Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Flipped zinc transporter 8 (ZnT8): a novel approach to characterize zinc transport and its possible relevance to type 2 diabetes
(USC Thesis Other)
Flipped zinc transporter 8 (ZnT8): a novel approach to characterize zinc transport and its possible relevance to type 2 diabetes
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
FLIPPED ZINC TRANSPORTER 8 (ZNT8): A NOVEL APPROACH TO
CHARACTERIZE ZINC TRANSPORT AND ITS POSSIBLE RELEVANCE TO TYPE
2 DIABETES
by
Beren Tomooka
A Thesis presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
MASTER OF SCIENCE
(PHYSIOLOGY & BIOPHYSICS)
August 2009
Copyright 2009 Beren Tomooka
ii
ACKNOWLEDGEMENTS
I would like to thank Dr. Robert H. Chow for offering me a position in his laboratory, for giving me the
chance to work on this exciting project, and for being a valuable mentor. I learned an incredible amount
during my one year under his guidance and I’m very thankful for the opportunity.
Of course, I would like to thank the members of the Chow Lab: Jung-Hwa Cho, Joe Johnson, Vivek
Shanmuganathan, Nancy Lee, Madison Zitting, Joyce Rohan, Matthew Behrend, and Andrew Weitz.
Throughout this project, they all offered tremendous support, important technical advice, and most
importantly, an enjoyable working environment.
I would also like to extend thanks to Dr. Derek Sieburth and members of the Sieburth lab: Jason Chan,
Krishnakali Dasgupta, Trevor Griffen, and Han Wang. They were very generous with their equipment and
supplies, and were a pleasure to work with.
Thank you also to the members of my thesis committee: Dr. Harvey Kaslow, Dr. Janos Peti-Peterdi, and
Dr. Richard Watanabe. I appreciated you taking the time out of your busy schedules.
Thanks always to my loving parents, Lee and Kris Tomooka, who never fail to offer unwavering support.
A very special thank you, of course, to my fiancé, Teresa McEvoy, for her love, her encouraging words,
and for always being there for me. Thanks also to my extended family: Teresa’s parents, Michael and
Stella McEvoy; Teresa’s brother, Michael, and his wife, Amy; and Teresa’s younger brother, Steven. I’m
very appreciative of your love and continued support. And finally, I will thank our Himalayan cat, Turk,
for helping me to relax.
iii
TABLE OF CONTENTS
Acknowledgements ii
List of Figures iv
Abstract v
1. Introduction 1
1.1 Role of zinc in insulin secretion 1
1.2 ZnT8 is critically involved in insulin secretion 5
1.3 Investigating zinc transport capabilities of ZnT8 7
2. Materials & Methods 9
2.1 Chemicals and reagents 9
2.2 Cell culture 9
2.3 Generation of the construct 10
2.4 Immunocytochemistry 12
2.5 Fluozin-3 imaging 13
3. Results 15
3.1 SP-ZnT8 R325 construct expresses RFP in transfected cells. 15
3.2 Prolactin signal peptide targets ZnT8 to the plasma membrane in a
flipped orientation. 15
3.3 Flipped ZnT8 is functional and mediates intracellular zinc uptake. 16
3.4 SP-ZnT8 R325 transected cells demonstrate greater intracellular zinc
accumulation than SP-ZnT8 R325W 20
4. Discussion 24
References 29
iv
LIST OF FIGURES
Figure 1. Suspected localization of known zinc transporters. 1
Figure 2. Cascade of events leading up to insulin exocytosis. 2
Figure 3. Insulin-zinc hexamers form dense cores in secretory vesicles. 3
Figure 4. Pulsatile insulin secretion rates at the portal vein. 4
Figure 5. Predicted membrane topology of ZnT8. 6
Figure 6. Illustration of the bicistronic construct encoding for flipped ZnT8 and
RFP. 12
Figure 7. Immunocytochemistry targeting the N-terminal tail of SP-ZnT8 R325. 17
Figure 8. Immunocytochemistry targeting the C-terminal tail of SP-ZnT8 R325. 18
Figure 9. Imaging of zinc uptake in cells expressing RFP using Fluozin-3-AM. 20
Figure 10. Fluozin-3 fluorescence accumulates at a faster rate in cells expressing
RFP during zinc exposure. 22
Figure 11. SP-ZnT8 R325 transfected HEK293T cells demonstrate significantly
elevated rates of zinc uptake. 23
Figure 12. SP-ZnT8 R325 transfected HEK293T cells are distributed at higher
rates of zinc uptake. 24
v
ABSTRACT
Zinc is an essential trace element in living organisms, and zinc content in pancreatic !-cells is amongst the
highest in the body. Zinc transporter 8 (ZnT8) is localized to the insulin vesicle membrane and is
responsible for transporting zinc from the cytoplasm into the vesicles. Inside the vesicles, zinc is
hypothesized to crystallize with insulin in a physiologically important process. Interestingly, recent
genome-wide association studies have identified a non-synonymous single nucleotide polymorphism (SNP)
in the gene encoding ZnT8 that is highly associated with the development of type 2 diabetes. However, to
date, no studies have been capable of analyzing the effects of the SNP on ZnT8 function. We have
designed a flipped ZnT8 that can be used to monitor cytoplasmic zinc influx using the fluorescent zinc
indicator, Fluozin-3-AM. This appears to be a promising method for directly analyzing the functional
consequences of the ZnT8 T2D-associated SNP.
1
1. INTRODUCTION
Zinc is an essential trace element in living organisms, and the bioavailability of zinc is regulated in part by
transmembrane zinc transport proteins. Zinc transporters are divided into two families: the ZIP (Zrt, Irt-
like protein) family and the CDF (Cation Diffusion Facilitator) family, which are distinguished by their
ability to facilitate cytoplasmic zinc influx or efflux, respectively. CDF zinc transporters (abbreviated as
ZnT) achieve cytoplasmic zinc efflux by moving zinc either out to the extracellular space or into
intracellular organelles. Figure 1 illustrates the typical subcellular locations of known ZIP and ZnT
members.
Figure 1. Suspected localization of known zinc transporters. Arrows indicate the direction of zinc
transport. ZIP members promote cytoplasmic influx and CDF (ZnT) members promote cytoplasmic efflux.
The red box highlights ZnT8. TGN, trans-Golgi network. (Modified from Kambe et al., 2004)
1.1 Role of zinc in insulin secretion
Of note, the zinc content in pancreatic !-cells is among the highest for cells in the human body, and it
appears to play a key physiological role in insulin secretion. Within !-cells, insulin is first synthesized as
preproinsulin, which includes a signal peptide at the N-terminus necessary for translocation into the
2
endoplasmic reticulum (ER). Once translation is completed, cleavage of the signal peptide results in
proinsulin, which is composed of three peptide chains termed A, B, and C (which are linearly arranged,
from N- to C-terminus, B-C-A) (Fig. 2A). In the reducing environment of the ER, proinsulin undergoes
conformational changes, and the A-chain and B-chain become linked to each other by two disulfide bonds.
Once proinsulin is transferred to the Golgi apparatus, it begins to associate with zinc ions, and forms the
zinc-proinsulin hexamer that consists of two Zn
2+
ions and six proinsulin molecules (Fig. 2B & Fig. 3B).
As these hexamer-containing vesicles (also called “granules”) bud off from the Golgi, they translocate to
the plasma membrane, where they await the appropriate signal to undergo secretion.
Figure 2. Cascade of events leading up to insulin exocytosis. The molecular events (left) corresponding
to the cellular events (right) are shown. Noteworthy events are also highlighted: (A) Pre-proinsulin is
synthesized in the rough ER; (B) Proinsulin begins to associate with zinc in the Golgi and forms a hexamer;
(C) As vesicles bud off from the Golgi, proinsulin is gradually converted to insulin by cleavage of the C-
peptide; (D) Zinc-insulin hexamers form a dense crystal in the core of the vesicles. (Modified from Emdin
et al., 1980)
3
During this time, the vesicles also go through a “maturation” process in which the proinsulin is converted
to biologically active insulin by removal of the C-chain (also termed the C-peptide) by intravesicular pro-
carboxypeptidases (Fig. 2C). In the zinc-proinsulin hexamer, the C-peptide resides on the periphery, while
the A- and B-chains are retained in the interior, near the zinc ions (Fig. 3A). This appears to be important
for two reasons: first, the C-peptide is readily accessible to peptidases in this conformation; and second, the
removal of the C-peptide drastically changes the solubility characteristics of the zinc-insulin hexamer.
Once the C-peptide is removed, the remaining A and B chains continue to associate with zinc in the same
hexameric conformation but the exterior of the molecule is now changed (Dodson & Steiner, 1998). While
the zinc-proinsulin hexamer is soluble, the zinc-insulin hexamer is highly insoluble and forms a dense
crystal in the core of the vesicle (Fig. 2D & Fig. 3C). The dense core is a low osmotic form of insulin,
meaning more insulin molecules can be stored per volume. The formation of this dense core is dependent
on the availability of zinc. The C-peptide is retained in vesicles in the more soluble “halo” surrounding the
dense core, which also includes other peptides and ions.
Figure 3. Insulin-zinc hexamers form dense cores in secretory vesicles. (A) The insulin monomer
(black) is shown in the context of the complete hexamer (light gray). Note that the B-Chain C-terminus
(B30) and A-Chain N-terminus (A1-3) are exposed on the surface and represent points of connection to the
C-peptide. (Taken from Dodson and Steiner, 1998) (B) Drawing of the zinc-insulin hexamer viewed along
the 3-fold hexamer symmetry axis. A round zinc ion is shown in the middle, surrounded by three insulin
molecules. (Taken from Dunn, 2005) (C) Electron micrograph demonstrating the zinc-insulin dense cores
of secretory granules in rat !-cells. (Taken from Michael et al., 1987)
4
Interestingly, our lab has previously shown that rat and human pancreatic !-cells secrete insulin in
heterogenous forms that disperse over a wide range of times, from milliseconds to minutes (Michael et al.,
2006). We hypothesize that the slowly dissolving forms represent the insoluble dense zinc-insulin crystals
from mature vesicles, while the rapidly dispersing forms most likely represent non-crystalline insulin that is
not associated with zinc, or the more soluble proinsulin-zinc hexamers arising from less mature vesicles.
The ratio of fast- and slow-release storage forms of insulin may be regulated, either by controlling the
amount of intravesicular zinc, or by the degree of proinsulin-to-insulin processing. Maintaining a mixed
population of insulin vesicles that dissolve at various rates may be a potential mechanism by which
pancreatic !-cells release insulin with a complex time course. When measured at the portal vein, insulin
release appears as a series of brief (~2 minute duration) pulses, separated by periods of approximately 5-10
minutes. Between pulses, the insulin release occurs at a constant, low, basal rate (Fig. 5). No secretion is
believed to take place between bursts, and we speculate that the slowly dissolving insulin cores are largely
responsible for ensuring the gradual release of insulin during the periods between pulses. It appears that
this constitutively available insulin may be necessary for proper metabolic function. Srivastava & Goren
(2003) reported that in the absence of constitutive insulin, primary cultures of mouse pancreatic islets failed
to exhibit the first-phase of insulin secretion in response to glucose.
Figure 4. Pulsatile insulin secretion rates at the portal vein. Insulin secretion appears as a series of
oscillating pulses that is superimposed over a sustained basal rate. The pulses may be due to the rapidly
dissolving insulin cores, while the continuous basal secretion may be due to the slowly dissolving insulin
cores. (Adapted from Song et al., 2000)
5
1.2 ZnT8 is critically involved in insulin secretion
ZnTs are predicted to be essential for facilitating the cytoplasmic efflux of zinc ions within the !-cells. To
date, ten isoforms of ZnTs have been identified, numbered 1-10 (Cousins et al., 2006). Common protein
structures that are shared amongst all isoforms include six transmembrane domains that are thought to form
the pore for zinc to pass through, and a histidine-rich region that is thought to bind the zinc ions. Most, if
not all, ZnTs appear to be expressed in numerous tissues. mRNAs from ZnT1-5 have been reported in
pancreatic islets (Clifford et al., 2000, Kambe et al., 2002). In particular, ZnT5 mRNA was found to be
abundantly expressed in !-cells and colocalized with insulin granules (Kambe et al., 2002). However,
ZnT5
-/-
mice were shown to maintain normal serum glucose and zinc levels (Inoue et al., 2002). It now
appears that ZnT5 functions primarily in the Golgi apparatus.
The recently-identified ZnT8 initially appeared to be expressed solely in the pancreas, and was thought to
be highly specific to the insulin vesicles (Chimienti et al., 2004). Using a ZnT8-eGFP (Green Fluorescent
Protein) fusion protein, it was concluded that ZnT8 mediates zinc transport from the cytosol to the inside of
insulin vesicles (Chimienti et al., 2006). More recent studies, though, have reported that expression of
ZnT8 can be found in glucagon-secreting pancreatic "-cells (Gyulkhandanyan et al., 2008), testis tissue
(Wenzlau et al., 2007), thyroid epithelium and the adrenal cortex (Murgia et al., 2008). Interestingly,
though, Chimienti et al. (2006) found that over-expression of ZnT8-eGFP in the insulinoma cell line INS-1
resulted in increased insulin secretion in response to glucose stimulation, suggesting a direct role for ZnT8
in insulin release. More recently, short hairpin RNA (shRNA) was used to knockdown ZnT8 mRNA
expression by >90% in INS-1 cells, resulting in reduced insulin content, decreased glucose-stimulated
insulin secretion, and fewer dense-core vesicles (Fu et al, 2009). Studies of ZnT8
-/-
mice also demonstrated
moderate impairments in insulin secretion (Pound et al., 2009).
Further evidence for the role of ZnT8 in insulin secretion comes from recent genome-wide association
studies (GWAS) that compare the genotypes of type 2 diabetes (T2D) patients and controls. Repeatedly,
across multiple independent studies, a single nucleotide polymorphism (SNP) in the gene SLC30A8 that
6
encodes ZnT8 was found to be highly associated with the incidence of T2D (Saxena et al., 2007; Scott et
al., 2007; Sladek et al., 2007; Steinthorsdottir et al., 2007; Zeggini et al., 2007). Meta-analysis of these
GWAS approximates the p-value significance at 1x10
-19
(Frayling, 2007). Notably, this SNP resides in the
protein-coding region of SLC30A8, and results in an amino acid change.
ZnT8 contains 369 amino acids, and is predicted to contain the six membrane-spanning segments that are
typical of ZnTs. The N- and C-terminal tails are thus on the same side of the plasma membrane, and
immunocytochemistry experiments suggest that the tails face the cytoplasm (Fig. 6). The SNP associated
with T2D is a C !T polymorphism that results in an arginine (R) to tryptophan (W) change at the 325
th
amino acid. Amino acid 325 is in the C-terminal tail, which spans from amino acids 267-369. It should be
noted that the risk allele is the major allele, encoding for R325. The major allele frequency was
approximately 70% in the populations studied (Frayling, 2007). In this thesis, ZnT8 R325 will denote the
risk allele, and ZnT8 R325W will denote the non-risk allele.
Figure 5. Predicted membrane topology of ZnT8. The predicted membrane topology based on
immunocytochemistry of ZnT8-eGFP transfected INS-1 cells. ZnT8 is predicted to mediate the transport
of zinc from the cytosol to the interior of insulin vesicles. The histidine-rich region between
transmembrane domains 4 & 5 is thought to bind zinc ions. Antibodies 9A and 9B target the regions
shown. Also shown is the approximate location of the non-synonymous SNP associated with T2D.
(Modified from Chimienti et al., 2006)
Results from the GWAS have certainly piqued interest in ZnT8, and subsequent studies have focused on
clinical measurements of glucose metabolism to correlate phenotype with the ZnT8 genotype. Both oral
glucose tolerance tests (OGTT) and intravenous glucose tolerance tests (IVGTT) have been used to obtain
7
an index of insulin secretion, but there have been conflicting results depending on the assay utilized.
Steinthorsdottir et al. (2007) found reduced insulin secretion by OGTT in subjects that were homozygous
carriers of ZnT8 R325. Similarly, Staiger et al. (2007) found reduced insulin secretion by OGTT as well as
IVGTT in homozygous carriers. However, Boesgaard et al. (2008) found no difference in insulin secretion
by OGTT, but did detect decreased first-phase insulin secretion by IVGTT in homozygous carriers. Also
notable, Kirchoff et al. (2008) reported that the ZnT8 SNP was associated with an impaired proinsulin to
insulin conversion. Thus, a number of clinical studies support the conclusions of the GWAS studies,
pointing to a T2D-associated phenotype in those people with the ZnT8 R325 genotype.
1.3 Investigating zinc transport by ZnT8
Evidence continues to mount that links ZnT8 R325 with !-cell dysfunction. However, to date, no studies
have been able to investigate the affect of the SNP on the function of ZnT8 – that is, does the SNP increase
or decrease the ability of ZnT8 to transport zinc?
We hypothesize that the ZnT8 R325 genotype reduces zinc transport into insulin vesicles. This would
lower zinc-insulin crystal formation, possibly decreasing the amount of insulin that could be stored in each
vesicle, and forcing the !-cell to “overwork” in order to secrete enough insulin to contend with elevated
glucose levels. The overwork hypothesis has been used in other contexts to explain !-cell dysfunction in
T2D. In addition, reduced zinc-insulin crystal formation could decrease the basal rate of insulin secretion
between pulses, which appears to be physiologically important as described earlier. These disturbances
could thus interfere with the normal release of insulin and ultimately predispose one to T2D.
Unfortunately, investigating zinc uptake into isolated insulin granules is complicated by their very small
size (diameter ~ 250 nm), and the fact the isolated granules are already loaded with high levels of zinc
(insulin vesicle [Zn2+] ~ 20-30 mM, Emdin et al., 1980). A more feasible option is to monitor zinc fluxes
in the cytosol, where intracellular [Zn
2+
] is typically in the pM to nM range.
8
Here, we propose to target ZnT8 to the plasma membrane and, in the process, “flip” its membrane
topology: regions that were facing the cytosol in the native protein now face the extracellular space, and
vice versa. In this flipped orientation, its function is also reversed, and it would be predicted to transport
zinc into the cytosol rather than out. By preloading the cytosol with a fluorescent zinc indicator, we can
monitor zinc uptake. Observing transport of extracellular zinc into the cytoplasm is a strategy previously
used for ZIP transporters (Muylle et al., 2006). In addition, flipping the orientation of a transmembrane
protein has also been carried out in other studies. Hu et al. (2003) utilized the bovine pre-prolactin signal
peptide to flip the three subunits of the Soluble NSF Attachment Protein Receptor (SNARE) complex.
While two of the subunits, Syntaxin and SNAP-25, are plasma membrane associated proteins, the other
subunit, VAMP, is a vesicular transmembrane protein similar to ZnT8. As VAMP was capable of retaining
functionality when redirected to the plasma membrane in a flipped orientation, we felt this approach would
also be feasible using ZnT8. As described below, we adopted the same technique to generate a construct
that expresses a flipped ZnT8 in the plasma membrane, and we characterized the transport properties of this
flipped transporter.
As our studies required imaging cytosolic fluorescent dye signal at the single-cell level, it was necessary to
be able to identify which cells were expressing the flipped ZnT8. Chimienti et al. (2004 & 2006) utilized a
ZnT8-eGFP fusion protein in their studies, but we were concerned that a fluorescent protein attached to the
transporter could alter zinc transport function. Thus, we generated a bicistronic vector, encoding for
separate expression of ZnT8 and a red fluorescent protein (to avoid fluorescence overlap with a green
fluorescent zinc dye). In our construct, the first cistron (flipped ZnT8) was expressed under a
cytomegalovirus (CMV) promoter, and the second cistron (Red Fluorescent Protein, or TagRFP) was
expressed under the control of an internal ribosome entry site (IRES) sequence.
To test our plasmid, we utilized AtT-20 and HEK293T cell lines primarily because of their ease of
transfection and lack of significant native zinc uptake. AtT-20 cells are derived from mouse anterior
pituitary cancer cells, and secrete high levels of adrenocorticotropic hormone (ACTH). HEK293T cells are
human embryonic kidney cells that were transformed into a cell line by adenovirus. While AtT20 cells
9
contain dense-core secretory vesicles, HEK293T cells do not. We demonstrate by immunocytochemistry
that the N-terminal tail of flipped ZnT8 is exposed on the cell surface in transfected cells, suggesting not
only that the protein reaches the plasma membrane, but also that its topology has been flipped.
To determine changes in cytoplasmic zinc concentrations, we used the membrane permeant Fluozin-3-AM
zinc dye as a fluorescent indicator. Fluozin-3-AM is a highly specific and sensitive zinc indicator, with a
K
d
~15 nM. Peak excitation and emission wavelengths are 494/516, giving it fluorescent properties similar
to GFP and allowing us to conduct two-color fluorescence imaging in conjunction with our TagRFP
protein. Our unique strategy then allows direct measurements and analysis of functioning ZnT8, and can
potentially clarify the consequences of the ZnT8 R325 gene variant that predisposes one to T2D.
2. MATERIALS AND METHODS
2.1 Chemicals and reagents
All chemicals were of reagent grade and from Sigma-Aldrich, unless otherwise noted. In particular, 1-
Hydroxypyridine-2-thione (Pyrithione), N,N,N!,N!-Tetrakis(2-pyridylmethyl)ethylenediamine (TPEN),
Probenecid, and Sulfinpyrazone were obtained from Sigma-Aldrich. Fluorescent dye Fluozin-3-AM was
obtained from Molecular Probes (Eugene, OR). Lipofectamine 2000 was obtained from Invitrogen.
Pluronic F-127 was obtained from Biotium, Inc. All DNA primers were ordered from Integrated DNA
Technologies (IDT).
2.2 Cell culture
AtT-20 cells were obtained from ATCC, USA, and grown in Dulbecco’s Modified Eagle’s Medium
(DMEM, CellGro) containing 4.5 g/L glucose, L-glutamine, sodium pyruvate, supplemented with 10%
horse serum, 1% penicillin-streptomycin, and pH adjusted to 8.0. HEK 293T cells were grown in modified
Improved Minimum Essential Medium (IMEM, Gibco) containing L-gluatamine, not containing
gentamycin-sulfate, and supplemented with 10% fetal bovine serum and 1% penicillin-streptomycin.
Culture medium for both cells was changed every 3-4 days, and cells were grown to 90-100% confluency.
Cells were then detached by treatment with 0.5 mM EDTA and 0.05% Trypsin in PBS, centrifuged, and
10
diluted and redistributed at either 1:10 (AtT-20) or 1:20 (HEK 293T). Both cell lines were incubated at
37°C in a 5% CO
2
-enriched atmosphere.
2.3 Generation of the construct
We obtained the plasmid encoding ZnT8-eGFP as a gift from Dr. Fabrice Chimienti.
The ZnT8 sequence was amplified by PCR using primers EcoRI-Zn5
(GGAATTCCATGGAGTTTCTTGAAAGAACGTATCTTGT) and BamHI-Zn3
(CGGGATCCCGCTAGTCACAGGGGTCTTCACAGAAAA) and the ZnT8-eGFP plasmid as a template.
The PCR product was digested with EcoRI and BamHI and then subcloned into the pDNR-CMV
(Clontech) parent vector at the multiple cloning site (MCS).
The bovine pre-prolactin signal peptide (SP) sequence
(ATGGACAGCAAAGGTTCGTCGCAGAAAGGGTCCCGCCTGCTCCTGCTGCTGGTGGTGTCAAA
TCTACTCTTGTGCCAGGGTGTGGTCTCCACC) with a Kozak translation initiation sequence
(GCCGCCACCATGG) was then introduced by QuikChange site-directed mutagenesis (Stratagene).
Briefly, a PCR primer was designed that contained three regions: the first region was complementary to the
18 bps of the pDNR parent vector before ZnT8 (CGACGGTACCGGACATAT), the second region
contained the SP and Kozak sequence (102 bps), and the third region was complementary to the first 30 bps
of ZnT8 (ATGGAGTTTCTTGAAAGAACGTATCTTGTG). The reverse complement of this primer was
also ordered, and PCR was performed according to the manufacturer’s protocol (Stratagene) using the high-
fidelity PfuUltra polymerase. The PCR product was then incubated with DpnI endonuclease, which digests
the parent methylated DNA leaving only the unmethylated daughter DNA containing the newly introduced
Kozak-SP sequence. After incubation, 2 "l were transformed into 200 "l DH5-" competent cells, where
endogenous enzymes ligate the daughter DNA strands. Cells were plated on Ampicillin LB agar plates,
and DNA was prepped from surviving colonies. The final DNA product was sequenced to confirm the
introduction of Kozak and SP sequences immediately before the ZnT8 sequence.
11
The next steps involved inserting the IRES and TagRFP sequences downstream of the ZnT8 sequence. The
IRES2 sequence was amplified by PCR using primers BamHI-ires2-5 (CCGGGATCCGCCCCTCTC) and
BamHI-ires2-3 (GCTCACCATGGATCCGTTGTGGCC) and the pIRES2-eGFP plasmid (Clontech) as a
template. The PCR product was digested with BamHI. The pDNR-CMV-ZnT8 construct was treated with
Calf Intestinal Alkaline Phosphatase (CIP) after BamHI digestion to prevent self-ligation. The IRES2 PCR
product was then subcloned downstream of and in-frame with the ZnT8 sequence. The TagRFP sequence
was amplified by PCR using primers XhoI-NPY-RFP5 (CTCAGATCTCGAGCTCAAGCTTCG) and
SpeI-NPY-RFP3 (GGCCGCACTAGTTCAATTAAGTTTGTGCCCCAGT) and the pTag-NPY-RFP
plasmid (generated in our lab by Dr. Joe Johnson using pTagRFP from Evrogen) as a template. The NPY-
RFP (Neuropeptide Y) fusion protein was initially planned to be included in our construct, but it was later
decided to include only the TagRFP protein. The PCR product was then digested with BamHI (a BamHI
cut site was located between NPY and RFP) and SpeI, and subcloned into a region downstream of and in-
frame with the IRES2 sequence. All steps were verified by agarose gel electrophoresis, PCR screening,
and finally, DNA sequencing. This final construct then contains
CMV promoter – Kozak – Signal Peptide – ZnT8 – IRES2 – TagRFP
sequence in the pDNR parent vector (Fig. 6). As the risk allele is the major allele, this construct contains
the T2D-associated SNP and will be referred to as “SP-ZnT8 R325”.
The R325W SNP was introduced to ZnT8-RFP by QuikChange site-directed mutagenesis using PCR
primers R325W-5 (GCTACAGCAGCCAGCTGGGACAGCCAAG) and R325W-3 (reverse complement
of R325W-5) and carried out as previously described. Correct mutagenesis was confirmed by DNA
sequencing. This construct will be referred to as “SP-ZnT8 R325W”.
12
Figure 6. Illustration of the bicistronic construct encoding for flipped ZnT8 and RFP. The CMV
promoter (purple) regulates expression of SP-ZnT8 (blue and green), while an IRES sequence (orange)
promotes expression of TagRFP (red).
2.4 Immunocytochemistry
AtT-20 or HEK 293T cells were grown in 6-well plates to ~80-90% confluency. Cells were then
transfected with the appropriate plasmid using Lipofectamine 2000 according to the manufacturer’s
protocol. The following day, cells were trypsinized, spun down, and resuspended. Poly-D-lysine coated 12
mm round coverslips were placed in a 24-well plate, and cells were plated at ~25-50% confluency.
48-72 hours after transfection, cells were analyzed in both permeabilized and non-permeabilized
conditions. All steps were carried out at room temperature. In the permeabilized protocol, cells were
washed 3x in PBS, followed by fixation in 4% PFA in PBS for 15 minutes. Cells were then washed 3x in
0.1% BSA, 0.25% NH
4
Cl in PBS, followed by permeabilization in 0.1% Triton X-100 in PBS for 10
minutes. Cells were then blocked in 5% goat serum, 3% BSA in PBS for 1 hour. Primary antibodies (9A
directed against amino acids 355-359 in the C-terminal tail or 9B directed against amino acids 34-49 in the
N-terminal tail) were a gift from Dr. Fabrice Chimienti. Primary antibodies were diluted 1:100 (9B) or
13
1:50 (9A) in blocking solution and incubated for 1 hour. Cells were then washed 3x in 0.1% BSA in PBS.
Goat anti-rabbit Alexa secondaries (Alexa 488 or Alexa 555) were then diluted 1:1000 in blocking solution
and incubated for 45 minutes in the dark. Cells were then washed 3x in 0.1% BSA in PBS, 3x in PBS, and
followed by a quick rinse with ddH
2
0 before mounting in VectaShield mounting solution.
In the non-permeabilized protocol, cells were washed 1x in PBS and then incubated in primary antibody
diluted in 0.1% BSA in PBS for 30 minutes. Cells were then washed 3x in 0.1% BSA in PBS, and
followed by fixation in 4% PFA for 15 minutes. Cells were then washed 3x in 0.1% BSA, 0.25% NH
4
Cl in
PBS, and then blocked as described above. Secondary antibody and subsequent washes were also as
described above.
Fluorescent images were taken on a Leica microscope using a 63x glycerol objective (courtesy of Dr. Janos
Peti-Peterdi). eGFP or Alexa 488 signal was excited by a 488 nm Argon laser. RFP or Alexa 555 signal
was excited by a 543 nm HeNe laser.
2.5 Fluozin-3 imaging
Similar to the immunocytochemistry experiments, AtT-20 or HEK 293T cells were grown in 6-well plates
to ~80-90% confluency. Cells were then transfected with the appropriate plasmid using Lipofectamine
2000 according to the manufacturer’s protocol. The following day, cells were trypsinized, spun down, and
resuspended. Poly-D-lysine coated 12 mm round coverslips were placed in a 24-well plate, and cells were
plated at ~25-50% confluency.
48-72 hours after transfection, cells were analyzed. 5 mM Fluozin-3-AM (resuspended in ~99.9% reagent
grade DMSO), 20% Pluronic F-127, 250 mM probenecid, and 25 mM sulfinpyrazone stock solutions were
prepared. The subconfluent cells that were plated on 12 mm round coverslips were then washed 2x in
external solution consisting of (in mM): 140 NaCl, 5 KCl, 2 CaCl
2
, 1 MgCl
2
, 10 HEPES, 5 D-glucose, pH
7.4, 290-300 mOsm. Fluozin-3-AM, Pluronic F-127, probenecid, and sulfinpyrazone stock solutions were
then diluted in external solution to a final concentration of 5 "M Fluozin-3-AM, 0.02% Pluronic F-127, 2.5
14
mM probenecid, and 250 "M sulfinpyrazone. After washing, the coverslips were incubated in the Fluozin
solution for 60 minutes at room temperature. Cells were then washed 2x in external solution, and then
incubated in 2.5 mM probenecid, 250 µM sulfinpyrazone in external solution for 10-15 minutes at 37°C to
complete probe de-esterification.
Coverslips were then placed in the microscope chamber for imaging. All images were taken at a frequency
of 0.2 Hz. Excitation wavelength light was set by a monochromator (Till Photonics, Germany) and
illumination was maintained for 250 msecs. Between exposures, the excitation wavelength was switched to
600 nm to reduce loss of signal due to photobleaching. RFP signal was excited at 543 nm, and captured
using a 567 dichroic and a 585 longpass emission filter. Fluozin-3 signal was excited at 488 nm, and
captured using a 505 dichroic and 525/50 bandpass emission filter. A perfusion system provided a
continuous flow of solution into one side of the chamber, while a suction pipette gently aspirated solution
from the other side (Fig. 9A). Zn
2+
-free external solution was first applied as a baseline fluorescent signal
was obtained for approximately 5 minutes. The perfusion buffer was then switched to external solution
containing 50 "M ZnCl
2
. Imaging was then conducted for approximately 10-15 minutes. At the end of this
period, we saturated the Fluozin-3 dye to obtain the maximum fluorescence signal, as required for
calibration of the dye. This perfusion buffer was a Ca
2+
and Mg
2+
free external solution containing (in
mM): 145 NaCl, 5.4 KCl, 10 HEPES, 5 D-glucose, pH 7.4, 290-300 mOsm. In addition, this solution
contained 50 "M ZnCl
2
with 50 "M Pyrithione as an ionophore. Cellular fluorescence was then imaged
until a stable fluorescent signal was reached, approximately 5 minutes. Finally, the perfusion buffer was
switched to Ca
2+
and Mg
2+
free external solution containing 100 "M TPEN, which is a membrane-permeant
heavy metal chelator (K
d
(Zn
2+
) ~ 10
-16
). TPEN chelates all Zn
2+
, providing the baseline fluorescent signal
of unbound Fluozin-3, also required for calibration.
Images were captured with a Roper Cascade 512B camera and taken on an Axiovert 100 Zeiss microscope
using a 20x/0.50 air objective, allowing multiple cells to be imaged in a single frame. MetaFluor software
was used to quantify the fluorescent signal, and data was later analyzed in Igor Pro, using custom-written
analysis routines.
15
3. RESULTS
3.1 SP-ZnT8 R325 construct expresses RFP in transfected cells.
After DNA sequencing confirmed our SP-ZnT8 R325 construct, AtT-20 and HEK293T cells were
transfected and imaged by fluorescence microscopy. Detectable red fluorescence demonstrated that the
RFP protein was being expressed, and was determined to be a valid indicator of successful plasmid
transfection.
3.2 Prolactin signal peptide targets ZnT8 to the plasma membrane in a flipped orientation.
To assess the targeting and membrane topology of our construct (SP-ZnT8), we performed
immunocytochemistry using rabbit anti-human ZnT8 antibodies generously provided by Dr. Fabrice
Chimienti. As a control, we first performed experiments on AtT-20 cells transfected with the ZnT8-eGFP
(non-flipped) plasmid provided by Dr. Chimienti. As expected (Chimienti et al., 2006), the N- and C-
terminal tails were accessible in permeabilized cells but not non-permeabilized cells (data not shown).
Antibody 9A appeared to bind with less affinity than antibody 9B as a smaller percentage of eGFP
expressing cells were immunostained by 9A than by 9B (data not shown).
To test our novel construct, SP-ZnT8 R325, we transfected HEK 293T cells and stained them in both
permeabilized and non-permeabilized settings. In permeabilized cells, immunofluorescent staining with
both 9A and 9B is demonstrated in RFP-positive cells (Fig. 7 & Fig. 8, top panels). Again, antibody 9A
demonstrated less specificity and strength of immunofluorescent signal (compare Fig. 7 & Fig. 8, top
panels). In addition, the antibody signal is detected in the cytosol, suggesting only a certain percentage is
targeted to plasma membrane. Hu et al. (2003) estimated that only ~5% of their flipped VAMP was
expressed in the plasma membrane, probably due to the endogenous unfolded protein response in the ER.
Punctate staining may indicate that the protein tends to form aggregates.
In non-permeabilized cells, the epitopes targeted by the antibodies should remain accessible in the flipped
ZnT8, in contrast to native ZnT8 where the epitopes are only accessible in permeabilized cells. Antibody
9B demonstrates definitive staining on the periphery of a subset of RFP-positive cells, suggesting the N-
16
terminus is exposed on the cell surface (Fig. 7, bottom panels). Antibody 9A generally demonstrated a lack
of antibody binding in non-permeabilized cells (Fig. 8, 3
rd
row of images). In a small subset of RFP-
positive cells, 9A immunofluorescence was detected although we were not able to rule out some
contribution from non-specific staining (Fig. 8, bottom row). As 9A showed less binding in positive
controls, it was not entirely unexpected that immunofluorescence in non-permeabilized cells was difficult
to detect. Unfortunately, limited amounts of antibody prevented complete antibody titration experiments.
It is also quite possible that the epitope targeted by antibody 9A is no longer easily accessible. This will be
further addressed in the Discussion section.
It is also apparent that the level of ZnT8 expressed varies from cell to cell, and that immunofluorescence
does not always correlate with RFP expression (data not shown). However, it is clear that in at least a
subset of RFP-positive cells, SP-ZnT8 R325 is targeted to the plasma membrane, and the N-terminal tail is
exposed on the cell surface.
3.3 Flipped ZnT8 is functional and mediates transmembrane zinc uptake.
To assess the function of flipped ZnT8, we used the membrane-permeant zinc dye, Fluozin-3-AM. This
acetoxymethyl derivative is nonpolar and able to pass through cell membranes. After entering a cell,
endogenous intracellular esterases cleave off the acetoxymethyl groups on Fluozin-3-AM, resulting in a
negatively charged molecule that is no longer membrane permeable and is retained inside the cell. We
found that addition of the anion transporter blockers probenecid and sulfinpyrazone greatly enhanced the
dye-loading of cells, and were routinely used in all experiments.
After dye-loading and equilibration, the coverslips were loaded in the microscope chamber. First, we
identified RFP-expressing cells. In the camera frame, we attempted to include >1 RFP-expressing single
cells and >1 non-RFP expressing single cells. Cell clusters were avoided, if possible. If cell clusters were
in the camera frame they were not included in the data analysis. Once the camera frame had been
established, it was not moved for the remainder of the experiment.
17
Figure 7. Immunocytochemistry targeting the N-terminal tail of SP-ZnT8 R325. Antibody 9B
(targeting N-terminal amino acids 34-49) was incubated with SP-ZnT8 R325 transfected HEK293T cells in
permeabilized (top panels) and non-permeabilized (bottom panels) conditions. RFP expression was present
in both conditions and suggested successful transfection. In permeabilized cells, in addition to
immunofluorescence in the cell periphery, some immunofluorescence appears to be in the cytosol
suggesting not all SP-ZnT8 R325 is targeted to the plasma membrane. In non-permeabilized cells,
detectable immunofluuorescence on the cell periphery suggests the N-terminus is exposed on the cell
surface.
18
Figure 8. Immunocytochemistry targeting the C-terminal tail of SP-ZnT8 R325. Antibody 9A
(targeting C-terminal amino acids 355-359) was incubated with SP-ZnT8 R325 transfected HEK293T cells
in permeabilized (top panels) and non-permeabilized (bottom panels) conditions. RFP expression was
present in both conditions, suggesting successful transfection. Antibody 9A was able to recognize the
ZnT8 protein in permeabilized cells although to a lesser extent than antibody 9B (see Fig. 7, top panels). In
non-permeabilized cells, there was generally no detectable binding (3
rd
row of images) although sparse
immunofluorescence was detected on the periphery of a small subset of RFP-expressing cells (bottom
panel).
19
The addition of the zinc ionophore, pyrithione, caused the Fluozin-3 fluorescence to saturate, giving an
indication of the amount of dye loaded within each cell. After measuring the saturated Fluozin-3 signal, we
added the high-affinity zinc chelator TPEN to obtain the Zn
2+
-free Fluozin-3 signal. Using the pyrithione
value as F
max
and the TPEN value as F
min
, we normalized the data from each cell by adjusting each
fluorescence trace such that F
min
=0 and F
max
=1.
Figure 9 shows the experimental design and selected images from an experiment on SP-ZnT8 R325
transfected HEK293T cells. Although difficult to appreciate in these images, the RFP-negative control cell
(con1) shows a slight increase in Fluozin-3 fluorescence after 50 "M ZnCl
2
exposure (Fig. 9E). In contrast,
RFP-expressing single cells (RFP1 and RFP2) clearly show observable fluorescent increases after 50 "M
ZnCl
2
exposure (Fig. 9E). This experiment also demonstrates that the amount of zinc uptake varies
amongst RFP-expressing cells, and does not appear to correlate with RFP expression (cell RFP1 had the
greatest elevation in zinc uptake, but less RFP fluorescence than cell RFP2). In other experiments, it was
also observed that not all RFP-expressing single cells demonstrated increased Fluozin-3 fluorescence.
These observations are consistent with the immunocytochemistry data in that RFP expression did not
always correlate with ZnT8 immunofluorescence (Fig. 7 and Fig. 8). However, we did conclude that in at
least some cells, there is concurrent expression of both RFP and ZnT8, and that the ZnT8 appears
functional and capable of mediating zinc uptake in the cytoplasm. More detailed data will be presented in
the next section.
During baseline recordings, all cells demonstrated nearly constant Fluozin-3 fluorescence. In some cases, a
slight decrease in signal was attributed to a photobleaching effect. It was also noted that the addition of
TPEN caused the fluorescent signal to rise transiently before falling at an exponential rate. It is speculated
that the TPEN caused the cells to shrink and change shape slightly, thus increasing the measured
fluorescent signal. This may need to be taken into account when calculating the F
min
value in future
studies, but was not considered in this paper.
20
Figure 9. Imaging of zinc uptake in RFP-expressing cells using Fluozin-3-AM. (A) Experimental
setup of the perfusion system. HEK293T cells were transfected with SP-ZnT8 R325 and plated on glass
coverslips for imaging. (B) Bright field image of the chosen cells prior to imaging. (C) RFP fluorescence
of the chosen cells prior to imaging. RFP1 & RFP2 express RFP, but con1 does not. (D), (E), (F). Fluozin
fluorescence of the chosen cells is shown at various time points: at baseline (D), after 50 "M ZnCl
2
exposure (E), and after 50 "M ZnCl
2
+ 50 "M pyrithione exposure (F). After ~ 750 seconds of exposure to
50 "M ZnCl
2
, zinc accumulates in RFP-expressing cells (E). Exposure of the cells to pyrithione saturates
the Fluozin-3 signal in all dye-loaded cells (F). The Fluozin-3 fluorescence data from this experiment is
plotted in Fig. 10. The cell cluster seen in the upper left of each image was not included in the data
analysis.
3.4 SP-ZnT8 R325 transfected cells demonstrate greater intracellular zinc accumulation than SP-ZnT8
R325W
In order to investigate the possible effects of the ZnT8 SNP on ZnT8 function, three plasmids were tested:
SP-ZnT8 R325, SP-ZnT8 R325W, and pTagRFP. pTagRFP was used as a control for transfection with an
RFP-expressing plasmid.
21
As stated earlier, we attempted to image a portion of the coverslip that contained both RFP-positive and
RFP-negative single cells. After normalizing the fluorescent recordings in each cell to the pyrithione-
induced F
max
and TPEN-induced F
min
, we plotted the data (Fig. 10A) and determined the maximum rate of
fluorescent rise (slope) during the 50 "M ZnCl2 exposure period (Fig. 10B). RFP-negative control cells
(n=16) had an average maximum slope of 5.01 ± 0.72 (arbitrary units, ± s.e. of the mean), pTagRFP-
positive cells (n=5) had an average maximum slope of 6.25 ± 1.78, SP-ZnT8-R325-positive cells (n=7) had
an average maximum slope of 12.24 ± 3.42, and SP-ZnT8-R325W-positive cells (n=11) had an average
maximum slope of 7.41 ± 1.98. Zinc uptake is observed in each condition (evident by the average upward
slopes), suggesting that zinc uptake may occur through endogenous ZIP transporters present on the cell
membrane, or through other ion channels.
To analyze the effects of our flipped ZnT8 constructs on zinc uptake, we first considered the maximum
slopes of each RFP-positive cell relative to the maximum slopes of RFP-negative cells in the same camera
frame (i.e. the maximum slope of each RFP-positive cell was divided by the maximum slope of RFP-
negative cells within the same experiment). If more than one RFP-negative cell was imaged in an
experiment, then the maximum slopes of each were averaged to generate a single value. By taking the
ratio, we normalized our data to RFP-negative control cells. The maximum slope ratio thus reflects the
change in Fluozin-3 fluorescence due to the transfected plasmid. A maximum slope ratio of 1 would then
denote no change in Fluozin-3 fluorescence relative to RFP-negative cells, while a ratio > 1 indicates
elevated fluorescent levels over controls.
The ratio of the slopes of RFP-positive over RFP-negative cells (denoted “maximum slope ratio”) was then
used to compare levels of zinc uptake between the three plasmids tested. Results were compiled from 12
separate experiments (4 using SP-ZnT8 R325, 5 using SP-ZnT8 R325W, and 3 using pTagRFP). In each
experiment, an average of 1.9 ± 0.3 (s.e. of the mean) RFP-positive cells and 1.3 ± 0.2 RFP-negative cells
were imaged. SP-ZnT8 R325-positive cells had an average maximum slope ratio of 5.87 ± 1.07 (s.e. of the
mean), SP-ZnT8 R325W-positive cells had an average maximum slope ratio of 1.42 ± 0.24, and pTagRFP-
positive cells had an average maximum slope ratio of 1.17 ± 0.26. Figure 11 plots the slope ratios of each
22
Figure 10. Fluozin-3 fluorescence accumulates at a faster rate in cells expressing RFP during zinc
exposure. HEK293T cells were transfected with SP-ZnT8 R325 and plated on coverslips for imaging.
The data shown is from the same experiment as the images shown in Fig. 9. (A) Fluozin-3 fluorescence
was normalized to F
min
(determined after TPEN exposure) and F
max
(determined after Pyrithione exposure).
The perfusion buffers were switched as indicated above the black bars. The letters refer to the time points
at which the images in Fig. 9 were taken. (B) The period during zinc exposure is magnified. The lines
give an example of how the maximum slopes were estimated.
23
RFP-positive cell imaged, categorized by plasmid transfected. Maximum slope ratios in SP-ZnT8 R325
transfected cells were statistically significantly higher than both SP-ZnT8 R325W (two-tailed t-test p-value
= 0.00556) and pTagRFP (one-tailed t-test p-value = 0.00211). No difference in zinc uptake was observed
between SP-ZnT8 R325W and pTagRFP transfected cells.
Figure 12 shows a frequency histogram of the slope ratios. pTagRFP and SP-ZnT8 R325W distributions
overlap, but the SP-ZnT8 R325 distribution is clearly shifted to the right.
Figure 11. SP-ZnT8 R325 transfected HEK293T cells demonstrate significantly elevated rates of zinc
uptake. Data was plotted for each experiment as shown in Fig. 10A and maximum slopes were calculated
for each cell as shown in Fig. 10B. RFP-positive slopes were divided by RFP-negative (control) slopes
within each experiment (this value is referred to as the maximum slope ratio). The data was plotted for
each of the three plasmids tested. This data represents the increase in zinc uptake due to the transfected
plasmid. pTagRFP ratios represent the zinc uptake for cells transfected only with TagRFP, as a control.
24
Figure 12. SP-ZnT8 R325 transfected HEK293T cells are distributed at higher rates of zinc uptake.
The maximum slope ratios were binned into 0.5 increments, and the number of cells falling into each bin
was counted for each of the three plasmids tested. pTagRFP and SP-ZnT8 R325W populations overlap, but
the SP-ZnT8 R325 population is clearly shifted to the right.
DISCUSSION
To date, many studies have correlated the ZnT8 R325 SNP to T2D and associated phenotypes. However,
none have looked at the effect of the non-synonymous SNP on the function of ZnT8. Here, we demonstrate
that a signal peptide targets ZnT8 to the plasma membrane and that the membrane topology appears to be
flipped. More importantly, though, it appears that this flipped ZnT8 is functional, and can facilitate the
movement of zinc across the plasma membrane. While the data is preliminary, this approach has the
potential to answer many key questions regarding the integral role of ZnT8 in insulin secretion and the
development of T2D.
The location of the ZnT8 SNP is interesting because it is in close proximity to several putative signaling
domains. A PROSITE database scan (http://www.expasy.ch/tools/scanprosite/) identified an N-
glycosylation site (NYS, 283-285) and a Casein Kinase II phosphorylation site (SPVD, 353-356) in the C-
terminal tail. These modification sites are important to consider because they may potentially influence
either the antigenicity of ZnT8, its ability to function as a zinc transporter, or both.
25
Our immunocytochemistry data suggests the N-terminal tail is exposed on the cell surface (Fig. 7), but is
inconclusive regarding the C-terminal tail (Fig. 8). Immunocytochemistry experiments need to be repeated
to confirm that the C-terminal tail is exposed on the cell surface in flipped ZnT8. One explanation for the
lack of conclusive binding by antibody 9A (which targets C-terminal amino acids 355-359) is that improper
protein folding in the C-terminus has masked the epitope. Normally, ZnT8 N- and C-terminals face the
cytosol, but in the flipped orientation they are exposed to the ER lumen during processing and may be
subject to atypical post-translation modifications. N-glycosylation occurs only in the lumen of the ER and
glycosylation could affect the epitope for binding of 9A antibody. To address this issue, we have
introduced a S285A (serine to alanine) mutation by QuikChange PCR that is expected to prevent
glycosylation from occurring.
The Casein Kinase II phosphorylation site (SPVD, 353-356) in the C-terminus overlaps with the epitope
targeted by 9A. Phosphorylation, or lack thereof, at this site may affect the antigenicity of the surrounding
region. The C-terminal tail of flipped ZnT8 would face the ER lumen, where it might not encounter the
normal kinases. To introduce a residue that would mimic phosphorylated serine at the phosphorylation site,
we performed a S353D (serine to asparate) mutation. Aspartate (D) carries a negative charge and has
previously been shown to successfully mimic phosphorylation-induced protein changes (Léger et al., 2007).
We have already introduced the glycosylation-site and phosphomimetic mutations alone and in
combination with each other. Future immunocytochemistry experiments utilizing these constructs will
address the importance of these modifications in exposing the C-terminal epitope.
As stated earlier, these modifications may also affect the function of ZnT8. Hu et al. (2003) found their
flipped SNARE proteins were only functional after introducing mutations to prevent artifactual
glycosylation from occurring. Even though our SP-ZnT8 R325 construct appears to be functional, future
Fluozin-3-AM experiments can address whether a potential glycosylation residue confounds the
observation of normal ZnT8-mediated zinc transport. In addition, phosphorylation may be necessary in
order for ZnT8 to be fully active. Fluozin-3-AM studies utilizing our phosphomimetic mutants can suggest
whether this signaling domain is important for ZnT8 under normal physiologic conditions.
26
A very important next step will be to estimate the number of zinc transporters in the plasma membrane.
SP-ZnT8 R325W transfected cells demonstrated significantly less zinc uptake than SP-ZnT8 R325
transfected cells, and was found to be no different than pTagRFP transfected control cells (Fig. 11).
However, this apparent lack of zinc uptake could be due to the fact there are less SP-ZnT8 R325W zinc
transporters in the plasma membrane. In that case, differences in zinc uptake could be observed even if the
individual zinc transporters operate at the same transport rate. As immunocytochemistry was not carried
out in cells transfected with SP-ZnT8 R32W, and as the N-terminal antibody (9B) appears capable of
binding the N-terminus, we would first like to perform immunocytochemistry on SP-ZnT8 R325W
transfected cells using antibody 9B. Comparing the immunofluorescence between the two SNP variants
can provide a very rough estimate of the their relative efficiency in targeting flipped ZnT8 to the plasma
membrane.
In future studies, though, we will need to make more precise estimations. In order to do this, we plan to
conduct Fluozin-3 imaging followed by quantitative immunofluorescence in those same cells. Using this
approach, we can make direct correlations between Zn
2+
influx and transporter density in the plasma
membrane. By resolving the zinc transport rate per transporter, we will have a reliable indicator of protein
function.
In order to verify that our flipped ZnT8 is functional, it will also be important to establish that it exhibits
properties expected of transporters. Zinc transport via zinc transporter proteins has previously been shown
to be time, temperature, and concentration dependent (Reyes 1996). Future studies will need to
demonstrate that our flipped ZnT8 exhibits those defining characteristics in our measurements. There is
also the concern that that the energy source that drives Zn
2+
transport via ZnT8 has not been identified.
Ohana et al. (2009) recently reported that zinc transport mediated by ZnT5 (a vesicle membrane associated
ZnT) is catalyzed by a H
+
/Zn
2+
exchange. Establishing the appropriate H
+
-gradient across the plasma
membrane may be necessary to observe true ZnT8 function.
27
Currently, we can make crude estimates regarding zinc transport rates. To convert observed fluorescence
changes to zinc concentration responses, we can apply the Grynkiewicz equation for non-ratiometric probes
(Grynkiewicz et al., 1985): [Zn
2+
] = K
d
* (F – F
min
)/(F
max
– F) where F
min
is the minimum signal after TPEN
treatment and F
max
is the maximum signal after pyrithione treatment. Although K
d
may need to be
intrinsically determined for our experimental system, we used the reported K
d
(15 nM) from Molecular
Probes in these preliminary calculations. Basal intracellular Zn
2+
levels were estimated at ~ 150 ± 100 pM
(n=39 cells). These levels are about an order of magnitude less than the 1.4 ± 0.2 nM that Muylle et al.
(2006) reported in fish hepatocytes based on Fluozin-3 fluorescence. Measures of zinc uptake in our data
were also very small, with the greatest change in zinc concentration estimated at ~3.5 nM over a period of
10 minutes. Interestingly, though, this was comparable to the reported average zinc increases of ~4 nM in
fish hepatocytes during the first 10 minutes of zinc exposure (Muylle et al., 2006). However, this
comparison is only informative if we assume the cells are of roughly the same volume. While volume of
fish hepatocyte cells were not estimated, HEK293 cells have been reported to have an average membrane
capacitance of 21 ± 4 pF (Chambard et al., 2003). If you assume a spherical cell, this translates to a cell of
diameter ~ 25.9 "m and volume ~ 9 pL. By taking the change in zinc concentration over time and
multiplying by the approximate volume, we estimate ~ 5.2 x 10
-23
moles of zinc ions are being transported
per second. Multiplying this by Avogadro’s number gives ~ 30 Zn
2+
ions transported into a cell per second
during zinc exposure. If we can quantify the number of zinc transporters in the membrane, we can
calculate the rate of Zn
2+
transport per transporter protein. It will be important to keep in mind that this rate
will not only depend on the number of zinc transporters, but also the external [Zn
2+
], the temperature, and
possibly the [H
+
] gradient. Again, this is a very crude estimate based on rough approximations, but it gives
an example of the type of per transporter data we can potentially glean from this experimental approach.
Regarding the data described here, we observed that the SP-ZnT8 R325 plasmid facilitates a significantly
elevated rise in Fluozin-3 fluorescence over both the SP-ZnT8 R325W and pTagRFP plasmids. As
discussed previously, this finding still requires extensive validation. However, if we suppose that this is a
valid result then the hypothesis proposed in our introduction is false. Our initial prediction was that SNP
R325 would reduce insulin secretion by reducing zinc transport into insulin vesicles, compared to SNP
28
R325W. How might increased zinc in insulin granules lead to !-cell defects? If more zinc is present in the
insulin granules within !-cells, then larger amounts of zinc would be secreted in conjunction with insulin.
Several studies have reported both paracrine and autocrine signaling by zinc. Kim et al. (2000) reported
that zinc may act as a paracrine effector in inducing islet cell death. In addition, zinc has been reported to
inhibit both glucagon secretion in "-cells (Franklin et al., 2005) and insulin secretion in !-cells (Bloc et al.,
2000) by activating K
ATP
channels. Exocytosis in these cells is normally stimulated by the inactivation of
K
ATP
channels. Thus, it is possible that excessive zinc content within insulin vesicles could have
downstream effects that could lead to improper regulation and islet cell dysfunction. In addition to playing
a direct role in Zn
2+
transport, we cannot rule out the possibility that the SNP variant changes other
biological aspects of the ZnT8 transporter. For example, the SNP may also alter the stability (lifetime) of
the protein or its appropriate targeting.
In conclusion, we have demonstrated that flipped ZnT8 is a novel and feasible approach for investigating
zinc transport and has the potential to directly analyze ZnT8-mediated zinc fluxes for the first time. Given
that the heterologous expression of plasma membrane proteins is typically inefficient, we were encouraged
by our results. While the findings from our initial studies are highly intriguing, we also realize they must
be interpreted cautiously. As stated earlier, much work is needed to confirm the presence of our flipped
ZnT8 in the plasma membrane, as well as to confirm it functions properly. However, if those key issues
are addressed, then this will be a very promising method capable of clarifying the biological pathways
leading to T2D, and in particular, revealing the connection between the ZnT8 R325 gene variant and !-cell
dysfunction.
29
REFERENCES
Bloc A, Cens T, Cruz H, Dunant Y. Zinc-induced changes in ionic currents of clonal rat pancreatic -cells:
activation of ATP-sensitive K+ channels. J Physiol. 2000 Dec 15;529 Pt 3:723-34.
Boesgaard TW et al. The common SLC30A8 Arg325Trp variant is associated with reduced first-phase
insulin release in 846 non-diabetic offspring of type 2 diabetes patients--the EUGENE2 study.
Diabetologia. 2008 May;51(5):816-20.
Chambard JM, Ashmore JF. Sugar transport by mammalian members of the SLC26 superfamily of anion-
bicarbonate exchangers. J Physiol. 2003 Aug 1;550(Pt 3):667-77.
Chimienti F, Devergnas S, Favier A, Seve M. Identification and cloning of a beta-cell-specific zinc
transporter, ZnT-8, localized into insulin secretory granules. Diabetes. 2004 Sep;53(9):2330-7.
Chimienti F, Devergnas S, Pattou F, Schuit F, Garcia-Cuenca R, Vandewalle B, Kerr-Conte J, Van
Lommel L, Grunwald D, Favier A, Seve M. In vivo expression and functional characterization of the zinc
transporter ZnT8 in glucose-induced insulin secretion. J Cell Sci. 2006 Oct 15;119(Pt 20):4199-206.
Clifford KS, MacDonald MJ. Survey of mRNAs encoding zinc transporters and other metal complexing
proteins in pancreatic islets of rats from birth to adulthood: similar patterns in the Sprague-Dawley and
Wistar BB strains. Diabetes Res Clin Pract. 2000 Aug;49(2-3):77-85.
Cousins RJ, Liuzzi JP, Lichten LA. Mammalian zinc transport, trafficking, and signals. J Biol Chem. 2006
Aug 25;281(34):24085-9.
Dodson G, Steiner D. The role of assembly in insulin's biosynthesis. Curr Opin Struct Biol. 1998
Apr;8(2):189-94. Review.
Dunn MF. Zinc-ligand interactions modulate assembly and stability of the insulin hexamer -- a review.
Biometals. 2005 Aug;18(4):295-303. Review.
Emdin SO, Dodson GG, Cutfield JM, Cutfield SM. Role of zinc in insulin biosynthesis. Some possible
zinc-insulin interactions in the pancreatic B-cell. Diabetologia. 1980 Sep;19(3):174-82.
Franklin I, Gromada J, Gjinovci A, Theander S, Wollheim CB. Beta-cell secretory products activate alpha-
cell ATP-dependent potassium channels to inhibit glucagon release. Diabetes. 2005 Jun;54(6):1808-15.
Frayling TM. Genome-wide association studies provide new insights into type 2 diabetes aetiology. Nat
Rev Genet. 2007 Sep;8(9):657-62. Review.
Fu Y, Tian W, Pratt EB, Dirling LB, Shyng SL, Meshul CK, Cohen DM. Down-regulation of ZnT8
expression in INS-1 rat pancreatic beta cells reduces insulin content and glucose-inducible insulin
secretion. PLoS ONE. 2009 May 25;4(5):e5679.
Gaither LA, Eide DJ. Eukaryotic zinc transporters and their regulation. Biometals. 2001 Sep-Dec;14(3-
4):251-70. Review.
Grynkiewicz G, Poenie M, Tsien RY. A new generation of Ca2+ indicators with greatly improved
fluorescence properties. J Biol Chem. 1985 Mar 25;260(6):3440-50.
Gyulkhandanyan AV, Lu H, Lee SC, Bhattacharjee A, Wijesekara N, Fox JE, MacDonald PE, Chimienti F,
Dai FF, Wheeler MB. Investigation of transport mechanisms and regulation of intracellular Zn2+ in
pancreatic alpha-cells. J Biol Chem. 2008 Apr 11;283(15):10184-97. Epub 2008 Feb 4.
30
Hu C, Ahmed M, Melia TJ, Söllner TH, Mayer T, Rothman JE. Fusion of cells by flipped SNAREs.
Science. 2003 Jun 13;300(5626):1745-9.
Inoue K, Matsuda K, Itoh M, Kawaguchi H, Tomoike H, Aoyagi T, Nagai R, Hori M, Nakamura Y, Tanaka
T. Osteopenia and male-specific sudden cardiac death in mice lacking a zinc transporter gene, Znt5. Hum
Mol Genet. 2002 Jul 15;11(15):1775-84.
Kambe T, Narita H, Yamaguchi-Iwai Y, Hirose J, Amano T, Sugiura N, Sasaki R, Mori K, Iwanaga T,
Nagao M. Cloning and characterization of a novel mammalian zinc transporter, zinc transporter 5,
abundantly expressed in pancreatic beta cells. J Biol Chem. 2002 May 24;277(21):19049-55.
Kambe T, Yamaguchi-Iwai Y, Sasaki R, Nagao M. Overview of mammalian zinc transporters. Cell Mol
Life Sci. 2004 Jan;61(1):49-68. Review.
Kim BJ, Kim YH, Kim S, Kim JW, Koh JY, Oh SH, Lee MK, Kim KW, Lee MS. Zinc as a paracrine
effector in pancreatic islet cell death. Zinc as a paracrine effector in pancreatic islet cell death. Diabetes.
2000 Mar;49(3):367-72.
Kirchhoff K, Machicao F, Haupt A, Schäfer SA, Tschritter O, Staiger H, Stefan N, Häring HU, Fritsche A.
Polymorphisms in the TCF7L2, CDKAL1 and SLC30A8 genes are associated with impaired proinsulin
conversion. Diabetologia. 2008 Apr;51(4):597-601.
Léger J, Kempf M, Lee G, Brandt R. Conversion of serine to aspartate imitates phosphorylation-induced
changes in the structure and function of microtubule-associated protein tau. J Biol Chem. 1997 Mar
28;272(13):8441-6.
Michael DJ, Ritzel RA, Haataja L, Chow RH. Pancreatic beta-cells secrete insulin in fast- and slow-release
forms. Diabetes. 2006 Mar;55(3):600-7.
Michael J, Carroll R, Swift HH, Steiner DF. Studies on the molecular organization of rat insulin secretory
granules. J Biol Chem. 1987 Dec 5;262(34):16531-5.
Murgia C, Devirgiliis C, Mancini E, Donadel G, Zalewski P, Perozzi G. Diabetes-linked zinc transporter
ZnT8 is a homodimeric protein expressed by distinct rodent endocrine cell types in the pancreas and other
glands. Nutr Metab Cardiovasc Dis. 2008 Dec 16.
Muylle FA, Adriaensen D, De Coen W, Timmermans JP, Blust R. Tracing of labile zinc in live fish
hepatocytes using FluoZin-3. Biometals. 2006 Aug;19(4):437-50.
Ohana E, Hoch E, Keisar C, Kambe T, Yifrach O, Hershfinkel M, Sekler I. Identification of the Zn2+
binding site and mode of operation of a mammalian Zn2+ transporter. J Biol Chem. 2009 Apr 14. [Epub
ahead of print]
Pound LD et al. Deletion of the mouse Slc30a8 gene encoding zinc transporter-8 results in impaired insulin
secretion. Biochem J. 2009 May 18. [Epub ahead of print]
Reyes JG. Zinc transport in mammalian cells. Am J Physiol. 1996 Feb;270(2 Pt 1):C401-10. Review.
Saxena R et al. Genome-wide association analysis identifies loci for type 2 diabetes and triglyceride levels.
Science. 2007 Jun 1;316(5829):1331-6.
Scott LJ et al. A genome-wide association study of type 2 diabetes in Finns detects multiple susceptibility
variants. Science. 2007 Jun 1;316(5829):1341-5.
31
Sladek R et al. A genome-wide association study identifies novel risk loci for type 2 diabetes. Nature. 2007
Feb 22;445(7130):881-5.
Song SH, McIntyre SS, Shah H, Veldhuis JD, Hayes PC, Butler PC. Direct measurement of pulsatile
insulin secretion from the portal vein in human subjects. J Clin Endocrinol Metab. 2000 Dec;85(12):4491-
9.
Srivastava S, Goren HJ. Insulin constitutively secreted by beta-cells is necessary for glucose-stimulated
insulin secretion. Diabetes. 2003 Aug;52(8):2049-56.
Staiger H, Machicao F, Stefan N, Tschritter O, Thamer C, Kantartzis K, Schäfer SA, Kirchhoff K, Fritsche
A, Häring HU. Polymorphisms within novel risk loci for type 2 diabetes determine beta-cell function.
PLoS ONE. 2007 Sep 5;2(9):e832.
Steinthorsdottir V et al. A variant in CDKAL1 influences insulin response and risk of type 2 diabetes. Nat
Genet. 2007 Jun;39(6):770-5.
Wenzlau JM, Juhl K, Yu L, Moua O, Sarkar SA, Gottlieb P, Rewers M, Eisenbarth GS, Jensen J, Davidson
HW, Hutton JC. The cation efflux transporter ZnT8 (Slc30A8) is a major autoantigen in human type 1
diabetes. Proc Natl Acad Sci U S A. 2007 Oct 23;104(43):17040-5.
Zeggini E et al. Replication of genome-wide association signals in UK samples reveals risk loci for type 2
diabetes. Science. 2007 Jun 1;316(5829):1336-41. Erratum in: Science. 2007 Aug 24;317(5841):1035-6.
Abstract (if available)
Abstract
Zinc is an essential trace element in living organisms, and zinc content in pancreatic β-cells is amongst the highest in the body. Zinc transporter 8 (ZnT8) is localized to the insulin vesicle membrane and is responsible for transporting zinc from the cytoplasm into the vesicles. Inside the vesicles, zinc is hypothesized to crystallize with insulin in a physiologically important process. Interestingly, recent genome-wide association studies have identified a non-synonymous single nucleotide polymorphism (SNP) in the gene encoding ZnT8 that is highly associated with the development of type 2 diabetes. However, to date, no studies have been capable of analyzing the effects of the SNP on ZnT8 function. We have designed a flipped ZnT8 that can be used to monitor cytoplasmic zinc influx using the fluorescent zinc indicator, Fluozin-3-AM. This appears to be a promising method for directly analyzing the functional consequences of the ZnT8 T2D-associated SNP.
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
A novel construct to study the pulsatility of insulin secretion in single cells, islets and whole pancreas
Asset Metadata
Creator
Tomooka, Beren H.
(author)
Core Title
Flipped zinc transporter 8 (ZnT8): a novel approach to characterize zinc transport and its possible relevance to type 2 diabetes
School
Keck School of Medicine
Degree
Master of Science
Degree Program
Physiology
Publication Date
08/09/2009
Defense Date
06/18/2009
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
Diabetes,fluozin,insulin,OAI-PMH Harvest,slc30a8,SNP,zinc transport,zinc transporter 8
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Chow, Robert H. (
committee chair
), Kaslow, Harvey R. (
committee member
), Peti-Peterdi, Janos (
committee member
), Watanabe, Richard M. (
committee member
)
Creator Email
beren.tomooka@stanfordalumni.org,btomooka@gmail.com
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-m2557
Unique identifier
UC1463829
Identifier
etd-Tomooka-3085 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-252558 (legacy record id),usctheses-m2557 (legacy record id)
Legacy Identifier
etd-Tomooka-3085.pdf
Dmrecord
252558
Document Type
Thesis
Rights
Tomooka, Beren H.
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Repository Name
Libraries, University of Southern California
Repository Location
Los Angeles, California
Repository Email
cisadmin@lib.usc.edu
Tags
fluozin
insulin
slc30a8
SNP
zinc transport
zinc transporter 8