Molecular Biology of the Cell
Vol. 21, 1282–1292, April 1, 2010
Manganese-induced Trafficking and Turnover of the
cis-Golgi Glycoprotein GPP130
Somshuvra Mukhopadhyay,* Collin Bachert,* Donald R. Smith,†
and Adam D. Linstedt*
*Department of Biological Sciences, Carnegie Mellon University, Pittsburgh, PA 15213; and†Department of
Microbiology and Environmental Toxicology, University of California at Santa Cruz, Santa Cruz, CA 95064
Submitted December 1, 2009; Revised January 20, 2010; Accepted January 21, 2010
Monitoring Editor: Keith E. Mostov
Manganese is an essential element that is also neurotoxic at elevated exposure. However, mechanisms regulating Mn
homeostasis in mammalian cells are largely unknown. Because increases in cytosolic Mn induce rapid changes in the
localization of proteins involved in regulating intracellular Mn concentrations in yeast, we were intrigued to discover that
low concentrations of extracellular Mn induced rapid redistribution of the mammalian cis-Golgi glycoprotein Golgi
phosphoprotein of 130 kDa (GPP130) to multivesicular bodies. GPP130 was subsequently degraded in lysosomes. The
Mn-induced trafficking of GPP130 occurred from the Golgi via a Rab-7–dependent pathway and did not require its transit
through the plasma membrane or early endosomes. Although the cytoplasmic domain of GPP130 was dispensable for its
ability to respond to Mn, its lumenal stem domain was required and it had to be targeted to the cis-Golgi for the Mn
response to occur. Remarkably, the stem domain was sufficient to confer Mn sensitivity to another cis-Golgi protein. Our
results identify the stem domain of GPP130 as a novel Mn sensor in the Golgi lumen of mammalian cells.
The subcellular levels of ions and other micronutrients such
as amino acids are closely regulated in eukaryotic cells.
Alterations in the levels of these small molecules often in-
duce rapid changes in the intracellular localization of pro-
teins involved in regulating their subcellular concentrations.
High levels of Cu lead to the translocation of the Cu-trans-
porting pumps ATP7A and ATP7B from the trans-Golgi
network (TGN) to the plasma membrane where they pump
Cu out of the cell (La Fontaine and Mercer, 2007) and an
autosomal recessive mutation in ATP7B gives rise to cellular
copper accumulation and Wilson’s disease (Kitzberger et al.,
2005). In yeast, under conditions of amino acid excess, the
general amino acid permease GAP1 is both endocytosed and
degraded from the plasma membrane, and newly synthe-
sized GAP1 transiting the Golgi is diverted to the vacuole
for degradation without reaching the cell surface at all (Scott
et al., 2004; Risinger and Kaiser, 2008). The yeast ferrichrome
receptor ARN1, in contrast, is up-regulated in the presence
of its substrate. In the absence of ferrichrome, ARN1 is
targeted for degradation from the TGN to the vacuole (Deng
et al., 2009). When low concentrations of ferrichrome are
present in the medium, small amounts of it get internalized
by fluid phase endocytosis and bind a high-affinity receptor
domain in ARN1. ARN1 bound to ferrichrome traffics to the
plasma membrane from the TGN, instead of the vacuole,
where it rapidly pumps ferrichrome into the cell (Deng et al.,
Regulation of the intracellular levels of Mn is particularly
important. Mn is an essential cofactor required for the cat-
alytic activity of diverse enzymes such as DNA and RNA
polymerases, mitochondrial superoxide dismutases, Golgi
enzymes required for N-linked glycosylation as well as cer-
tain decarboxylases and kinases (Missiaen et al., 2004). In the
CNS, the majority of cellular Mn occurs as a cofactor in the
astrocyte enzyme glutamine synthetase (Aschner et al.,
1999). Although Mn is required in diverse cellular processes,
exposure to high doses of Mn is toxic. In humans, over
exposure to Mn results in the development of a Parkinson-
like syndrome called manganism in occupationally exposed
individuals (Olanow, 2004), whereas environmental expo-
sure to Mn has been associated with cognitive and behav-
ioral effects in children (Wright et al., 2006; Bouchard et al.,
2007). Unlike Parkinson’s disease, there is no medical man-
agement to treat or control the onset and progression of
manganism (Olanow, 2004). Thus, elucidating the subcellu-
lar mechanisms that regulate Mn homeostasis and mediate
its cytotoxic effects is of high clinical significance.
Studies of Mn homeostasis in yeast indicate the impor-
tance of regulated intracellular trafficking of Mn transport-
ers. Smf1p, which is a member of the natural resistance-
associated macrophage protein (NRamp) family of metal
transporters, pumps extracellular Mn into the yeast cytosol
(Culotta et al., 2005). Pmr1p, a P-type Ca- and Mn-transport-
ing ATPase, then pumps cytosolic Mn into the lumen of the
Golgi for use by enzymes of the glycosylation pathway.
Under normal conditions, intracellular levels of Mn are reg-
ulated by adjustment of surface levels of Smf1p through
endocytosis and ubiquitin-mediated targeting to the vacuole
for degradation (Culotta et al., 2005). At toxic Mn levels,
This article was published online ahead of print in MBoC in Press
on February 3, 2010.
Address correspondence to: Adam D. Linstedt (linstedt@andrew.
Abbreviations used: Dyn, dynamin; EE, early endosome; EEA, early
endosome antigen; GFP, green fluorescent protein; GPP130, Golgi
phosphoprotein of 130 kDa; GP73, Golgi phosphoprotein of 73 kDa;
Lamp, lysosome-associated membrane protein; MVB, multivesicu-
lar body; TGN, trans-Golgi Network; WT, wild type.
1282 © 2010 by The American Society for Cell Biology
newly synthesized Smf1p protein transiting the Golgi is also
diverted to the vacuole for degradation (Jensen et al., 2009)
but Pho84p, an Pitransporter with low affinity for Mn,
persists on the surface and pumps large amounts of Mn into
the cytosol as metal-phosphate complexes (Culotta et al.,
2005). The excess cytosolic Mn is removed as Pmr1p pumps
it into the Golgi so that it will be secreted (Culotta et al.,
2005). In addition, some cytosolic Mn may be sequestered in
the yeast vacuole as a putative Mn pump has been identified
in the vacuole and its loss leads to enhanced sensitivity to
Mn (Gitler et al., 2009).
In mammals, mechanisms of Mn homeostasis and toxicity
are not well understood. Multiple transporters mediate in-
tracellular Mn uptake including splice isoforms of divalent
metal transporter (DMT)1, which is a homologue of yeast
Smf1p, the solute-carrier 39 metal transporter member ZIP8,
transferrin receptor, various calcium channels, and the glu-
tamate ionotropic receptor (Au et al., 2008). Some of these
transporters are localized to intracellular compartments,
such as DMT1 in lysosomes (Tabuchi et al., 2000), but it is not
known whether increased Mn exposure alters their localiza-
tion. Surprisingly, none of these transporters are specific for
Mn and even the relative contribution of each transporter to
the overall extent of uptake is not known (Au et al., 2008).
Moreover, although the human homologue of the Golgi local-
ized yeast Pmr1p pump secretory pathway Ca ATPase 1
(SPCA1) has been cloned (Sudbrak et al., 2000), its role in Mn
detoxification, if any, has not been established (Missiaen et
al., 2004). Additionally, while exposure to elevated levels of
Mn enhances oxidative stress, induces apoptosis, fragments
the Golgi apparatus, compromises mitochondrial function,
changes cell cycle progression and increases production of
mediators of inflammation like prostaglandins in mamma-
lian cells in culture (Towler et al., 2000; Roth et al., 2002; Zhao
et al., 2008; Milatovic et al., 2009), the molecular mechanisms
by which elevated Mn causes cytotoxicity are still not well
Here, we report that Mn induces the rapid trafficking of
the cis-Golgi glycoprotein Golgi phosphoprotein of 130 kDa
(GPP130) to multivesicular bodies (MVBs) with its subse-
quent degradation in lysosomes. GPP130 is an integral mem-
brane protein with a large lumenal domain (Linstedt et al.,
1997). The protein cycles within the Golgi and from the
Golgi to endosomes (Bachert et al., 2001) and GPP130 deple-
tion interferes with endosome to Golgi retrieval of specific
proteins that traffic in the bypass pathway (Natarajan and
Linstedt, 2004). Our studies define the pathway of Mn-
induced redistribution of GPP130 and identify the GPP130
cis-acting elements necessary and sufficient for the response.
MATERIALS AND METHODS
Cell Culture and Transfections
HeLa cells were grown in minimum essential medium (MEM) with 100
IU/ml penicillin-G and 100 ?g/ml streptomycin (Fisher Scientific, Hanover
Park, IL) supplemented with 10% fetal bovine serum (Atlanta Biologicals,
Lawrenceville, GA). For Mn treatment, freshly prepared MnCl2(Fisher Sci-
entific) was added for the times and concentrations described in individual
figures. DNA transfections were performed using the JetPEI transfection
reagent (Genesee Scientific, San Diego, CA) according to the manufacturer’s
protocol. Cultures were routinely transfected 24 h after plating and used for
experiments 24 h after transfection. RNA transfections were performed using
the Oligofectamine transfection reagent (Invitrogen, Carlsbad, CA) according
to the manufacturer’s protocol. The small interfering RNA (siRNA) used
targeted the sequence AAAAGCCAACACGAGGAGCTA in GPP130 cDNA
and has been described previously by us (Natarajan and Linstedt, 2004). In the
gene replacement experiments (Puthenveedu and Linstedt, 2004), cells were
transfected with the replacement GPP130 construct 48 h after siRNA trans-
fection by using JetPEI as described above. The replacement constructs had
three silent mutations in the sequence targeted by the siRNA. These silent
mutations were introduced using the forward primer 5?-CATCAGATGTT-
GAAAAGCCAACATGAAGAGCTCAAGAAA CAGCACAGTG-3? and the
QuikChange mutagenesis kit (Stratagene, La Jolla, CA). The mutated base
pairs in the replacement construct are underlined.
A fragment of GPP130 extending from amino acids 1-435 was nondirectionally
cloned into the Hind-3 site of the pEGFP-N3 vector (enhanced green fluorescent
protein) (Clontech, Mountain View, CA). This construct (GPP1301-435-GFP) was
used to create a deletion construct which had residues 1-247 of GPP130
(GPP1301-247-GFP) using a polymerase chain reaction (PCR)-based loop-out
modification of the QuikChange protocol. Further deletions were created
using the GPP1301-247-GFP construct. The deletion constructs created were
as follows: GPP130?176-247-GFP, GPP130?88-247-GFP, GPP130?36-175-GFP,
GPP130?36-87-GFP, GPP130?88-175-GFP, and GPP130?2-11-GFP. For chimeric
Golgi phosphoprotein of 73 kDa (GP73)-GPP130 constructs, a fragment of
GP73 extending from amino acid 1-184 (the end of its predicted coiled-coil
stem domain) was inserted upstream of GPP1301-247between the Nhe1 and
Xho1 sites of pEGFP-N3. The cytoplasmic and transmembrane domains of
GPP130 (amino acids 1-35 of GPP130) were then deleted to obtain proper
membrane topology. This construct was G73-G13036-247-GFP. Further dele-
tions in the GPP130 domains in the chimeric constructs were subsequently
performed as described for the parent GPP1301-247-GFP above to obtain G73-
GFP, G73-G13088-175-GFP, and G73-G13036-87_176-247-GFP. For the chimeras,
the amino acids of GPP130 present in the construct are expressed in their
respective names. GFP-tagged Rab5 wild-type (WT), Rab5 S34N and Rab5
Q79L were from Allan Levey, Emory University, Atlanta, GA (Volpicelli et al.,
2001);GFP-tagged Rab7 WT, Rab7 T22N, and Rab7 Q67L were from Marlyn
Farquhar (University of California at San Deigo, La Jolla, CA) and have been
described previously (Bucci et al., 2000), whereas hemagglutinin (HA)-tagged
dynamin (Dyn) II WT and Dyn II K44A were from Sandra Schmid (The
Scripps Institute, La Jolla, CA) (Damke et al., 1994). The GFP tag was removed
from the Rab5 Q79L construct by the loop out technique described above.
Epidermal Growth Factor (EGF) Uptake and Degradation
The EGF uptake and degradation assay was performed as described previ-
ously (Raiborg et al., 2001, Kauppi et al., 2002). In brief, cells were transferred
to MEM containing 2% bovine serum albumin (BSA). Alexa Fluor 555-EGF
complex (Invitrogen) was used at a final concentration of 200 ng/ml. For the
uptake experiments, cells were loaded with EGF for 15 min at 37°C and then
fixed with 3% paraformaldehyde (PFA) and processed for immunofluores-
cence microscopy. For the degradation assays, cells were loaded with EGF for
60 min at 37°C, then washed and chased for a further 60 min at 37°C in
EGF-free medium. At the end of the chase, cells were permeabilized using
phosphate-buffered saline (PBS) containing 0.03% saponin and 2% BSA at
room temperature for 5 min. The Alexa Fluor dye is not degraded by lyso-
somal hydrolases and persists even after EGF itself is degraded. The saponin
permeabilization before fixation releases the free Alexa tag from the cell
without affecting undegraded Alexa Fluor-EGF complexes. After permeabili-
zation, cells were fixed with 3% PFA and mounted for microscopy (Raiborg et
al., 2001, Kauppi et al., 2002).
Cells were fixed with 3% PFA and immunofluorescence analyses was per-
formed essentially as described previously (Yadav et al., 2008; Sengupta et al.,
2009). Alexa Fluor 488-, Alexa Fluor 594-, Alexa Fluor 568-, or Cy5-tagged
secondary antibodies were used (Invitrogen). Images were captured using a
spinning-disk confocal scan head microscope equipped with three-line laser
and independent excitation and emission filter wheels (PerkinElmer Life and
Analytical Sciences, Boston, MA) and a 12-bit Orca ER digital camera
(Hamamatsu Photonics, Bridgewater, NJ) mounted on an Axiovert 200 mi-
croscope with a 100?, 1.4 numerical aperture oil immersion objective (Carl
Zeiss, Thornwood, NY). Sections at 0.5-?m spacing were acquired using
Imaging Suite software (PerkinElmer Life and Analytical Sciences). All qual-
itative images depicted in the figures are maximum value projections of the
Immunofluorescence images were analyzed using ImageJ (http://rsbweb.nih.
gov/ij). To measure fluorescence per cell, average value projections were
created from individual Z-stacks, and background was subtracted from each
image separately. The mean fluorescence per cell was calculated by drawing
the outline of the cell and using the Analyze Measure function in ImageJ. To
measure total number of endosomes per cell, maximum value projections
were created, and the projected images were uniformly thresholded after
background subtraction. Total numbers of cytoplasmic objects ranging be-
tween 10 and 100 pixels were calculated per cell using the Analyze Particle
function of ImageJ. Line profiles were obtained using the RGB Profiler, and
Pearson’s coefficients were calculated using the Intensity Correlation Analysis
Mn-induced Degradation of GPP130
Vol. 21, April 1, 20101283
plug-ins of ImageJ, respectively. All quantitative fluorescence data are from a
single experiment, and all experiments have been replicated at least three
The methylthiazolyldiphenyl-tetrazolium bromide (MTT) assay for cell via-
bility was performed as described previously (Denizot and Lang, 1986; Zhao
et al., 2008). HeLa cells were exposed to 0–16 mM Mn for 12 h, washed with
PBS, and adjusted to 0.05% MTT (wt/vol, EMD Chemicals, Inc., Gibbstown,
NJ) in PBS for 2 h at 37°C. Cells were then lysed using 500 ?l of 0.1 N
hydrochloric acid in isopropanol and 1% Triton X-100. Absorption was sub-
sequently measured at 570 nm. Annexin-V-fluorescein isothiocyanate (FITC)
(Sigma-Aldrich, St. Louis, MO) was used as described by the manufacturer.
HeLa cells were treated with 500 ?M MnCl2for 0–24 h. Subsequently, cells
were incubated with Annexin-V-FITC used at a final dilution of 1:100 in buffer
containing 100 mM HEPES/NaOH, pH 7.5, 1.4 M NaCl, and 25 mM CaCl2for
15 min at room temperature and mounted immediately for microscopy.
For immunoblot assays cells were harvested in Tris-SDS lysis buffer (25 mM
Tris-Cl, pH 6.7, 1.5% SDS, 0.8% ?-mercaptoethanol, 1 mM dithiothreitol, and
1 mM polymethyl sulfonyl fluoride). Immunoblotting was performed as
described previously (Yadav et al., 2008; Sengupta et al., 2009). Immunoblots
were quantitated using ImageGauge software (Fujifilm, Valhalla, NY).
Antibodies and Other Reagents
Polyclonal (pAb) and monoclonal (mAb) antibodies against GPP130 and
anti-HA mAb have been described previously (Linstedt et al., 1997; Puri et al.,
2002). Unless indicated, all experiments used the anti-GPP130 pAb. The
anti-GPP130 mAb binds GPP130 between amino acids 278-433 (Linstedt et al.,
1997) and did not recognize any of the exogenously expressed GFP-tagged
constructs used in this study. Monoclonal antibodies against early endosome
antigen (EEA)1, ? tubulin, and lysosome-associated membrane protein
(Lamp)2 were from BD Biosciences (San Jose, CA), Sigma-Aldrich, and Ab-
cam (Cambridge, MA), respectively. All chemicals were from Sigma-Aldrich
unless otherwise specified.
Comparisons between any two groups were performed using two-tailed
Student’s t test assuming equal variances. For comparing multiple groups at
the same time, single-factor analysis of variance with the Tukey–Kramer post
hoc test was used (NCSS 2007 software; NCSS, Kaysville, UT). An asterisk in
bar graphs indicates p ? 0.05.
The cis-Golgi Glycoprotein GPP130 Is Rapidly and
Specifically Degraded in Response to Mn
In a test of the cellular response to 500 ?M Mn, we observed
that GPP130 levels were significantly reduced during an 8-h
time course (Figure 1A). During the same period, the levels
of a related cis-Golgi protein, GP73 (Puri et al., 2002), and the
loading control tubulin were unaffected. Quantification in-
dicated that GPP130 loss exceeded 65% at the 8-h time point
(Figure 1B). In accordance with a previous report (Zhao et
al., 2008), cell viability analyzed using the MTT assay and
Annexin-V staining was unaffected during this period (data
not shown). Specific loss of GPP130 was also confirmed by
immunofluorescence, and this analysis indicated that Golgi
morphology remained normal (Figure 1C). Furthermore,
neither the alkaline earth metals Ca, Mg, and Sr nor the
transition metals Cd, Cu, Co, Fe, and Zn induced loss of
GPP130 (data not shown). Concentrations of Mn as low as
100 ?M were sufficient to induce GPP130 degradation in
both HeLa and AF-5 GABAergic cells (Supplemental Figure
S1), whereas Mn-induced changes in cell cycle proteins and
inflammatory mediators are not evident at concentrations
?500 ?M (Zhao et al., 2008; Milatovic et al., 2009). Loss of
GPP130 in AF-5 cells is significant because it is the GABA-
enriched neurons of the globus pallidus that are primarily
affected in manganism (Sanchez et al., 2006; Crooks et al.,
Mn-induced Degradation of GPP130 Occurs in the
The t1/2of Mn-induced degradation of GPP130 was much
shorter than its ?18-h half-life (Linstedt et al., 1997), suggest-
ing that Mn increases GPP130 degradation rather than de-
creasing its synthesis. Indeed, within 5 min of Mn addition,
GPP130 began to redistribute from the Golgi to peripheral
punctate structures and by 4 h all the remaining GPP130 was
present in these structures (Figure 2A). To test whether the
redistributed GPP130 was en route to lysosomes, we first ex-
pressed Rab7-GFP, a marker of MVBs, and exposed the cells to
Mn for 2 h. Many GPP130 peripheral punctae were colabeled
with Rab7-GFP (Figure 2B). At higher magnifications, Rab7-
GFP was evident in a ring pattern lining the MVBs, with
GPP130 either colocalized with the ring or present within the
MVBs lumen lined by Rab7, strongly suggesting that GPP130
was undergoing internalization from the limiting membrane of
MVBs (Figure 2C).
Additional evidence for the internalization of GPP130 into
the lumen of MVBs was obtained by using a guanosine
triphosphate (GTP)-restricted version of Rab5 (Rab5 Q79L),
which produces giant MVBs by causing fusion of early
endosomes (EEs) and MVBs (Stenmark et al., 1994). Indeed,
GPP130 was clearly detectable within the lumen of the Rab5
Q79L-induced giant MVBs (Figure 2D, 4- and 8-h time
points). In fact, small amounts of GPP130 could be detected
within the lumen of these MVBs as early as 30 min after Mn
(Figure 2D). The intralumenal localization of GPP130 within
MVBs was confirmed by colabeling with EEA1, an EE-local-
ized protein that lines the limiting membrane of the giant
Rab5 Q79L-induced MVBs (Rosenfeld et al., 2001; Volpicelli
et al., 2001; Rink et al., 2005). As expected, GPP130 was
clearly detected inside the limiting membrane marked by
both EEA1 and Rab5 Q79L (Supplemental Figure S2). Fur-
thermore, 4 h after Mn, a large number of the cytoplasmic
GPP130 punctae also colabeled with Lamp2, a lysosomal
HeLa cells were treated with 500 ?M MnCl2for the times indicated
and subjected to immunoblot analyses to detect GPP130, GP73, and
?-tubulin. (B) Levels of GPP130 and GP73 were quantified at the 0-
and 8-h time points and normalized to tubulin levels (mean ? SE,
n ? 3; p ? 0.05). (C) HeLa cells on coverslips were untreated or
treated with 500 ?M MnCl2for 8 h, fixed, and costained using
anti-GPP130 (using the anti-GPP130 mAb) and anti-GP73 antibod-
ies. Bar, 10 ?m.
Mn induces the degradation of GPP130 in HeLa cells. (A)
S. Mukhopadhyay et al.
Molecular Biology of the Cell 1284
marker (Figure 2E), indicating that GPP130 eventually traf-
ficked to lysosomes.
To confirm that GPP130 was indeed degraded in lyso-
somes, we used pretreatment with the lysosomal protease
inhibitors leupeptin and pepstatin (Nicoziani et al., 2000).
Although cultures treated with Mn alone showed substan-
tial loss of GPP130, pretreatment with the inhibitors blocked
the degradation of GPP130 after Mn, and these cells exhib-
ited a dramatic accumulation of GPP130 in peripheral punc-
tae (Figure 3A). Thus, under these conditions, GPP130 pre-
sumably trafficked to lysosomes but was not degraded due
to the inhibitors. Note that cells pretreated with the inhibi-
tors but no Mn exhibited a slightly increased level of GPP130
in peripheral puncta, indicating that the inhibitors were
effective in blocking constitutive turnover of GPP130. Immu-
noblotting confirmed the ability of the lysosomal inhibitors
to block the Mn-induced degradation of GPP130 (Figure 3B).
Eight hours after Mn, GPP130 levels decreased to ?35% of
starting levels in control cultures. However, in cultures pre-
treated with the inhibitors, it remained ?80% of starting
levels (Figure 3C). To rule out any contribution of protea-
some-mediated degradation, cells were pretreated with the
proteasome inhibitor lactacystin, and this did not block the
Mn-induced degradation of GPP130 (data not shown). Thus,
in response to Mn, GPP130 traffics out of the Golgi to MVBs
and lysosomes where it is degraded.
Endocytosis and Early Endosome Trafficking Are Not
Required for Mn-induced GPP130 Degradation
In general, proteins traffic from the Golgi to lysosomes using
one of three well-characterized pathways: 1) Golgi to plasma
membrane (PM), followed by endocytosis to EEs and traf-
ficking via MVBs to lysosomes; 2) Golgi to EEs followed by
trafficking to MVBs and lysosomes; or 3) Golgi directly to
MVBs and then to lysosomes (Ihrke et al., 2004; Scott et al.,
2004, Piper and Luzio, 2007). To test whether the Mn-in-
lysosomes in response to Mn. (A) HeLa cells
were treated with 500 ?M MnCl2for the times
indicated, fixed, and costained to detect
GPP130 and giantin. Bar, 10 ?m. (B and C)
HeLa cells were transfected with GFP-tagged
Rab7 WT, and 24 h after transfection they were
treated with 500 ?M MnCl2for 2 h and stained
to detect GPP130. Arrowheads indicate over-
lap between GPP130 and Rab7-GFP. Bar, 4 ?m
(B) and 2 ?m (C). (D) HeLa cells were trans-
fected with Rab5 Q79L-GFP, exposed to 500
?M MnCl2for the times indicated, and subse-
quently stained to detect GPP130. Expression
of Rab5 Q79L did not alter the basal localiza-
tion of GPP130 (data not shown). Bar, 10 ?m;
insets, 5?. (E) HeLa cells were treated with 500
?M MnCl2for 4 h, fixed, and costained to
detect GPP130 and Lamp2. Arrowheads indi-
cate overlap between GPP130 and Lamp2. Bar,
GPP130 redistributes to MVBs and
after Mn. (A) HeLa cells were pretreated with-
out or with leupeptin (100 ?g/ml) and pepsta-
tin (50 ?g/ml) for 24 h (Nicoziani et al., 2000),
and then the media were adjusted to 0 or 500
?M MnCl2for 8 h. Coimmunostaining was
used to detect GPP130 and giantin. Bar, 10 ?m.
(B) The experiment was also analyzed by im-
munoblot using antibodies against GPP130
and ?-tubulin. (C) Quantitation of GPP130 lev-
els normalized to tubulin levels with GPP130
levels in the absence of Mn as 100% (mean ?
SE, n ? 3; p ? 0.05).
GPP130 is degraded in lysosomes
Mn-induced Degradation of GPP130
Vol. 21, April 1, 2010 1285
duced trafficking of GPP130 to MVBs required its transit
through the PM, we used a dominant-negative version of the
GTPase dynamin II (Dyn II K44A), which is an effective
inhibitor of most types of endocytosis (Conner and Schmid,
2003). Significantly, dominant-negative dynamin II did not
have any discernible effect on the loss of GPP130 in response
to 8 h of Mn exposure; yet, as expected, it was effective in
inhibiting internalization of fluorescent EGF (Supplemental
Because endocytosis was not required, we next used a
dominant-negative Rab5 construct (Rab5 S34N), which in-
hibits trafficking to/from EEs (Stenmark et al., 1994), to test
whether GPP130 might traffic through EEs en route to deg-
radation. However, trafficking to EEs did not seem necessary
because Rab5 S34N failed to block Mn-induced GPP130
degradation, whereas as a control, it did block degradation
of internalized fluorescent EGF (Supplemental Figure S4).
To further test the role of EEs, we disrupted the microtubule
network by using nocodazole. Maturation of EEs into MVBs
requires a functional microtubule network (Gruenberg et al.,
the isolated Golgi ministacks present in nocodazole-treated
cells remain transport competent (Hirschberg et al., 1998; Thy-
berg and Moskalewski, 1999, Yadav et al., 2008). Indeed, even
though the Golgi was fragmented, loss of GPP130 in response
to Mn seemed normal in nocodazole-treated cells (Figure 4A).
Unlike the transient transfections described above, nocodazole
treatment affects all cells in the population, thus allowing con-
firmation by immunoblotting. GPP130 degradation was clearly
evident by immunoblot analysis, and there was no statistical
difference between residual GPP130 levels in control and no-
codazole-treated cells (Figure 4, B and C). In contrast, EGF
degradation was significantly impaired under these conditions
(Figure 4D), with ?70% of the endocytosed EGF resisting
degradation after nocodazole, compared with ?10% in con-
trols (Figure 4E). Thus, the studies in this section indicate that
Mn-induced GPP130 redistribution to lysosomes is indepen-
dent of both endocytosis and EE trafficking, suggesting that
GPP130 takes a direct path from the Golgi to MVBs en route to
Mn-induced Degradation of GPP130 Is Rab7 Dependent
The small GTPase Rab7 is required for trafficking from
MVBs to lysosomes and for the formation of functionally
mature lysosomes (Bucci et al., 2000, Vanlandingham and
Ceresa, 2009). Dominant-negative Rab7 (Rab7 T22N) inhibits
MVB-to-lysosome trafficking (Bucci et al., 2000; Vanlanding-
ham and Ceresa, 2009), causing MVBs to enlarge (Vanland-
ingham and Ceresa, 2009), but it does not affect endocytosis
or trafficking through EEs (Papini et al., 1997; Vitelli et al.,
1997, Bucci et al., 2000, Vanlandingham and Ceresa, 2009).
Indeed, in accordance with previous work (Papini et al.,
1997), EGF was efficiently degraded in cells expressing Rab7
WT, but its degradation was blocked by expression of Rab7
T22N (Supplemental Figure S5). To test the role of Rab7 in
Mn-induced GPP130 degradation, we examined cells ex-
pressing Rab7 WT and T22N, neither of which altered the
basal localization of GPP130 in the Golgi. GPP130 degrada-
tion was unaffected by Rab7 WT, but in cells expressing
Rab7 T22N, GPP130 dramatically accumulated in large pe-
ripheral puncta after Mn (Figure 5A). In cells expressing
Rab7 WT, the levels of GPP130 decreased to ?25% of start-
ing levels after 8 h of Mn. In cells expressing Rab7 T22N the
block GPP130 degradation. (A) HeLa cells were pretreated for 3 h
with or without nocodazole (1 ?g/ml; Yadav et al., 2008), and then
the media were adjusted to 0 or 500 ?M MnCl2for 8 h. Coimmu-
nostaining was with antibodies against GPP130 and giantin. Bar, 10
?m. (B and C) The experiment was also analyzed by immunoblot
using antibodies against GPP130 and ?-tubulin and quantified with
GPP130 levels normalized to tubulin and in the absence of Mn
expressed as 100% (mean ? SE, n ? 3; p ? 0.05). (D) HeLa cells were
treated with nocodazole, and then the EGF degradation assay was
performed. Bar, 10 ?m. (E) Quantitation of the mean EGF fluores-
cence per cell with levels at time 0 expressed as 100% (mean ? SE,
n ? 12 cells/condition/time point; p ? 0.05).
Depolymerizing microtubules with nocodazole does not
is Rab7 dependent. (A) HeLa cells were trans-
fected with GFP-tagged Rab7 WT or Rab7 T22N,
and 24 h after transfection they were treated
with 500 ?M MnCl2for 8 h or left untreated.
Cells were then fixed, stained, and imaged to
detect GPP130 and GFP. Bar, 10 ?m. (B) Quan-
titation of the mean GPP130 fluorescence per cell
with levels in the absence of Mn normalized to
100% (mean ? SE, n ? 12 cells/construct/time
point; p ? 0.05).
Mn-induced degradation of GPP130
S. Mukhopadhyay et al.
Molecular Biology of the Cell1286
levels of GPP130 remained at ?70% of the starting Golgi
signal (Figure 5B). Thus, in the presence of dominant-nega-
tive Rab7, Mn-induced the trafficking of GPP130 out of the
Golgi but its degradation was blocked. The accumulation of
GPP130 in large cytoplasmic puncta in Mn-treated cells ex-
pressing Rab7 T22N is consistent with it being trapped in an
enlarged MVB compartment.
Overall, extracellular Mn induces the redistribution of
GPP130 from the Golgi to the interior of MVBs en route to
degradation in lysosomes. The dependence of the trafficking
on Rab7 in combination with its independence from dy-
namin, Rab5, and microtubules indicates that the trafficking
is direct from the Golgi to MVBs.
pensable. (A) Schematic of full-length GPP130 and the deletion
constructs lacking the lumenal acidic domain with and without the
cytoplasmic domains. Position and residue numbers of cytoplasmic
(C), transmembrane (TM), stem, and acidic domains are indicated.
“M” indicates the first methionine. (B) HeLa cells were transfected
with GPP1301-247-GFP. Twenty-four hours after transfection, they
were exposed to 100 ?g/ml cycloheximide for 2 h and then adjusted
to 0 or 500 ?M MnCl2for 2 h. Cells were then fixed, stained, and
imaged to detect GPP130 and GFP. Bar, 10 ?m. (C) HeLa cells were
siRNA-transfected to knockdown endogenous GPP130. After 2 d,
the cells were retransfected with an RNAi-immune version of
GPP1301-247-GFP, and after 24 h they were treated with cycloheximide
and Mn as described above. The cells were then fixed, stained, and
acidic domain) and GFP. Bar, 10 ?m. (D) HeLa cells were transfected
with GPP130?2-11-GFP, treated with cycloheximide and Mn as de-
scribed above, and processed to detect GPP130 and GFP. Bar, 10 ?m.
The GPP130 acidic and cytoplasmic domains are dis-
sensitivity to Mn. (A) Schematic of the deletion constructs used in
this figure. All deletions were made on GPP1301-247-GFP. (B) Cells
were transfected with the indicated constructs, and after 24 h they
were exposed to cycloheximide for 2 h then adjusted to 500 ?M Mn
for 2 h. After processing, GPP130 and GFP were imaged in the same
cells. Thresholding was used to enhance visualization of endo-
some with identical thresholds applied to all constructs in a given
channel. Bar, 10 ?m. (C) Quantitation of the number of GFP
endosomes per cell with (gray) and without (black) Mn for each
GPP130 construct. GPP1301-247-GFP was also analyzed after rescue
(Res.). The Mn-induced increase in GFP endosomes per cell was
statistically significant for GPP1301-247-GFP, GPP1301-247-GFP res-
cue, and GPP130?2-11-GFP (mean ? SE, n ? 25 cells for each con-
struct with and without Mn; p ? 0.05) but not the other stem deleted
constructs (p ? 0.05).
Deletions in the stem domain of GPP130 abolish its
Mn-induced Degradation of GPP130
Vol. 21, April 1, 20101287
The Acidic Domain of GPP130 Is Not Required for
GPP130 is a single-pass transmembrane protein with a short
cytoplasmic domain and a two-part lumenal domain in which
the first 210 residues form a coiled-coil stem domain and the
final 451 residues form a domain highly enriched in acidic
amino acids. One possibility was that the acidic residues inter-
acted with Mn and somehow altered the function of the coiled-
coil stem domain, which contains the sequence features that
are necessary and sufficient for GPP130 Golgi localization
(Bachert et al., 2001). To test the role of the acidic domain, we
deleted it by replacing it with GFP (Figure 6A). The resulting
construct, GPP1301-247-GFP, was localized to the Golgi, and
upon exposure to Mn for 2 h, redistributed to endosomal
structures that colabeled with endogenous GPP130 (Figure 6B;
dispensable. To rule out the possibility that GPP1301-247-GFP
was simply associating with endogenous GPP130, we carried
out rescue after RNA interference (RNAi) (Puthenveedu and
Linstedt, 2004, Natarajan and Linstedt, 2004). Cells expressing
GPP1301-247-GFP, but undetectable endogenous GPP130, were
readily identified using GFP fluorescence to detect GPP1301-247-
GFP and an antibody against the acidic domain to specifically
detect endogenous GPP130 (see Materials and Methods; Linstedt
et al., 1997). In these cells GPP1301-247-GFP was Golgi localized
and redistributed to endosomes upon Mn addition (Figure 6C;
see Figure 7C for quantification).
Mn-induced GPP130 Trafficking and Internalization into
MVBs Occurs without the GPP130 Cytoplasmic Domain
Because the acidic domain was not required, we next focused
attention on the cytoplasmic domain. Both AP3 and GGA
adaptors mediate sorting of proteins from the Golgi directly to
MVBs, and each depends on cytoplasmic domain sorting sig-
nals. The AP3-binding sequence motif YXX? (Ihrke et al., 2004)
is not present in the 12-residue GPP130 cytoplasmic domain,
but this domain does contain two lysines that, in principle,
could be ubiquinated, creating a GGA binding site (Scott et al.,
2004; Piper and Luzio, 2007; Risinger and Kaiser, 2008). To test
the role of the cytoplasmic domain, it was deleted from
GPP1301-247-GFP to create GPP130?2-11-GFP (Figure 6A). Re-
markably, GPP130?2-11-GFP was targeted to the Golgi and
redistributed to endosomes upon Mn addition (Figure 6D), as
did a GPP130 construct in which the two cytoplasmic lysines
were substituted to alanines (data not shown). We also con-
firmed that GPP130?2-11-GFP was responsive to Mn in the
absence of endogenous GPP130 (data not shown).
Furthermore, when cotransfected with Rab5 Q79L?GFP,
GPP130?2-11-GFP was present within the lumen of the Rab5-
induced giant MVBs (detected using EEA1) after Mn (Sup-
plemental Figure S6), indicating that even internalization
into MVBs was independent of the GPP130 cytoplasmic
domain. Thus, the cytoplasmic domain of GPP130, and by
extension any possible posttranslational modification of this
domain such as ubiquitination, was not required for the
Mn-induced trafficking of GPP130 to MVBs. In sum, neither
the acidic nor the cytoplasmic domains mediate Mn-induced
down-regulation of GPP130, suggesting a role for the trans-
membrane or stem domains.
Stem Domain Deletions Alter Intra-Golgi Localization
and Block Mn Sensitivity
The coiled-coil stem domain of GPP130 mediates its Golgi
targeting (Bachert et al., 2001). Given that our investigation of
the cis-acting elements underlying the Mn response required a
Golgi-targeted construct before Mn treatment, we were fortu-
change the intra-Golgi localization of GPP130.
(A) HeLa cells were transfected with the indi-
cated constructs, exposed to cycloheximide for
4 h, and processed to detect endogenous
GPP130 (using the anti-acidic domain mAb)
and GFP. Line plots to assess overlap between
endogenous GPP130 (red) and GFP (green)
signals accompany individual immunofluores-
cence panels. Bar, 10 ?m. (B) Plot of the Pear-
son’s coefficient for colocalization between
GFP and endogenous GPP130 signals for the
indicated constructs. Pearson’s coefficient for
GPP130?2-11-GFP was significantly greater
than all the other constructs (mean ? SE, n ?
10 cells/construct; p ? 0.05).
Deletions in the stem domain
S. Mukhopadhyay et al.
Molecular Biology of the Cell1288
nate that the stem domain was subdivided previously and
found to contain independently acting Golgi localization de-
terminants within residues 36-87 and 176-245 (Bachert et al.,
2001). Intriguingly, the intervening sequence 88–175 confers
(Bachert et al., 2001). To test the role of these subdomains, each
was deleted, singly or in combination, from GPP1301-247-GFP
(Figure 7A). The resulting constructs each contained at least
one Golgi determinant and were Golgi-localized (Supplemen-
tal Figure S7). Nevertheless, none of these constructs re-
sponded to Mn additions, whereas, as a positive control, the
GPP130?2-11-GFP again did (Figure 7B). The lack of a Mn
response in the stem-deleted constructs was not a failure of the
cells to respond to Mn because, in the same cells, endogenous
GPP130 redistributed to endosomes (Figure 7B). Quantification
clearly supported the conclusion that each of these deletions
ure 7C). These assays took advantage of the presence of en-
dogenous GPP130 to confirm an intact cellular Mn response,
which necessitated treatment with cycloheximide to limit over-
expression and mistargeting of the transfected proteins. To
avoid the cytotoxic effects of cycloheximide, the assays were
carried out at the 2-h time point, leaving open the possibility
that the mutations merely delayed rather than blocked the Mn
response. To carryout the analysis at a later time point and
avoid the cytotoxic effects of prolonged cycloheximide treat-
ment, we repeated the experiment with two siRNA-immune
constructs after GPP130 knockdown that limited the total
amount of GPP130 present and allowed for proper targeting of
the transfected constructs in the absence of cycloheximide.
Even after 8 h of Mn treatment, the Mn-insensitive construct
GPP130?88-247-GFP remained stably localized to the Golgi,
whereas the control construct GPP1301-247-GFP redistributed
and was degraded leaving only residual levels (Supplemental
Figure S8). Thus, deletion in the GPP130 stem domain caused
a profound block in its Mn responsiveness.
A possible explanation for the puzzling result that all stem
domain deletions blocked Mn sensitivity emerged when the
targeting of the constructs was more precisely determined.
Taking advantage of the specific recognition of endogenous
GPP130 by the antibody against the acidic domain, we car-
ried out colocalization analysis of endogenous GPP130 with
the stem-deleted constructs in cells before the addition of
Mn. Whereas the GPP130 responsive construct GPP130?2-11-
transfected with chimeric GP73-GPP130 constructs and treated with cycloheximide for 4 h and processed to detect endogenous GPP130
(using anti-acidic domain mAb) and GFP. Line plots for the overlap between endogenous GPP130 (red) and GFP (green) channels are shown.
Bar, 10 ?m. (C) Plot of the Pearson’s coefficient for colocalization between GFP and endogenous GPP130 signals for the indicated constructs.
The Pearson’s coefficient for the GPP130?2-11-GFP construct is also included for comparison. Except G73-G13036-175-GFP, there was no
significant difference between the Pearson’s coefficients for GPP130?2-11-GFP and the chimeric constructs (mean ? SE, n ? 10 cells/construct;
p ? 0.05).
Chimeric GP73-GPP130 constructs are targeted to the cis-Golgi. (A) Schematic of the chimeric constructs. (B) HeLa cells were
Mn-induced Degradation of GPP130
Vol. 21, April 1, 20101289
GFP colocalized with endogenous GPP130, the stem-deleted
constructs did not (Figure 8A). For each of these constructs,
the GFP signal was adjacent to the endogenous GPP130
staining. The difference was particularly striking when the
endogenous staining took on ring-like patterns. The GFP
signal was clearly inside these rings. Line plots of the fluo-
rescent intensity also showed the lack of colocalization (Fig-
ure 8A). Furthermore, quantitative analysis using the Pear-
son’s coefficient confirmed that there was a statistically
significant decrease in the colocalization of the stem-deleted
constructs with endogenous GPP130 compared with that of
GPP130?2-11-GFP (Figure 8B). It is typical for the staining of
cis- or medial-Golgi markers to ring that of trans-Golgi or
TGN markers under these fixation conditions (Puri et al.,
2002), suggesting that the deletions in the stem domain of
GPP130 had caused the proteins to shift their localization
toward the TGN. Indeed, both the stem-deleted constructs
and the TGN marker golgin97 (Yoshino et al., 2003) were
found inside the same Golgi ring structures (data not
shown). Thus, deletions in the stem domain unexpectedly
displaced the protein from the cis-Golgi, perhaps by weak-
ening interactions required for GPP130 retrieval, and this
correlated with a loss of sensitivity to Mn.
GPP130 Stem Domain Confers Mn Sensitivity to a
Related cis-Golgi Protein
Based on the idea that cis-Golgi targeting in conjunction with
one or more of the GPP130 stem subdomains might be needed
for the Mn response, we took advantage of the fact that GP73
is localized to the cis-Golgi and shares features with GPP130,
including a coiled-coil stem domain that targets GP73 to the
cis-Golgi (Puri et al., 2002). Thus, a set of constructs was gen-
erated in which the stem subdomains of GPP130 were placed,
singly or in combinations, after the stem domain of GP73 and
before a C-terminal GFP (Figure 9A). Colocalization analysis
identical to that described above was carried out on these
constructs and indicated that the constructs, with one excep-
tion, were indeed localized to the cis-Golgi. The GFP fluores-
cence was strikingly similar to endogenous GPP130 staining,
and the two yielded correlated line plots (Figure 9B). Further-
more, quantitative analyses using Pearson’s coefficients con-
firmed that, with one exception, there was no statistical differ-
ence between the colocalization of the chimeric constructs and
endogenous GPP130 compared with that of GPP130?2-11-GFP
(Figure 9C). Even the one construct that yielded noncolocaliza-
tion (G73-G13036-175-GFP) was still likely to have been targeted
to the cis-Golgi. The fluorescence of this construct defined rings
surrounding endogenous GPP130, suggesting that endoge-
nous GPP130 may have been pushed out of the cis-Golgi by
this construct. Competitive interactions in Golgi targeting of
GPP130 and GP73 have been observed previously (Puri et al.,
2002). Regardless, most, if not all, the chimeric constructs were
of the stem subdomains was sufficient to confer Mn sensitivity.
To test for Mn sensitivity, cells expressing each construct
were exposed to Mn for 2 h, and redistribution of the con-
struct as well as endogenous GPP130 in the same cells was
determined. Remarkably, constructs with any two of the
three subdomains yielded clear redistribution, whereas con-
structs with individual subdomains did not (Figure 10, A
and B). Endogenous GPP130 responded even in cells
where a construct did not thus confirming that the cellular
Mn response was intact. Although the N-terminal seg-
ment of the stem domain, residues 36-87, failed to respond
in isolation, it seemed to contribute the most. Of the three
constructs with combinations of the subdomains, those with
the N-terminal subdomain (G73-G13036-175-GFP and G73-
G13036-87_176-247-GFP) yielded a 100% response rate, and the
level of response, as judged by the number of peripheral
punctae per cell, was high (Figure 10B). In contrast, for the
construct lacking this subdomain (G73-G13088-247-GFP), the
response rate was only 65% and the number of puncta per
cell was significantly lower (Figure 10B).
In sum, the stem domain of GPP130 represents a transfer-
able determinant for Mn-sensitive targeting to MVBs from
Our study identifies the cis-Golgi glycoprotein GPP130 as a
Mn-responsive protein. Extracellular Mn triggered exit of
HeLa cells were transfected with GP73-GPP130 chimeric constructs,
and after 24 h they were treated with cycloheximide for 2 h and
adjusted to 500 ?M Mn for 2 h. The cells were processed to detect
GPP130 and GFP. Images depicted are uniformly thresholded to max-
imize visualization of endosomes. Bar, 10 ?m. (B) Quantitation of the
for the indicated GP73-GPP130 chimeric constructs. The Mn-induced
increase in GFP endosomes per cell was statistically significant for
G73-G13036-175-GFP, G73-G13088-247-GFP, and G73-G13036-87_176-247-
GFP (mean ? SE, n ? 25 cells for each construct with and without Mn;
p ? 0.05) but not the other chimeric constructs (p ? 0.05).
Chimeric GP73-GPP130 constructs respond to Mn. (A)
S. Mukhopadhyay et al.
Molecular Biology of the Cell1290
GPP130 from the Golgi and trafficking directly to MVBs
where it was internalized to intralumenal vesicles and ulti-
mately degraded by lysosomal hydrolases. The Mn-induced
trafficking of GPP130 was initiated within minutes and re-
quired its lumenal stem domain. Proper targeting of this
stem domain to the cis-Golgi seemed to be important in
conferring Mn sensitivity to GPP130. The stem domain was
also sufficient in that appending it to another cis-localized
Golgi protein caused the resulting chimera to traffic out of
the Golgi in response to Mn. These findings show that Mn
alters the intracellular trafficking of a mammalian protein
and identify, within the protein, a novel transferable Mn
Although we considered the possibility that GPP130 re-
distribution in response to Mn was caused by a block in
bypass pathway trafficking, this is unlikely. As stated,
GPP130 cycles out of the Golgi and is retrieved from EEs via
the bypass pathway (Linstedt et al., 1997; Bachert et al., 2001;
Puri et al., 2002; Natarajan and Linstedt, 2004). If Mn were to
induce a block in the bypass pathway retrieval of GPP130,
GPP130 might be diverted from EEs to MVBs and over time,
undergo degradation. However, three lines of evidence argue
that targeting of GPP130 to MVBs after Mn was not secondary
to a block in bypass pathway retrieval. First, trafficking from
EEs to MVBs, the pathway for diversion out of the bypass
pathway, was not required for Mn-induced trafficking of
GPP130 to MVBs. Second, GPP130 residues 88-175, which are
required for cycling in the bypass pathway, were not required
for Mn sensitivity. Third, Mn did not change the localization of
other proteins trafficking in the bypass pathway such as GP73.
Because Golgi localization of bypass pathway proteins is pH
sensitive, the last point also implies that Mn did not alter
tion of GP73 argues that levels of GPP130 remaining after Mn
were sufficient to support bypass pathway function (Natarajan
and Linstedt, 2004). Thus, Mn did not inhibit the bypass path-
way but rather induced a change in the sorting of GPP130
directing it from the Golgi to MVBs.
The Mn responsiveness of GPP130 was mapped to its lume-
nal stem domain. This domain, when localized to the cis-Golgi,
represented a transferable Mn-sensitive signal for trafficking
from the Golgi to MVBs. Our results raise three provocative
questions. First, how could the stem domain of GPP130 medi-
ate Mn-sensitive trafficking? Second, what could be the possi-
ble physiological relevance of the Mn-induced degradation of
GPP130? And third, why does the stem domain of GPP130
have to be targeted to the cis-Golgi to be responsive to Mn?
We suggest the possibility that, in the presence of Mn, the
GPP130 stem domain binds one or more proteins in the cis-
Golgi. The GPP130 binding partner could be a cis-Golgi resi-
dent or newly synthesized cargo transiting through the Golgi.
Formation of the hypothetical complex might block Golgi lo-
calization signals while exposing lysosomal targeting signals,
thereby targeting the entire complex to lysosomes for degra-
dation. Because the cytoplasmic domain of GPP130 is dispens-
able, it is likely that the binding partner contributes cytoplas-
mically disposed signals for interactions with specific adaptors
leading to trafficking to, and internalization in, MVBs.
Given the physiological context, it is conceivable that the
putative GPP130 binding partner is a Mn transporter under-
going homeostatic regulation. GPP130 could potentially bind a
newly synthesized pool of a Mn transporter transiting through
the Golgi en route to the plasma membrane and target it for
degradation. This would be analogous to yeast cells diverting
newly synthesized Mn transporter to the vacuole to inhibit
further uptake of Mn at the plasma membrane (Jensen et al.,
2009). GPP130 could also potentially regulate the trafficking of
the newly synthesized pool of SPCA1, the trans-Golgi–local-
ized Mn and Ca transporter (Behne et al., 2003), in a Mn-
dependent manner. Thus, GPP130 might conceivably interact
with any Mn-related activity, known or unknown, entering the
cis-Golgi and, in the presence of excess Mn, divert it to lyso-
somes as a means of down-regulation.
In this scenario, GPP130 would act as a targeting cofactor
conferring part of the cellular Mn response. This would be
similar to the situation for the yeast high-affinity iron per-
mease Ftr1, which associates with its cofactor Fet3 (Stearman
et al., 1996). Fet3 is an oxidase that uses Cu2?to convert Fe3?
to Fe2?so that Ftr1 can internalize Fe2?. In the absence of
Cu2?, the complex is trapped intracellularly, but when Fet3
binds Cu2?the complex of Fet3/Ftr1 traffics to the plasma
membrane (Gaxiola et al., 1998).
Deletions of part of the GPP130 stem domain caused an
apparent shift in localization from cis-cisternae to a more
trans-position and also a loss of Mn sensitivity, which could
be restored by the cis-targeting information in GP73. A va-
riety of observations suggest that the steady-state localiza-
tion of GPP130 in the cis-Golgi depends largely on retrieval
within the Golgi and from endosomes (Linstedt et al., 1997;
Puri et al., 2002; Bachert et al., 2001). Thus, it is likely that
deletions in the stem domain, which contain GPP130-target-
ing determinants (Bachert et al., 2001), reduce the efficiency
of its retrieval causing a shift in its localization to a more
trans-compartment. Although it is unclear why the Mn re-
sponse was restricted to cis-cisternae, if the GPP130 binding
partner proposed to bind GPP130 in elevated Mn were
cis-localized, trans-localized GPP130 constructs would fail to
form the complex needed for the altered trafficking.
Overall, our studies identify a Mn-responsive protein in
mammalian cells, describe its trafficking route, and begin to
elucidate sequence requirements within the molecule for the
response. This discovery may lead to a better understanding
of intracellular Mn homeostasis and the pathobiology of
Mn-induced neurodegeneration in humans.
We thank Ritika Tewari for performing the Annexin-V-FITC staining and
Tina Lee and Manojkumar Puthenveedu for comments on the manuscript.
Funding was provided by the National Institutes of Health grant R01 GM-
084111 (to A.D.L.).
Aschner, M., Vrana, K. E., and Zheng, W. (1999). Manganese uptake and distri-
bution in the central nervous system (CNS). Neurotoxicology 20, 173–180.
Au, C., Benedetto, A., and Aschner, M. (2008). Manganese in eukaryotes.
Neurotoxicology 29, 569–576.
Bachert, C., Lee, T. H., and Linstedt, A. D. (2001). Lumenal endosomal and
Golgi-retrieval determinants involved in pH-sensitive targeting of an early
Golgi protein. Mol. Biol. Cell 12, 3152–3160.
Behne, M. J., Tu, C. L., Aronchik, I., Epstein, E., Bench, G., Bikle, D. D., Pozzan
T., and Mauro, T. M. (2003). Human keratinocyte ATP2C1 localizes to the
Golgi and controls Golgi Ca2?stores. J. Invest. Dermatol. 121, 688–694.
Bouchard, M., Laforest, F., Vandelac, L., Bellinger, D., and Mergler, D. (2007).
Hair manganese and hyperactive behaviours: pilot study of school-age chil-
dren exposed through tap water. Environ. Health Perspect. 115, 122–127.
Bucci, C., Thomsen, P., Nicoziani, P., McCarthy, J., and van Deurs, B.(2000).
Rab 7: a key to lysosome biogenesis. Mol. Biol. Cell 11, 467–480.
Conner, S. D., and Schmid, S. L. (2003). Regulated portals of entry into the cell.
Nature 422, 37–44.
Crooks, D., Welch, N., and Smith, D. R. (2007). Low-level manganese expo-
sure alters glutamate metabolism in GABAergic AF5 cells. Neurotoxicology
Mn-induced Degradation of GPP130
Vol. 21, April 1, 20101291
Culotta, V. C., Yang, M., and Hall, M. D. (2005). Manganese transport and Download full-text
trafficking: lessons learned from Saccharomyces cerevisiae. Eukaryot. Cell 4, 1159–
Damke, H., Baba, T., Warnock, D. E., and Schmid, S. L. (1994). Induction of
mutant dynamin specifically blocks endocytic coated vesicle formation. J. Cell
Biol. 127, 915–934.
Deng, Y., Guo, Y., Watsom, H., Au, W. C., Shakoury-Elizeh, M., Basrai, M. A.,
Bonifacino, J. S., and Philpott, C. C. (2009). Gga2 mediates sequential ubiq-
uitin-independent and ubiquitin-dependent steps in the trafficking of ARN1
from the trans-Golgi network to the vacuole. J. Biol. Chem. 284, 23830–23841.
Denizot, F., and Lang, R. (1986). Rapid colorimetric assay for cell growth and
survival. Modifications to the tetrazolium dye procedure giving improved
sensitivity and reliability. J. Immunol. Methods 22, 271–277.
Gaxiola, R. A., Yuan, D. S., Klausner, R. D., and Fink, G. R. (1989). The yeast
CLC channel functions in cation homeostasis. Proc. Natl. Acad. Sci. USA 95,
Gitler, A. D., et al. (2009). alpha-Synuclein is part of a diverse and highly
conserved interaction network that includes PARK9 and manganese toxicity.
Nat. Genet. 41, 308–315.
Gruenberg, J., Griffiths, G., and Howell, K. E. (1989). Characterization of the
early endosome and putative endocytic carrier vesicles in vivo and with an
assay of vesicle fusion in vitro. J. Cell Biol. 108, 1301–1316.
Hirschberg, K., Miller, C. M., Ellenberg, J., Presley, J. F., Siggia, E. D., Phair, R.,
and Lippincott-Schwartz, J. (1998). Kinetic analysis of secretory protein traffic
and characterization of Golgi to plasma membrane transport intermediate in
living cells. J. Cell Biol. 143, 1485–1503.
Ihrke, G., Kyttala, A., Russell, M.R.G., Rous, B. A., and Luzio, J. P. (2004).
Differential use of two AP-3-mediated pathways by lysosomal membrane
proteins. Traffic 5, 946–962.
Jensen, L. T., Carroll, M. C., Hall, M. D., Harvey, C. J., Beese, S. E., Culotta,
V. C. (2009). Down-regulation of a manganese transporter in the face of metal
toxicity. Mol. Biol. Cell 20, 2810–2819.
Kauppi, M., Simonsen, A., Bremnes, B., Vieira, A., Callaghan, J., Stenmark, H.,
and Olkkonen, V. M. (2002). The small GTPase Rab22 interacts with EEA1 and
controls endosomal membrane trafficking. J. Cell Sci. 115, 899–911.
Kitzberger, R., Madl, C., and Ferenci, P. (2005). Wilson disease. Metab. Brain
Dis. 20, 295–302.
La Fontaine, S., and Mercer, J. (2007). Trafficking of the copper-ATPases,
ATP7A and ATP7B: role in copper homeostasis. Arch. Biochem. Biophys. 463,
Linstedt, A. D., Mehta, A., Suhan, J., Reggio, H., and Hauri, H. P. (1997).
Sequence and overexpression of GPP130/GIMPc: evidence for saturable pH-
sensitive targeting of a type II early Golgi membrane protein. Mol. Biol. Cell
Milatovic, D., Zaja-Milatovic, S., Gupta, R. C., Yu, Y., and Aschner, M. (2009).
Oxidative damage and neurodegeneration in manganese-induced neurotox-
icity. Toxicol. Appl. Pharmacol. 240, 219–225.
Missiaen, L., Raeymaekers, L., Dode, L., Vanoevelen, J., Baelen, K. V., Parys, J. B.,
Callewaert, G., Smedt, H. D., Segaert, S., and Wuytack, F. (2004). SPCA1 pumps
and Hailey-Hailey disease. Biochem. Biophys. Res. Commun. 322, 1204–1213.
Natarajan, R., and Linstedt, A. D. (2004). A cycling cis Golgi protein mediates
endosome-to-Golgi trafficking. Mol. Biol. Cell 15, 4798–4806.
Nicoziani, P., Vilhardt, F., Llorente, A., Hilout, L., Courtoy, P. J., Sandvig, K.,
and van Deurs, B. (2000). Role of dynamin in late endosome dynamics and
trafficking of cation-independent mannose-6-phosphate receptor. Mol. Biol.
Cell 11, 481–495.
Olanow, C. W. (2004). Manganese-induced Parkinsonism and Parkinson’s
disease. Ann. NY Acad. Sci. 1012, 209–223.
Papini, E., Satin, B., Bucci, C., de Bernard, M., Telford, J. L., Manetti, R.,
Rappouli, R., Zerial, M., and Montecucco, C. (1997). The small GTPase rab7 is
essential for cellular vacuolation induced by Helicobacter pylori cytotoxin.
EMBO J. 16, 15–24.
Piper, R. C., and Luzio, J. P. (2007). Ubiquitin-dependent sorting of integral
membrane proteins for degradation in lysosomes. Curr. Opin. Cell Biol. 19,
Puri, S., Bachert, C., Fimmel, C. J., and Linstedt, A. D. (2002). Cycling of early
Golgi proteins via the cell surface and endosomes upon lumenal pH disrup-
tion. Traffic 3, 641–653.
Puthenveedu, M. A., and Linstedt, A. D. (2004) Gene replacement reveals
p115/SNARE interactions are essential for Golgi biogenesis. Proc. Natl. Acad.
Sci. USA 101, 1253–1256.
Raiborg, C., Bache, K. G., Mehlum, A., Stang, E., and Stenmark, H. (2001). Hrs
regulates clathrin to early endosomes. EMBO J. 20, 5008–5021.
Rink, J., Ghigo, E., Kalaidzidis, Y., and Zerial, M. (2005). Rab conversion as a
mechanism of progression from early to late endosomes. Cell 122, 735–749.
Risinger, A. L., and Kaiser, C. A. (2008). Different ubiquitin signals act at the
plasma membrane and the Golgi to direct GAP1 trafficking. Mol. Biol. Cell 19,
Rosenfeld, J. L., Moore, R. H., Zimmer, K. P., Allizar-Foster, E., Dai, W., Zarka,
M. N., and Knoll, B. (2001). Lysosome proteins are redistributed during
expression of a GTP-hydrolysis-defective rab5a. J. Cell Sci. 114, 4499–4508.
Roth, J. A., Horbinski, C., Higgins, D., Lein, P., and Garrick, M. D. (2002).
Mechanisms of manganese-induced rat pheochromocytoma (PC12) cell death
and cell differentiation. Neurotoxicology 23, 147–157.
Sanchez, J., et al. (2006). GABAergic lineage differentiation of AF5 neural
progenitor cells in vitro. Cell Tissue Res. 324, 1–8.
Scott, P. M., Bilodeau, P. S., Zhdankina, O., Winistorfer, S. C., Hauglund, M. J.,
Allaman, M. M., Kearney, W. R., Robertson, A. D., Boman, A. L., and Piper,
R. C. (2004). GGA proteins bind ubiquitin to facilitate sorting at the trans Golgi
network. Nat. Cell Biol. 6, 252–259.
Sengupta, D., Trushel, S., Bachert, C., and Linstedt, A. D. (2009). Organelle
tethering by a homotypic PDZ interaction underlies formation of the Golgi
membrane network. J. Cell Biol. 186, 41–55.
Stearman, R., Yuan, D. S., Yamaguchi-Iwai, Y., Klausner, R. D., and Dancis, A.
(1996). A permease-oxidase complex involved in high-affinity iron uptake in
yeast. Science 271, 1552–1557.
Stenmark, H., Parton, R. G., Steele-Mortimer, O., Lutcke, A., Gruenberg, J.,
and Zerial, M. (1994). Inhibition of Rab5 GTPase activity stimulates mem-
brane fusion in endocytosis. EMBO J. 13, 1287–1296.
Sudbrak, R., et al. (2000). Hailey-Hailey disease is caused by mutations in
ATP2C1 encoding a novel Ca2? pump. Hum. Mol. Genet. 9, 1131–1140.
Tabuchi, M., Yoshimori, T., Yamaguchi, K., Yoshida, T., and Kishi, F. (2000)
Human NRamp2/DMT1, which mediates iron transport across endosomal
membranes, is localized to late endosomes and lysosomes in HEp-2 cells.
J. Biol. Chem. 275, 22220–22228.
Thyberg, J., and Moskalewski, S. (1999). Role of microtubules in the organi-
zation of the Golgi complex. Exp. Cell Res. 246, 263–279.
Towler, M. C., Prescott, A. R., James, J., Lucocq, J. M., and Ponnambalam, S.
(2000). The manganese cation disrupts membrane dynamics along the secre-
tory pathway. Exp. Cell Res. 259, 167–179.
Vanlandingham, P. A., and Ceresa, B. P. (2009). Rab7 regulates late endocytic
trafficking downstream of multivesicular body biogenesis and cargo seques-
tration. J. Biol. Chem. 284, 12110–12124.
Vitelli, R., Santillo, M., Lattero, D., Chiariello, M., Bufilco, M., Bruni, C. B., and
Bucci, C. (1997) Role of the small GTPase Rab7 in the endocytic pathway.
J. Biol. Chem. 272, 4391–4397.
Volpicelli, L. A., Lah, J. J., and Levey, A. I. (2001). Rab5-dependent trafficking
of the m4 muscarinic acetylcholine receptor to the plasma membrane, early
endosomes, and multivesicular bodies. J. Biol. Chem. 276, 47590–47598.
Wright, R. O., Amarasiriwardena, C., Woolf, A. D., Jim, R., and Bellinger,
D. C. (2006). Neuropsychological correlates of hair arsenic, manganese, and
cadmium levels in school-age children residing near a hazardous waste site.
Neurotoxicology 27, 210–216.
Yadav, S., Puri, S., and Linstedt, A. D. (2008). A primary role for Golgi
positioning in directed secretion, cell polarity and wound healing. Mol. Biol.
Cell 20, 1728–1736.
Yoshino, A., Bieler, B. M., Harper, D. C., Cowan, D. A., Sutterwala, S., Gay,
D. M., Cole, N. B., McCaffery, J. M., and Marks, M. S. (2003). A role for GRIP
domain proteins and/or their ligands in structure and function of the trans
Golgi network. J. Cell Sci. 116, 4441–4454.
Zhao, P., Zhong, W., Ying, X., Yuan, Z., Fu, J., and Zhou, Z. (2008). Manganese
chloride-induced G0/G1 and S phase arrest in A549 cells. Toxicology 250, 39–46.
S. Mukhopadhyay et al.
Molecular Biology of the Cell1292