(176g) Metabolomic and Proteomic Analysis of Metabolic Enzyme Disruption in Adipocytes

Sims, J. K., Tufts University
Lee, K., Tufts University
Jayaraman, A., Texas A&M University
Choi, K., Texas A&M University
Manteiga, S., Tufts University

and Proteomic Analysis of Metabolic Enzyme Disruption in Adipocytes

Obesity results from a chronic imbalance in caloric intake and expenditure.
The excess calories are stored mainly as intracellular lipids (triglycerides,
TG), leading to an expansion of body fat or adipose tissue through increases in
fat cell (adipocyte, AD) size and number. One approach to controlling body fat
could be to intervene in the metabolic processes of the adipose tissue that
directly contribute to lipid accumulation and cell growth. This approach
requires a systematic investigation of adipose tissue-specific metabolism,
preferably in isolation from confounding systemic influences. The goal of the
present study is to investigate the effect of targeted knockdown of lipid
pathway enzymes through metabolomic and proteomic analyses.

and Methods:
3T3-L1 preadipocytes (PAs) were in grown in 12-well plates and
induced to differentiate into ADs using a standard hormone cocktail.
Experiments were conducted in both 2D planar culture and 3D collagen culture.
For 3D culture, the induced cells were detached, concentrated, and mixed into
an ice-cold collagen pre-polymer solution with (coculture) or without ECs
(monoculture). The collagen cell suspensions were added into 12-well plates and
allowed to gel. The cells were exposed to siRNA by lipofection. Fresh siRNA was
added every 48 h. Gene expression, TG content, and droplet distribution were
quantified using qRT-PCR, enzymatic assay, and image analysis, respectively. Intracellular
and extracellular metabolite concentrations were measured using in-house
developed LC-MS/MS protocols, and changes in expression levels of metabolic
enzymes were analyzed using isobaric tags (iTRAQ).

Results and Discussion: Metabolic
proteins targeted for knockdown were selected from different stages of lipid
accumulation: upstream synthesis (breakdown and transformation of glycolysis
intermediates into precursors of fatty acid synthesis), downstream synthesis
(formation of triglyceride from synthesized precursors), and droplet stability
(regulation of intracellular lipases and formation of larger, more stable lipid
droplets). Based on our prior work characterizing the metabolic flux
distribution in differentiating ADs, pyruvate carboxylase (PCX) and isocitrate
dehydrogenase (IDH) were identified as targets for upstream synthesis,
diglyceride acyltransferase (DGAT) for downstream synthesis, and fat-specific
protein 27 (FSP) and perilipin (PLIN1) for droplet stability. We first tested
the efficiency of knockdown in 2D planar culture. Each of the targets we tested
led to significant reduction in triglyceride accumulation (Fig. 1). To
understand the efficacy of the lipofection in 3D culture, we first characterized
the effect of PLIN1 knockdown, as PLIN1 expression is specific to ADs and
correlates with lipid droplet stability. As expected, TG accumulation was
reduced in both monoculture and coculture (Fig. 2A). Results of image analysis
showed significant reduction in average lipid droplet size in both 3D mono- and
coculture (Fig. 2B). The decrease in the average droplet size reflected a shift
in the distribution of droplet sizes (Fig. 2C). The siRNA treatment did
not affect the variance in the lipid droplet sizes (see error bars),
indicating knockdown was achieved. Interestingly, PLIN1 knockdown did not
significantly affect the TG content (not shown). Using metabolite uptake and
release data in conjunction with a stoichiometric model of AD metabolism, we calculated
the impact of the knockdowns on major metabolic pathway fluxes. The flux data
were combined with enzyme concentration data to determine the rate controlling
steps in TG accumulation using metabolic control analysis as a framework. Lastly,
we performed pairwise knockdown of enzymes. We investigated knocking down
enzymes in two different stages of accumulation (e.g. PCX and DGAT) as well as
two enzymes in the same stage (e.g. PCX and IDH). As expected, we found that
combined knockdown led to a greater reduction in triglyceride accumulation.

Conclusions: We report the
effects of metabolic gene knockdowns in a 3D coculture system capturing the
cell-cell interactions found in adipose tissue. We have shown successful
knockdown of several metabolic proteins, and each lead to a reduction in
accumulation of triglycerides; however, inhibition of PCX (an upstream step)
had the greatest effect. Image analysis of lipid droplets provides an efficient
and non-invasive method of monitoring TG accumulation and cellular
hypertrophy.  Further work is warranted to identify additional silencing
targets to limit lipid accumulation and adipose tissue growth, to better
understand the metabolic consequences of knockdown and to continue to improve
our 3D coculture system. Furthermore, results from MFA and MCA will be used to
focus future experiments on key regulatory steps and identify which proteins
could be used as potential therapeutic targets.

KD 1

Fig 1 Identification of knockdown
targets in 2D planar monoculture.

KD 2

Fig 2 (A) Triglyceride accumulation
in 3D mono- and coculture treated with non-targeting (control) or
anti-perilipin siRNA. Error bars are SEM. (B) Average lipid droplet size in 3D
mono- and coculture treated with non-targeting (control) or anti-perilipin
siRNA. Error bars are SEM. *Significantly different from respective control (n~1,000
droplets from 8 different images), p=0.0005 and p=0.03, respectively. (C)
Histogram of droplet size in 3D monoculture.