Single Quantum Dot Tracking Reveals Serotonin Transporter Diffusion

May 22, 2018 - Serotonin transporter (SERT) terminates serotonin signaling in the brain by enabling rapid clearance of the neurotransmitter. SERT dysf...
0 downloads 0 Views 13MB Size
Subscriber access provided by UNIVERSITY OF TOLEDO LIBRARIES

Letter

Single Quantum Dot Tracking Reveals Serotonin Transporter Diffusion Dynamics are Correlated with Cholesterol-Sensitive Threonine 276 Phosphorylation Status in Primary Midbrain Neurons Danielle Marie Bailey, Mackenzie Alice Catron, Oleg Kovtun, Robert L. Macdonald, Qi Zhang, and Sandra J. Rosenthal ACS Chem. Neurosci., Just Accepted Manuscript • DOI: 10.1021/acschemneuro.8b00214 • Publication Date (Web): 22 May 2018 Downloaded from http://pubs.acs.org on May 23, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Chemical Neuroscience

1 1

Single Quantum Dot Tracking Reveals Serotonin Transporter Diffusion Dynamics

2

are Correlated with Cholesterol-Sensitive Threonine 276 Phosphorylation Status

3

in Primary Midbrain Neurons

4

Danielle M. Bailey1,2,5, Mackenzie A. Catron7, Oleg Kovtun1,5, Robert L. Macdonald8, Qi

5

Zhang2, and Sandra J. Rosenthal1-6*

6

Departments of 1Chemistry, 2Pharmacology, 3Chemical and Biomolecular Engineering,

7

4

8

6

9

8

Physics and Astronomy, 5Vanderbilt Institute of Nanoscale Science and Engineering, Vanderbilt Institute of Chemical Biology, 7Neuroscience Graduate Program, Neurology, Vanderbilt University, Nashville, TN, *corresponding author:

10

[email protected]

11

Abstract

12

Serotonin transporter (SERT) terminates serotonin signaling in the brain by enabling

13

rapid clearance of the neurotransmitter. SERT dysfunction has been associated with a

14

variety of psychiatric disorders, including depression, anxiety, and autism. Visualizing

15

SERT behavior at the single molecule level in endogenous systems remains a

16

challenge. In this study, we utilize quantum dot (QD) single particle tracking (SPT) to

17

capture SERT dynamics in primary rat midbrain neurons. Membrane microenvironment,

18

specifically membrane cholesterol, plays a key role in SERT regulation and has been

19

found to affect SERT conformational state. We sought to determine how reduced

20

cholesterol content affects both lateral mobility and phosphorylation of conformationally-

21

sensitive threonine 276 (Thr276) in endogenous SERT using two different methods of

22

cholesterol manipulation, statins and methyl-β-cyclodextrin. Both chronic and acute

23

cholesterol depletion increased SERT lateral diffusion, radial displacement along the

ACS Paragon Plus Environment

ACS Chemical Neuroscience 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 25

2 1

membrane, mobile fraction, and Thr276 phosphorylation levels. Overall, this work has

2

provided new insights about endogenous neuronal SERT mobility and its associations

3

with membrane cholesterol and SERT phosphorylation status.

4

Key words: quantum dots, serotonin transporter protein, threonine 276

5

Introduction

6

The serotonin transporter protein (SERT) controls the amplitude and duration of

7

synaptic serotonin signaling by enabling rapid clearance of the neurotransmitter back

8

into presynaptic terminals. SERT is enriched in the axonal arborizations and

9

accumulates in presynaptic specializations where it sequesters synaptic serotonin by

10

the Na+/Cl--dependent alternating access mechanism.1,2 SERT is the primary target for

11

selective serotonin reuptake inhibitors (SSRIs), a main class of antidepressants, and is

12

sensitive to psychostimulants and drugs of abuse such as cocaine and 3,4-

13

methamphetamine.3 SERT dysfunction and coding variations have been linked to many

14

psychiatric disorders, including autism spectrum disorder (ASD), depression, anxiety,

15

obsessive-compulsive disorder (OCD), and bipolar disorder.4 Relevance to brain

16

disorders and therapeutic utility of SERT have led to rigorous investigation of its function

17

and regulation4.

18

These research efforts have been primarily driven by biochemical methods

19

measuring ensembled behavior of many SERTs. Elucidating SERT behavior at the

20

single protein level remains a challenge, especially in endogenous neuronal systems.

21

Quantum

22

candidates for probing single SERT properties due to their unique photophysical

23

characteristics. QDs exhibit broad absorption with narrow, size-tunable Gaussian

dots

(QDs),

nanometer-sized

semiconductor

ACS Paragon Plus Environment

crystals,

are

excellent

Page 3 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Chemical Neuroscience

3 1

emission profiles, have high quantum yields, are easily functionalized, and perhaps

2

most importantly, are extremely resistant to photobleaching.5,6 Additionally, QD-tagged

3

proteins can be tracked with 5-10 nm localization accuracy at video rates, making QD-

4

based single particle tracking (SPT) a highly sensitive method.7 The Rosenthal lab

5

pioneered the use of ligand-conjugated QD SPT for a variety of neuronal proteins,

6

including SERT, norepinephrine transporter (NET), dopamine transporter (DAT), γ-

7

aminobutyric acid receptor (GABA).7-10 In this study, we sought to capture SERT

8

dynamics in an endogenous neuronal system with an antibody specifically targeted to

9

the fourth extracellular loop of SERT along with secondary antibody-conjugated QDs.

10

Since

our

previous

antagonist-based

ligand

approach

could

affect

intrinsic

11

conformational dynamics of SERT, we chose to use the off-site antibody tag for these

12

experiments. This report offers the first visualization of single SERT diffusion dynamics

13

in living rat midbrain neurons using QD SPT.

14

SERT function and expression are under strict control in part through a variety of

15

biochemical and biophysical events, including phophorylation4, membrane microdomain

16

localization10,11 and conformation shifts.12 Because SERT is a transmembrane protein,

17

the lipid environment, especially cholesterol, plays a crucial role in modulating its

18

activity.11,13-15 Statins, which inhibit the production of 3-hydroxy-3-methylglutaryl-

19

Coenzyme A reductase (HMG-CoA reductase) in cholesterol synthesis, have been

20

shown to enhance the effects of antidepressants such as imipramine and fluoxetine in

21

behavioral studies, demonstrating the importance of cholesterol for SERT function.16,17

22

At the subcellular level, SERT has been shown to associate with cholesterol-rich lipid

23

rafts and rely on these microdomains for efficient serotonin uptake.11 We have

ACS Paragon Plus Environment

ACS Chemical Neuroscience 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 25

4 1

previously shown that SERT appears to be laterally confined within membrane

2

microdomains in an immortalized RN46A cell line, demonstrating that SERT can exist in

3

both confined and more freely diffusing states. These states were found to be sensitive

4

to both cholesterol depletion and protein kinase G (PKG) activation, leading to an

5

increase in diffusion rate.10

6

Both experimental evidence and simulations in heterologous expression systems

7

have shown that a conserved cholesterol binding site (CHOL1) appears to be located

8

along SERT transmembrane domain (TM) 1a, TM5, and TM7, and cholesterol binding

9

to CHOL1 promotes the outward-facing conformation, accelerating serotonin uptake.18

10

When cholesterol binding to CHOL1 was disrupted, the SERT conformation equilibrium

11

shifted toward inward-facing, which provided more accessibility to the PKG-sensitive

12

threonine 276 (Thr276) residue as it lies on the cytoplasmic side of the TM5 helix.18 As

13

such, Thr276 is more susceptible to phosphorylation when inward-facing.12 While

14

several groups have interrogated the contribution of cholesterol and kinase activity to

15

SERT function and regulation, a lack of consensus persists due to heterogeneity of the

16

SERT molecular phenotype across these studies, which likely stem from differences in

17

expression density and cell host.4,13,18-20

18

cholesterol content affects both lateral mobility and phosphorylation of Thr276 in

19

endogenous SERT in primary midbrain neuron cultures using two different methods of

20

cholesterol manipulation, statins and MβCD. These alterations in membrane

21

composition might deleteriously impact SERT efficiency as a molecular transducer of

22

antidepressant effects and disrupt proper SERT axonal targeting, compromising brain

23

serotonin homeostasis.

We sought to determine whether lowered

ACS Paragon Plus Environment

Page 5 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Chemical Neuroscience

5 1

Results and Discussion:

2

PKG activation mobilizes SERT along the membrane of midbrain neurons.

3

Strong evidence suggests that Thr276 is the cGMP-sensitive phosphorylation site that is

4

activated by a PKG cascade.12,21-22 This pathway is upregulated in patients with the

5

autism-associated Gly56Ala SERT coding variant, suggesting that phosphorylation of

6

Thr276 is clinically important.23,24 However, individual differences in SERT behavior can

7

be missed with traditional ensemble methods. We have found that Gly56Ala, which is

8

hyperphosphorylated and resistant to further 8-Br-cGMP stimulation21, displayed

9

increased membrane mobility and diffusion, which led us to examine whether

10

phosphorylation status is linked to membrane mobility (submitted).25 Serotonergic

11

neurons are largely derived from the midbrain raphe nuclei26, thus for these experiments

12

the midbrain was isolated from postnatal day 1 rats, digested, dissociated, and cultured

13

onto coverslips. To examine SERT behavior at the single molecule level, endogenous

14

single neuronal SERT complexes were labeled with QDs using a two-step labeling

15

approach: specifically targeted monoclonal primary antibodies were bound followed by

16

commercial 655 QDs (Figure 1). Antibody specificity was confirmed by eliminated QD

17

binding after blocking the antibody binding site with a peptide. QD nonspecific binding

18

was nearly eliminated with the inclusion of 4% bovine serum albumin (BSA). Because

19

we are interested in comparing diffusion between conditions and because previous

20

groups have shown that divalent antibodies do not significantly alter diffusion dynamics,

21

we used full IgG in these studies.27-29 As the QD-IgG conjugates are relatively large

22

compared to the tagged SERT, it is also important to note that the viscous drag on a QD

23

bound to a protein was previously shown to be dominated by the mobility of the target

ACS Paragon Plus Environment

ACS Chemical Neuroscience 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 25

6 1

anchored in the membrane.30,31 QDs provide the distinct advantage of longer imaging

2

timescales, thus time series were taken over the course of 1 minute at 11-15 Hz. To

3

ensure surface SERTs were being tracked, all imaging took place immediately after

4

labeling within 20 minutes of the QD labeling step. Endocytosis was evident in longer

5

imaging experiments through the observation of QD clustering in endosomes, which

6

was also seen in our previous work.10 Blinking was used to verify single QDs were being

7

analyzed.

8

Under basal conditions, endogenous SERT showed a bimodal mobility

9

distribution with the majority of proteins (62%) falling within the confined regime, likely a

10

manifestation of SERT partitioning into lipid microdomains. These results are consistent

11

with behavior seen previously in our lab in the RN46A cell line, where SERT colocalized

12

with

13

superresolution STORM imaging of the homologous DAT by Gether and colleagues

14

using antibody-dye conjugates showed that DAT is largely clustered (about 60%) along

15

the plasma membrane of midbrain neurons32, consistent with our percentage of

16

confined serotonin transporters. In another endogenous system, clustering of SERT has

17

been observed on the surface of lymphocytes.33 Next, we examined the effects of PKG

18

activation on cell surface trafficking of SERT in midbrain neurons by applying 8-Br-

19

cGMP. 8-Br-cGMP stimulation significantly increased the diffusion rate of SERT by 2-

20

fold (Figure 2A-C). This increase in diffusion rate is also accompanied by a population

21

shift from confined to more freely diffusing (64%). We then analyzed the lateral

22

displacements of single SERTs over 5 seconds, which represents the extent of

23

mobilization for each trajectory. In addition to increased diffusion coefficients, we also

membrane

microdomain-associated

GM1

ganglioside.10

ACS Paragon Plus Environment

Moreover,

recent

Page 7 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Chemical Neuroscience

7 1

observed a significant increase in radial displacement upon 8-Br-cGMP treatment

2

(Figure 3). We thus determined that 8-Br-cGMP stimulation led to increased diffusion

3

rate, displacement along membrane, and mobile fraction compared to basal conditions.

4

Mevastatin-

and

methyl-beta-cyclodextrin-treated

neurons

exhibit

5

increased SERT mobility. Membrane cholesterol plays an important role in regulating

6

SERT, along with a variety of transmembrane proteins, and can affect both the

7

membrane fluidity as well as specific protein conformations.11,13-15,34 Molecular

8

dynamics simulations in a lipid raft membrane model revealed six potential SERT

9

cholesterol binding sites that all had functional significance.35 More recently,

10

experimental evidence validated this notion, showing that cholesterol can bind to a few

11

particular “hot spots” on SERT overlapping with those predicted binding sites.18 One

12

such site was found to be conformationally-sensitive to cholesterol concentration levels

13

and was near transmembrane domains 1, 5, and 7, which is near the same functional

14

domain that contains Thr276.18 Reportedly, a more inward-facing conformation was

15

stabilized when cholesterol binding at this site was disrupted. These changes also

16

enhanced ligand binding, which induced a more inward-facing conformation. As the

17

inward-facing conformation favors SERT phosphorylation at

18

Thr276, we hypothesize that lowering cholesterol content will increase lateral mobility

19

and Thr276 phosphorylation.18

20

We acutely extracted cholesterol from the membrane with methyl-beta-

21

cyclodextrin (MβCD), an established method for removing cellular surface cholesterol.36

22

In addition to direct cholesterol extraction, we also chronically depleted cholesterol with

23

mevastatin. Statins are used clinically to reduce the risk of cardiovascular disease by

ACS Paragon Plus Environment

ACS Chemical Neuroscience 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 25

8 1

inhibiting production of the HMG-CoA reductase in cholesterol synthesis, which reduces

2

overall cholesterol serum levels.16 These findings are particularly relevant as many

3

statins are blood-brain-barrier permeable, and treatment with statins has led to lowered

4

cholesterol levels in the brain, which could have implications for transmembrane

5

proteins regulated by cholesterol.17 In one example, statins have been shown to alter

6

the dynamics of serotonin 1a receptors by increasing the mobile fraction along the

7

membrane.37 Additionally, statins were shown to decrease membrane viscosity and

8

overall brain cholesterol content, decrease SERT activity, as well as induce behavioral

9

changes in rats.17 Thus, we wanted to explore the effects of statins on endogenous

10

neuronal SERT at the single protein level.

11

Similar to results for 8-Br-cGMP stimulated SERT, both chronic and acute

12

cholesterol depletion significantly increased the diffusion rate of SERT in comparison to

13

control (Figure 2). Acute cholesterol depletion had the greatest effect among all

14

conditions, which is expected since MβCD sequesters cholesterol directly from the

15

membrane, potentially disrupting membrane microdomains and SERT cholesterol

16

binding. Acute cholesterol depletion shifted the SERT populations the most, with 65% of

17

proteins being in the freely diffusing population. Chronic cholesterol depletion caused a

18

subtler population shift than the other treatments, with 47% of proteins still falling in the

19

confined regime and 53% being freely diffusing. Depleting cholesterol also significantly

20

increased cell surface SERT radial displacement along the membrane as seen in 5

21

second displacement plots for both chronic and acute conditions (Figure 3). We ensured

22

that our treatments were depleting cholesterol by the well-accepted filipin staining

23

(Figure 4).38 Both conditions had significantly less cholesterol than control.

ACS Paragon Plus Environment

Page 9 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Chemical Neuroscience

9 1

Acute and chronic cholesterol depletion increase phosphorylation of Thr276.

2

Since these increases in membrane mobility were well aligned with our hypothesis, we

3

then sought to understand if these increases in diffusion coefficient from cholesterol

4

depletion were also coupled with increased Thr276 phosphorylation. With recent

5

evidence showing that the inward-facing conformation of SERT provides access to

6

Thr276, as well as our data showing similar diffusion behavior between PKG-stimulated

7

and cholesterol-depleted conditions, we hypothesized that lowering the cholesterol

8

content would increase Thr276 phosphorylation. To test this, we used western blot to

9

compare Thr276 phospho-SERT levels across conditions. In agreement with our

10

hypothesis, both chronically and acutely depleting cholesterol significantly increased

11

Thr276 phosphorylation levels by over 2.5- and 3-fold respectively (Figure 5). 8-Br-

12

cGMP-stimulated conditions were included as a positive control, as this treatment is

13

known to increase Thr276 phosphorylation levels and exhibited a 1.62-fold change

14

increase compared to control. These data provide the first evidence that lowered

15

cholesterol content affects a critical regulatory posttranslational pathway that is also

16

coupled with increased SERT membrane mobility.

17

In conclusion, we have successfully labeled endogenous SERT with QDs to study

18

the surface mobility of individual SERTs in live midbrain neurons. We established that 8-

19

Br-cGMP-mediated PKG activation led to increased SERT surface mobility in its native

20

environment. We also showed that lowering membrane cholesterol, both acutely and

21

chronically, increased SERT surface mobility as well as SERT displacement along the

22

membrane. In addition, we demonstrated increased phosphorylation of Thr276 in

23

endogenous SERT upon both chronic and acute cholesterol depletion by over 2.5 fold.

ACS Paragon Plus Environment

ACS Chemical Neuroscience 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 25

10 1

We hypothesize that cholesterol-sensitive Thr276 activation may either directly drive

2

transporter cell surface reorganization or arise from transporter redistribution away from

3

compartments that restrict phosphorylation. Another possibility is that dynamic

4

destabilization of SERT is a direct consequence of the transporter conformational

5

equilibrium shift that inadvertently provides more access to the Thr276 site. Unraveling

6

the causal relationship between SERT phosphorylation status, membrane microdomain

7

residence, and surface trafficking will be the central goal of our future studies. Overall,

8

this work has provided further understanding of the role of cholesterol in endogenous

9

SERT dynamic membrane organization and has provided a method for visualizing

10

single SERT in this more complex primary neuron system.

11

Materials/Methods:

12

Cell Culture:

13

All procedures were in accordance with the Vanderbilt University Institutional Animal

14

Care and Use Committee standards. Primary midbrain cultures were prepared as

15

described previously.39 Midbrain was dissected from P1 Sprague-Dawley rats and

16

digested with trypsin-EDTA (Gibco). Tissue was then gently dissociated into single cells

17

in Hank’s Buffer Salt Solution (HBSS, Sigma). Dissociated cells were resuspended in

18

Minimal Essential Media (MEM, Life Technologies) containing 27 mM glucose, 2.4 mM

19

NaHCO3, 1.25 µM transferrin, 2 mM L-glutamine, 4.3 µM insulin, and 10%/vol fetal

20

bovine serum (FBS, Omega). Neurons were plated onto round 12 mm glass coverslips

21

coated with matrigel (BD Biosciences) and allowed to incubate at 37°C with 5% CO2.

22

After 1 hour, 1 mL media was added to each coverslip. After 24 hours, 1 mL 4-AraC

ACS Paragon Plus Environment

Page 11 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Chemical Neuroscience

11 1

was added in cell media that also contained 1%/vol B27 supplement (Sigma) to prevent

2

glia growth. Experiments were performed between DIV 10 and 20.

3

Treatments:

4

For acute cholesterol depletion, cells were incubated with 5 µM methyl-β-cyclodextrin

5

(MβCD, Fisher Scientific) at 37°C for 30 min. At this concentration, no morphology

6

changes were seen. For chronic cholesterol depletion, cells were incubated with 50 µM

7

mevastatin (Sigma) for 16 hours according to previously published methods.37 Again, at

8

this incubation concentration and time, cell morphology was not compromised. For PKG

9

activation, cells were incubated with 100 µM 8-Br-cGMP (Sigma) for 30 min. to 1 hour.

10

Widefield Epifluorescence Microscopy:

11

Images were obtained on an Olympus IX-81 epifluorescence microscope with an Andor

12

EMCCD camera. Images were collected using a 60X (filipin) or 100X (QD tracking) oil

13

immersion objective lens. QDs and fluorophores were excited with a 405 nm arc lamp.

14

QD 655 signal was collected with a 660/40 nm emission filter. Filipin signal was

15

collected with a 460/50 emission filter. For live cell imaging, cells on coverslips were

16

mounted in an imaging chamber and imaged at 37°C in Tyrode’s solution (150 mM

17

NaCl, 4 mM KCl, 2 mM MgCl2, 2 mM CaCl2, 10 mM HEPES buffer, 10 mM glucose in a

18

final volume of 1 L deionized water, pH 7.35) for the duration of the experiment. Single

19

QD tracking was performed at a frame rate of 11-15 Hz for 60 seconds, and data were

20

obtained within 20 minutes of the final wash step after QD labeling.

21

Live Cell QD Tracking:

22

Cells were incubated with 5 µg/mL primary SERT antibody (Advanced Targeting

23

Systems) in warm fluorobrite media (Thermo Fisher) at 37°C and 5% CO2 for 6 minutes.

ACS Paragon Plus Environment

ACS Chemical Neuroscience 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 25

12 1

Cells were washed once and incubated with 0.1 nM anti-mouse secondary antibody-

2

conjugated 655 QDs (Invitrogen) in warm 4% bovine serum albumin (BSA, Sigma)

3

fluorobrite media for 4 minutes. Cells were washed 5 times with warm Tyrode’s solution

4

and immediately mounted in Tyrode’s onto a heated microscope stage.

5

Trajectory Analysis:

6

Analysis was performed on data from at least three separate cultures and at least three

7

separate fields of view per culture. Stacks of individual TIFF images were processed

8

using ImageJ. Positions and trajectories of the QD-labeled SERT proteins were

9

determined using modified MATLAB routines developed by Daniel Blair and Eric

10

Dufresne (The Matlab Particle Tracking Code Repository). The subpixel localization

11

accuracy of the central QD position was determined to be ±10-15 nm.40 The uncertainty

12

(∆σ) is dependent of the width (which is about wide-field diffraction limit, 250 nm) and

13

the signal to noise ration (SNR), defined as I0/(σbg2+σI02)1/2, where I0 is maximum signal

14

above background, σbg2 is the variance of the background intensity values, and σI02 is

15

the variance of the maximum signal intensity above the background, giving ∆σ ≈250/SNR

16

(nm). All analyzed trajectories were at least 5 seconds in length. Trajectory analysis was

17

performed using custom MATLAB routines to determine mean square displacement

18

(MSD) using the following formula:

19 

MSD nΔt =  −   [ −   +  −   ] 

20

where N is the total number of frames, n∆t is the time interval in which the MSD is

21

calculated, and xi and yi are positions of the particles in the trajectories. Diffusion

ACS Paragon Plus Environment

Page 13 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Chemical Neuroscience

13 1

coefficients were determined by fitting MSD to the equation 4Dtα, where D is the

2

diffusion coefficient and α is the anomolous diffusion parameter. Statistical significance

3

for diffusion coefficient analysis was determined with Kolmogorov-Smirnov statistical

4

test. Stastitical significance for 5 second radial displacements was determined with

5

Student’s t-test. To determine the confined and freely diffusing populations, bimodal

6

fitting was performed in MATLAB, and the D threshold value (0.0082 µm2/s) was

7

determined to be where the two fits intersected. This D value is consistent with other

8

methods of determining the confined threshold.41

9

Western Blot:

10

Cells were washed with cold PBS and lysed with CelLytic M (Sigma) containing

11

protease and phosphatase inhibitors (Sigma) for 15 minutes. The lysate for each

12

condition was collected and centrifuged at 4°C at maximum speed for 10 minutes.

13

Protein concentration of collected supernatant was assayed using Bio-Rad dye reagent

14

(Bio-Rad; #500-0006) on a Beckman Coulter DU 530 spectrophotometer. Samples were

15

then normalized according to their protein concentration, denatured, and subjected to

16

gel

17

electrophoresis) (Bio-Rad) precast gels. Two sets of each sample were run side-by-side

18

in one gel to split later, one for blotting of phospho-SERT, and one for blotting of total

19

SERT. Protein was then transferred to polyvinylidene difluoride membranes (Millipore).

20

Membranes were cut in half vertically to separate the two samples. Antibodies used

21

were polyclonal anti-SLC6A4 (goat; 1:500; Sigma #SAB2502028) to detect total SERT

22

protein

23

PhosphoSolutions #P1700-276) to detect phospho-SERT. Monoclonal anti-GAPDH

electrophoresis

and

using

polyclonal

10%

serotonin

Mini-PROTEAN

transporter

TGX

Thr276

ACS Paragon Plus Environment

(polyacrylamide

antibody

(rabbit;

gel

1:500;

ACS Chemical Neuroscience 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 25

14 1

antibody (mouse; 1:6000; Millipore #CB1001) was used as a loading control. After

2

incubation with primary antibodies, IRDye (LI-COR Biosciences) conjugated secondary

3

antibodies were used, and the signals were detected using the Odyssey Infrared

4

Imaging System (LI-COR Biosciences).

5

Western Blot Analysis:

6

The integrated intensity value of each specific band was calculated using the Image

7

Studio Lite 5.2 software (LI-COR Biosciences). Each target band’s intensity was

8

normalized to each lane’s GAPDH signal, and phospho-SERT levels were calculated as

9

a percentage of the total SERT level from that sample. Data was collected from 7

10

experiments and at least three separate animal cultures. Statistical analysis was

11

conducted in GraphPad Prism 5. Statistical analyses comparing control and

12

experimental conditions were performed using Student’s t-test.

13

Filipin Staining:

14

Cells were fixed in PBS containing 4% formaldehyde for 30 minutes, washed 3 times

15

with PBS, and stained with 0.05 mg/mL filipin (Sigma) in PBS for 2 hours at room

16

temperature. Coverslips were mounted on slides and imaged on an Olympus IX-81

17

microscope as described above. Data was collected from three separate coverslips and

18

at least 150 neurons per condition. ROIs containing neurons were manually selected,

19

and average ROI intensities were measured in ImageJ. Background was measured for

20

each image and subtracted from the mean intensity values for the cells. Statistical

21

analyses comparing control and experimental conditions were performed using

22

Student’s t-test and one-way ANOVA with post-hoc Dunnett’s test.

23

ACS Paragon Plus Environment

Page 15 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Chemical Neuroscience

15 1

Supporting Information:

2

We have included an example time-lapse image series demonstrating diffusion of

3

endogenous serotonin transporter on the surface of primary midbrain neurons.

4 5

Contributions:

6

D.M.B. and S.J.R. conceived and planned the experiments. D.M.B. generated

7

monolayer neuronal culture, carried out single quantum dot tracking experiments,

8

analyzed tracking data, performed filipin staining experiments, generated figures, and

9

carried out the statistical analysis. Q.Z. provided the fluorescence microscope and

10

animals. M.A.C. carried out the western blot experiments. D.M.B, M.A.C, O.K., R.L.M.,

11

Q.Z., and S.J.R. contributed to the interpretation of the results. D.M.B. took the lead in

12

writing the manuscript. All authors critically edited the manuscript.

13 14

Acknowledgments:

15

This work was supported by NIH Grant EB003728 to SJR. DMB was supported by the

16

National Science Foundation Graduate Research Fellowship under Grant No. DGE-

17

1445197. The authors would like to gratefully acknowledge Dr. Jerry Chang for

18

MATLAB routines and helpful discussion, and Kristina Kitko for microscope expertise.

19 20

Figure Legends

21 22 23 24 25 26 27

Figure 1: QD labeling scheme. QDs are bound to SERT with antibodies in a two-step labeling protocol (antibody and QD size are not to scale). First, primary antibody is bound, followed by secondary antibody-conjugated QDs. QDs are specific to endogenous neuronal SERT as shown with a peptide block (bottom panel). When peptide was present, QD bnding was eliminated. Differential interference contrast (DIC) and QD655 channels are overlaid to demonstrate binding along neurites. Scale bar, 10 µm.

ACS Paragon Plus Environment

ACS Chemical Neuroscience 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 25

16 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46

Figure 2: Representative trajectories (a) and diffusion coefficients for endogenous midbrain SERT under (b) control, (c) 8-Br-cGMP, (d) mevastatin, and (e) MβCD treated neurons (scale bar, 10 µm). Diffusion coefficients are significantly higher when neurons are stimulated and cholesterol depleted (0.0211 ± 0.0030 µm2/s for control, 0.0457 ± 0.0040 µm2/s for 8-Br-cGMP-treated, p