosylation factor 6; ARNO, ADP-ribosylation factor nucleotide site opener; AS, Angelman syndrome;
ASK1, Apoptosis signal-regulating kinase 1; BDNF, Brain-derived neurotrophic factor; cCASP3, Cleaved caspase 3
(Asp175); ChFP, Monomeric Cherry fluorescent protein; CS, Christianson syndrome; DMSO, Dimethylsulfoxide;
EEA1, Early endosome antigen 1; EGF, Epidermal growth factor; EGFR, Epidermal growth factor receptor;
EndoH, Endoglycosidase H; ER, Endoplasmic reticulum; ERAD, Endoplasmic reticulum-associated degradation;
ESCRT, Endosomal sorting complex required for transport; FITC, Fluorescein isothiocyanate; FRIA, Fluorescence
ratiometric image analysis; GAPDH, Glyceraldehyde-3-phosphate dehydrogenase; GFP, Green fluorescent protein;
HA, Influenza virus hemagglutinin epitope; IRE1, Inositol-requiring kinase 1; JNK, c-Jun N-terminal kinase; Lamp-
2, Lysosome-associated membrane protein 2; M.I.F., Median intensity fluorescence; NHE6/SLC9A6, Sodium/proton
exchanger isoform 6; NMDAR, N-methyl-D-aspartate receptor; NT-4, Neurotrophin-4; p75NTR, p75 neurotrophin
receptor; PI, Propidium iodide; PNGaseF, Peptide-N-glycosidase F; SDS-PAGE, Sodium dodecyl sulfate polyacrylamide
gel electrophoresis; siRNA, small interfering ribonucleic acids; Tf, Transferrin; TfR, Transferrin receptor;
TrkB, Tropomyosin-related kinase receptor B; Ub, Ubiquitin; UBE3A/E6AP, E3 ubiquitin ligase 3A; UPR, Unfolded
protein response; WT, Wild-type; XLID, X-linked intellectual disability; ΔES, Deletion of Glu287-Ser288
Background NHE6 belongs to the endomembrane subclass of

Christianson syndrome (CS) is a recently described mammalian alkali cation/proton exchangers that are
neurodevelopmental and progressively neurodegenera- sorted differentially to discrete organelles along the
tive disorder characterized by moderate to severe in- secretory and degradative pathways [19, 20]. In the case
tellectual disability, epilepsy, mutism, truncal ataxia, of NHE6, it is widely expressed but most abundant in
hyperkinesis, happy demeanor, postnatal microcephaly excitable tissues such as brain, heart, and skeletal
and is often accompanied by one or more secondary muscle. In non-neuronal cells, the bulk of the trans-
symptoms (e.g., autistic behaviour, eye movement dys- porter accumulates preferentially in a transferrin recep-
function, hypotonia, gastroesophageal reflux, low height tor (TfR)-enriched recycling endosomal compartment
and/or weight, high pain threshold, motor regression, [21–23] where it has been implicated in the regulation
cerebellar vermis and brain stem atrophy as well as of vesicular pH and trafficking [24, 25] and maintenance
neuronal cell loss) [1–3]; a phenotype that partially of epithelial cell polarity [24]. In the central nervous sys-
mimics Angelman syndrome (AS) [4, 5]. However, the tem, NHE6 is expressed in all regions and is particularly
underlying pathogenic gene variants are different. high in the hippocampus, cortex, and Purkinje cell layer of
Whereas AS is caused by defects in the E3 ubiquitin ligase the cerebellum [26–28] (also see Allen Brain Atlas; http//
UBE3A/E6AP gene located at chromosome 15q11.2 [6, 7], mouse.brain-map.org; [29]). Within hippocampal neurons,
CS was recently found to arise from mutations in the gene NHE6 is present in both the soma and neurites, where it
encoding the pH-regulating solute carrier (Na+, K+)/H+ overlaps with known early and recycling endosomal
exchanger NHE6/SLC9A6 located at chromosome Xq26.3 markers and is especially prevalent in vesicles containing
[2, 3, 8, 9]. The frequency of CS amongst X-linked devel- the glutamatergic α-amino-3-hydroxy-5-methyl-4-isoxa-
opmental brain disorders has been estimated at approxi- zolepropionic acid receptor (AMPAR) subunit GluA1 [27]
mately 1 to 2 % [3, 10–12], which places NHE6 amongst and the neurotrophin tropomyosin-related kinase receptor
the more commonly mutated genes associated with X- B (TrkB) which preferentially binds brain-derived neuro-
linked intellectual disability (XLID) [13–17]. The des- trophic factor (BDNF) and neurotrophin-4 (NT-4) [28].
criptions of autistic behaviour in patients with CS also Like AMPAR [30] and TrkB [31], NHE6 is recruited into
parallels recent findings of a significant reduction in dendritic spines (i.e., the postsynaptic compartment of the
NHE6 gene expression in postmortem cerebral cortex majority of excitatory synapses) upon induction of N-me-
from patients with idiopathic autism compared to control thyl-D-aspartate receptor (NMDAR)-dependent long term
tissue [18], further implicating an important role for this potentiation, thereby implicating its involvement in learn-
transporter in disorders of cognitive development. As with ing and memory processes [27]. In a NHE6 null mouse
most X-linked disorders, mutations in NHE6 profoundly model, hippocampal neurons showed reduced axonal and
affect males whereas female carriers are either asymptom- dendritic arborization and decreased circuit activity [28]
atic or display a milder phenotype [1, 3, 8]. as well as a progressive loss of cerebellar Purkinje cells
[26], further linking its participation in processes under- polyclonal anti-HA, monoclonal anti-glyceraldehyde-3-
lying neuronal growth, maturation, function and survival. phosphate dehydrogenase (GAPDH) and monoclonal
A variety of inherited and de novo mutations in NHE6 anti-GFP antibodies were obtained from Abcam Inc.
have recently been identified in CS patients, including (Cambridge, MA); mouse monoclonal anti-Flag M2
frameshifts, nonsense, missense and deletions, although antibody was from Sigma; rabbit polyclonal anti-GFP
their precise consequences on neuronal function have antibody was from Life Technologies. Mouse monoclo-
yet to be elucidated [2, 3, 8, 9, 11, 12, 32–41]. The nal anti-human transferrin antibody was from Invitro-
majority of these mutations cause premature translation gen. Mouse monoclonal anti-β-tubulin was from Sigma.
termination and loss-of-function. However mutations Mouse monoclonal anti-ubiquitin antibody (P4D1) was
that do not truncate the protein may have different con- from Santa Cruz Biotechnology. Rabbit polyclonal anti-
sequences on its membrane targeting and activity which, cleaved caspase-3 (Asp175) antibody was purchased
in conjunction with other genetic factors, may account from Cell Signaling Technology (kindly provided by Dr.
for the phenotypic diversity of patients with CS [3]. For Peter Siegel, McGill University). Horseradish peroxidase-
example, Garbern et al. [9] described an in-frame exci- conjugated secondary IgG antibodies, as well as FITC-
sion of three amino acids (370Trp-Ser-Thr372, ΔWST) in conjugated goat anti-mouse secondary Fab were
the predicted ninth transmembrane helix of NHE6 that purchased from Jackson ImmunoResearch Laboratories
caused the protein to be largely retained in the en- (West Grove, PA). All Alexa Fluor® conjugated secondary
doplasmic reticulum (ER) [42]. While these patients antibodies were purchased from Molecular Probes
displayed many of the pathophysiological and behav- (Eugene, OR). Alpha-minimum essential medium (α-
ioural features ascribed to CS, they were distinguished MEM), fetal bovine serum, penicillin/streptomycin, and
by atypical accumulation of the microtubule protein tau trypsin-EDTA were purchased from Wisent (Saint-Bruno,
in neuronal and glial cells of cortical and sub-cortical QC, Canada). The DMEM/F12 medium was from Corn-
regions and widespread neuronal loss, features that are ing. All other chemical and reagents were obtained from
characteristic of adult-onset neurodegenerative disorders BioShop Canada (Burlington, ON, Canada), Sigma or
such as Alzheimer’s disease [9]. Patients with this Fisher Scientific and were of the highest grade available.
mutation also exhibited only occasional mild microceph-
aly, modest gross motor disability and absence of the Recombinant DNA constructs and mutagenesis
dysmorphic features attributed to the Christianson and The long transcript splice-variant of human NHE6
Angelman-like syndromes. (NHE6v1; NCBI refseq NM_001042537) was cloned from
Here, we examine one of the original CS mutations a human brain Matchmaker™ cDNA library (Clontech)
identified by Gilfillan et al. [8] that results in an in-frame using PCR methodology and was engineered to contain
deletion of two highly conserved amino acids the influenza virus hemagglutinin (HA) (YPYDVPDYAS)
(p.E255_S256del, ΔES) in the predicted seventh trans- epitope at its extreme C-terminal end. This construct was
membrane helix of a short splice-variant of NHE6 (i.e., termed wild-type NHE6HA (NHE6HA-WT) and inserted
NHE6v2). This mutant variant was found to be unstable into the HindIII and XbaI sites of the mammalian expres-
[43], though the precise mechanism of its pathogenicity sion vector pcDNA3 (Invitrogen), as described previously
remains poorly understood. In our study, we established [42]. NHE6HA was then used as a template to engineer the
relevant non-neuronal and neuronal cell model systems to following mutations by PCR mutagenesis: double deletion
gain greater insight into the molecular and cellular conse- mutation of amino acids E287 and S288 (ΔE287/S288,
quences provoked by this mutation, but using the equally ΔES), the conservative double substitution E287Q/S288A,
abundant longest splice-variant of NHE6 (i.e., and the single mutations E287A, E287Q, and S288A.
p.E287_S288del in NHE6v1) as a template for study. Our The same template (NHE6HA) was also used to
results show that though the mutant protein assembles introduce a triple Flag epitope (AAADYKDDDDKGD
properly as a homodimer and is sorted to the plasma YKDDDDKGDYKDDDDKAAA) in the first extracel-
membrane, its oligosaccharide maturation, half-life and lular loop immediately after residue Met53. First, PCR
steady-state abundance are greatly reduced. In addition, was used to engineer an in-frame NotI restriction site
the membrane trafficking of the ΔES mutant and certain after M53, followed by the introduction of annealed
other associated cargo are also impaired, ultimately elicit- primers representing the 3xFlag epitope, which gener-
ing cell death in both non-neuronal and neuronal cells. ated a construct termed 3FNHE6HA. This construct
was further used as a template to introduce the
Methods ΔE287/S288, E287Q/S288A, E287Q, and S288A muta-
Antibodies and reagents tions using PCR mutagenesis.
Mouse monoclonal anti-hemagglutinin (HA) antibody Green fluorescent protein (GFP) C-terminal-tagged
was purchased from Covance Inc. (Berkeley, CA); rabbit forms of NHE6 WT and ΔES mutant were constructed
by insertion between the XhoI and HindIII restriction inhibitor cocktail (Roche Diagnostics). Lysates were
sites of the pAcGFP1-N1 vector (BD Biosciences incubated for 30 min on a rocker at 4 °C, and then
Clontech, Palo Alto, CA). Insertion of the different centrifuged for 20 min at 4 °C to pellet the nuclei and
epitope tags in the various positions did not alter the cellular debris. Twenty μg of protein from the resulting
biochemical properties or cellular distribution of ex- supernatants were eluted in sodium dodecyl sulfate
ogenous NHE6 compared to the endogenous protein (SDS)-sample buffer (50 mM Tris–HCl, pH 6.8, 1 %
[42]. All constructs were sequenced to insure that no SDS, 50 mM dithiothreitol, 10 % glycerol, 1 % bromo-
additional mutations were introduced during PCR. phenol blue), and subjected to 9 % SDS-polyacrylamide
gel electrophoresis (SDS-PAGE), then transferred to
Cell culture polyvinylidene fluoride (PVDF) membranes (Millipore,
Chinese hamster ovary AP-1 cells [44], HeLa, and Nepean, Ontario, Canada) for immunoblotting. The
HEK293 cells were cultured in α-MEM supplemented membranes were blocked with 5 % non-fat skim milk
with 10 % fetal bovine serum, penicillin (100 units/mL), for 1 h, then incubated with the specified primary anti-
streptomycin (100 μg/mL), and 25 mM NaHCO3 bodies (mouse monoclonal HA 1:5000, anti-β-tubulin
(pH 7.4). Human neuroblastoma SH-SY5Y cells were 1:10,000, anti-ubiquitin 1:2000, or GAPDH 1:50,000) in
cultured in high glucose Dulbecco’s Modified Eagle PBS containing 0.1 % Tween 20, followed by extensive
Medium (DMEM)/Ham’s F12 medium supplemented washes and incubation with goat anti-mouse horseradish
with 10 % fetal bovine serum. peroxidase (HRP)-conjugated secondary antibody for
Primary cultures of mouse hippocampal neurons were 1 h. Immunoreactive bands were detected using Western
prepared from post-natal day (PD) 0–2 day C57BL/6 Lightning TM Plus-ECL blotting detection reagents (Per-
and L17 transgenic mice as previously described [27]. kin Elmer Inc., Waltham, MA).
The L17 mice line express membrane-targeted enhanced
GFP (mGFP) under the control of a Thy1.2 promoter Endoglycosidase treatments
cassette in a subset of hippocampal neurons, allowing To obtain post-nuclear supernatants of AP-1 cells
the visualization of cell soma and other neuronal struc- transiently expressing NHE6HA WT or ΔES (24-h trans-
tures. To prepare cultures, the pups were decapitated, fection), cells grown in 10-cm dishes were washed with
their brains were removed, and the hippocampi were ice-cold PBS, collected in 1 ml PBS, and then pelleted at
dissected out. These hippocampi were maintained in 10,000 × g for 4 min at 4 °C. Pellets were resuspended in
chilled HBSS supplemented with 0.1 M HEPES buffer 300 μl sucrose solution (250 mM sucrose, 10 mM
and 0.6 % glucose, then digested with 165 U papain for HEPES-NaOH, 1 mM EDTA, pH 7.5), and passaged 15
20 min at 37 °C. Neurons and glia were dissociated by times through a 26.5-gauge needle. The nuclei and insol-
trituration and suspended in DMEM supplemented with uble cell debris were sedimented at 700 × g for 15 min
1 % penicillin-streptomycin, 10 % FBS, and 0.6 % at 4 °C. The resulting post-nuclear supernatants were
glucose. Cells were then plated onto poly-D-lysine- treated with endoglycosidases, according to the manu-
coated 10 mm coverslips at an approximate density of facturer’s recommendations. First, glycoproteins were
12,000 cells/cm2 and placed in an incubator at 37 °C. denatured in 1x denaturing buffer (0.5 % SDS, 1 % β-
Twenty-four hours later, plating media was then re- mercaptoethanol) at 100 °C for 10 min. Samples were
placed with Neurobasal-A growth media supplemented then divided equally and treated with either Peptide N-
with 2 % B-27 supplement, 1 % GlutaMAX, and 1 % glycosidase F (PNGase F) (750 units, New England
penicillin-streptomycin. Cultures were then fed every 3– Biolabs, Mississauga, ON, Canada) or Endo-β-N-acetyl-
4 days and maintained at 37 °C in a humidified environ- glucosaminidase H (Endo H) (750 units, New England
ment of 95 % air, 5 % CO2. Biolabs, Mississauga, ON, Canada). The enzymes were
added to 30 μl reactions and incubated overnight at
Western blotting 37 °C. Next day, samples were diluted with two-fold
For western blot analyses, AP-1, HeLa or SH-SY5Y cells concentrated SDS-PAGE sample buffer, incubated for
were grown in 10-cm dishes and transiently transfected 30 min at room temperature, briefly centrifuged, and
with 5 μg of plasmid DNA encoding NHE6HA wild-type analyzed by SDS-PAGE and western blotting with
or mutant constructs using Lipofectamine2000™ (Invitro- monoclonal anti-HA antibody.
gen) according to the manufacturer’s recommended
procedure. Cell lysates were prepared following 6 to Cell surface biotinylation
48 h post-transfection (as indicated in the figure legends) AP-1 cells expressing NHE6HA WT or ΔES constructs
by washing cells twice on ice with ice-cold PBS, followed were cultured in 10-cm dishes to sub-confluence, placed
by scraping in 0.5 mL of lysis buffer (0.5 % NP40/0.25 % on ice and washed three times with ice-cold PBS con-
sodium deoxycholate/PBS supplemented with protease taining 1 mM MgCl2 and 0.1 mM CaCl2, pH 8.0 (PBS-
CM). Next, cells were incubated at 4 °C for 30 min with Tris–HCl, pH 6.8, 1 % SDS, 50 mM dithiothreitol,
the membrane-impermeable reagent N-hydroxysulfosuc- 10 % glycerol, 1 % bromophenol blue). The immuno-
cinimydyl-SS-biotin (0.5 mg/mL) (ThermoScientific, precipitated proteins, along with aliquots representing
Rockford, IL). Cells were washed and incubated twice in the total lysate were subjected to 9 % SDS-
quenching buffer (50 mM glycine in PBS-CM) for 7 min polyacrylamide gel electrophoresis (SDS-PAGE), then
each on ice to remove unreacted biotin. After two transferred to polyvinylidene fluoride (PVDF) mem-
more washes in PBS-CM, the cells were lysed for branes (Millipore, Nepean, Ontario, Canada) for im-
30 min on ice, and then centrifuged at 16,000 × g for munoblotting with a mouse monoclonal anti-ubiquitin
20 min at 4 °C to remove insoluble cellular debris. A antibody (Santa Cruz Biotechnology). The membranes
fraction of the resulting supernatant was removed and were then stripped and reblotted with a mouse mono-
this represents the total fraction. The remaining clonal anti-HA antibody (Covance).
supernatant was incubated with 100 μL of 50 % Neu- AP-1 cells were cultured in 3.5-cm dishes and transfected
trAvidin® Agarose Resin slurry (Fisher Scientific, with NHE6HAWT or ΔES using Lipofectamine2000™ (Invi-
Whitby, ON, Canada) in lysis buffer overnight at 4 °C trogen). Twenty-four hours after transfection, cells were
to extract biotinylated membrane proteins. The pro- treated with DMSO as control, the proteasomal inhibitors
teins were then resolved by SDS-PAGE and analyzed MG-132 (40 μM) or lactacystin (30 μM), the lysosomal
by western blotting. inhibitors leupeptin/pepstatin (100 μg/ml) or chloro-
quine (500 μM), all in the presence of cycloheximide
Stability of NHE6 WT and ΔES mutant (150 μg/mL) to prevent new protein synthesis. Cellular
To determine the stability of wild-type and mutant lysates were obtained after 4 and 8 h of treatment and
NHE6, AP-1 cells were grown in 10-cm dishes and equal amounts of proteins were subjected to SDS-
transfected with NHE6HA wild-type or ΔES mutant con- PAGE and western blotting with a mouse monoclonal
structs. Twenty-four hours post-transfection, the cells anti-HA antibody. The membranes were also immuno-
were transferred to 6-well plates and further grown for blotted with mouse monoclonal anti-GAPDH antibody
24 h. To inhibit new protein synthesis, cells were treated as a loading control.
with cycloheximide (150 μg/mL) in α-MEM supple-
mented with 10 % FBS and penicillin/streptomycin for Co-immunoprecipitation
up to 24 h. At specific time points, cells were lysed, To test the hypothesis that WT and ΔES mutant can
protein concentrations were measured, and equal qu- heterodimerize, HeLa cells were cultured in 10-cm
antities of protein were subjected to SDS-PAGE and dishes and transfected with a total of 8 μg of expression
immunoblotting with mouse monoclonal anti-HA and plasmid DNA containing NHE6HAWT, NHE6v1GFPΔES
anti-GAPDH antibodies. The intensity of the bands was or empty vector (pCMV), either singly or in combination
quantified by densitometry of X-ray films exposed in the (4 μg each), using FuGene6 (Promega), according to
linear range and analyzed using ImageJ software. the manufacturer’s instructions. Forty hours post-
transfection, lysates were prepared by washing the
Ubiquitination and protein degradation cells twice with ice-cold PBS and scraping them into
To examine ubiquitination of NHE6 WT and ΔES, AP-1 500 μl of cell lysis buffer (0.5 % Nonidet-P40, 0.25 %
cells expressing NHE6HAWT or ΔES were lysed 24 h sodium deoxycholate, and protease inhibitors in PBS,
after transfection in lysis buffer containing 10 mM N- pH 7.4). Cell lysates were rocked at 4 °C for 30 min
ethylmaleimide (Sigma). Equal amounts of total pro- and centrifuged for 20 min at 16,000 × g to pellet
tein were pre-cleared for 2 h on Protein G-Sepharose cellular debris. Supernatants were pre-cleared for 2 h
beads™ (GE Healthcare). After removing a small frac- on Protein G-Sepharose beads™ (GE Healthcare). The
tion of the pre-cleared cell lysate for immunoblotting, beads were removed by brief centrifugation and a
the remaining lysate was divided into two equal frac- fraction of the cell lysate was removed for immuno-
tions: one fraction was used to immunoprecipitate the blotting. Rabbit polyclonal anti-HA antibody (5 μg)
NHE6 protein overnight at 4 °C with a rabbit poly- was added to the remaining cell lysates and incubated
clonal anti-HA antibody (Abcam); the second fraction with gentle rocking overnight at 4 °C. Subsequently, a
was subjected to immunoprecipitation with a nonspe- 50 % slurry of Protein G-Sepharose beads was added
cific rabbit IgG antibody (Southern Biotech) as a to each tube and incubated with the immunopreci-
negative control. Next day, the lysates were incubated pitates for 2 h at 4 °C. The immunoprecipitated pro-
with a 50 % slurry of Protein G-Sepharose beads for teins, as well as aliquots of initial lysates were
3 h at 4 °C. The beads were subsequently washed and resolved by SDS-PAGE, transferred to polyvinylidene
the immunoprecipitated proteins were eluted in so- fluoride membranes, and immunoblotted with the in-
dium dodecyl sulfate (SDS)-sample buffer (50 mM dicated mouse monoclonal antibodies.
Fluorescence-based endocytosis assay 10 mM NaCl, 20 mM Hepes or 20 mM MES, 1 mM

AP-1 cells were grown in 10-cm dishes and transiently MgCl2, and 0.1 mM CaCl2) with 10 μM nigericin, 10 μM
transfected with 6 μg of 3FNHE6HA WT or ΔES DNA monensin, 0.4 μM bafilomycin and 20 μM carbonyl
constructs. Twenty-four h post-transfection, cells were cyanide m-chlorophenyl hydrazone (CCCP) and recording
transferred to 12 well-plates and grown for an additional the fluorescence ratios.
12 to 24 h. Cells were chilled on ice, washed with ice-
cold PBS-CM, pH 7.4, blocked in 10 % goat serum/ Calcium-phosphate transfection
PBS-CM, and then incubated with a mouse monoclonal Primary neurons were transfected by calcium phosphate-
anti-Flag antibody (1:3000) (Sigma) on ice. Internaliza- mediated transfection [46]. Briefly, at 10–12 DIV, cover-
tion of the bound antibody was initiated by incubating slips were transferred into a 35-mm dish filled with
the cells with warm (37 °C) α-MEM for the indicated warmed preconditioned growth media. For four cover-
time points and terminated by placing the plates on ice. slips, 4 μg of DNA plasmid was mixed with 50 μl 250 mM
Cell were washed and labeled with goat anti-mouse CaCl2 solution, which was then added to 50 μl 2x HEPES-
HRP-conjugated secondary antibody (1:1000) (GE Health- buffered phosphate solution, pH 7.05 to form DNA-
care). After extensive washes with PBS-CM, cells were tagged calcium phosphate precipitate. This was added
treated on ice with Amplex® Red reagent (Invitrogen). Ali- dropwise to each dish of coverslips, which were then incu-
quots were transferred to 96-well plates and fluorescence bated at 37 °C for 90 min. Afterwards, 33 μl of 0.3 M MES
readings were taken with a POLARstar OPTIMA (BMG hydrate solution, pH 5.5, was added to each dish, which
Labtech. Inc, Offenburg, Germany) plate reader using acidified the media to dissolve any remaining precipitate.
544-nm excitation and 585-nm emission wavelengths. All The coverslips were returned to their original plates with
experiments were performed in triplicates and repeated at fresh growth media, and cultures were then maintained at
least three times. Results were expressed as a percentage 37 °C for 48 h before being processed and fixed.
of the fluorescence recorded prior to internalization, after Cultures prepared from C57BL/6 mice were trans-
subtraction of the value measured with mock-transfected fected with GFP and NHE6ChFPWT and NHE6ChFPΔES.
cells. Results are shown as mean ± standard error of the The L17 cultures, which already express myristoylated
mean (S.E.M.). GFP (mGFP), were instead transfected with
NHE6ChFPWT and NHE6ChFPΔES and incubated with
Measurement of endosomal pH Tf-AF633 (100 μg/ml) for 1 h at 37 °C and then fixed.
AP-1 cells were transfected with 3FNHE6HAWT or ΔES
and cell surface resident proteins were labeled on ice Immunofluorescence confocal microscopy
with mouse monoclonal anti-Flag antibody (1:2000 To examine the internalization of NHE6 into transferrin-
dilution, Sigma-Aldrich) and FITC-conjugated goat anti- containing endosomes, AP-1 cells were grown in 10-cm
mouse secondary Fab fragment (1:1000 dilution, Jackson dishes and transfected with 3FNHE6HA WT or ΔES
ImmunoResearch Laboratories, West Grove, PA). Cells mutant. Twenty-four hours post-transfection, the cells
were then washed in PBS-CM and chased for the indi- were transferred to fibronectin-coated glass coverslips and
cated time points in growth medium at 37 °C. Endosomal grown for an additional 12 to 24 h (as specified in text).
pH was measured by single-cell fluorescence ratiometric For the zero time point, cells were serum-depleted for 1 h,
imaging analysis (FRIA) on an Axiovert 100 inverted incubated with 10 μg/ml Alexa Fluor® 488-conjugated
fluorescence microscope (Carl Zeiss MicroImaging, transferrin (Tf-AF488) for 45 min in serum-free media at
Toronto, ON, Canada) at room temperature equipped 37 °C, and then placed on ice. Cells were then labelled for
with a Hamamatsu ORCA-ER 1394 (Hamamatsu, Japan) 1 h on ice with mouse monoclonal anti-Flag antibody
cooled CCD camera and a Planachromat (63 × NA 1.4) (1:2000) in PBS-CM/10 mM glucose/10 mM HEPES,
objective essentially as described previously [45]. Image pH 7.4, washed 3 times with PBS supplemented with
acquisition and FRIA were performed with MetaFluor 1 mM MgCl2 and 0.1 mM CaCl2 (PBS-CM), and then
software (Molecular Devices, Downingtown, PA). Images incubated on ice with goat anti-mouse Alexa Fluor® 568-
were acquired at 490 ± 5 and 440 ± 10 nm excitation conjugated secondary antibody (1:1300). After extensive
wavelengths, using a 535 ± 25-nm emission filter. In each washes with PBS, cells were fixed in 2 % paraformalde-
experiment, the pH of 100–1000 vesicles was determined. hyde/PBS for 20 min at room temperature and
Mono- or multipeak Gaussian distributions of vesicular mounted onto glass slides. For the 60 min endocytosis
pH values were obtained with Origin 7.5 software (Origi- time point, after serum depletion, cells were labeled on
nLab, Northampton, MA). Calibration curves have been ice with the primary anti-Flag and secondary Alexa
performed as previously described [45]. Briefly, in situ Fluor® 568-conjugated antibodies, and then chased with
calibration was performed by clamping the vesicular pH serum free media at 37 °C for 1 h. For transferrin label-
between 4 and 7.5 in K+-rich medium (135 mM KCl, ling, the chasing media contained 10 μg/ml Tf-AF488
during the last 45 min of chase. Cells were subse- Topographical order of neuronal morphology was
quently fixed and mounted. Cells were examined by performed on 3D confocal images of primary hippocam-
laser scanning confocal microscopy using the ZEN soft- pal neurons prepared from C57BL/6 mice. Images were
ware of a Zeiss LSM 710 Meta equipped with a PMT analyzed using the FilamentTracer program (Bitplane
detector, with images acquired using a 63 × /1.4 N.A. AG, Zurich, Switzerland), which semi-automatically de-
oil immersion objective lens. tects 3D neuronal GFP-labeled filament structures and
Primary hippocampal neuronal cultures were fixed calculates parameters such as the number of branch
with 4 % PFA/0.1 M PB, pH 7.4 (Sigma Aldrich) for points, total dendrite length and area.
15 min at room temperature and washed with 0.1 M PB.
C57BL/6 cultures immunoprocessed for cleaved Flow cytometry
caspase-3 (cCASP3) were first permeabilized for 1 min To measure transferrin and EGF uptake by flow cytome-
in 0.2 % Triton X-100/0.1 M PB and blocked for 1 h at try, HeLa cells were transfected with GFP alone, NHE6GFP
room temperature in 0.2 % Triton X-100/1 % HIHS/ WT or ΔES mutant using FuGene6 (Promega). Forty-
0.1 M PB before being incubated with a primary rabbit eight h after transfection, the cells were serum-depleted
polyclonal antibody against cleaved caspase-3 (Asp175) for 2 h, and then incubated with Alexa Fluor® 633-
(cCASP3) (Cell Signaling Technology, 1:300) diluted in conjugated transferrin (Tf-Alexa633, 10 μg/mL) or Alexa
blocking solution overnight at 4 °C. After subsequent Fluor® 647-conjugated EGF (EGF-Alexa647, 100 ng/ml) for
washing, cells were incubated with a secondary goat 5 min at 37 °C, in the absence or presence of 10-fold
anti-rabbit DyLight 649-conjugated secondary antibody excess unlabelled transferrin or EGF, respectively, followed
(Jackson laboratories, 1:1000) diluted in 1 % HIHS/ by washes to remove unbound transferrin or EGF. Cells
0.1 M PB for 45 min at room temperature, washed were detached from the plates by trypsinization and 5 μL
again, and mounted. All coverslips were mounted onto of the cell viability dye 7-amino-actinomycin D (7-AAD,
SuperFrost (Menzel-Glaser) microscope slides using eBioscience) was added to each cell suspension. Cells were
UltraMount fluorescence mounting medium (Dako) and analyzed by flow cytometry using a FACS Aria Sorter
left to dry overnight at room temperature in the dark. (Becton Dickinson, San Jose, CA). A gate was set around
The soma of transfected neurons were then examined the GFP-positive cells and the amount of Tf-Alexa633 or
for cCASP3 staining to assess apoptosis. EGF-Alexa647 taken up by 104 GFP-expressing live cells
Mounted primary hippocampal cultures were im- (i.e., 7-AAD negative) was measured using the BD FACS
aged using a Leica SP2 confocal microscope. Images Diva software.
were acquired using 40X and 63X HCXPL APO oil-
immersion objectives (NAs 1.25 and 1.4, respectively). siRNA Knockdown
GFP was imaged using a 488 nm Ar laser line; HeLa cells were cultured in 6-well plates and transfected
mCherry was imaged using a 543 nm HeNe laser line, with 100 nM control siRNA pool #1 (scrambled siRNA)
and Alexa Fluor 647, Alexa Fluor 633 and DyLight or SMARTpool® NHE6 siRNA (Dharmacon, Lafayette,
649 were imaged using the 633 nm HeNe laser line. CO) using Dharmafect1 transfection reagent (Dharmacon)
Channels were acquired sequentially to prevent spec- according to the manufacturer’s recommended protocol.
tral overlap of fluorophores. Optical sections of 300– Seventy-two hours post-transfection, the cells were
500 nm were taken and frame averaged 3X at low serum-starved for 1.5 h and then incubated with Tf-AF633
resolution or line-averaged 2X at high resolution to for 20 min at 37 °C. 105 cells were analyzed by flow
improve the signal-to-noise ratio. Images were first cytometry per experiment.
deconvolved using Huygen’s Essential software by
using a full maximum likelihood extrapolation algo- Apoptosis assays
rithm (Scientific Volume Imaging), and 3D images Apoptosis was measured using three independent
were compiled as maximum intensity projections methods: (1) a flow cytometry assay that measures
using Imaris software (Bitplane Ag). Colocalization changes in plasma membrane asymmetry (annexin V-allo-
analyses between NHE6 and AF-Tfn were determined phycocyanin conjugate (Annexin V-APC) binding to
using the ImarisColoc algorithm, which generated a phosphatidylserine) and permeability (propidium iodide,
new channel (coloc) containing voxels representing PI); (2) a luminescent assay to detect caspase 3/7 activity;
pixels of channel overlap. This also automatically cal- and (3) an immunofluorescence-immunocytochemistry
culated Mander’s coefficient M1-M2 values between assay to detect activated cleaved caspase-3 (Asp175).
NHE6 and Tf-AF633 channels from a set threshold. For flow cytometry, AP-1 cells were grown in 6-cm
Thresholds were applied relatively consistently be- dishes and transfected with 4 μg of GFP, NHE6v1GFP WT
tween cells in order to remove subjectivity during the or ΔES using Lipofectamine2000 (Invitrogen). Forty-eight
analysis. h post-transfection, the cells were washed twice with
warm (37 °C) PBS, detached using Cell Dissociation Results

Buffer (GIBCO), and then collected by centrifugation. Post-translational processing and stability of NHE6ΔES
The cell pellets were re-suspended in PBS and labeled is impaired
using the Annexin V-APC Apoptosis Detection Kit A foundational study by Gilfillan et al. [8] identified four
(eBioscience) according to the manufacturer’s instruc- mutations in NHE6 (ΔV176-R201, H203fsX59, ΔE287-
tions. After labeling, cells were placed on ice, and 1 × 104 S288, R500X in the longest splice-variant of NHE6,
GFP-positive cells were examined on a BD™LSR II flow NHE6v1, as illustrated in Fig. 1a) in patients affected by
cytometer and the percentage of AnnexinV-positive cells a severe form of X-linked intellectual disability originally
(i.e., early apoptotic), PI-positive (i.e., necrotic), and identified in a multigenerational South African family [1]
AnnexinV/PI – double positive cells (i.e., late apoptotic), - since termed Christianson syndrome. The most subtle
was determined. mutation identified in these patients that would produce
Caspase 3/7 activity was measured using the lumines- a near-intact protein was the two amino acid in-frame
cent Caspase-Glo® 3/7 assay (Promega). AP-1 cells were deletion ΔE287-S288 (ΔES). These amino acids are highly
transfected in 10 cm dishes with 8 μg of GFP vector, conserved in all human NHE isoforms (Additional file 1:
NHE6v1GFP wild-type or ΔES using Lipofectamine2000 Figure S1).
(Invitrogen). Twenty-four hours after transfection, cells To assess the consequences of ΔES deletion on NHE6
were sorted using a Becton Dickinson FACSAria Sorter biosynthesis, the mutation was engineered into NHE6v1
and 5 × 104 GFP-positive cells were plated in each well which contained an HA-epitope at its C-terminus (called
of a 96-well plate. Twenty-four hours later, the assay was NHE6v1HA). This construct was then examined in a
performed according to the manufacturer’s instructions subline of Chinese hamster ovary cells (AP-1 cells) that
and luminescence was read in white-walled 96-well expresses negligible levels of endogenous NHE6 [27].
luminometer plates. Previous light microscopy analyses indicated that human
To measure apoptosis in primary hippocampal neu- NHE6 is properly sorted to the plasma membrane and
rons, immunocytochemistry was performed on fixed recycling endosomes in this cell line [42]. We now show
neurons transfected with cytosolic GFP alone or co- by electron microscopy that the transporter further par-
transfected with GFP and NHE6v1ChFP WT or ΔES titions to microvilli and other small plasma membrane
using a primary rabbit polyclonal antibody against projections (Additional file 1: Figure S2) present in
cleaved caspase-3 (Asp175) (cCASP3) (Cell Signaling Chinese hamster ovary cells [47]; a pattern that mimics
Technology, 1:300) and a secondary goat anti-rabbit the trafficking of NHE6 to dendritic spines of hippocam-
DyLight 649-conjugated secondary antibody (1:1000) pal pyramidal neurons [27]. Thus, these cells provide an
from Jackson laboratories. The soma of transfected ideal tractable model system to study the wild-type
neurons were then examined for cCASP3 staining to (WT) as well as mutant NHE6 transporters in isolation
assess apoptosis. of each other. In addition to AP-1 cells, some experi-
ments were also performed in human neuroblastoma
Real-time PCR (RT-PCR) SH-SY5Y cells which do express NHE6 natively [27].
HeLa cells were grown in 6-well plates and trans- Since NHEs assemble as homodimers [48–50], this
fected with 100 nM siRNA control pool #1 (scram- would allow us to examine the impact of the mutation
bled siRNA) or SMARTpool® NHE6 siRNA (Dharmacon) on the biosynthesis and post-translational processing of
using Dharmafect1 transfection reagent (Dharmacon) either a homodimer (ΔES/ΔES) or potentially heterodi-
according to the manufacturer’s recommended proto- mer (WT/ΔES) in AP-1 and SH-SY5Y cells, respectively.
col. Total RNA was extracted with TRIzolreagent (Invi- Indeed, biochemical and cellular analyses showed that
trogenCanada Inc.). Real-time PCRs were performed the WT and ΔES proteins can form a heterodimer pro-
using IQ SYBR Green Supermix (Bio-Rad Laboratories tein complex by co-immunoprecipitation assays and par-
Inc., ON) and analyzed with Gene Expression Analysis tially colocalize in transfected cells (Additional file 1:
for iCycler iQ® Real-Time PCR Detection System Figure S3). As presented in Fig. 1b and c and consistent
(BioRad). with earlier findings [21, 42], analyses of total cell lysates
from AP-1 and SH-SY5Y cells showed that the WT
Statistical analyses transporter migrated identically in both preparations as
The data represent the mean ± the standard error of multiple bands that reflect the oligomeric and post-
the mean (S.E.M.) and statistical analyses were per- translational oligosaccharide state of the transporter: (1)
formed by using the Student’s t-test or a one-way a slower migrating, high molecular weight band (~200–
analysis of variance (ANOVA) followed by a Tukey 250 kDa) representing the fully-glycosylated dimeric
test. A minimum p-value of <0.05 was considered form of the exchanger that does not fully dissociate
significant. under SDS-PAGE conditions; (2) a major diffuse, fully-
Fig. 1 Expression and post-translational processing of NHE6 ΔES mutant protein is impaired in transfected AP-1 and SH-SY5Y cells. a Schematic
drawing of the predicted membrane topology of mammalian NHE6v1 (based on sequence alignment and transmembrane organization of NHE1
proposed by Wakabayashi et al. [127] and Nygaard et al. [128]) and locations of mutations (red shading) identified by Gilfillan et al. [8] in patients
with Christianson syndrome. The blue shading in the second extracellular loop (EL2) represents the additional 32 amino acids (residues 145–176)
present in the NHE6v1 splice-variant. Two predicted N-glycosylation sites within EL2 are also illustrated. b AP-1 and c, SH-SY5Y cells were transiently
transfected (24 h) with NHE6HA WT or ΔES mutant. Total cell lysates of WT and ΔES-transfectants were examined by SDS-PAGE. The immunoblots were
probed with a mouse monoclonal anti-HA antibody (α-HAm) to detect NHE6v1. NHE6v1 migrates as multiple bands: slower migrating high molecular
weight bands representing the fully-glycosylated (fg) and core-glycosylated (cg) dimeric forms of the exchanger (~200 and 175 kDa, respectively) that
do not fully dissociate under SDS-PAGE conditions and faster migrating fully-glycosylated (fg, ~100 kDa) and core-glycosylated (cg, ~70 kDa) and
unglycosylated (ug, ~65 kDa) forms of the monomeric protein. To control for protein loading, the blots were reprobed with a mouse monoclonal
anti-GAPDH antibody (α-GAPDHm). d, e To confirm the nature of the oligosaccharide modifications of the NHE6 bands, AP-1 cells
transiently transfected (24 h) with WT or ΔES constructs were lysed in non-detergent buffers and post-nuclear supernatants were left
untreated or incubated with either endoglycosidase H (EndoH), which cleaves only asparagine-linked mannose-rich oligosaccharides (i.e.,
core-glycosylated) but not more highly processed complex oligosaccharides (i.e., fully-glycosylated), or peptide-N-glycosidase F (PNGaseF)
which cleaves between the innermost N-acetylglucosamine and asparagine residues of all oligosaccharide structures (i.e., high mannose,
hybrid, and complex). The lysates were then subjected to SDS-PAGE and immunoblotting with an α-HAm antibody. The data are representative of
three independent experiments
glycosylated form(s) of the monomeric protein (~100– monomers. By contrast, the level of expression of the ΔES
120 kDa); and (3) minor faster migrating, lower molecular mutant was noticeably reduced compared to WT in both
weight bands characteristic of the newly synthesized core- cell lines, but especially in AP-1 cells. Densitometric
glycosylated (~70 kDa) or unglycosylated (~65 kDa) analysis of the immunoblots (using multiple exposures of
the immunoblots to ensure the signal intensities of the to promote the proper maturation of NHE6v1, though
bands were within the linear range of the X-ray film) indi- Glu287 appears more critical.
cated that the levels of expression of the ΔES mutant in The marked reduction in abundance of the ΔES
AP-1 and SH-SY5Y cells were 22.1 % ± 3.2 (n = 3) and mutant might be due to impaired biosynthetic matur-
79.8 % ± 4.0 (n = 3) of WT levels (normalized to GAPDH ation, reduced stability, or a combination of the two. To
expression), respectively. Unlike WT, in both cell lines the obtain an approximate measure of their biosynthetic
ΔES mutant migrated predominantly as a single band at maturation, AP-1 cells were transiently transfected with
the level of the immature core-glycosylated or unglycosy- NHE6v1HA WT or ΔES constructs, and cellular lysates
lated monomers. However, fainter diffuse bands corre- were obtained at periodic intervals over 48 h. Quantita-
sponding to the predicted fully-glycosylated monomer tive immunoblot analysis revealed that the WT protein
(~100–120 kDa) and/or the fully- (~250 kDa) and core- is efficiently processed to the fully-glycosylated mature
glycosylated (~175 kDa) dimers were visible, suggesting form (monomer and dimer) within 12 h, such that at
that a fraction of the mutant protein is capable of under- 36 h most of the transiently synthesized protein is fully
going further processing and oligosaccharide maturation. processed (Fig. 2a, c). On the contrary, while protein
The similar patterns of expression of ΔES in both cell lines production and dimer assembly of the ΔES mutant are
also indicate that though the overall abundance of the similar to WT during the first 6–12 h, the subsequent
ΔES mutant was greater in the SH-SY5Y cells, possibly post-translational maturation of the protein is deficient,
due to stabilizing effects mediated by dimerization with en- as revealed by a marked reduction in the addition of
dogenous WT NHE6, the relative oligosaccharide process- complex sugars at 18 h which is more apparent in the
ing of the exogenous ΔES mutant remained compromised. dissociated monomeric form (Fig. 2b, d). Moreover, the
To confirm the post-translational oligosaccharide state abundance of ΔES is greatly decreased at 36 and 48 h
of the proteins, cell lysates of AP-1 cells transiently ex- after transfection relative to WT, suggesting that it might
pressing NHE6v1HA WT or ΔES constructs were prepared be subject to more rapid degradation.
using detergent-free buffers and then subjected to treat- In order to estimate the half-lives of the respective
ment with either endoglycosidase H (EndoH), which proteins, AP-1 cells were transfected with WT or ΔES
cleaves only asparagine-linked mannose-rich oligosaccha- for 24 h and then treated with cycloheximide for an
rides (i.e., core-glycosylated) but not more highly proc- additional 2 to 24 h in order to block de novo protein
essed complex oligosaccharides (i.e., fully-glycosylated), or synthesis while tracking the fate of the previously syn-
peptide-N-glycosidase F (PNGaseF) which cleaves be- thesized transporters. As shown in the immunoblots
tween the innermost N-acetylglucosamine and asparagine presented in Fig. 3a and quantified in Fig. 3b, the WT
residues of all oligosaccharide structures (i.e., high man- protein was relatively stable even after 24 h of treatment
nose, hybrid, and complex). As shown in the immunoblots (t1/2 > 24 h). By contrast, the ΔES mutant was rapidly
presented in Fig. 1d and e for WT and ΔES, respectively, degraded with a calculated half-life of ~2.5 h. As a
EndoH treatment decreased the size of only the core- loading control, the expression of the glycolytic enzyme
glycosylated dimeric (~175 → ~165 kDa) and monomeric glyceraldehyde-3-phosphate dehydrogenase (GAPDH)
(~70 → ~65 kDa) proteins for both WT and ΔES, whereas was measured and found to be invariant over the treat-
the more diffuse fully-glycosylated dimeric (~250 kDa) ment period. The enhanced degradation of the ΔES mu-
and monomeric (~100–120 kDa) bands were unaffected. tant could arise by shuttling the defective protein to
Conversely, PNGaseF removed essentially all the oligosac- proteasomes and/or lysosomes, pathways previously
charide moieties for both WT and ΔES, resulting in implicated in the proteolysis of the shorter NHE6v2-ΔES
smaller dimeric and monomeric bands that migrated at splice variant [43]. Incubating the cells with MG132 or
~165 and ~65 kDa, respectively. lactacystin, two widely used inhibitors of the proteaso-
To assess the relative contribution of Glu287 and mal machinery, concurrently with cycloheximide did not
Ser288 to the observed processing behaviour, we engi- noticeably alter the levels of NHE6 WT over a subse-
neered the following point mutations into NHE6v1: the quent 8-h period compared to diluent (dimethylsulfox-
single conservative (i.e., E287Q) and non-conservative ide, DMSO) controls (Fig. 3c), whereas they markedly
(i.e., E287A or S288A) substitutions as well as the double abrogated the decline in the levels of ΔES (Fig. 3d).
substitution E287Q/S288A. The electrophoretic profiles Similarly, incubating the transfected cells with lysosomal
of the conservative E287Q as well as the S288A mutants inhibitors leupeptin/pepstatin or chloroquine did not
appear similar to WT (Additional file 1: Figure S4). By noticeably affect the abundance of the WT transporter,
contrast, mutants containing the double E287Q/S288A whereas degradation of the ΔES mutant was lessened
and single non-conservative E287A substitutions showed significantly by either treatment regimen. These results
reduced abundance, though not as severe as ΔES. Collect- reveal that deletion of amino acids Glu287 and Ser288 in
ively, these data suggest both residues work synergistically NHE6v1HA drastically decreases the stability of the
Fig. 2 Biosynthetic maturation of NHE6 is reduced for the ΔES mutant. AP-1 cells were transiently transfected with a NHE6v1HA WT or b ΔES
and lysed at the indicated time points over a 48 h period. Equal amounts of proteins were subjected to SDS-PAGE and immunoblotting with a
monoclonal anti-HA antibody (α-HA). The identities of the various NHE6 bands are as described in the legend to Fig. 1. For the ΔES immunoblot
in panel B, a longer X-ray film exposure (18X) of the 36 h and 48 h time points is also shown. The same immunoblots were also probed with a
monoclonal anti-β-tubulin antibody as a loading control. c-d Densitometric quantification of the relative abundances of the monomeric and
dimeric forms of WT or ΔES was assessed using ImageJ software and expressed as ratios of fully glycosylated/total protein (fg/total). For
quantification, multiple exposures of the immunoblots were taken to ensure the signal intensities of the bands were within the linear
range of the X-ray film. Data are shown as mean ± standard error of the mean (S.E.M.) of four different experiments
protein. Moreover, this degradation appears to be per- NHE6 monomers and dimers that would increase their
formed by two pathways; initially by the proteasomal molecular weight, extending from ~75 to >250 kDa.
machinery indicating that disposal of the mutant protein
is mediated by the endoplasmic reticulum-associated Cell surface abundance and internalization kinetics of
degradation (ERAD) pathway [51, 52] and subsequently NHE6ΔES are reduced in AP-1 cells
by the peripheral (i.e., plasma membrane and endo- Although NHE6 WT accumulates in a perinuclear
somes) protein quality control machinery that sorts con- recycling endosomal compartment, a minor fraction (i.e.,
formationally impaired membrane proteins that escape ~5-10 % of total NHE6) resides at the plasma membrane
the ERAD pathway to lysosomes (i.e., endosomal sorting as the transporter transits along the recycling endosomal
complex required for transport (ESCRT)–dependent pathway [22, 42]. To investigate whether the ΔES
lysosomal degradation) [53, 54]. mutant can traffic to the cell surface, plasmalemmal
A common molecular signature of misfolded proteins localization was measured biochemically in transfected
targeted for degradation by the ERAD or peripheral quality AP-1 cells using a cell surface biotinylation assay [58].
control machinery is enhanced multi-monoubiquitination For these experiments, a triple Flag epitope-tag was
or polyubiquitination [52, 55–57]. To test biochemically for inserted in the first extracellular loop of WT and ΔES
increased levels of ubiquitin, NHE6v1HA WT and ΔES were (3FNHE6v1HA-WT and 3FNHE6v1HA-ΔES) as illustrated
transiently expressed (24 h) in AP-1 cells, followed by im- Fig. 5a (upper panel). Insertion of epitopes in this
munoprecipitation with a rabbit polyclonal antibody against position has no discernible effects on the processing,
the HA-epitope (α-HAp) and immunoblotting with a trafficking and function of the transporter [22, 42]. As
monoclonal anti-ubiquitin antibody (α-Ubm). As illustrated shown in the immunoblot in Fig. 5a (lower panel), the
in Fig. 4a and quantified by densitometry in Fig. 4b, the extracted biotinylated cell surface fraction of the WT
levels of ubiquitination of ΔES were markedly increased protein was markedly higher compared to the ΔES
(~13-fold) relative to WT on a protein mass basis, as mutant, and essentially only fully-glycosylated trans-
revealed by stripping the ubiquitin-probed immunoblot porters (intact dimer and dissociated monomer) were
and reblotting with a monoclonal anti-HA antibody detected in the cell surface-enriched fraction. The ΔES
(α-HAm) to detect NHE6v1HA. The visibly diffuse signals mutant protein also reached the cell surface mainly in
for ubiquitin represent various levels of ubiquitination of its dimeric fully-glycosylated form, although a minor
Fig. 3 Stability of NHE6 is diminished for ΔES mutant. a AP-1 cells were transiently transfected with NHE6v1HA WT or ΔES mutant for 24 h and
then treated with 150 μg/mL cycloheximide for the indicated time points, lysed and analysed by SDS-PAGE and immunoblotting with a mouse
monoclonal anti-HA (α-HAm) antibody. Equal amounts of proteins were loaded, as shown by probing the membranes with a monoclonal anti-GAPDH
antibody (α-GAPDHm). b Quantitative analysis by densitometry of NHE6v1 WT and ΔES protein abundance (normalized to GAPDH levels) as a function
of time in the presence of cycloheximide. Values represent the mean ± S.E.M. of three separate experiments. c-d AP-1 cells were transiently transfected
with NHE6v1HA WT (c) or ΔES (d) for 24 h and then treated with 150 μg/mL cycloheximide for the indicated time points in the presence of DMSO
(vehicle), the proteasomal inhibitors MG-132 (40 μM) or lactacystin (LC, 30 μM) (left panels), or the lysosomal inhibitors leupeptin/pepstatin (LeuP,
100 μg/ml) or chloroquine (CQ, 500 μM) (right panels). Cellular lysates were analysed by immunoblotting with a mouse monoclonal α-HAm antibody.
Membranes were also probed with a mouse monoclonal α-GAPDHm antibody as a loading control
amount of the core-glycosylated protein was also were performed with various organellar markers using
detected. The non-biotinylated fractions of WT and detergent permeabilized cells. As expected, the WT trans-
ΔES, representative of their intracellular pools that com- porter highly co-localized with internalized Alexa Fluor®
prise ~90–95 % of total expression, were comparable to 594-conjugated transferrin (Tf-AF594), a marker of recyc-
their respective abundances in the total cell lysates, as ling endosomes that is internalized in an adaptor protein
expected. To ensure that the extracted cell surface 2 (AP2)/clathrin-dependent manner. Conversely, the
biotinylated proteins were not contaminated with intra- fluorescence signals for the ΔES mutant showed minimal
cellular proteins, the immunoblots were probed simul- overlap with Tf-AF594 and, furthermore, the intracellular
taneously with an antibody to GAPDH. GAPDH was accumulation of Tf-AF594 in cells expressing ΔES was
readily detected in the total cell lysates and non- visibly reduced relative to neighboring untransfected cells
biotinylated fractions, but not in the plasmalemmal frac- (Fig. 5c). More quantitatively, by calculating the Mander’s
tions; thereby confirming the selective enrichment of cell overlap coefficient, a statistical parameter that describes
surface proteins. The presence of NHE6v1 WT and ΔES the degree of channel overlap that is not dependent upon
at the plasma membrane was further confirmed visually correlated intensity, the degree of colocalization was
by imaging of fixed, non-permeabilized AP-1 cells significantly reduced for the ΔES mutant compared to
(Fig. 5b). These analyses also showed that the cell WT (i.e., WT-Tf, 0.80 ± 0.05, versus ΔES-Tf, 0.36 ± 0.03,
surface distribution of NHE6 was discontinuous or p < 0.01). Additional subcellular localization analyses
punctate. The basis for this is unclear, but may reflect showed that immunostaining of ΔES mutant had min-
sites of exocytosis and/or endocytosis though other imal overlap with signals for the transfected ER marker
explanations are also possible. KDELGFP [59] and endogenous early endosomal marker
To further characterize the subcellular distribution of EEA1 [60], whereas it overlapped strongly with the
NHE6v1 WT and ΔES, dual immunolabelling experiments transfected late endosomal/multivesicular body marker
Fig. 4 Ubiquitination of NHE6 is increased for the ΔES mutant. a AP-1 cells transiently expressing NHE6v1HA WT or ΔES for 24 h were lysed and
total levels of NHE6 were analyzed by immunoblotting with a mouse monoclonal anti-HA (α-HAm) antibody. GAPDH was measured as a loading
control (panel 1, far left). The level of ubiquitinated proteins in the total cell lysates (TCL) was examined using a mouse monoclonal anti-ubiquitin
(α-Ubm) antibody (panel 2). The TCL were subjected to immunoprecipitation with a non-specific rabbit polyclonal IgG (α-IgGp) antibody (panel 3)
or a rabbit polyclonal anti-HA (α-HAp) antibody (panel 4) and analyzed with a mouse monoclonal α-Ubm antibody to detect the ubiquitination
levels of NHE6 WT versus ΔES. The membrane from panel 4 was then stripped and reprobed with a mouse monoclonal α-HAm antibody to
examine the total amount of NHE6 WT and ΔES retrieved by immunoprecipitation (panel 5). b Quantitative analysis by densitometry of the ratio
of ubiquitinated to total protein abundance of NHE6v1 WT and ΔES in the immunopreciptated pellets. Values represent the mean ± S.E.M. of three
separate experiments. Statistical significance was assessed using a Student’s t-test, * p < 0.01
Rab7GFP [61], suggesting that at 36 h post-transfection panels). After 60 min of internalization, WT was
the bulk of the defective transporter was being re- highly concentrated in the perinuclear Tf-AF488-con-
directed towards the lysosomal degradative pathway taining recycling endosomal pool (Fig. 6a, upper
(Additional file 1: Figure S5). panel, second row). By comparison, the punctate
Having established that a fraction of 3FNHE6HA signals for ΔES mutant and Tf-AF488 were more
ΔES can reside at the cell surface and intracellular dispersed throughout the cell after 60 min of internal-
vesicles, we next assessed whether its rate of internalization rather than coalescing into a more compact
ization was different from that of the WT protein. To perinuclear compartment, but nevertheless were partially
this end, 3FNHE6v1HA WT and ΔES were transiently overlapping. However, similar to results described in
expressed in AP-1 cells and their internalization was Fig. 5c, the accumulation of Tf-AF488 by the ΔES-
examined both visually by confocal microscopy and expressing cells was visibly diminished compared to
quantitatively using a cell-based enzyme-linked im- neighboring non-transfected cells (Fig. 6a, lower panel,
munosorbent assay [62]. For image analysis, the cells second row). Quantitative measurements of NHE6 intern-
were preincubated for 45 min with Alexa Fluor® 488- alization using a cell-based enzyme-linked immuno-
conjugated transferrin (Tf-AF488) to label the recycling sorbent assay showed that the WT transporter was
endosomal compartment, then placed on ice and endocytosed more rapidly than the ΔES mutant (Fig. 6b).
incubated with primary mouse monoclonal anti-Flag Fitting the data to a first order exponential decay function
antibody and Alexa Fluor® 568-conjugated goat anti- yielded time constants of 2.76 ± 0.48 and 4.81 ± 2.68 min
mouse secondary antibody to label cell surface for WT and ΔES, respectively. Collectively, these results
3FNHE6v1HA , followed by internalization of the la- indicate that the sequence 287Glu-Ser288 is important for
belled pool at 37 °C for 60 min. Confocal microscopy the efficient biosynthetic maturation and stability of
analysis of these cells revealed the presence of both NHE6, which in turn seemingly affects the internalization
WT and ΔES at the cell surface before the initiation of NHE6-containing recycling endosomes and associated
of endocytosis (Fig. 6a, first row of upper and lower vesicular cargo.
Fig. 5 Subcellular distribution of NHE6 ΔES is altered in AP-1 cells. a Plasma membrane location of NHE6v1 as measured biochemically using a
cell-surface biotinylation assay. To detect cell surface expression of NHE6v1, a triple Flag epitope-tag was inserted into the first predicted extracellular
loop of NHE6v1HA (3FNHE6v1HA), as illustrated in the upper panel. AP-1 cells were transiently transfected with 3FNHE6v1HA WT or ΔES for 36 h and cell
surface proteins were labeled with biotin as described in ‘Material and Methods’. Total cell lysates (TCL) were prepared and a small portion representing
the total fraction was removed. The remaining supernatants containing equal amounts of total protein for WT and ΔES were loaded onto NeutrAvidin®
Agarose beads to purify the biotinylated cell surface proteins from the non-biotinylated (intracellular) proteins. For the TCL and the remaining
non-biotinylated fractions, aliquots containing 20 μg and 80 μg protein for WT and ΔES, respectively, were examined by Western blotting (left
and right lower panels, respectively). For the plasma membrane fraction, 25 % and 100 % of the biotinylated proteins extracted from the total
cell lysates of WT and ΔES transfectants, respectively, were subjected to Western blotting (middle lower panel). All immunoblots were probed
with mouse monoclonal anti-HAm antibody to detect NHE6 and anti-GAPDHm antibody to assess the enrichment of the biotinylated fraction,
as GAPDH is a cytosolic protein. b Confocal fluorescence microscopy and transmitted light images of fixed non-permeabilized AP-1 cells
showing surface expression of 3FNHE6v1HA WT or ΔES. Scale bars represent 5 μM. c AP-1 cells were transiently transfected with 3FNHE6v1HA
WT (upper panels) or ΔES (lower panels). Thirty-six h after transfection, cells were loaded with Alexa Fluor594-labelled transferrin (Tf-AF594,
10 μg/mL) for 45 min, fixed in 4 % paraformaldehyde, permeabilized, mounted onto glass slides and then examined by confocal microscopy.
Footprints of the transfected cells are indicated as white dotted outlines and were derived from the transmitted light images (far left panels).
Scale bars represent 10 μm
NHE6-mediated stimulation of transferrin, but not EGF, receptor (TfR). To determine the effect of the ΔES
uptake is lost in cells expressing the ΔES mutant mutant on internalization of Tf more quantitatively, we
The above imaging results indicated that the uptake of developed a flow cytometry-based assay to measure the
Tf-AF488 by ΔES-expressing cells was impaired relative uptake of red Alexa Fluor® 633-conjugated Tf (Tf-AF633).
to untransfected neighbouring cells. This observation is For these experiments, HeLa cells were used instead of
consistent with a previous study [25] which reported AP-1 cells because the signal to noise ratio of Tf-AF633
that siRNA knockdown of NHE6 expression in HeLa uptake was much greater (4-fold) in HeLa versus AP-1
cells also decreased internalization of the transferrin cells at the early linear phase of endocytosis (i.e., 5 min
uptake) (Fig. 7a), presumably due to the higher plasma

membrane and total cellular abundance (3- to 4-fold) of
the TfR in HeLa cells (Fig. 7b). To this end, internaliza-
tion of Tf-AF633 was monitored by flow cytometry in live
HeLa cells transiently expressing GFP alone or GFP-
tagged constructs of NHE6 WT or ΔES (Fig. 7c). Live
cells were distinguished from non-viable cells by their
ability to exclude the membrane-impermeant fluorescent
dye 7-amino actinomycin D (7-AAD). The median
intensity fluorescence (M.I.F.) of internalized Tf-AF633
was measured in 104 live (i.e., 7-AAD negative) GFP-
positive HeLa cells. As illustrated in Fig. 7d, cells
expressing WTGFP exhibited a significant increase in Tf-
AF633 uptake compared to GFP-only expressing cells
(GFP: 100 ± 6 %; WTGFP: 145 ± 5 %; p < 0.01, one-way
ANOVA followed by a Tukey test). However, this stimu-
lation was not evident in cells expressing ΔESGFP and, if
anything, showed a slight, albeit statistically non-
significant, decrease to 85 ± 7 % (p = 0.26) of GFP con-
trol levels, suggesting that the mutant is defective in en-
hancing the uptake of Tf-AF633 and possibly other cargo
internalized via AP2/clathrin-dependent endocytosis. To
validate that the measured fluorescent signal was due to
receptor-mediated uptake of Tf-AF633 rather than non-
specific bulk endocytosis, parallel competition experi-
ments were performed in the presence of 10-fold excess
unlabeled Tf. As shown in Fig. 7e, the fluorescent signal
was dramatically reduced to background levels in the
presence of competing unlabeled Tf, suggesting that the
measurements are indicative of receptor-mediated endo-
Fig. 6 Rate of endocytosis of NHE6 ΔES is reduced in AP-1 cells. cytosis of fluorescently-labelled Tf. Moreover, reducing
a AP-1 cells were transiently transfected with 3FNHE6v1HA WT (upper
panels) or ΔES (lower panels) for 36 h, serum-depleted and then
endogenous NHE6 levels in HeLa cells using siRNA
incubated with Alexa Fluor488-conjugated transferrin (Tf-AF488) for (Fig. 7f ) decreased Tf-AF633 uptake by ~23 % (77 ± 2;
45 min. Cells were then placed on ice and incubated with primary p < 0.01, one-sample Student’s t-test) relative to con-
mouse monoclonal anti-Flag antibody and Alexa Fluor568 conjugated trols (Fig. 7g). Hence, up- or down-regulation of NHE6
secondary antibody, and either fixed in 2 % paraformaldehyde expression reciprocally modulates the endocytosis of the
immediately (0 min time point) or after incubation at 37 °C for
60 min. Footprints of the cells within the field of view are indicated
ligand-bound TfR.
as white dotted outlines and were derived from the transmitted Mechanistically, the elevated uptake of Tf-AF633 in
light images (far left panels). After 60 min of internalization, the NHE6GFP-WT expressing HeLa cells could arise from an
signals for the WT transporter highly colocalized with those for enhanced rate of NHE6-dependent endocytosis due to
Tf-AF488, whereas the ΔES mutant showed only partial overlap (the the elevated abundance of NHE6. However, overexpres-
white arrows heads indicate some overlapping punctate signals).
Scale bars represent 10 μm. b Kinetics of endocytosis measured in
sion of NHE6GFP-WT in HeLa cells may also increase
AP-1 cells transiently expressing 3FNHE6v1HA WT or ΔES using a the abundance of the TfR at the cell surface due to
cell-based enzyme-linked immunosorbent assay (ELISA). Thirty-six enhanced recycling or exocytosis of the receptor and
hours post-transfection, surface NHE6 was labeled on ice using a thereby increase net Tf-AF633 uptake, an effect that is
mouse monoclonal anti-Flag antibody, followed by internalization abolished by the ΔES mutation. To examine this latter
of the Flag-labeled NHE6 at 37 °C for the indicated time points.
Remaining cell surface NHE6 was labeled with a goat anti-mouse
hypothesis, TfR cell surface levels were examined using a
HRP-conjugated secondary antibody and detected using the biotinylation assay in HeLa cells expressing GFP, WTGFP
fluorescence Amplex®Red substrate. Data points represent mean ± or ΔESGFP. As shown in Fig. 7h, the levels of TfR at the
S.E.M. of four different experiments, each done in triplicate. Statistical cell surface were significantly higher (~1.6 fold) in HeLa
significance (p < 0.05) was evaluated using a paired two-tailed cells expressing WTGFP as compared to cells expressing
Student’s t-test and indicated by an asterisk
GFP alone. By contrast, in cells expressing ΔESGFP, no
enhanced recruitment of TfR at the plasma membrane
was observed. Thus, NHE6 expression levels appear to
Fig. 7 Uptake of transferrin is impaired in HeLa cells expressing NHE6 ΔES. a Comparison of uptake of transferrin-Alexa Fluor633 (Tf-AF633; 10 μg/
ml, 5 min) in AP-1 vs. HeLa cells. Significance was measured using a one-sample Student’s t-test; *p < 0.05. b Surface expression of transferrin
receptor (TfR) in AP-1 vs. HeLa cells was measured by cell surface biotinylation. A representative immunoblot shows TfR and GAPDH expression in
AP-1 and HeLa cells (left panel). Quantification of surface and total TfR relative to total GAPDH expression from three different experiments; values
are normalized to the TfR/GAPDH ratio in AP-1 cells (right panel). c Transient expression of GFP, NHE6v1GFP WT or ΔES in HeLa cells after 48 h.
Representative immunoblot probed with a polyclonal anti-GFP (α-GFPP) antibody. d Uptake of Tf-AF633 was monitored in HeLa cells expressing
GFP, NHE6v1GFP WT or ΔES. Median fluorescence intensity (M.I.F.) of Tf-AF633 was measured in 104 GFP-positive cells by flow cytometry. Data were
normalized and represent mean ± S.E.M. of eight different experiments. Significance was established using a one-way ANOVA followed by a Tukey
test, **p < 0.01. e Uptake of Tf-AF633 in HeLa cells expressing GFP, NHE6v1GFP WT or ΔES in the absence (−) or presence (+) of 10-fold excess
unlabeled Tf. The M.I.F. of Tf-AF633 was measured in 104 GFP-positive cells by flow cytometry. Data represent mean ± S.E.M. of three different ex-
periments, **p < 0.01. f qPCR of NHE6 mRNA levels in HeLa cells treated for 72 h with non-target (scrambled) siRNA or NHE6 siRNA. g Uptake of
Tf-AF633 was monitored in HeLa cells expressing non-target siRNA or NHE6 siRNA for 72 h. The M.I.F. of Tf-AF633 was measured in 105 cells by flow
cytometry. Data were normalized and represent mean ± S.E.M. of three different experiments. Significance was established using a one-sample
Student’s t-test, **p < 0.01. h Cell surface levels of TfR in HeLa cells transiently expressing GFP, NHE6v1GFP WT or ΔES were determined by cell
surface biotinylation. Representative immunoblot of TfR expression at the cell surface and the total fraction is shown in the left panel. Quantification
by densitometry of cell surface/total TfR levels from three different experiments is shown in the right panel. Values were normalized to TfR amounts
present in GFP-expressing cells and represent the mean ± S.E.M. Significance was established using a one-sample Student t-test, ** p <0.01, * p <0.05
upregulate the surface abundance and recycling of that the EGF receptor (EGFR) could follow different
certain cargo. internalization pathways depending on the concentration
While NHE6 is seemingly important for the internal- of its ligand EGF. Accordingly, at low concentrations of
ization of Tf-TfR, the study by Xinhan et al. [25] also EGF (1–2 ng/ml), the EGFR was endocytosed via a
indicated that this effect was somewhat selective, as up- clathrin-dependent route, whereas at higher ligand
take of other clathrin-mediated cargo such as the epider- concentrations (10–100 ng/ml), the receptor partitioned
mal growth factor (EGF) at a concentration of 1 ng/ml roughly equally between clathrin-coated pits and ca-
was unaffected. However, an earlier study [63] proposed veolae, suggesting that it can be internalized by both
clathrin- and lipid raft-dependent mechanisms [63, 64]. consistent with earlier findings by Xinhan et al. [25] and
While subsequent studies have disputed these latter confirm that NHE6 regulates the internalization of some,
findings [65], we nevertheless tested whether NHE6 but not all, clathrin (or potentially caveolin)-dependent
could internalize the ligand-activated EGFR at higher cargo, at least in HeLa cells.
EGF concentrations that involve not only clathrin medi-
ated endocytosis, but could also potentially favour a The pH of NHE6ΔES-containing endosomes fails to
caveolae-dependent route, and whether the ΔES mutant alkalinize
impairs this function. To this end, we measured the The impaired uptake of Tf-AF633 in cells expressing
uptake of EGF-AF647 at the higher concentration of NHE6ΔES suggested that recycling endosomal function
100 ng/ml in HeLa cells. As shown in Fig. 8a, the signals has been compromised. Since proper acidification of
for neither NHE6 WTGFP nor ΔESGFP overlapped with organelles is an important determinant of membrane
those for the EGFR labelled with EGF-AF647 by image trafficking [66–68] and because NHE6 is believed to
analysis. Consistent with this observation, overexpress- operate as an alkalinizing mechanism to counter the
ing WTGFP or ΔESGFP did not alter the uptake of EGF- acidification generated by the vacuolar H+-ATPase [25],
AF647 when compared to control cells expressing GFP we measured the vesicular pH (pHv) of NHE6 WT- and
alone (Fig. 8b). The specificity of the EGF fluorescent ΔES-containing endosomes by fluorescence ratiometric
signal was validated by competition experiments in the image analysis (FRIA) [69]. To this end, AP-1 cells were
presence of 10-fold excess unlabeled EGF (Fig. 8c). Thus, transfected with 3FNHE6v1HA WT or ΔES. Thirty-six h
even at high concentrations of EGF, NHE6 does not post-transfection, cell surface NHE6 molecules were
colocalize with the EGF-EGFR complex. These data are labeled on ice with a primary mouse monoclonal anti-
Fig. 8 Expression of NHE6 does not affect trafficking of the EGF receptor in HeLa cells. a HeLa cells were transiently transfected with NHE6v1GFP
WT (upper panels) or ΔES (lower panels). Forty-eight h after transfection, cells were loaded with Alexa Fluor647-labelled EGF (EGF-AF647, 100 ng/mL)
for 5 min, fixed in 4 % paraformaldehyde, permeabilized, mounted onto glass slides and then examined by confocal microscopy. Footprints of
the transfected cells within the field of view are indicated as white dotted outlines and were derived from the transmitted light images (far left
panels). Scale bars represent 10 μm. b Uptake of EGF-AF647 (100 ng/ml, 5 min) in 104 GFP-positive HeLa cells expressing GFP, NHE6v1GFP WT or
ΔES measured by flow cytometry. Data were normalized and represent mean ± S.E.M. of five different experiments. c Uptake of EGF-AF647 (100 ng/ml,
5 min) in HeLa cells expressing GFP, NHE6v1GFP WT or ΔES in the absence (−) or presence (+) of 10-fold excess unlabeled EGF (1 μg/ml). Median
fluorescence intensity (M.I.F.) of EGF-AF647 was measured in 104 GFP-positive cells by flow cytometry and values represent the mean ± S.E.M. of three
different experiments. Significance was established using a one sample Student t-test, **p < 0.001
Flag antibody, followed by incubation with a Fab second- compartment (pHv ~ 5.37 ± 0.25) (Fig. 9c) more consist-
ary antibody conjugated to the pH-sensitive ratiometric ent with late endosomes and lysosomes, suggesting that
dye fluorescein isothiocyanate (FITC). Cells were then its’ alkalinizing function was compromised. As a control,
incubated in cell culture media at 37 °C for 30 min or we also measured the luminal pH of recycling endosomes
60 min and individual vesicles were analyzed by FRIA. in untransfected AP-1 cells using FITC-conjugated Tf as a
The pH calibration curve is shown in Fig. 9a and an pH-sensitive probe. After 30 min of internalization, the
example of the pH distribution of newly formed recyc- pHv was 6.26 ± 0.14 (mean ± S.E.M., n = 4) and remained
ling endosomes as a function of time is presented in at that level after 60 min (Additional file 1: Figure S6).
Fig. 9b, with the median vesicular pH values from mul- These values are intermediate between those obtained for
tiple experiments summarized in Fig. 9c. After 30 min of the WT- and ΔES-transfected cells. Collectively, these
internalization, both WT and ΔES were predominantly data further confirm an important role for NHE6 in the
targeted to a compartment with median pHv values of late-phase alkalinization of recycling endosomes.
~6.25 ± 0.35 and 5.80 ± 0.39, respectively, consistent with
accumulation in early/sorting endosomes, though in the NHE6 ΔES expression enhances apoptosis in AP-1 cells
case of ΔES-expressing cells a minor fraction of the During the course of these studies, we noted that the
transporter was also detected in more acidic vesicles morphology of a signficant proportion of AP-1 cells
(pHv ~5.0) (Fig. 9b). At 60 min after internalization, the expressing the ΔES mutant appeared smaller, more
WT protein was found in a more alkaline vesicular pool rounded, showed surface blebbing and exhibited exten-
(median pHv ~6.50 ± 0.09), corresponding to the re- sive loss of filamentous actin stress fibers (Fig. 10a,
cycling endosomal compartment, whereas the bulk of Additional file 1: Figure S5), all features consistent with
the ΔES mutant protein accumulated in a very acidic cells undergoing apoptosis [70–72]. To examine this
Fig. 9 Over-acidification of endosomes containing NHE6 ΔES. AP-1 cells were transiently transfected with 3FNHE6HA WT and ΔES and endosomal
delivery was assessed 36 h post-transfection. Anti-Flag M2 primary and FITC-conjugated Fab secondary antibodies were bound to live cells for 1 h
on ice. The temperature was raised to 37 °C for 30 or 60 min and endosomal pH was measured by fluorescence ratio imaging (FRIA). a In situ
calibration of FITC fluorescence as a function of vesicular pH was performed by clamping the vesicular pH between 4 and 7.5 as described in
“Methods”. b Measurement of vesicular pH as a function of time upon internalization of 3FNHE6HA WT and ΔES. The pH values represent the
mean ± S.E.M. and the number of vesicles (~250 to 1000 vesicles) analyzed from a representative experiment are shown. c Graphical plot of the
median vesicular pH as a function of time derived from four experiments (mean ± S.E.M.)
Fig. 10 Expression of NHE6 ΔES enhances apoptosis in AP-1 cells. a Confocal microscopy of fixed AP-1 cells expressing 3FNHE6v1HA WT (upper
panels) or ΔES (lower panels). NHE6 was labelled with a mouse monoclonal anti-HAm antibody and an Alexa Fluor488-conjugated goat anti-mouse
secondary antibody. Actin filaments were labelled with rhodamine-phalloidin and the nuclei were stained with DAPI. b Transient expression
(48 h) of GFP, NHE6v1GFP WT or ΔES in AP-1 cells. Representative immunoblot probed with a polyclonal anti-GFP antibody (α-GFPP). c Flow
cytometry analysis of AP-1 cells transfected with GFP alone, NHE6GFP WT or ΔES. Forty-eight hours after transfection, cells were labeled with
Annexin V-APC and propidium iodide (PI) and 104 GFP-positive cells were examined by flow cytometry for each transfectant. Annexin V- and
PI- double negative cells represent viable cells. Cells taking up only PI (PI+) are indicative of dead cells; Annexin V+ positive cells indicate early
apoptotic cells whereas Annexin V+ and PI+ double positive cells represent late apoptotic cells. Results are shown as mean ± S.E.M. of eight
independent experiments. Significance was determined using a paired two-tailed Student t-test, **p < 0.01. d AP-1 cells transiently expressing
GFP alone, NHE6GFP WT or ΔES were isolated by cell sorting and then assayed for caspase 3/7 activity as described in “Materials and Methods”.
Data were normalized to values for GFP-expressing cells and displayed as mean ± S.E.M. of four independent experiments, each done in triplicate.
Significance was determined using a paired two-tailed Student t-test, *p < 0.05
possibility, AP-1 cells were transfected with GFP as a with GFP alone or GFP and either fluorescent mCherry
control, GFP-tagged NHE6v1 WT or ΔES (Fig. 10b). (ChFP)-tagged constructs of NHE6 WT or ΔES
Forty-eight hours post-transfection, the cells were (NHE6ChFP-WT and NHE6ChFP-ΔES, respectively) and
incubated in the presence of the fluorescent annexin V- visualized by confocal microscopy after 48 h. Following
allophycocyanin conjugate (Annexin V-APC) and propi- fixation and mounting, multiple z-stack optical sections
dium iodide (PI) and analyzed by flow cytometry to of 300–500 nm were taken and frame averaged 3X at
determine the fraction of apoptotic cells. Annexin V is a low resolution or line-averaged 2X at high resolution to
Ca2+-dependent phospholipid-binding protein with high improve the signal-to-noise ratio. Images were then
affinity for phosphatidylserine which is normally present deconvolved by using a full maximum likelihood ex-
in the inner leaflet of the plasma membrane [73]. Propi- trapolation algorithm Huygens deconvolution software
dium iodide is a fluorescent molecule whose signal is (SVI), and 3D images were compiled as maximum inten-
enhanced 20- to 30-fold upon binding to double- sity projections using Imaris software (Bitplane AG). As
stranded DNA and RNA but generally cannot cross the shown in Fig. 11a, representative neurons expressing
intact plasma membrane of viable cells. During the early control GFP alone or GFP and NHE6ChFP-WT exhibited
stages of apoptosis, phosphatidylserine is translocated to extensive dendritic arborization, whereas those express-
the outer leaflet of the plasma membrane, where it is ing GFP and NHE6ChFP-ΔES displayed an apparent
now accessible for binding to an Annexin V-APC probe. reduction in higher-order dendritic branching, though
However, the integrity of the plasma membrane at this the number of primary dendrites (originating from the
stage is maintained, so PI cannot penetrate inside the soma) was similar to controls (GFP: 5.67 ± 0.95; WT:
cells. As such, Annexin V-APC positive and PI negative 4.86 ± 0.59; ΔES: 4.14 ± 0.44; p > 0.05). Using Filament-
cells are considered to be early apoptotic. In the later Tracer Imaris software, there were significant reductions
stages of apoptosis, as well as in necrosis (unregulated in total dendritic length, surface area and number of
cell death) or necroptosis (programmed necrosis) [74], branch points of the neurons expressing NHE6ChFP-ΔES
the plasma membrane becomes leaky, allowing PI to (~50 % for each parameter; p < 0.01 one-way ANOVA
enter the cells and to bind to nucleic acids, so Annexin followed by a Tukey test) compared to control GFP or
V-APC/PI-double positive cells are late apoptotic or GFP and NHE6ChFP-WT (Fig. 11b-d).
necrotic, whereas PI-only positive cells are considered To investigate the vesicular nature of the NHE6-
necrotic [75]. GFP-positive cells (104 cells per experi- positive puncta, dual labelling experiments were per-
ment) were analyzed and a significantly higher propor- formed using Tf-AF633 to mark recycling endosomes.
tion of early and late apoptotic cells were detected Visual analysis revealed that within the soma, the signals
among the ΔES-expressing cells compared to cells ex- for NHE6 WT closely overlapped with those for Tf-
pressing wild-type NHE6GFP (Fig. 10c). To further sub- AF633, though this spatial relationship diminished
stantiate the activation of an apoptotic pathway, which is markedly for the ΔES mutant (Fig. 12a). Quantitative
defined as a caspase-dependent form of regulated cell calculation of the Mander’s overlap coefficient M1-M2
death [76, 77], we measured caspase 3 and 7 activity indicated that the degree of colocalization was sig-
using the luminescent Caspase-Glo® 3/7 assay from Pro- nificantly reduced for the ΔES mutant compared to WT
mega. To this end, GFP-positive AP-1 cells transiently (i.e., M1; WT-Tf, 51.9 % ± 5.1, versus ΔES-Tf, 35.0 % ±
expressing GFP, NHE6GFP WT or ΔES were isolated 4.4, p < 0.05; M2; Tf:WT, 78.7 % ± 2.7 versus Tf-ΔES,
by fluorescence-activated cell sorting (FACS) 24 h 44.6 % ± 1.7, p < 0.01) (Fig. 12b). These data suggest that
post-transfection and grown for an additional 24 h. the ΔES mutant is being partitioned away from the
As shown in Fig. 10d, the activation of caspase 3 and recycling endosomal pool, results consistent with those
7 is significantly higher (~3-fold) in cells expressing obtained for AP-1 cells.
the NHE6GFP-ΔES mutant compared to GFP or Given the above observations, we next investigated the
NHE6GFP-WT, consistent with the flow cytometry viability of the primary neurons transiently expressing
analyses. GFP alone, NHE6ChFP-WT or NHE6ChFP-ΔES by immu-
nolabelling for the activated, cleaved form of caspase-3
The distribution of the NHE6ΔES mutant is altered in (caspase-3 (Asp175); cCASP3) as a convenient indicator
primary mouse hippocampal neurons of apoptotic cell death. As shown in Fig. 13, the per-
The above results show that NHE6ΔES is mislocalized centage of neurons expressing GFP alone or GFP and
in non-neuronal cells. To investigate whether the subcel- NHE6ChFP-WT that co-stained for cCASP3 was low in
lular distribution of the mutant protein is similarly each case (~ < 20 %), whereas the percentage of trans-
altered in neurons, primary cultures of differentiated fected cells expressing NHE6ChFP-ΔES that were positive
hippocampal pyramidal neurons (10–12 days in vitro, for cCASP3 increased 3-fold, indicative of pronounced
DIV) prepared from C57BL/6 mice were transfected cell death.
Fig. 11 Expression of NHE6ΔES decreases the complexity of dendritic arborisation in primary hippocampal neurons. a Confocal images of mouse
primary hippocampal neurons transfected with cytosolic GFP alone or co-transfected with GFP and monomeric cherry fluorescent protein-tagged
NHE6 (NHE6ChFP) WT or ΔES. Forty-eight h post-transfection, cells were fixed in 4 % paraformaldehyde, mounted onto glass slides, and examined
by confocal microscopy. Images show each channel individually. b-d Quantification of parameters related to neuronal branching, including the
sums of total dendritic length (b) and area (c), as well as the total number of branch points per cell (d), using the FilamentTracer plug-in
module from Imaris Software. Transfection of NHE6ChFP-ΔES appeared to reduce the extent of neuronal arborisation, as can be discerned
from the representative images in a. Data from four experiments is represented as mean ± S.E.M. values. *: p < 0.01, one-way ANOVA with
Tukey post-hoc test. Scale bar: 60 μm
Discussion respectively. Activation of ERAD indicates that newly

Patients with CS possess a spectrum of core and second- synthesized mutant proteins have an increased propen-
ary clinical symptoms [1, 3, 8, 9], the variability of which sity to misfold and undergo multi-monoubiquitination
may depend on the severity of the mutation in NHE6, or polyubiquitination and retrotranslocation from the
functional overlap with other pH-regulating solute car- ER to the cytoplasm for degradation by the proteasome
rier proteins, or genetic-modifier effects. In the present [51, 52]. Consistent with this notion, the overall state of
study, we examined in detail the molecular and cellular ubiquitination of the ΔES mutant was significantly
consequences of one of the originally identified CS mu- elevated (~13-fold) compared to WT. By light micros-
tations that causes an in-frame deletion of two highly copy, there was limited accumulation of the mutant pro-
conserved amino acids in the predicted seventh trans- tein in the ER, suggesting that the ERAD system was not
membrane helix of the transporter, using the longest overly saturated due to ectopic expression under our
splice-variant as a template (i.e., NHE6v1Δ287ES288). experimental conditions and could adequately handle
Compared to the WT protein, excision of these two resi- the misfolded mutated proteins. The remaining fraction
dues markedly reduced its rate of post-translational mat- of mutated NHE6, however, was able to undergo pos-
uration to complex oligosaccharides, as revealed by a ttranslational modification and assembly as a fully-
greater accumulation of its core-glycosylated form rela- glycosylated homodimer and, furthermore, was delivered
tive to the total ΔES population in transient transfection correctly to the plasma membrane. However, upon
assays in both fibroblastic and neuroblastoma cell lines. internalization, the ΔES mutant was redirected away
This correlated with a significant reduction (~10-fold) in from the recycling endosomal compartment as it showed
its half-life compared to WT, effects that could be minor overlap with markers of recycling endosomes
partially prevented by inhibitors of both proteasomal- (transferrin) and early endosomes (i.e., EEA1), and in-
and lysosomal-mediated proteolysis; indicative of stead accumulated in Rab7-associated late endosomes/
processing via the endoplasmic reticulum (i.e., ERAD) multivesicular bodies prior to degradation in lysosomes.
and peripheral (i.e., ESCRT) quality control pathways, Consistent with the microscopic evidence, degradation
Fig. 12 Expression of NHE6ΔES causes a reduction in transferrin uptake and colocalization in primary hippocampal neurons. a Confocal images of
the cell bodies of primary hippocampal neurons transfected with GFP (green) and NHE6ChFP WT or ΔES (red) and incubated with Alexa Fluor633-tagged
transferrin (Tf-AF633) (pseudo-coloured blue) to assess endocytotic transferrin uptake and colocalization. Images show each channel individually, with
merged images of the NHE6ChFP and Tf-AF633 channels. b Quantitative summary of mean ± S.E.M. thresholded Mander’s coefficients, a measure of
colocalization, between Tf-AF633 and NHE6ChFP from four experiments. By the Mander’s coefficient, the majority of Tf-AF633 was colocalized with
NHE6ChFP WT than the inverse. The degree of colocalization was decreased with the NHE6ChFP ΔES mutant. *: p < 0.0001; **: p < 0.05, independent
Student’s t-test, two-tailed. Scale bar: 10 μm
of this population of transporters could be partly attenu- bands by Western blotting which is considerably smaller
ated by lysosomal inhibitors. These findings are compar- than its predicted molecular mass of ~100 kDa, and
able to the lysosomal degradation of other misfolded unlike the multiple core- and fully-glycosylated mono-
plasma membrane proteins that escape the ERAD path- mers and dimers detected in the present study using the
way, such as certain mutant Cl− channels (i.e., cystic longer NHE6v1 variant. Moreover, the GFP-tagged
fibrosis transmembrane regulator, CFTR) responsible for NHE6v2-Δ255ES256 protein accumulated mainly in the
cystic fibrosis [78, 79] and defective K+ channels (i.e., endoplasmic reticulum, with only a minor fraction de-
human ether-a-go-go-related, hERG) that cause long QT tected in early endosomes but not recycling endosomes.
syndrome 2 [80], and highlight the importance of a The basis for these differences is unclear, but may relate
secondary peripheral quality control mechanism to elim- to technical variances or the biochemical natures of the
inate the accumulation of improperly folded proteins. respective splice-variants.
These findings partially corroborate those of an earlier Aside from the intrinsic instability of the NHE6v1-
study that showed that the analogous mutation in the ΔES protein, its rate of internalization from the cell
shorter NHE6v2 splice-variant (i.e., Δ255ES256) also surface of AP-1 cells was also significantly reduced
caused the protein to be highly unstable and rapidly compared to its WT counterpart. By microscopy, ΔES-
degraded by the proteasome and lysosome [43]. How- expressing AP-1 cells showed visibly reduced uptake and
ever, in this latter study, the GFP-tagged WT and accumulation of fluorescently-labelled Tf compared to
Δ255ES256 proteins were detected as single ~70 kDa neighboring untransfected cells. Comparable results
Fig. 13 Expression of NHE6ΔES induces apoptotic cell death of primary hippocampal neurons. a Representative confocal images of primary
hippocampal neurons transfected with cytosolic GFP alone or GFP co-transfected with either NHE6ChFP-WT or NHE6ChFP-ΔES. Forty-eight h
post-transfection, cells were fixed, permeabilized, blocked, and assessed for apoptosis by immunostaining for cleaved caspase-3 (cCASP3, blue).
For each transfection condition, an overview is presented of the entire transfected neuron with the GFP, ChFP, and cCASP3 channels merged
(middle panels) with higher magnification cut-away images of the area around the cell soma (indicated by the white square) with each channel
displayed separately (right panels). For the co-transfected cells, the signals for the ChFP-tagged NHE6 constructs are also shown separately (left
panels). As noted in the images, cCASP3 was also detected in nontransfected cells, which could include not only neurons, but also astrocytes and
glia. Hence, to estimate the background level of apoptotic cells per field of view, companion cultures in each preparation were fixed, permeabilized
and immunolabelled for cCASP3 and stained with propidium iodide to mark the nuclei in order to calculate total cell density/field of view. Under each
treatment condition, the average number of cells per field of view ranged from 350 to 400 and the percentage of apoptotic cells per field of view for
GFP, GFP + NHE6ChFP-WT or GFP + NHE6ChFP-ΔES was 9.8 % ± 0.7, 9.7 % ± 1.0 and 9.1 % ± 0.8 (mean ± S.E.M.), respectively. Hence, ~10 % of the cells in
the background were apoptotic under each condition. b Quantitative representation of the percentages of cCASP3-positive (cCASP3+) neurons of
examined GFP or GFP- and NHE6ChFP transfected cells within each condition for four separate experiments. Data are presented as the mean ± S.E.M.
(total transfected cells examined ranged from 21–25 cells per condition). Compared to transfection with GFP alone or GFP + NHE6ChFP-WT, significantly
more neurons transfected with GFP + NHE6ChFP-ΔES were cCASP3-positive. *: p < 0.01, one-way ANOVA with Tukey post-hoc test. Scale bar: whole cell
images, 60 μm; high magnification images, 10 μm
were also obtained in HeLa cells using a flow cytometry- latter difference was not statistically significant. One
based assay where the net uptake of the labelled Tf-TfR possibility for the apparent lack of a strong dominant-
complex was significantly lower in ΔES compared to negative effect of the ΔES mutant in HeLa cells
WT transfected cells, and slightly lower that values compared to AP-1 cells is that its level of expression
obtained for GFP-transfected control cells, though the might not be sufficient to completely suppress the
actions of the endogenous NHE6 WT transporter, which also for its catalytic activity. The increased endosomal
is well expressed in HeLa cells [23] but negligible in AP- acidification would also be consistent with the partition-
1 cells. For this reason, we also performed a siRNA ing of ΔES-containing vesicles and any associated cargo
knockdown (>95 %) of NHE6 in HeLa cells to validate towards the endo-lysosomal degradative pathway.
its involvement in endocytosis of Tf-TfR complexes, and Acidification has long been recognized as an import-
indeed we observed a small, but statistically significant, ant determinant of vesicular biogenesis, trafficking and
depressive effect (~23 %, Fig. 7g), consistent with earlier function [67, 86, 87, 89]. While the roles of acidification
findings [25]. By contrast, other clathrin-mediated cargo on enzymatic processing of proteins along the secretory
such as the EGF-bound EGFR was unaffected. The mo- and degradative pathways are well appreciated, the pre-
lecular basis for the differential regulation of clathrin- cise mechanisms by which intraorganellar pH is sensed
dependent cargo by NHE6 is unknown, but may relate and transmitted to the cytoplasmic molecular machinery
to the recruitment of different endocytic adaptor pro- that controls vesicular events such as budding, coat
teins and associated accessory proteins. Unlike the acti- formation, sorting and fusion are less well understood.
vated Tf-TfR which is highly dependent on the AP2 However, emerging evidence indicates that endosomal
adaptor complex for internalization [81], the EGF-EGFR pH-regulators themselves can serve as both pH-sensors
is less restricted and can bind to AP2 as well as alternate and scaffolds to recruit components of the vesicular
endocytic adaptors such as epsin-1 [82] and Grb2 [83, trafficking machinery. Recent studies by Marshansky
84] and then is preferentially sorted to the lysosome for and colleagues [90, 91] have shown that two distinct
degradation. Hence, NHE6 appears to play a role in the subunits of the transmembrane V0 complex of the
endocytosis of a discrete subpopulation of clathrin- vacuolar H+-ATPase, the c- and a2-subunits, directly re-
dependent cargo that is preferentially targeted to recyc- cruit the small GTPase Arf6 (ADP-ribosylation factor 6)
ling endosomes in HeLa cells and presumably other cell and its associated guanine nucleotide exchange factor
types as well. In addition, we found that overexpression ARNO (ADP-ribosylation factor nucleotide site opener),
of NHE6 WT, but not ΔES, also increased the abun- respectively, in an intra-endosomal pH-dependent man-
dance of TfR at the plasma membrane. Based on these ner; interactions that are critical for endosomal trafficking
data, we propose that NHE6 elevates the net uptake of between the early and late endosomal compartments. This
Tf not only by enhancing endocytosis but also in part by process is seemingly selective, as it does not appear to in-
promoting the exocytosis and steady-state cell surface fluence membrane trafficking along the recycling endoso-
abundance of the TfR, and that this upregulation is mal pathway. This is intriguing, but unexpected, since
deficient in cells expressing the ΔES mutant. previous findings had also linked Arf6 to the recycling
The deficit in Tf uptake in ΔES-expressing cells also pathway [92–94]. This suggests that other endosomal pH-
correlated with aberrant over-acidification of endosomes regulatory transporters, such as NHE6, may play a more
relative to those in WT-expressing cells. Using an prominent role in directing vesicular trafficking along the
immunological-based approach that selectively targeted recycling endosomal pathway and that this process is im-
a pH-sensitive fluorescent probe to the lumen of NHE6- paired in the ΔES-expressing cells.
containing vesicles in AP-1 cells, we found that WT- In addition to disrupting recycling endosomal pH and
containing vesicles initially acidified (i.e., pHv ~6.25 ± trafficking, we found that expression of the NHE6 ΔES
0.35) followed by a gradual alkalization (i.e., pHv ~6.50 mutant in AP-1 cells elicited morphological and bio-
± 0.09) over a 60 min period. This biphasic pH fluctu- chemical changes symptomatic of programmed’apopto-
ation is consistent with previous reports of pH transients tic’ cell death [95, 96], as revealed by (1) disassembly of
along the recycling endosomal pathway [85–87]. By the filamentous actin network accompanied by cell
contrast, the ΔES-containing vesicles became progres- rounding and retraction, (2) plasma membrane blebbing,
sively more acidic throughout the measurement period phospholipid flipping (i.e., external exposure of phospha-
(i.e., pHv ~ 5.37 ± 0.25). This suggests that the catalytic tidylserine) and permeabilization, and (3) elevated activ-
activity of the mutant was compromised and unable to ities of caspases 3 and 7. Similarly, ectopic expression of
counter the H+ influx driven by the vacuolar H+-ATPase, ΔES in primary hippocampal neurons showed aberrant
resulting in a net increase in the luminal H+ concentra- subcellular distribution of ΔES-containing endosomes
tion. This loss-of-function is perhaps not unexpected as and pronounced neurodegeneration, as manifested by
mutation of the analogous glutamate residue (i.e., E262) significantly reduced dendritic length, surface area and
in the plasmalemmal-type NHE1 isoform was also found number of secondary branch points as well as increased
to significantly decrease its total cellular and plasma activation of caspase 3; features consistent with regu-
membrane abundance as well as intrinsic catalytic activ- lated cell death [97]. These observations complement
ity (~20 % of wild-type activity) [88]. Thus, this glutam- earlier in vitro studies showing that manipulations that
ate residue is critical not only for protein stability, but disrupt the molecular machinery involved in recycling
endosomal trafficking at dendritic spines of hippocampal between prosurvival growth factor receptors and proa-
neurons also cause pronounced morphological changes, poptotic death- or dependence-receptors is chronically
including decreased dendritic spine size and density and perturbed [109–112].
impair long-term potentiation [98, 99]. Indeed, we have While loss of NHE6 function may disrupt trophic or
recently shown that NHE6 exhibits a high degree of activity-dependent survival signals leading to cell deteri-
colocalization with vesicles containing the glutamatergic oration and death, ER stress [113] may also be another
AMPA GluA1-containing receptor in dendrites and important factor that reduces cell function and viability
dendritic spines of hippocampal neurons, suggestive of a in cells possessing the ΔES mutant. The mutant protein
role for NHE6-containing vesicles in synapse formation, might trigger the unfolded protein response (UPR), an
maturation, and plasticity [27]. intricate homeostatic process that results in the arrest of
The above findings are also consistent with in vivo general protein translation, while simultaneously per-
observations of progressive neurodegeneration in NHE6 mitting enhanced production of molecular chaperones
null mice [26, 28] and Christianson syndrome patients involved in protein folding, and increased protein po-
[1, 3, 8, 9]. In the null mouse model, both mutant male lyubiquitination and export of misfolded proteins to the
(Nhe6-/Y) and homozygous female (Nhe6−/−) knockout cytoplasm for proteasome-mediated degradation [114, 115].
mice exhibit impaired endo-lysosomal function (i.e., Consistent with this process, the ΔES mutant is highly
aberrant accumulation of GM2 ganglioside and choles- ubiquitinated and its degradation can be partly blocked by
terol) in subpopulations of neurons within the amygdala, proteasomal inhibitors. However, like many neurodegenera-
hippocampus, hypothalamus and cerebral cortex as well tive diseases that arise from prolonged impaired protein
as pronounced formation of axonal spheroids and folding, such as Alzheimer’s, Parkinson’s and Hunting-
degeneration of cerebellar Purkinje cells, features typical ton’s disease, the UPR is not always sufficient to rescue
of many lysosomal storage diseases [26, 100]. Further- the cell and apoptosis will be induced. Prolonged ER
more, hippocampal and cortical pyramidal neurons of stress is known to activate several kinases, including
Nhe6 knockout mice examined in vivo and in vitro glycogen synthase kinase-3β [116] and inositol-
displayed morphological and functional abnormalities requiring kinase 1 (IRE1) which activates apoptosis
typified by enhanced endosomal acidification, reduced signal-regulating kinase 1 (ASK1) that, in turn, stimu-
axonal and dendritic arborization, decreased synapse lates c-Jun N-terminal kinase (JNK) [113, 117], ultim-
density and maturation, and impaired circuit activity [28]. ately leading to caspase activation and cell death. Post-
These changes correlated with marked decreases in the mitotic neurons are especially susceptible to ER stress,
levels of total and phospho-activated forms of the neuro- as they are not protected from the accumulation of
trophin receptor TrkB, effects that could be largely misfolded proteins through the dilution of the ER fol-
mitigated by pharmacological inhibitors of lysosomal pro- lowing cell division [118]. This may also account for the
teolysis or by chronic incubation with the exogenous TrkB higher percentage of cell death observed in ΔES-
ligand BDNF [28]. Moreover, immunohistochemical stain- transfected primary hippocampal neurons (i.e., 60 %)
ing revealed substantial colocalization of NHE6 and TrkB compared to immortalized AP-1 fibroblastic cells (i.e.,
in endosomes in the perinuclear region and along growing ~30 %), at least under our experimental conditions.
axons and dendrites of hippocampal neurons. Thus, loss
of NHE6 was proposed to lead to excess degradation of Conclusions
the TrkB receptor (and possibly neurotransmitter recep- To conclude, our results provide new insight into the
tors such as AMPAR) and attenuation of downstream sig- molecular mechanisms by which disruption of NHE6
nalling due to over-acidification of the endosome activity impairs recycling endosomal trafficking and
compartment and sorting to lysosomes. Impairment in promotes neurodegeneration in the context of CS. These
TrkB signaling has also been implicated in the develop- analyses provide a framework for future investigations of
ment of Angelman Syndrome [101], a disorder that bears other NHE6 mutations and potential avenues for thera-
many features in common with Christianson Syndrome peutic interventions aimed at modulating the trafficking
[8]. Hence, these findings suggest that disruption of endo- of NHE6-dependent recycling endosomal cargo, such as
somal trafficking that promotes neurotrophin receptor- TrkB and AMPAR, thereby mitigating cell dysfunction
mediated prosurvival signals (e.g., via TrkB) [31, 102–104] and damage in CS. These findings may also be relevant
may shift the balance towards neurotrophin receptor- to our understanding of other neurodevelopmental or
mediated proapoptotic signals (e.g., via p75NTR) [105–108] neurodegenerative disorders such as autism [119–123],
leading to neuronal cell death. Analogous perturbations fragile X syndrome [124, 125] and Alzheimer’s disease
of plasma membrane/endomembrane-triggered signal- [126] where aberrations in recycling endosomal-associated
ling pathways may also apply in non-neuronal cells ex- cargo and signaling events have been implicated as
pressing the NHE6-ΔES mutant when the equilibrium contributing factors.
Additional file 3. Pescosolido MF, Stein DM, Schmidt M, El Moufawad AC, Sabbagh M, Rogg
JM, et al. Genetic and phenotypic diversity of NHE6 mutations in
Christianson syndrome. Ann Neurol. 2014;76:581–93.
Additional file 1: Supplementary data. (PDF 915 kb)
4. Angelman H. Syndrome of coloboma with multiple congenital
abnormalities in infancy. Br Med J. 1961;1:1212–4.
Acknowledgements 5. Williams CA, Angelman H, Clayton-Smith J, Driscoll DJ, Hendrickson JE, Knoll
We are grateful for the services provided by McGill Life Sciences Advanced JH, et al. Angelman syndrome: consensus for diagnostic criteria. Angelman
BioImaging Facility and the Flow Cytometry Facility, the Electron Microscopy syndrome foundation. Am J Med Genet. 1995;56:237–8.
Facility of the Department of Pharmacology and Therapeutics (McGill 6. Sutcliffe JS, Jiang YH, Galijaard RJ, Matsuura T, Fang P, Kubota T, et al. The
University), and Genome Québec for DNA sequencing. E6-Ap ubiquitin-protein ligase (UBE3A) gene is localized within a narrowed
Angelman syndrome critical region. Genome Res. 1997;7:368–77.
Funding 7. Kishino T, Lalande M, Wagstaff J. UBE3A/E6-AP mutations cause Angelman
This work was supported by research funds from the Canadian Institutes of syndrome. Nat Genet. 1997;15:70–3.
Health Research to JO (MOP-111191), RAM (MOP-86724) and GLL (MOP-66996) 8. Gilfillan GD, Selmer KK, Roxrud I, Smith R, Kyllerman M, Eiklid K, et al. SLC9A6
and a Canada Research Chair to GLL. AYLG is the recipient of an Alexander mutations cause X-linked mental retardation, microcephaly, epilepsy, and
Graham Bell Canada Graduate Scholarship. ataxia, a phenotype mimicking Angelman syndrome. Am J Hum Genet.
2008;82:1003–10.
9. Garbern JY, Neumann M, Trojanowski JQ, Lee VM, Feldman G, Norris JW, et
Availability of data and materials
al. A mutation affecting the sodium/proton exchanger, SLC9A6, causes
All data generated or analysed during this study are included in this
mental retardation with tau deposition. Brain. 2010;133:1391–402.
published article and its supplementary information file (Additional file 1).
Additional file 1: Figure S1. Sequence alignment of the different human NHE 10. Tarpey PS, Smith R, Pleasance E, Whibley A, Edkins S, Hardy C, et al. A systematic,
isoforms in the region encompassing the E287-S288 mutation. Figure S2. large-scale resequencing screen of X-chromosome coding exons in mental
Detection of NHE6 at the plasma membrane and endosomes in transfected retardation. Nat Genet. 2009;41:535–43.
AP-1 cells by transmission electron microscopy. Figure S3. NHE6v1 wild-type 11. Schuurs-Hoeijmakers JH, Vulto-van Silfhout AT, Vissers LE, van de
and ΔES can form a complex in intact cells. Figure S4. Expression of wild-type Vondervoort II, van Bon BW, de Ligt J, et al. Identification of pathogenic
and mutant NHE6v1 constructs in transfected AP-1 cells. Figure S5. Subcellular gene variants in small families with intellectually disabled siblings by exome
localization of NHE6v1 ΔES in transfected AP-1 cells. Figure S6. Measurement of sequencing. J Med Genet. 2013;50:802–11.
recycling endosomal pH in AP-1 cells. 12. Tzschach A, Grasshoff U, Beck-Woedl S, Dufke C, Bauer C, Kehrer M, et al.
Next-generation sequencing in X-linked intellectual disability. Eur J Hum
Genet. 2015.
Authors’ contributions
13. Inlow JK, Restifo LL. Molecular and comparative genetics of mental
AI, AB, and CM performed molecular, biochemical and data analyses; HB and
retardation. Genetics. 2004;166:835–81.
AI performed the vesicular pH measurements; AI, AYLG and JR performed
14. Vaillend C, Poirier R, Laroche S. Genes, plasticity and mental retardation.
the immunofluorescence imaging studies; AI, RAM, GLL and JO designed,
Behav Brain Res. 2008;192:88–105.
supervised and coordinated different aspects of the experiments and data
15. Gécz J, Shoubridge C, Corbett M. The genetic landscape of intellectual
analyses; AI and JO wrote the manuscript; all authors critically discussed
disability arising from chromosome X. Trends Genet. 2009;25:308–16.
results, revised and approved the manuscript.
16. Ropers HH. Genetics of early onset cognitive impairment. Annu Rev
Genomics Hum Genet. 2010;11:161–87.
Authors’ information
17. Lubs HA, Stevenson RE, Schwartz CE. Fragile X and X-linked intellectual
Not applicable.
disability: four decades of discovery. Am J Hum Genet. 2012;90:579–90.
18. Schwede M, Garbett K, Mirnics K, Geschwind DH, Morrow EM. Genes for
Competing interests endosomal NHE6 and NHE9 are misregulated in autism brains. Mol
The authors declare that they have no competing interests. Psychiatry. 2013;19:277–9.
19. Brett CL, Donowitz M, Rao R. Evolutionary origins of eukaryotic sodium/
Consent for publication proton exchangers. Am J Physiol Cell Physiol. 2005;288:C223–39.
Not applicable. 20. Orlowski J, Grinstein S. Na+/H+ exchangers. Compr Physiol. 2011;1:2083–100.
21. Miyazaki E, Sakaguchi M, Wakabayashi S, Shigekawa M, Mihara K. NHE6
Ethics approval and consent to participate protein possesses a signal peptide destined for endoplasmic reticulum
All procedures for animal handling were approved by the McGill membrane and localizes in secretory organelles of the cell. J Biol Chem.
University Facility Animal Care Committee (FACC) and carried out in full 2001;276:49221–7.
compliance with the Policies and Guidelines of the Canadian Council on 22. Brett CL, Wei Y, Donowitz M, Rao R. Human Na+/H+ exchanger isoform 6 is
Animal Care (CCAC). found in recycling endosomes of cells, not in mitochondria. Am J Physiol
Cell Physiol. 2002;282:C1031–41.
Author details 23. Nakamura N, Tanaka S, Teko Y, Mitsui K, Kanazawa H. Four Na+/H+
1
Department of Physiology, McGill University, Bellini Life Sciences Bldg., Rm, exchanger isoforms are distributed to Golgi and post-Golgi compartments
166, 3649 Promenade Sir-William-Osler, Montreal, QC H3G 0B1, Canada. and are involved in organelle pH regulation. J Biol Chem. 2005;280:1561–72.
2
Department of Pharmacology and Therapeutics, McGill University, Montreal, 24. Ohgaki R, Matsushita M, Kanazawa H, Ogihara S, Hoekstra D, Van IJzendoorn
Canada. SC. The Na+/H+ exchanger NHE6 in the endosomal recycling system is
involved in the development of apical bile canalicular surface domains in
Received: 24 February 2016 Accepted: 27 August 2016 HepG2 cells. Mol Biol Cell. 2010;21:1293–304.
25. Xinhan L, Matsushita M, Numaza M, Taguchi A, Mitsui K, Kanazawa H. Na+/H
+
exchanger isoform 6 (NHE6/SLC9A6) is involved in clathrin-dependent
References endocytosis of transferrin. Am J Physiol Cell Physiol. 2011;301:C1431–44.
1. Christianson AL, Stevenson RE, van der Meyden CH, Pelser J, Theron 26. Stromme P, Dobrenis K, Sillitoe RV, Gulinello M, Ali NF, Davidson C, et
FW, van Rensburg PL, et al. X linked severe mental retardation, al. X-linked Angelman-like syndrome caused by Slc9a6 knockout in
craniofacial dysmorphology, epilepsy, ophthalmoplegia, and cerebellar mice exhibits evidence of endosomal-lysosomal dysfunction. Brain.
atrophy in a large South African kindred is localised to Xq24-q27. 2011;134:3369–83.
J Med Genet. 1999;36:759–66. 27. Deane EC, Ilie AE, Sizdahkhani S, Das Gupta M, Orlowski J, McKinney RA.
2. Schroer RJ, Holden KR, Tarpey PS, Matheus MG, Griesemer DA, Friez MJ, Enhanced recruitment of endosomal Na+/H+ Exchanger NHE6 into dendritic
et al. Natural history of Christianson syndrome. Am J Med Genet A. spines of hippocampal pyramidal neurons during NMDA receptor-
2010;152A:2775–83. dependent long-term potentiation. J Neurosci. 2013;33:595–610.
28. Ouyang Q, Lizarraga SB, Schmidt M, Yang U, Gong J, Ellisor D, et al. 51. Meusser B, Hirsch C, Jarosch E, Sommer T. ERAD: the long road to
Christianson syndrome protein NHE6 modulates TrkB endosomal signaling destruction. Nat Cell Biol. 2005;7:766–72.
required for neuronal circuit development. Neuron. 2013;80:97–112. 52. Raasi S, Wolf DH. Ubiquitin receptors and ERAD: a network of pathways to
29. Lein ES, Hawrylycz MJ, Ao N, Ayres M, Bensinger A, Bernard A, et al. the proteasome. Semin Cell Dev Biol. 2007;18:780–91.
Genome-wide atlas of gene expression in the adult mouse brain. Nature. 53. Raiborg C, Stenmark H. The ESCRT machinery in endosomal sorting of
2007;445:168–76. ubiquitylated membrane proteins. Nature. 2009;458:445–52.
30. Lu W, Man H, Ju W, Trimble WS, MacDonald JF, Wang YT. Activation of 54. Apaja PM, Xu H, Lukacs GL. Quality control for unfolded proteins at the
synaptic NMDA receptors induces membrane insertion of new AMPA plasma membrane. J Cell Biol. 2010;191:553–70.
receptors and LTP in cultured hippocampal neurons. Neuron. 2001;29:243–54. 55. Goder V. Roles of ubiquitin in endoplasmic reticulum-associated protein
31. Huang SH, Wang J, Sui WH, Chen B, Zhang XY, Yan J, et al. BDNF- degradation (ERAD). Curr Protein Pept Sci. 2012;13:425–35.
dependent recycling facilitates TrkB translocation to postsynaptic density 56. Barriere H, Nemes C, Du K, Lukacs GL. Plasticity of polyubiquitin recognition
during LTP via a Rab11-dependent pathway. J Neurosci. 2013;33:9214–30. as lysosomal targeting signals by the endosomal sorting machinery. Mol
32. Hu H, Wrogemann K, Kalscheuer V, Tzschach A, Richard H, Haas SA, et Biol Cell. 2007;18:3952–65.
al. Mutation screening in 86 known X-linked mental retardation genes 57. Piper RC, Dikic I, Lukacs GL. Ubiquitin-dependent sorting in endocytosis.
by droplet-based multiplex PCR and massive parallel sequencing. Hugo Cold Spring Harb Perspect Biol. 2014;6.
J. 2009;3:41–9. 58. Le Bivic A, Real FX, Rodriguez-Boulan E. Vectorial targeting of apical and
33. Madrigal I, Fernandez-Burriel M, Rodriguez-Revenga L, Cabrera JC, Marti M, basolateral plasma membrane proteins in a human adenocarcinoma
Mur A, Mila M. Xq26.2-q26.3 microduplication in two brothers with epithelial cell line. Proc Natl Acad Sci U S A. 1989;86:9313–7.
intellectual disabilities: clinical and molecular characterization. J Hum Genet. 59. Dayel MJ, Hom EF, Verkman AS. Diffusion of green fluorescent protein in the
2010;55:822–6. aqueous-phase lumen of endoplasmic reticulum. Biophys J. 1999;76:2843–51.
34. Takahashi Y, Hosoki K, Matsushita M, Funatsuka M, Saito K, Kanazawa H, et 60. Wilson JM, de Hoop M, Zorzi N, Toh BH, Dotti CG, Parton RG. EEA1, a
al. A loss-of-function mutation in the SLC9A6 gene causes X-linked mental tethering protein of the early sorting endosome, shows a polarized
retardation resembling Angelman syndrome. Am J Med Genet B distribution in hippocampal neurons, epithelial cells, and fibroblasts. Mol
Neuropsychiatr Genet. 2011;156B:799–807. Biol Cell. 2000;11:2657–71.
35. Tzschach A, Ullmann R, Ahmed A, Martin T, Weber G, Decker-Schwering O, 61. Vanlandingham PA, Ceresa BP. Rab7 regulates late endocytic trafficking
et al. Christianson syndrome in a patient with an interstitial Xq26.3 deletion. downstream of multivesicular body biogenesis and cargo sequestration.
Am J Med Genet A. 2011;155A:2771–4. J Biol Chem. 2009;284:12110–24.
36. Piton A, Gauthier J, Hamdan FF, Lafreniere RG, Yang Y, Henrion E, et al. 62. Glozman R, Okiyoneda T, Mulvihill CM, Rini JM, Barriere H, Lukacs GL.
Systematic resequencing of X-chromosome synaptic genes in autism N-glycans are direct determinants of CFTR folding and stability in secretory
spectrum disorder and schizophrenia. Mol Psychiatry. 2011;16:867–80. and endocytic membrane traffic. J Cell Biol. 2009;184:847–62.
37. Riess A, Rossier E, Kruger R, Dufke A, Beck-Woedl S, Horber V, et al. 63. Sigismund S, Woelk T, Puri C, Maspero E, Tacchetti C, Transidico P, et al.
Novel SLC9A6 mutations in two families with Christianson syndrome. Clathrin-independent endocytosis of ubiquitinated cargos. Proc Natl Acad
Clin Genet. 2013;83:596–7. Sci U S A. 2005;102:2760–5.
38. Bosemani T, Zanni G, Hartman AL, Cohen R, Huisman TA, Bertini E, Poretti A. 64. Sigismund S, Argenzio E, Tosoni D, Cavallaro E, Polo S, Di Fiore PP.
Christianson syndrome: spectrum of neuroimaging findings. Clathrin-mediated internalization is essential for sustained EGFR
Neuropediatrics. 2014;45:247–51. signaling but dispensable for degradation. Dev Cell. 2008;15:209–19.
39. Mignot C, Heron D, Bursztyn J, Momtchilova M, Mayer M, Whalen S, et al. 65. Kazazic M, Roepstorff K, Johannessen LE, Pedersen NM, van Deurs B, Stang
Novel mutation in SLC9A6 gene in a patient with Christianson syndrome E, Madshus IH. EGF-induced activation of the EGF receptor does not trigger
and retinitis pigmentosum. Brain Dev. 2013;35:172–6. mobilization of caveolae. Traffic. 2006;7:1518–27.
40. Zanni G, Barresi S, Cohen R, Specchio N, Basel-Vanagaite L, Valente EM, et al. 66. Johnson LS, Dunn KW, Pytowski B, McGraw TE. Endosome acidification and
A novel mutation in the endosomal Na+/H+ exchanger NHE6 (SLC9A6) receptor trafficking: bafilomycin A1 slows receptor externalization by a mechanism
causes Christianson syndrome with electrical status epilepticus during involving the receptor’s internalization motif. Mol Biol Cell. 1993;4:1251–66.
slow-wave sleep (ESES). Epilepsy Res. 2014;108:811–5. 67. Weisz OA. Acidification and protein traffic. Int Rev Cytol. 2003;226:259–319.
41. Redin C, Gerard B, Lauer J, Herenger Y, Muller J, Quartier A, et al. Efficient 68. Casey JR, Grinstein S, Orlowski J. Sensors and regulators of intracellular pH.
strategy for the molecular diagnosis of intellectual disability using targeted Nat Rev Mol Cell Biol. 2010;11:50–61.
high-throughput sequencing. J Med Genet. 2014;51:724–36. 69. Barriere H, Apaja P, Okiyoneda T, Lukacs GL. Endocytic sorting of CFTR
42. Ilie A, Weinstein E, Boucher A, McKinney RA, Orlowski J. Impaired variants monitored by single-cell fluorescence ratiometric image analysis
posttranslational processing and trafficking of an endosomal Na+/H+ (FRIA) in living cells. Methods Mol Biol. 2011;741:301–17.
exchanger NHE6 mutant (D370WST372) associated with X-linked intellectual 70. Hacker G. The morphology of apoptosis. Cell Tissue Res. 2000;301:5–17.
disability and autism. Neurochem Int. 2014;73:192–203. 71. Maeno E, Ishizaki Y, Kanaseki T, Hazama A, Okada Y. Normotonic cell
43. Roxrud I, Raiborg C, Gilfillan GD, Stromme P, Stenmark H. Dual degradation shrinkage because of disordered volume regulation is an early prerequisite
mechanisms ensure disposal of NHE6 mutant protein associated with to apoptosis. Proc Natl Acad Sci U S A. 2000;97:9487–92.
neurological disease. Exp Cell Res. 2009;315:3014–27. 72. Bortner CD, Cidlowski JA. The role of apoptotic volume decrease and ionic
44. Rotin D, Grinstein S. Impaired cell volume regulation in Na+-H+ homeostasis in the activation and repression of apoptosis. Pflugers Arch.
exchange-deficient mutants. Am J Physiol. 1989;257:C1158–65. 2004;448:313–8.
45. Barriere H, Lukacs GL. Analysis of endocytic trafficking by single-cell 73. Andree HA, Reutelingsperger CP, Hauptmann R, Hemker HC, Hermens WT,
fluorescence ratio imaging. Curr Protoc Cell Biol. 2008;Chapter 15:1–21. Willems GM. Binding of vascular anticoagulant alpha (VAC alpha) to planar
46. Jiang M, Chen G. High Ca2+-phosphate transfection efficiency in low-density phospholipid bilayers. J Biol Chem. 1990;265:4923–8.
neuronal cultures. Nat Protoc. 2006;1:695–700. 74. Vandenabeele P, Galluzzi L, Vanden Berghe T, Kroemer G. Molecular
47. Porter K, Prescott D, Frye J. Changes in surface morphology of Chinese mechanisms of necroptosis: an ordered cellular explosion. Nat Rev Mol Cell
hamster ovary cells during the cell cycle. J Cell Biol. 1973;57:815–36. Biol. 2010;11:700–14.
48. Fafournoux P, Noël J, Pouysségur J. Evidence that Na+/H+ exchanger 75. Vermes I, Haanen C, Steffens-Nakken H, Reutelingsperger C. A novel assay
isoforms NHE1 and NHE3 exist as stable dimers in membranes with a high for apoptosis. Flow cytometric detection of phosphatidylserine expression
degree of specificity for homodimers. J Biol Chem. 1994;269:2589–96. on early apoptotic cells using fluorescein labelled Annexin V. J Immunol
49. Hisamitsu T, Pang T, Shigekawa M, Wakabayashi S. Dimeric interaction Methods. 1995;184:39–51.
between the cytoplasmic domains of the Na+/H+ exchanger NHE1 revealed 76. Galluzzi L, Aaronson SA, Abrams J, Alnemri ES, Andrews DW, Baehrecke EH,
by symmetrical intermolecular cross-linking and selective co- et al. Guidelines for the use and interpretation of assays for monitoring cell
immunoprecipitation. Biochemistry. 2004;43:11135–43. death in higher eukaryotes. Cell Death Differ. 2009;16:1093–107.
50. Hisamitsu T, Ammar YB, Nakamura TY, Wakabayashi S. Dimerization Is 77. Galluzzi L, Bravo-San Pedro JM, Vitale I, Aaronson SA, Abrams JM, Adam D,
Crucial for the Function of the Na+/H+ Exchanger NHE1. Biochemistry. et al. Essential versus accessory aspects of cell death: recommendations of
2006;45:13346–55. the NCCD 2015. Cell Death Differ. 2015;22:58–73.
78. Sharma M, Pampinella F, Nemes C, Benharouga M, So J, Du K, et al. 103. Leal G, Afonso PM, Salazar IL, Duarte CB. Regulation of hippocampal
Misfolding diverts CFTR from recycling to degradation: quality control at synaptic plasticity by BDNF. Brain Res. 1621;2015:82–101.
early endosomes. J Cell Biol. 2004;164:923–33. 104. Huang EJ, Reichardt LF. Neurotrophins: roles in neuronal development and
79. Okiyoneda T, Barriere H, Bagdany M, Rabeh WM, Du K, Hohfeld J, et al. function. Annu Rev Neurosci. 2001;24:677–736.
Peripheral protein quality control removes unfolded CFTR from the plasma 105. Coulson EJ, Reid K, Bartlett PF. Signaling of neuronal cell death by the
membrane. Science. 2010;329:805–10. p75NTR neurotrophin receptor. Mol Neurobiol. 1999;20:29–44.
80. Apaja PM, Foo B, Okiyoneda T, Valinsky WC, Barriere H, Atanasiu R, et al. 106. Zagrebelsky M, Holz A, Dechant G, Barde YA, Bonhoeffer T, Korte M. The
Ubiquitination-dependent quality control of hERG K+ channel with acquired p75 neurotrophin receptor negatively modulates dendrite complexity and
and inherited conformational defect at the plasma membrane. Mol Biol Cell. spine density in hippocampal neurons. J Neurosci. 2005;25:9989–99.
2013;24:3787–804. 107. Friedman WJ. Neurotrophins induce death of hippocampal neurons via the
81. Motley A, Bright NA, Seaman MN, Robinson MS. Clathrin-mediated p75 receptor. J Neurosci. 2000;20:6340–6.
endocytosis in AP-2-depleted cells. J Cell Biol. 2003;162:909–18. 108. Troy CM, Friedman JE, Friedman WJ. Mechanisms of p75-mediated death of
82. Kazazic M, Bertelsen V, Pedersen KW, Vuong TT, Grandal MV, Rodland MS, et hippocampal neurons. Role of caspases. J Biol Chem. 2002;277:34295–302.
al. Epsin 1 is involved in recruitment of ubiquitinated EGF receptors into 109. Schenck A, Goto-Silva L, Collinet C, Rhinn M, Giner A, Habermann B, et al.
clathrin-coated pits. Traffic. 2009;10:235–45. The endosomal protein Appl1 mediates Akt substrate specificity and cell
83. Jiang X, Huang F, Marusyk A, Sorkin A. Grb2 regulates internalization of EGF survival in vertebrate development. Cell. 2008;133:486–97.
receptors through clathrin-coated pits. Mol Biol Cell. 2003;14:858–70. 110. Schutze S, Tchikov V, Schneider-Brachert W. Regulation of TNFR1 and CD95
84. Johannessen LE, Pedersen NM, Pedersen KW, Madshus IH, Stang E. signalling by receptor compartmentalization. Nat Rev Mol Cell Biol. 2008;9:655–62.
Activation of the epidermal growth factor (EGF) receptor induces formation 111. Bredesen DE, Mehlen P, Rabizadeh S. Receptors that mediate cellular
of EGF receptor- and Grb2-containing clathrin-coated pits. Mol Cell Biol. dependence. Cell Death Differ. 2005;12:1031–43.
2006;26:389–401. 112. Thibert C, Fombonne J. Dependence receptors: mechanisms of an
85. Yamashiro DJ, Tycko B, Fluss SR, Maxfield FR. Segregation of transferrin to a announced death. Cell Cycle. 2010;9:2085–91.
mildly acidic (pH 6.5) para-Golgi compartment in the recycling pathway. 113. Tabas I, Ron D. Integrating the mechanisms of apoptosis induced by
Cell. 1984;37:789–800. endoplasmic reticulum stress. Nat Cell Biol. 2011;13:184–90.
86. Mellman I, Fuchs R, Helenius A. Acidification of the endocytic and exocytic 114. Doyle KM, Kennedy D, Gorman AM, Gupta S, Healy SJ, Samali A. Unfolded
pathways. Annu Rev Biochem. 1986;55:663–700. proteins and endoplasmic reticulum stress in neurodegenerative disorders. J
87. Maxfield FR, McGraw TE. Endocytic recycling. Nat Rev Mol Cell Biol. Cell Mol Med. 2011;15:2025–39.
2004;5:121–32. 115. Viana RJ, Nunes AF, Rodrigues CM. Endoplasmic reticulum enrollment in
88. Ding J, Rainey JK, Xu C, Sykes BD, Fliegel L. Structural and functional Alzheimer’s disease. Mol Neurobiol. 2012;46:522–34.
characterization of transmembrane segment VII of the Na+/H+ exchanger 116. Brewster JL, Linseman DA, Bouchard RJ, Loucks FA, Precht TA, Esch EA,
isoform 1. J Biol Chem. 2006;281:29817–29. Heidenreich KA. Endoplasmic reticulum stress and trophic factor withdrawal
89. van Weert AW, Dunn KW, Gueze HJ, Maxfield FR, Stoorvogel W. Transport activate distinct signaling cascades that induce glycogen synthase kinase-3
from late endosomes to lysosomes, but not sorting of integral membrane beta and a caspase-9-dependent apoptosis in cerebellar granule neurons.
proteins in endosomes, depends on the vacuolar proton pump. J Cell Biol. Mol Cell Neurosci. 2006;32:242–53.
1995;130:821–34. 117. Urano F, Wang X, Bertolotti A, Zhang Y, Chung P, Harding HP, Ron D.
90. Hurtado-Lorenzo A, Skinner M, El AJ, Futai M, Sun-Wada GH, Bourgoin Coupling of stress in the ER to activation of JNK protein kinases by
S, et al. V-ATPase interacts with ARNO and Arf6 in early endosomes transmembrane protein kinase IRE1. Science. 2000;287:664–6.
and regulates the protein degradative pathway. Nat Cell Biol. 118. Roussel BD, Kruppa AJ, Miranda E, Crowther DC, Lomas DA, Marciniak
2006;8:124–36. SJ. Endoplasmic reticulum dysfunction in neurological disease. Lancet
91. Merkulova M, Bakulina A, Thaker YR, Gruber G, Marshansky V. Specific motifs Neurol. 2013;12:105–18.
of the V-ATPase a2-subunit isoform interact with catalytic and regulatory 119. Correia CT, Coutinho AM, Sequeira AF, Sousa IG, Lourenco VL, Almeida JP, et al.
domains of ARNO. Biochim Biophys Acta. 1797;2010:1398–409. Increased BDNF levels and NTRK2 gene association suggest a disruption of
92. Souza-Schorey C, van Donselaar E, Hsu VW, Yang C, Stahl PD, Peters PJ. BDNF/TrkB signaling in autism. Genes Brain Behav. 2010;9:841–8.
ARF6 targets recycling vesicles to the plasma membrane: insights from an 120. Scattoni ML, Martire A, Cartocci G, Ferrante A, Ricceri L. Reduced social
ultrastructural investigation. J Cell Biol. 1998;140:603–16. interaction, behavioural flexibility and BDNF signalling in the BTBR T+ tf/J
93. Radhakrishna H, Al-Awar O, Khachikian Z, Donaldson JG. ARF6 requirement strain, a mouse model of autism. Behav Brain Res. 2013;251:35–40.
for Rac ruffling suggests a role for membrane trafficking in cortical actin 121. Koh JY, Lim JS, Byun HR, Yoo MH. Abnormalities in the zinc-metalloprotease-
rearrangements. J Cell Sci. 1999;112(Pt 6):855–66. BDNF axis may contribute to megalencephaly and cortical hyperconnectivity
in young autism spectrum disorder patients. Mol Brain. 2014;7:64.
94. Prigent M, Dubois T, Raposo G, Derrien V, Tenza D, Rosse C, et al. ARF6
122. Purcell AE, Jeon OH, Zimmerman AW, Blue ME, Pevsner J. Postmortem brain
controls post-endocytic recycling through its downstream exocyst complex
abnormalities of the glutamate neurotransmitter system in autism.
effector. J Cell Biol. 2003;163:1111–21.
Neurology. 2001;57:1618–28.
95. Taylor RC, Cullen SP, Martin SJ. Apoptosis: controlled demolition at the
123. Carlson GC. Glutamate receptor dysfunction and drug targets across models
cellular level. Nat Rev Mol Cell Biol. 2008;9:231–41.
of autism spectrum disorders. Pharmacol Biochem Behav. 2012;100:850–4.
96. Galluzzi L, Bravo-San Pedro JM, Kroemer G. Organelle-specific initiation of
124. Lauterborn JC, Rex CS, Kramar E, Chen LY, Pandyarajan V, Lynch G, Gall CM.
cell death. Nat Cell Biol. 2014;16:728–36.
Brain-derived neurotrophic factor rescues synaptic plasticity in a mouse
97. Bredesen DE, Rao RV, Mehlen P. Cell death in the nervous system. Nature.
model of fragile X syndrome. J Neurosci. 2007;27:10685–94.
2006;443:796–802.
125. Louhivuori V, Vicario A, Uutela M, Rantamaki T, Louhivuori LM, Castren E, et
98. Park M, Salgado JM, Ostroff L, Helton TD, Robinson CG, Harris KM, Ehlers
al. BDNF and TrkB in neuronal differentiation of Fmr1-knockout mouse.
MD. Plasticity-induced growth of dendritic spines by exocytic trafficking
Neurobiol Dis. 2011;41:469–80.
from recycling endosomes. Neuron. 2006;52:817–30.
126. Gao L, Tian M, Zhao HY, Xu QQ, Huang YM, Si QC, et al. TrkB activation
99. Petrini EM, Lu J, Cognet L, Lounis B, Ehlers MD, Choquet D. Endocytic by 7, 8-dihydroxyflavone increases synapse AMPA subunits and
trafficking and recycling maintain a pool of mobile surface AMPA receptors ameliorates spatial memory deficits in a mouse model of Alzheimer’s
required for synaptic potentiation. Neuron. 2009;63:92–105. disease. J Neurochem. 2016;136:620–36.
100. Walkley SU, Vanier MT. Secondary lipid accumulation in lysosomal disease. 127. Wakabayashi S, Pang T, Su X, Shigekawa M. A novel topology model of the
Biochim Biophys Acta. 1793;2009:726–36. human Na+/H+ exchanger isoform 1. J Biol Chem. 2000;275:7942–9.
101. Cao C, Rioult-Pedotti MS, Migani P, Yu CJ, Tiwari R, Parang K, et al. 128. Nygaard EB, Lagerstedt JO, Bjerre G, Shi B, Budamagunta M, Poulsen KA, et
Impairment of TrkB-PSD-95 Signaling in Angelman Syndrome. PLoS Biol. al. Structural modeling and electron paramagnetic resonance spectroscopy
2013;11, e1001478. of the human Na+/H+ exchanger isoform 1, NHE1. J Biol Chem. 2011;286:
102. Ip NY, Li Y, Yancopoulos GD, Lindsay RM. Cultured hippocampal 634–48.
neurons show responses to BDNF, NT-3, and NT-4, but not NGF.
J Neurosci. 1993;13:3394–405.

NHE6 Paper

Uploaded by

Copyright:

Available Formats

NHE6 Paper

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

NHE6 Paper

Uploaded by

Copyright:

Available Formats

Ilie et al.

Molecular Neurodegeneration (2016) 11:63

RESEARCH ARTICLE Open Access

A Christianson syndrome-linked deletion

(Continued from previous page)

Background NHE6 belongs to the endomembrane subclass of

Fluorescence-based endocytosis assay 10 mM NaCl, 20 mM Hepes or 20 mM MES, 1 mM

warm (37 °C) PBS, detached using Cell Dissociation Results

uptake) (Fig. 7a), presumably due to the higher plasma

Discussion respectively. Activation of ERAD indicates that newly

You might also like