Translocation of GPAT4 from the endoplasmic reticulum to the mitochondria may alter
efficiency of TAG synthesis in liver cells
By: Audra Goldstein
Dr. Amanda Suchanek
Dr. Rosalind Coleman
Senior Honors Thesis
Biology
The University of North Carolina at Chapel Hill
March 21, 2017
Approval: Approved
Dr. Rosalind Coleman, PI
Dr. Amy Maddox, Reader
Dr. Bob Bourret , Reader
Goldstein 1
Abstract:
Synthesis of triacylglycerol (TAG) is critical for the formation of lipoproteins and lipid
droplets and is the main constituent of mammalian fat. The rate of TAG synthesis is controlled
by the enzyme glycerol-3-phosphate acyltransferase (GPAT). GPAT4, an isoform of GPAT, is
found predominantly in the endoplasmic reticulum (ER), while GPAT1 is found in the
mitochondria. Although GPAT4 and GPAT1 both produce lysophosphatidic acid (LPA), it is not
known whether the output of phospholipids whose synthesis was initiated in the outermitochondrial membrane (OMM) differ from those that are initiated in the ER. It was
hypothesized that GPAT4 and GPAT1 have non-interchangeable outputs within the cell and that
the introduction of GPAT4 to the mitochondria would not affect TAG output. To determine if
there was a difference between the outputs of GPAT 1 and GPAT4, primary liver cells were
infected with adenovirus to express GPAT4 on the mitochondria (Ad-Tom20-GPAT4,
“mtGPAT4-FLAG”). We measured GPAT enzyme activity in total particulate prepared from
primary mouse hepatocytes and total particulate prepared from a mouse hepatoma cell line, Hepa
1-6, infected with these adenoviruses. The introduction of mtGPAT4-FLAG had an inconclusive
change in activity in primary mouse hepatocytes while it had a doubling effect on enzyme
specific activity in Hepa 1-6 cells as compared to the GFP control adenovirus (Ad-T20-GFPBirA, “GFP control adenovirus”). These results indicate conflicting information about the
introduction of mtGPAT4-FLAG. Future experiments within Hepa 1-6 cells or possibly whole
animal models should be considered for further analysis on the effect of incorporating
mtGPAT4-FLAG. Understanding what affects the efficiency of TAG synthesis is important for
providing a better understanding of diseases that are influenced by TAG production and
maintenance, such as diabetes mellitus, lipodystrophy, and heart disease.
Goldstein 2
Introduction:
Within the glycerol phosphate pathway for triacylglycerol (TAG) synthesis, it is not
known whether the output of phospholipids whose synthesis is initiated in the outermitochondrial membrane (OMM) differs from those that are initiated in the endoplasmic
reticulum (ER). GPAT1 is found in the OMM whereas GPAT4 is found in the ER membrane. It
is hypothesized that GPAT4 and GPAT1 have sufficiently different outputs that cannot be used
interchangeably within the cell. To test this hypothesis, we infected Hepa1-6 cells and primary
mouse hepatocytes with adenovirus to express a FLAG-tagged GPAT4 on the mitochondria
(mtGPAT4-FLAG), in addition to endogenous ER GPAT4. This was done because the presence
of mitochondrial GPAT4 conveys important data on the environment of the cell, the possibility
of mitochondria associated membrane (MAM) interaction, and the TAG synthesis pathway.
Triacylglycerol Synthesis
Critical for the formation of lipoproteins and lipid droplets, GPAT is the rate-limiting
enzyme in the synthesis of TAG in the glycerol phosphate pathway. GPAT begins the process of
TAG synthesis by acylating glycerol-3-phosphate (G3P). This acylation occurs on the ER and on
the OMM, depending on the active protein isoforms within the cell. In mammalian cells, there
are four isoforms of GPAT, 1-4 [1]. GPAT1 and GPAT2 are present on the OMM whereas
GPAT3 and GPAT4 are present on the ER. GPAT4 is a 54-kDa transmembrane protein and
expression is highest in brown adipose tissue and testis, and moderate in the liver, kidney, brain,
and adipose tissue [1]. GPAT1 is a 94-kDa transmembrane protein that esterifies newly formed
acyl-CoAs, which diminishes their availability to bind to carnitine palmitoyltransferase-I (CPT1) and prevents the conversion of acyl-CoA to acyl-carnitines. If these acyl-carnitines were
formed, they would be imported into the mitochondria for beta-oxidation [1,2]. GPAT1
Goldstein 3
expression levels are highest in adipose and liver tissue and can account for 30-50% of total
activity in liver cells, indicating a large impact in TAG synthesis [1].
A key way to differentiate among the GPAT isoforms is that GPATs 2-4 are sensitive to
chemical inactivation by N-ethylmaleimide (NEM) whereas GPAT1 is NEM resistant. NEM is
an alkene derived from maleic acid and is reactive towards thiols. For GPATs 2-4, NEM
modifies the active site cysteine, rendering it non-functional. NEM is unable to form a disulfide
bond with GPAT1 because GPAT1 does not have cysteine in the G3P binding site [1].
Acylation of G3P yields LPA, which is a substrate for acylation by the 1-acylglycerol-3phosphate acyltransferase (AGPAT) family of ER-localized enzymes, producing phosphatidic
acid (PA). PA can be converted into diacylglycerol (DAG) by lipin enzymes in the ER, which is
ultimately converted to TAG by diacylglycerol acyltransferase (DGAT). Within liver cells,
GPAT1 and GPAT4 are responsible for the majority of TAG synthesis [1].
Mitochondria-Associated Membrane
The mitochondria associated membrane (MAM) is the outer-mitochondrial membrane
that is directly in contact with the ER membrane [3]. The MAM is enriched with proteins
involved with lipid metabolism enzymes, calcium-handling proteins, and is involved with TAG
synthesis [3]. Recent proteomic analyses of the MAM have revealed several proteins involved in
TAG synthesis; specifically proteins in the glycerol phosphate pathway, DGAT2 and AGPAT4
and AGPAT5 [2,3].
To test this possible MAM interaction within the rate limiting step of TAG synthesis,
Hepa 1-6 cells were infected with two adenoviruses (Ad-Tom20-GPAT4 “mtGPAT4-FLAG”
and Ad-T20-GFP-BirA, “GFP control adenovirus”) to determine if there was a significant
difference in the GPAT activity when mitochondrial GPAT4 was introduced to the cell line. The
Goldstein 4
mtGPAT4-FLAG adenovirus incorporated DNA that directed GPAT4 enzymes to the OMM and
that were tagged with FLAG. The GFP control adenovirus was used to direct GFP to the OMM.
In addition, the adenovirus infections were performed using primary mouse hepatocytes to gain
insight in a more physiologically relevant model system.
Methods:
Cell Culture and Infection
A mouse hepatoma derived cell-line, Hepa 1-6, and primary mouse hepatocytes were
used for adenovirus infection. Cells were maintained in a sterile environment and grown at 37° C
and 5% CO2. The Hepa 1-6 cells were allowed to grow to 80% confluence in complete medium
[10% FBS and DMEM (1X) solution [4.5g/L D-glucose, L-glutamine, 110 mg/L sodium
pyruvate] in 6-well plates. The following day, the Ad-T20-GFP-BirA and the Ad-Tom20GPAT4-FLAG adenoviruses were added to appropriate plates. The multiplicities of infection
(MOIs) tested were 0, 50, 100, 200, and 500 PFU/mL for the Hepa 1-6 cells. In the case of
primary mouse hepatocytes, which do not grow in culture, the cells were plated at 80% capacity
on a 6-well plate (~2.5x106 cells). The MOIs tested for the primary hepatocytes were 0, 10, 30,
and 50 PFU/mL. Cells were incubated with adenovirus for 24 hours, at which point adenovirus
was removed, and cells were incubated in complete medium for an additional 24 hours. 48 hours
after infection, cells were washed twice with 1 mL cold 1x PBS. The plates were placed on ice
and the cells were scraped and transferred into a glass homogenizing vessel. The cells were
homogenized using a Teflon-glass motor driven homogenizer. Then the mixture was transferred
into ultracentrifuge tubes. The tubes were centrifuged at 100,000xg for 1 hour at 4° C. The
supernatant was aspirated and the total particulate fraction was re-homogenized in 1 mL Medium
Goldstein 5
1 [250 mM sucrose, 10 mM tris pH 7.5, 1 mM EDTA]. The total particulate was aliquotted and
stored at -80° C for future analyses.
Western Blot
Proteins within the Hepa 1-6 liver total particulate and mouse primary hepatocyte liver
total particulate were separated by size using a 10% acrylamide SDS-PAGE gel and transferred
to PVDF membranes. Gels were run in 1x SDS-PAGE buffer. Equal amounts of protein (20 μg)
were loaded in each well in buffer containing 2-mercaptoethanol (0.02%) after boiling for 5
minutes. Each sample was then loaded in to the gel and 10 μL of Precision Plus Protein Dual
Color Standards ladder was loaded. The gel was run at 100 V for 1.5 hours.
Proteins were transferred to PVDF membranes in 10mM CAPS buffer (pH 11) for 40
minutes at 90V. Membranes were blocked in 5% milk/TBS/0.1%Tween for 1 hour, and then
washed in 1x TBS/Tween 5 times for 5 minutes each. Mouse monoclonal Anti-Flag primary
antibody was diluted 1:1000 and added to 10 mL of 1xTBS/Tween with 5% BSA. The
membrane was probed overnight at 4°C with gentle agitation. The membrane was then washed 3
times for 5 minutes each with TBS/Tween. Secondary antibody (goat anti-mouse-IgG-HRP) was
diluted 1:5000 in 10 mL 1xTBS/Tween and 5% milk. The membrane was probed at room
temperature for 1.5 hours with gentle agitation. The membrane was then developed using the
Pierce ECL chemiluminescent substrate (ThermoFisher Scientific, #32132) and imaged using the
Versadoc Molecular Imager.
GPAT Assay: Enzyme Activity and Enzyme Kinetics
GPAT activity was measured within total membrane fractions by radiolabeling the output
of PA. For each sample, water and GPAT mix [76 mM Tris pH 7.5, 160 mM MgCl2, 1 mg/mL
BSA, 8 mM NaF, 1mM DTT, 82.5 uM palmitoyl-coA, and 800 uM [3H]glycerol-3-phosphate]
Goldstein 6
were added to each tube. Each total membrane fraction of MOI 0, 50, 100, 200, and 500 for the
Hepa 1-6 and MOI 0, 10, 30 for the primary hepatocytes were thawed and diluted to 1 μg/ μL in
Medium 1 on ice. For NEM+, samples were incubated with 2mM NEM for 15 minutes on ice.
The start times were staggered to correspond with the start of each GPAT assay. After the
protein-NEM mixes incubated for 15 minutes or if the samples were NEM-, 0, 10 μg, 15 μg or
20 μg of protein was added to the reaction mix at staggered intervals. Immediately after the
addition of the protein, the samples were gently vortexed. Once the reaction was incubated for 10
minutes, the reaction was stopped by adding 0.6 mL of 1% perchloric acid and 3 mL of
chloroform: methanol (1:2) in that order and vortexed gently. The tubes were allowed to sit again
for 5 minutes. The phases were then broken by adding 1 mL of 1% perchloric acid and 1 mL of
chloroform. The solution was capped and vortexed vigorously, then centrifuged for 5 min at
3000 rpm at 4°C. The upper phase was aspirated into a radioactive waste flask in the fume hood
and the solution was washed in 2 mL of 1% perchloric acid. The tube was capped and vortexed
vigorously and then centrifuged for 5 minutes at 3000 rpm at 4°C. This process was repeated for
the second wash. After the second wash, the upper phase was aspirated into the radioactive waste
in the fume hood and exactly 1 mL of the lower phase solution was transferred in to a
scintillation vial and allowed to dry overnight. After the samples were dried, 4 mL of Ecolite (+)
scintillation cocktail were added, the vials were capped and vortexed, and the samples were read
for tritium in the Wallac L5C 1409 liquid scintillation counter.
Results
In order to help determine possible MAM interaction and how to impact TAG efficiency,
Hepa 1-6 cells were initially infected with both types of adenoviruses. This was done to
determine if these adenoviruses could properly infect the cells. Once proper expression of
Goldstein 7
GPAT4 or GFP was determined, primary mouse hepatocytes were used to represent a more
physiologically relevant model. The enzyme specific activity of GPAT was analyzed and
compared in both infected cells types, in order to see how mtGPAT4 plays a role in Hepa 1-6
cells and primary mouse hepatocytes.
After infecting of the Hepa 1-6 cells with adenovirus, the cells were visually inspected to
check for overall health and to see if there was an increase in lipid droplet formation. At all
MOI’s, the cells seemed healthy and, within the Ad-Tom20-GPAT4-FLAG infected cells, there
seemed to be a greater number of lipid droplets. The Western blot indicated that infection was
successful because the samples containing the FLAG-tagged GPAT4 (lanes 6-8) showed a
gradient dose dependent increase in expression with increased MOI. No expression of FLAG
was seen in GFP control cells (lanes 1-5). In the lower panel, β-actin expression was equal across
all samples, indicating equal loading.
Figure 1: Western blot of Hepa 1-6 cells of Ad-Tom20-GPAT4-FLAG infected cells and AdT20-GFP-BirA infected cells
The GPAT enzyme activity assay indicates a higher GPAT specific activity with the AdTom20-GPAT4 as compared to the Ad-T20-GFP-BirA infected cells. The highest activity was
Goldstein 8
recorded in the NEM- mtGPAT4-FLAG cells, with an MOI of 500. The specific activity was 2.5
nmol/mg/min. As shown in Figure 2, the specific activity of GPAT increased as the MOI
increased within the Hepa 1-6 cells. NEM treatment of the total membrane fractions had very
little effect on total GPAT activity. This indicates that the majority of enzyme activity is coming
from NEM sensitive GPAT. As shown in Figure 2 (solid bars), NEM resistant GPAT only makes
up a small portion of total GPAT activity. This observation means that GPAT1 is a minor
contributor to total GPAT activity in Hepa 1-6 cells. Therefore, with the introduction of
mtGPAT4-FLAG into the Hepa 1-6 cells and the steady increase of enzyme specific activity at
each MOI, mtGPAT4-FLAG may be playing a role in enzyme activity and efficiency.
Specific Activity (nmol/mg/min)
3
2.5
2
Total GPAT
1.5
NEM R
NEM S
1
0.5
0
MOI 0
MOI 50
MOI 100
MOI 200
MOI 500
Figure 2: GPAT Activity mtGPAT Infected Hepa 1-6 Cells. Measured total GPAT, NEM
resistant, and NEM sensitive specific activity in Ad-Tom20-GPAT4 infected cells (n = 2,
technical replicates).
In Hepa 1-6 cells infected with Ad-T20-GFP-BirA, initial GPAT activity is high, but
significantly decreases after an MOI of 200. Pre-treatment with NEM had very little effect on
total GPAT activity (Figure 3, hatched bars). As with the mtGPAT4-FLAG infected cells, this
indicates that GPAT1 is only contributing a small portion of total GPAT activity within the Hepa
Goldstein 9
1-6 infected cells. NEM sensitive GPAT activity decreased in a dose dependent manner as MOIs
increased (Figure 3, dotted bars). NEM resistant GPAT activity was unaffected by the increased
MOIs with this adenovirus (Figure 3, solid bars). Overall, total GPAT activity in mtGPAT4FLAG infected cells was higher across all MOIs tested as compared to GFP control adenovirus
(Figures 2 and 3). This indicates that mtGPAT4-FLAG is functional and contributing to the
GPAT activity within these cells.
Specifiic Activity (nmol/mg/min)
0.8
0.7
0.6
0.5
Total GPAT
0.4
NEM R
NEM S
0.3
0.2
0.1
0
MOI 0
MOI 50
MOI 100
MOI 200
MOI 500
Figure 3: GPAT Activity GFP Infected Hepa 1-6 Cells. Measured total GPAT, NEM resistant
and NEM sensitive specific activity in Ad-T20-GFP-BirA infected cells (n = 2, technical
replicates).
To gain insight of the role of mtGPAT4-FLAG in more physiologically relevant model,
we repeated the viral infections in freshly isolated primary mouse hepatocytes. The MOIs of 10,
30, and 50 were used because there was a very small difference between the enzyme specific
activity between MOI 50 and MOI 200 with the Hepa 1-6 when they were infected with the
mtGPAT4-FLAG adenovirus, so higher MOIs were not deemed necessary. The Western blot
analyses indicated that infection was inefficient and, for the most part unsuccessful (Figure 4).
Within the mtGPAT4-FLAG infected cells, only a faint GPAT4 band is present at all MOIs, as
Goldstein 10
shown in Figure 4, upper left panel. No endogenous GPAT4 expression (Figure 4, upper right
panel) and faint expression of GFP (Figure 4, middle right panel) were observed in the GFP
control adenovirus infected cells. This again shows that infection was inefficient and not
particularly successful. A strong presence of β-actin was consistent across all MOIs and
adenovirus infection types. This indicates that protein loading was the same across all samples
and that cellular protein is present.
Figure 4: Western blot of primary hepatocytes of Ad-Tom20-GPAT4-FLAG (left) infected cells
and Ad-T20-GFP-BirA (right) infected cells
The GPAT activity of the Ad-T20-GFP-BirA and Ad-Tom20-GPAT4 infected primary
hepatocyte cells are shown in Figure 5, where specific activity was very low in all samples.
There was no clear relationship with cells infected with the mtGPAT4-FLAG adenovirus as
compared to the GFP control. This was not altogether surprising when considering the Western
blot data (Figure 4). As the MOI for the GFP control adenovirus increased, the specific activity
of total GPAT, NEM resistant GPAT (GPAT1), and NEM sensitive (GPAT4) all decreased.
NEM treatment of the total membrane fractions had a significant impact on GPAT activity where
GPAT1 contributed close to one-third to half of total specific activity with the primary
hepatocytes membranes. This is in contrast to the results from the Hepa 1-6 cells (Figure 2 and
Goldstein 11
Figure 3), which showed a minimal contribution of GPAT1 to total activity. Therefore, within
primary mouse hepatocytes, GPAT1 contributes a greater amount of activity than in Hepa 1-6
cells.
Specific Activity (nmol/mg/min)
1.2
1
0.8
Total GPAT
0.6
NEM R
NEM S
0.4
0.2
0
novirus mtGPAT mtGPAT mtGPAT GFP MOI GFP MOI GFP MOI
MOI 10 MOI 30 MOI 50
10
30
50
Figure 5: GPAT Activity in mtGPAT and GFP Primary Mouse Hepatocytes. Measured total
GPAT, NEM resistant and NEM sensitive specific activity in Ad-Tom20-GPAT4 (mtGPAT4FLAG) and Ad-T20-GFP-BirA (GFP) primary mouse hepatocytes (n=1).
Lastly, the enzyme kinetics of the infected primary mouse hepatocytes were examined in
order to better understand how the efficiency and specificity of the GPAT4 enzyme under both
viral conditions. A MOI of 30 was chosen for both Hepa 1-6 and primary mouse hepatocytes
because there was similar activity between the GFP control adenovirus infected cells and the
mtGPAT4-FLAG adenovirus infected cells (compare Figure 2,3, and 5). The relationship
between specific activity and G3P of both types of infected cells is shown in Figure 6.
Both types of cells displayed a similar enzyme specific activity when considering their
dependence to G3P, which is shown by similar slopes between the two types of cells. This, once
again, is not all that surprising considering that the infection using both types of adenoviruses
Goldstein 12
were unsuccessful in hepatocytes (Figure 4) and that the specific activity between the two types
of primary mouse hepatocytes were similar (Figure 5, hatched bars). Interestingly, the G3P
dependence graph (Figure 6) was a linear graph, indicating that saturation of the enzyme with the
substrate did not take place.
Specific Activity (nmol/mg/min)
2.5
2
1.5
GFP MOI 30
1
GPAT MOI 30
0.5
0
0
100
200
300
[G3P] uM
400
500
Figure 6: Glycerol-3-Phosphate Dependence of Infected Primary Hepatocytes. Adenovirus
infected primary hepatocytes dependence on glycerol-3-phosphate. Both types of cells displayed
a similar, linear relationship of activity with increasing concentration so [3H]-glycerol-3phosphate.
The data from the graph in Figure 6 was then converted in to a Lineweaver-Burk graph of
the infected primary hepatocytes (Figure 7), in order to determine the Km and Vmax of the GPAT
enzymes. The results are shown in Table 1. Although there was a similarity in specific activity
between the two types of infected cells, there was a significant difference between how these
infected cells reacted with the substrate of G3P. As compared to the GFP control infected cells,
the mtGPAT4-FLAG cells had a greatly reduced Vmax and Km. When infected with mtGPAT4FLAG, the maximal velocity of GPAT was greatly reduced while the affinity for the substrate
Goldstein 13
was greatly increased. This may indicate that the incorporation of mtGPAT4-FLAG is affecting
the GPAT efficiency, even though it is present in very small amounts.
60
50
1/Specific Activity
40
30
GPAT MOI 30
GFP MOI 30
20
10
0
0
-10
0.02
0.04
0.06
0.08
0.1
0.12
0.14
1/[G3P]
Figure 7: Lineweaver-Burk of Infected Primary Hepatocytes. Enzyme kinetics of GPAT in
the mtGPAT4-FLAG and GFP infected primary mouse hepatocytes.
Table 1: The kinetics of the mtGPAT4 infected and GFP infected primary mouse hepatocytes.
Vmax (μM/sec)
Km (μM)
mtGPAT4-FLAG
0.33
33
MOI 30
GFP control MOI
1.9
750
30
Discussion
Overall, with the infection of Ad-Tom20-GPAT4 adenovirus, the Hepa 1-6 cells showed
nearly two times greater specific activity than the Ad-T20-GFP-BirA infected Hepa 1-6 cells. In
determining the outcomes, several assumptions were made. First that LPA output from the two
isoforms cannot be transferred across the MAM and second, that the cell is in a glucose saturated
state, so competition for substrate in the form of acyl-CoAs is solely proximity dependent, rather
Goldstein 14
than concentration dependent. With the first assumption, this implies that the mtGPAT4-FLAG
output of LPA cannot be transferred to the ER because it is somehow inherently different than
ER produced LPA. Therefore, it would be expected that there would be little to no increase in
TAG synthesis. In the second assumption, considering that GPAT1 uses the same acyl-CoA
input as CPT-1, which brings acyl-CoA in to the mitochondria for beta-oxidation; there was a
possibility that incorporating mtGPAT4-FLAG may cause a crowding-out effect of GPAT1
protein, causing a shift in acyl-CoA to CPT-1 competition. GPAT1 esterifies acyl-CoA, which
lessens the competition for acyl-CoA to serve as the substrate for CPT-1. With the incorporation
of mtGPAT4-FLAG, which does not esterify acyl-CoA, the competition for input of acyl-CoA
would shift toward the non-esterified input. Therefore, mtGPAT4-FLAG and CPT-1 would be
utilizing acyl-CoA before GPAT1 could esterify it. Considering these possibilities, saturating the
cell solves this problem, so that input is in abundance for full activity of all membrane proteins
that utilize acyl-CoA and G3P.
A greater amount of enzyme specific activity of NEM-sensitive GPAT indicates more
LPA output. Although my results in Hepa 1-6 cells indicates that mtGPAT4-FLAG is functional
in producing the greater output LPA, it does not indicate if there was a greater amount of TAG
produced within the cell. The cells seemed to contain more lipid droplets, but further analysis is
necessary to determine if there was an actual increase in the mtGPAT4-FLAG infected cells. If
there were an increase in TAG synthesis within the cells, then this would indicate that there may,
be a MAM interaction in the transferring of LPA to the ER for processing. This would disagree
with my hypothesis made earlier.
But when looking at the primary hepatocytes, specific activity was very low and there
was no clear relationship with the incorporation of mtGPAT4-FLAG adenovirus. In fact, the
Goldstein 15
expression of the mtGPAT4-FLAG was shown to be very small. This low expression and low
activity indicates that the incorporation of both adenoviruses had minimal effects on protein
activity. The low activity within the cells was not altogether unexpected. In other experiments,
when primary mouse hepatocytes were put in culture, the overall specific activity was very low
[5]. Furthermore, the primary mouse hepatocytes that were infected were in culture for 48 hours,
rather than in 16 hours like pervious experiments [5]. Since these cells were isolated and infected
for 48 hours, the significant loss of specific activity was expected. But even with low activity,
theoretically, the incorporation of mtGPAT4-FLAG should still show some increase in activity.
Here it did not. In fact, the results show that there is possibly no interaction between the
mitochondria and the ER for the TAG synthesis pathway. Further analysis and reassessment
must take place within an appropriate model, such as a whole animal model, to show any
conclusive results of how the pathway interacts with the mitochondria.
The idea that further analysis is needed is reinforced when looking at the GPAT kinetics.
When mtGPAT4-FLAG is incorporated into the primary hepatocytes, the Km, at 33, turned out to
be lower than in the GFP infected cells, at 750. This is a promising result for future study
because it indicates that when mtGPAT4-FLAG is present, the affinity of the substrate
significantly increases. Primary hepatocytes with mtGPAT4-FLAG only require 33 uM of
substrate in order to achieve half the enzyme’s active sites to be occupied by the substrate. The
GFP control infected cells, on the other hand, required 750 uM of substrate to do the same. But
with this increased affinity in the mtGPAT4-FLAG infected cells, there was a trade-off. The
Vmax of the mtGPAT4-FLAG cells was 5.85 times slower than the GFP infected cells. This
means that the rate of the enzyme-catalyzed reaction has been significantly reduced by the viral
infection. These results are perplexing. The increased affinity to substrate with the incorporation
Goldstein 16
of mtGPAT4-FLAG makes sense, in that more GPAT4 active sites are available, and they are in
a new location, so the probability of the substrate falling into an active site increases. However, a
lower Vmax is surprising because, as explained in the next paragraph, the BirA tag might be
affecting the activity of the cells infected with it, and therefore may be inhibiting overall enzyme
activity. This lowered Vmax may be explained by the difference in location in the ER and the
mitochondria. Since GPAT4 is normally located on the ER, something on the OMM may be
inhibiting the rate that the enzyme can catalyze the reaction.
The GPAT activity of Ad-T20-GFP-BirA infected cells show a significant decrease as
MOIs increased in both the Hepa 1-6 cells and the primary mouse hepatocytes (Figure 3 and
Figure 5). Although the reason for this is unclear, it may be that the structure of BirA at high
MOIs is interfering with GPAT function. This interference could have possibly been by the
structure of BirA interfering with the enzyme or by interference with transcriptional regulation,
but further work needs to be conducted to figure out the true reason why high concentrations of
BirA has negative effects on the enzyme specific activity within both types of cells.
Conclusion
Although this data indicated proper infection of the Hep 1-6 cells and that mtGPAT4FLAG was functional, we still do not know if there was a significant increase in TAG synthesis.
The Western blot (Figure 1) indicated that Hepa 1-6 took up both mtGPAT4-FLAG and GFP
control adenoviruses and that the enzyme specific activity of the Hepa 1-6 cells with mtGPAT4FLAG did increase (Figure 2). But when the model was changed to primary mouse hepatocytes,
infection was inefficient (Figure 4) and enzyme specific activity showed no applicable change
except with GFP control adenovirus with increasing MOI (Figure 5). Lastly, the GPAT enzyme
kinetics for the primary mouse hepatocytes, when mtGPAT4-FLAG was incorporated in the
Goldstein 17
cells, had a lower Vmax and a lower Km as compared to cells with the GFP control adenovirus
(Table 1). This indicates that incorporating mtGPAT4-FLAG into primary mouse hepatocytes
may change the affinity for substrate within the infected cells. Overall, an appropriate model
system is needed to continue this experiment, in order to provide information about possible
MAM interaction within the TAG synthesis pathway. This appropriate model might be in a
whole animal where one could infect a mouse with adenovirus and collect primary mouse
hepatocytes from the infected mouse. This may have a more informative outcome because the
primary mouse hepatocytes would not need to be put in to culture, and therefore any potential
loss in activity due to time in culture may be minimized. This research is significant for
providing a better understanding of diseases that are influenced by TAG production and
maintenance, such as diabetes mellitus, lipodystrophy, and heart disease.
Acknowledgements
A special thank you to my PI, Dr. Rosalind Coleman, for giving me the opportunity to
conduct research within her lab and for her guidance throughout the 2 years I have been a part of
her team. Her mentorship has been priceless in my growth as an undergraduate student. I would
also like to thank Dr. Zengying Wu and Dr. Amanda Suchanek for teaching me the skills
required to perform the experiments I conducted in my research. Their patience and kindness
was greatly appreciated. I would also like to thank the rest of the Coleman Lab for their advice
and encouragement throughout this process.
Goldstein 18
Works Cited
1. Takeuchi, K., and K. Reue. "Biochemistry, physiology, and genetics of GPAT, AGPAT,
and lipin enzymes in triglyceride synthesis." AJPendo: Endocrinology and Metabolism
296.6 (2009). Web.
2. M.R. Gonzalez-Baro, R.A. Coleman, Mitochondrial acyltransferases and
glycerophospholipid metabolism, Biochimica et Biophysica Acta (2016),
http://dx.doi.org/10.1016/j.bbalip.2016.06.023
3. Raturi, A., & Simmen, T. Where the endoplasmic reticulum and the mitochondrion tie the
knot: the mitochondria-associated membrane (MAM). Biochimica et Biophysica Acta,
(2013) 1883, 213–224. doi:10.1016/j.bbamcr.2012.04.013
4. Ohba, Y., Sakuragi, T., & Kage-Nakadai, E. (2013). Mitochondria-type GPAT is required
for mitochondrial fusion. EMBO Journal, 32(9), 1265–1279. Retrieved from
http://emboj.embopress.org/content/32/9/1265.long
5. Wendel, Angela A, Daniel E Cooper, Debora M Muoio, and Rosalind A Coleman. 0920-13. “ Glycerol-3-phosphate Acyltransferase (GPAT)-1, but Not GPAT4, Incorporates
Newly Synthesized Fatty Acids into Triacylglycerol and Diminishes Fatty Acid
Oxidation*.” Thejournalofbiologicalchemistry 288 (38): 27299–27306.
Goldstein 19