Based on the article below:

Title: High fat diet and Endocannabiods.

Question: Please write a summary of the article, with a deeper understanding of the important informations on the high fat diet and endcannabiods. And list the Advantage and disavantage of a high fat diet and endocannabiods.

Article:

1. Introduction

Obesity is a growing public health concern that increases the risk of

inflammatory and metabolic disorders such as type 2 diabetes, fatty

liver, and pulmonary inflammation [1,2]. The incidence of obesity has

drastically increased over the past few decades. In a nationally representative

survey (National Health and Nutrition Examination

Survey, 2014) of adults in the US, the prevalence of obesity was 35%

among men and 40% among women, where linear trends significantly

increased for women between 2005 and 2014 [3]. The prolonged and

excessive inflammation associated with obesity has also been associated

with increases in certain cancers, cardiovascular disease, and Alzheimer's

disease [4,5]. While the mechanisms linking obesity and metabolic

disorders are not fully understood, several studies suggest that

alterations in lipid-mediated metabolism play a significant role [1,2,6].

These studies have led to the hypothesis that change in the blood lipidome

can be exploited to identify lipid markers as prognostic indicators

for obesity and type 2 diabetes [7].

Lipids are a diverse subset of biomolecules that are not only responsible

for energy storage and structural regulation, but also participate in complex signaling networks whose disruption results in

the pathogenesis of obesity and other ailments. A few studies have

identified several lipid and lipoprotein abnormalities among obese

patients [8,9]. For example, Hu, et al. reported decreases in HDL cholesterol

along with altered triglyceride levels in nondiabetic obese patients

[9]. Additionally, the role of dietary fat in obesity and influence

of fatty food intake on inflammatory responses are well-established

[4,10]. Obesity-associated inflammation is not restricted to impaired

lipid metabolism, but is also strongly linked with type 2 diabetes, as

obesity is associated with insulin resistance, which heightens the risk

for metabolic syndrome [11,12].

The higher prevalence of type 2 diabetes among adults supports the

assertion that aging is the precursor to insulin resistance [13–16]. While

insulin resistance, type 2 diabetes, and metabolic syndrome have been

studied within the context of age to some extent, including by us [17],

the effect of age on the lipidome and/or age-obesity interactions have

not received significant attention. However, two more recent studies

demonstrate that age exerts appreciable, lipid species-specific effects on

the brain lipidome [18] and, importantly, that age has profound effects

on the female reproductive system (oocytes) lipidome [19]. These

studies support the hypothesis that age-obesity interactions alter the

lipidome. In comparison to the lack of knowledge on the effect of

obesity and/or age on the plasma lipidome as a whole, it is known that

obesity can alter the levels of select lipid species. For example, increased

circulating endocannabinoids, especially 2-arachidonoylglycerol

(2-AG), have been associated with obesity in both humans and

laboratory animals, i.e [20,21]; however, it is not known how the endocannabinoid

system is altered within the context of age in the face of

high-fat diet consumption. The endocannabinoid (EC) system participates

in the control of lipid and glucose metabolism and dysregulation

of this system can occur following unbalanced energy intake [22]. Such

dysregulation often results in overactivity across various organs involved

in energy homeostasis such as intra-abdominal adipose tissue

[23]. Over-activation of the endocannabinoid system has been shown to

promote insulin resistance [6]. Additionally, the essential role of the EC

system in adipogenesis and lipogenesis has been reviewed in detail by

Silvestri and Marzo, et al. [22,23].

In this study, we further investigated the effects of dietary fat consumption,

age, and their interaction at the level of the lipidome using

shotgun lipidomics with electrospray ionization-mass spectrometry

(ESI-MS). Because of the paucity of data and the linear increase of female

obesity among US women in the most recent decade [3], we assessed

the blood lipid profiles of female C57BL/6 mice following HFDconsumption

for short (6 weeks), long (22 weeks), and prolonged

(36 weeks) periods to evaluate the persisting effects of feeding. To

compare lipid alterations with metabolic and liver regulation, markers

of liver homeostasis were assessed and correlations between them and

indices of glucose tolerance and insulin sensitivity with the blood lipidome

were determined. Circulating and liver levels of the two major

endocannabinoids, 2-arachidonoylglycerol (2-AG) and anandamide

(AEA), were also measured to determine the effects of HFD-consumption

and age on the endocannabinoid system.

2. Materials and methods

2.1. Animals

Experiments were performed with female C57BL/6 mice (Harlan,

Indianapolis, IN). The mice were housed (4–5/cage) and maintained at

22–24 °C with food and water available ad libitum on a 12 h light/dark

cycle in an AAALAC accredited facility throughout the study. All experimental

procedures were in accord with the latest National Institutes

of Health (NIH) guidelines and the study was approved prior to initiation

by the Institutional Animal Care and Use Committee (IACUC) of

the University of Georgia.

2.2. Dietary treatment

The diets and dietary treatment are described in detail in our recent

publication [17], where the body weight changes, metabolic and behavioral

effects of the same experimental paradigm are reported.

Briefly, 6–7 weeks old female mice weighing 16.0 ± 0.20 g

(mean ± SEM) were randomly divided into two groups (n = 8/group/

time point) and placed on either a low-fat diet (LFD; 10% kcal from fat,

D12450J, Research Diets, Inc., New Brunswick, NJ) or a high-fat diet

(HFD; 60% kcal from fat, D12492, Research Diets) for either 6, 22, or

36 weeks. The LFD diet (3.85 kcal/g) consisted of 70% carbohydrate,

20% protein, 10% fat, of which 23.5% were saturated fatty acids [SFA],

29.7% monounsaturated fatty acids [MUFA], and 46.8% polyunsaturated

fatty acids [PUFA]) (Suppl. Table 4). The HFD diet

(5.24 kcal/g) consisted of 20% carbohydrate, 20% protein, 60% fat, of

which 32.2% were SFA, 35.9% MUFA, 31.9% PUFA (Suppl. Table 4).

2.3. Blood, plasma, and liver tissue collection

Mice were sacrificed at three time points (6, 22 and 36 weeks);

considerations for the selection of these time points are explained in

detailed in Krishna, et al., body weights (BW) were recorded and liver

was collected and quickly frozen at −80 °C [17]. Blood (1 ml) was

collected via cardiac puncture and immediately split into two aliquots:

500 μl was placed in Na citrate-containing tubes, mixed thoroughly,

and the plasma was separated by centrifugation (3500 Å~g; 10 min;

4 °C). Harvested plasma was then aliquoted and placed at −80 °C until

its use for endocannabinoid and esterase activity analyses as described

in detail below. The other 500 μl of blood was immediately mixed, by

vortexing, with 1 ml of methanol:water (1.0:0.4 v/v) and then placed at

−80 °C until lipid extraction as described below. Liver (6, 22, and

36 weeks) samples were used for qPCR and endocannabinoid analyses.

2.4. Glucose tolerance test (GTT) and insulin sensitivity test (IST) areas

under the curve (AUCs)

Glucose tolerance (GTT) and insulin sensitivity (IST) tests were

performed after 5, 20 and 33 weeks on respective diets as described in

our recent study [17]. We used the blood glucose integrated areas

under the curve (AUC) in the GTT and IST tests, as calculated using the

trapezoidal method [24], to determine if mice's response to oral glucose

challenge or to insulin correlates with specific lipid metabolites (described

below).

2.5. Bligh-dyer blood lipid extraction

Phospholipids were extracted using chloroform and methanol according

to the method of Bligh and Dyer [25]. Briefly, blood samples

designated for lipidomics analysis were suspended in 1.25 ml of methanol

and 1.25 ml of chloroform. Tubes were vortexed for 30 s, allowed

to sit for 10 min on ice, centrifuged (213 Å~g; 5 min), and the

bottom chloroform layer was transferred to a new test tube. The extraction

steps were repeated a second time and the chloroform layers

combined. The collected chloroform layers were dried under nitrogen,

reconstituted with 50 μl of methanol: chloroform (3:1 v/v), and stored

at −80 °C until analysis.

2.6. Lipid phosphorus assay

Lipid phosphorus was quantified using the phosphorus assay [26].

400 μl of sulfuric acid (5 M) was added to lipid extracts (10 μl) in a glass

test tube, and heated at 180–200 °C for 1 h. 100 μl of 30% H2O2 was

then added to the tube while vortexing, and heated at 180–200 °C for

1.5 h. 4.6 ml of reagent (1.1 g ammonium molybdate tetrahydrate in

12.5 ml sulfuric acid in 500 ml ddH20) was added and vortexed,

followed by 100 μl of 15% ascorbic acid and vortexing. The solution

was heated for 7–10 min at 100 °C, and a 150 μl aliquot was used to

measure the absorbance at 830 nm.

2.7. Phospholipid characterization with electrospray ionization-mass

spectrometry (ESI-MS)

Lipid extract samples (500 pmol/μl) were prepared by reconstitution

in chloroform: methanol (2:1, v/v). ESI-MS was performed as described

previously [27–29] using a Trap XCT ion-trap mass spectrometer

(Agilent Technologies, Santa Clara, CA) with a nitrogen drying

gas flow-rate of 8 l/min at 350 °C and a nebulizer pressure of 30 psi.

The scanning range was from 200 to 1000 m/z on 5 μl of the sample

scanned in positive and negative ion mode for 2.5 min with a mobile

phase of acetonitrile: methanol: water (2:3:1) in 0.1% ammonium formate.

As described previously [30], qualitative identification of individual

phospholipid molecular species was based on their calculated

theoretical monoisotopic mass values, subsequent MS/MS analysis, and

their level normalized to either the total ion count (TIC) or the most

abundant phospholipid.

MSnth fragmentation was performed on an Agilent Trap XCT iontrap

mass spectrometer equipped with an ESI source. Direct injection

from the HPLC system was used to introduce the analyte. The nitrogen

drying gas flow-rate was 8.0 l/min at 350 °C. The ion source and ion

optic parameters were optimized with respect to the positive molecular

ion of interest. Initial identification was typically based on the loss of

the parent head group followed by subsequent analysis of the lysophospholipid.

In the event that neutral loss scanning could not confirm

the species, the tentative ID was assigned based on the m/z value and

the LIPIDMAPS database (http://www.lipidmaps.org).

2.8. Multivariate statistical analysis of blood lipids

Multivariate principal component analysis (PCA) was performed

using MetaboAnalyst 3.0 (http://www.metaboanalyst.ca/). Automatic

peak detection and spectrum deconvolution was performed using a

peak width set to 0.5. Analysis parameters consisted of interquartile

range filtering and sum normalization with no removal of outliers from

the dataset. Features were selected based on volcano plot analysis and

were further identified using MS/MS analysis. Significance for volcano

plot analysis was determined based on a fold change threshold of 2.00

and p ≤0.05. Following identification, total ion count was used to

normalize each parent lipid level, and the change in the relative

abundance of that phospholipid species as compared to its control was

determined. This method is standard for lipidomic analysis as reported

in our previous studies [27,29].

2.9. Liver endocannabinoid (2-AG and AEA) levels

2-AG and AEA were extracted from liver using a modification of the

method of Kingsley and Marnett (2007) [74]. In brief, ~0.05–0.1 g of

frozen liver tissue (exact weight recorded) was Dounce homogenized in

2:2:1 v/v/v ethyl acetate:hexane:0.1 M potassium phosphate (pH 7.0)

[total volume 5 ml; supplemented with butylated hydroxytoluene and

triphenylphosphine, 0.05% w/v each (antioxidants)] containing deuterated

standards for 2-AG and AEA (5.6 pmol d8-AEA and 518 pmol d8-

2-AG). The mixture was vortexed (1 min) and centrifuged to separate

organic and aqueous phases (1400 Å~g, 10 min). The organic layer was

removed, dried under a stream of N2 and residues dissolved in 2:2:1 v/

v/v water:methanol:isopropanol (200 μl). After filtration (0.1 μm),

10 μl of the resolubilized lipid was injected onto a Acquity UPLC BEH

C18 column (2.1Å~ 50 mm, 1.7 μm) equipped with VanGuard precolumn

(2.1 Å~ 5 mm, 1.7 μm). The mobile phase was a blend of solvent

A (2 mM ammonium acetate/0.1% acetic acid in water) and solvent B

(2 mM ammonium acetate/0.1% acetic acid in methanol). Analytes are

eluted with the following gradient program: 0 min (95% A, 5% B),

0.5 min (95% A, 5% B), 5 min (5% A, 95% B), 6 min (5% A, 95% B),

7 min (95% A, 5% B), 8 min (95% A, 5% B). The flow rate was 0.4 ml/

min and the entire column eluate was directed into a Thermo Quantum

Access triple quadrupole mass spectrometer (heated electrospray ionization

in positive ion mode). Single reaction monitoring (SRM) of each

analyte was as follows: 2-AG, [M +NH4]+ m/z 396.3 > 287.3; 2-AGd8,

[M+NH4]+ m/z 404.3 > 295.3; AEA, [M+ H]+ m/z 348 > 62;

AEA-d8, [M+H]+ m/z 356 > 63. Endocannabinoids were quantified

by measuring the area under each peak in comparison to the deuterated

standards and normalized on tissue weight.

2.10. Plasma endocannabinoid (2-AG and AEA) levels

Plasma levels of the two endocannabinoids 2-arachidonoylglycerol

(2-AG) and anandamide (N-arachidonoylethanolamine; AEA) were determined

using mass spectrometry. First, 50 μl of mouse plasma was

placed into a glass vial. Deuterated standards, 6 pmol AEA-d8 and

52 pmol 2-AG-d8 were added to each sample, followed by 2 ml of ethyl

acetate for extraction. The mixture was vortexed (1 min) and centrifuged

at 1400 g for 10 min. The organic layer (~1.5 ml) was transferred

into a clean glass vial and was dried under a stream of N2. The

residues were reconstituted in 1:1 v/v water: methanol (100 μl). After

filtration (0.1 μm), 10 μl of samples was injected onto an Acquity UPLC

system (Waters, Milford, MA) coupled to a TSQ Quantum Ultra tandem

mass spectrometer equipped with a heated electrospray (H-ESI) source

(Thermo Fisher). Chromatographic separation was carried out using an

Acquity UPLC BEH C18 column (2.1 mmÅ~ 100 mm, 1.7 μm) equipped

with a VanGuard precolumn (2.1 mmÅ~ 5 mm, 1.7 μm) at 40 °C using

column oven. The mobile phases used were water containing 0.1%

acetic acid (A) and methanol containing 0.1% acetic acid (B). Mobile

phase gradient conditions were as follows: hold at 15% A and 85% B for

0.5 min, linear increase of B to 95% in 2 min, hold at 95% B for 4 min,

decrease of B to 5% in 1 min and re-equilibrate for 2 min at the starting

conditions. The overall run time was 10 min and flow rate was 0.2 ml/

min. Eluate from the LC was directly electrosprayed into the mass

spectrometer using an electrospray ionization interface in the positive

mode. MS conditions were set as follows: spray voltage = 3500 V, vaporizer

temperature =350 °C, sheath gas= 25 units, auxiliary gas =2

and capillary temperature = 350 °C. Samples were run in positive

single reaction monitoring (SRM) mode and the following precursor-toproduct

ion transitions were used for quantification: 2-AG, [M+ H]+

m/z 379.2 > 287.1; AEA, [M+ H]+ m/z 348.2 > 287.2; 2-AG-d8,

[M + H]+ m/z 387.2 > 292.3; AEA-d8, [M+H]+ m/z 356.2 > 294.2. Scan time was

0.2 s per SRM, and the scan width was

m/z 0.01. Optimum collision energy and S-lenses conditions were determined

for each compound by using autotune software for each

analyte by post-column infusion of the individual compounds into a

50% A/50% B blend of the mobile phase being pumped at a flow rate of

0.2 ml/min. Xcalibur software was employed for data acquisition and

processing. For quantification, each calibration standard was prepared

ranging from 1 to 1000 nM by fortifying phosphate-buffered saline with

stock standards of 2-AG or AEA prepared in methanol. Quality control

samples were prepared at a concentration of 50 nM for each endocannabinoid

in triplicate. Weighted calibration curves were constructed

using 1/x as a weighting factor for 2-AG and AEA, respectively.

2.11. Plasma esterase activity

Plasma esterase activity was determined using the substrate paranitrophenyl

valerate, as previously described [31]. Production of p-nitrophenol

liberated from pNPV was monitored at 405 nm on a spectrophotometer

[32]. An extinction coefficient of 13 cm−1 mM−1 [33]

was used to convert the slopes of each activity curve to specific enzyme

activities. All enzymatic reaction rates were corrected for non-enzymatic

hydrolysis rates as we have described previously [31].

2.12. Real-time quantitative PCR (qPCR)

qPCR was performed on liver samples as described in [17,34].

Briefly, total liver RNA (20 mg tissue) was isolated using a GeneJET ™

RNA Purification Kit (Thermo Fisher Scientific, Pittsburgh, PA) and

quantified using a Take 3 plate and an Epoch microplate spectrophotometer

(Bio-Tek, Winooski, VT). RNA was converted to cDNA

using qScript cDNA SuperMix (Quanta Bioscience, Gaithersburg, MD)

and a Peltier thermal cycler (Bio-Rad, Hercules, CA). Using 3 ng of

cDNA per sample (with each sample run in duplicate), expression of

peroxisome proliferator-activated receptor alpha (PPARα), peroxisome

proliferator-activated receptor gamma (PPARγ), hepatic fatty acid

transporter (CD36), fatty acid synthase (FAS), acetyl-CoA carboxylase

(ACC), stearoyl-CoA desaturase (SCD), monoacylglycerol lipase (MGL),

cannabinoid receptor type 1 (CB1), cannabinoid receptor type 2 (CB2)

and 18S, was determined by qPCR using mouse-specific primers (Real

Time Primers, Elkins Park, PA) and SYBR Green-based master mix

(Qiagen, Valencia, CA). Amplifications were performed on a Mx3005P

qPCR machine (Stratagene) and treatment differences were calculated

as a fold change by the ΔΔ Ct method with 18S used as a house-keeping

gene (HKG) as previously reported [17,34]. Correlation between the

results from the liver qPCR and the blood lipidome was investigated as

described below.

2.13. Regression analysis between lipid features, GTT/IST, plasma

endocannabinoids, and liver homeostasis

GraphPad Prism for windows version 5.04 (GraphPad Software,

Inc., La Jolla, CA) was used for all correlation analyses comparing the

differential lipid expression via the relative abundance of features to

multiple endpoints used in the study such as liver mRNA expression

values, plasma endocannabinoid levels, and AUC values from GTT/IST.

Reported correlations meet the fairly stringent cutoff correlation coefficient

criteria of R2 ≥0.6.

2.14. Statistical analysis

All statistical analyses were compiled using GraphPad Prism for

windows version 5.04 (GraphPad Software, Inc., La Jolla, CA). For all

analysis, the experimental unit was individual animals and samples

from a total of 6–8 animals/diet/time point were assessed. For all

analyses, significance was set at p ≤ 0.05 where data are expressed as

mean ± SEM based on t-test for pairwise analysis and/or ANOVA

analysis (two-factor repeated-measures with Bonferroni post hoc test).

3. Results

3.1. Morphometric, GTT, IST, and intestinal permeability data

The lipid changes reported below were correlated to glucose tolerance

(GTT) and insulin sensitivity (IST) outcomes, which were recently

reported [17]. Briefly, HFD-fed mice were significantly heavier than

LFD mice at all three time points, and had impaired glucose tolerance.

Interestingly, HFD-feeding had the greatest effect on glucose tolerance

rather than insulin sensitivity [17]. On the other hand, while HFD decreased

IST at the earlier time points, this trend was not seen at the end

of the study due to age-dependent decreases in IST in LFD-fed mice

[17]. HFD-consumption by the female C57BL/6 mice also increased the

gastrointestinal permeability, more so after longer feeding durations

[17].

3.2. Multivariate analysis of lipidome

Multivariate, unsupervised principal component analysis (PCA) of

spectral data comparing high-fat diet (HFD) and low-fat diet (LFD)

consumption showed distinct clustering within the blood lipidome

where diet and age were the major effectors (Fig. 1, Fig. 2). Scores plots

of all groups, for both positive (Fig. 1A) and negative ion mode

(Fig. 2A), demonstrated a striking separation between 6-week vs. 22-

and 36-week treatment groups. The separation of populations occurred

regardless of dietary treatment indicating a significant role for age on

the lipidome. However, HFD-consumption altered lipid profiles within

each respective time point where 6-week treatment (Fig. 1C, Fig. 2C)

had the most pronounced effect. Urine lipid profiles demonstrated a

similar trend (Suppl. Fig. 1). The baseline lipidome was also assessed

with PCA analysis comparing blood lipid profiles of ~5–6-week-old

mice to 6-week HFD/LFD-fed mice, indicating group differences along

with variability in the baseline lipidome, which diminished following

6 weeks of dietary treatment (Suppl. Fig. 4).

Volcano plots identified and ranked potentially important features

based on fold change and statistical significance level for age- (Fig. 3,

Suppl. Fig. 2) and diet-related (Fig. 4, Suppl. Fig. 3) effects. Age-related

pairwise comparisons within dietary treatment groups for short vs.

long/prolonged (6-week vs. 22- and 36-week) (Fig. 3A-3B, Suppl.

Fig. 2A–B) yielded the greatest number of features while 22-week vs.

36-week (Fig. 3C, Suppl. Fig. 2C) resulted in very few. Based on the

number of features altered, diet-related pairwise comparisons of LFDvs.

HFD-consumption indicated that HFD-induced alterations were

most robust following a short consumption period (Fig. 4A) with fewer

alterations following longer periods of exposure (Fig. 4B, C).

3.3. Phospholipid species

Diet- and age-dependent alterations elevated the relative abundances

of phospholipid (PL) species in blood lipid profiles of HFD-fed

mice after 6 weeks of consumption (Fig. 5A), and presented the most

changes following 22 weeks of treatment where HFD-fed mice had

decreased relative abundances of various PL species, differing from

those species affected after 6 weeks on the diet (Fig. 5A). None of the

age−/diet- changes persisted after 36 weeks of HFD feeding (Fig. 5A).

Significance analysis of lipid features was performed for all time

points within each dietary group (i.e. 6 weeks vs. 22 weeks of LFD

feeding) to identify age−/diet- alterations. These age−/diet- features

were subsequently excluded during analysis of LFD vs. HFD treatment

to characterize the effects of HFD feeding alone. Interestingly, the effects

of HFD-consumption alone did not appear following 6 weeks of

feeding (Fig. 6A). 22-week consumption demonstrated the greatest effect,

much like that observed in features altered due to diet−/age-,

although differences in PL relative abundances were bidirectional and

did not show a class-specific uniform trend (Fig. 6A). Following

36 weeks of feeding, the relative abundance of only one feature (m/z

578.3) changed due to HFD-consumption alone (Fig. 6A). MS/MS

analysis and neutral loss scanning was performed to validate phospholipid

class identities (Suppl. Table 1–2).

3.4. Fatty acyl species

Fatty acyl (FA) species altered based on diet- and age-related effects

demonstrated bidirectional effects at all three time points where the

majority of features identified were in blood profiles of mice following

22 weeks of treatment (Fig. 5B). One feature (m/z 562.8) persisted

between short- and long-term feeding, demonstrating an increase in

relative abundance in HFD-fed animals after 6-weeks followed by a

decrease after 22-weeks (Fig. 5B).

Diet alone altered the blood lipidome of the 6-week treatment group

in which HFD-consumption decreased the relative abundances of FAs

by ~2-fold (Fig. 6B). This was also observed in urine lipid profiles of

the 6-week treatment group where three features (m/z 337, m/z 385,

m/z 381) detected and altered in blood were also detected in urine with

comparable magnitudes of differences between LFD- and HFD-fed mice

(Fig. 6B). The effects of HFD-consumption alone did not persist after 22-

weeks and 36-weeks (Fig. 6B). MS/MS analysis and neutral loss scanning

was performed to validate fatty acyl class identities (Suppl.

Table 1–2).

3.5. Glycerolipid species

Most diet- and age-dependent feature alterations in glycerolipids

occurred after short- (6 weeks) and long-term feeding (22 weeks) where

changes were primarily bidirectional across groups (Fig. 5C). 22-week

feeding displayed a trend of general decreases in glycerolipid (GL) relative

abundances in HFD-fed mice (Fig. 5C). We identified one feature

that was altered after 36-weeks of dietary treatment, (m/z 708.6),

which was decreased in HFD-mice (Fig. 5C).

Resembling the pattern observed across FA species (Fig. 6B), the

majority of features altered due to HFD-consumption alone were

identified in the 6-week treatment group. Further, GL features demonstrated

species-specific increases and decreases in both blood and

urine profiles (Fig. 6C). Long-term feeding affected a few GL features

with net decreases in HFD-mice, which were no longer present after

36 weeks (Fig. 6C). MS/MS analysis and neutral loss scanning was

performed to validate glycerolipid class identities (Suppl. Table 1–2).

3.6. Liver endocannabinoids

Given that many of the species altered in HFD-fed mice were

phospholipids containing polyunsaturated fatty acids (PUFAs), and

because many of the fatty acyls correlated to derivatives of fatty acids,

we focused on arachidonic acid-containing metabolites, including 2-

arachidonoylglycerol (2-AG) and N-arachidonoylethanolamide (AEA).

Upon release, these endocannabinoids target cannabinoid receptors

(CB1 and CB2) and work together to play a role in energy homeostasis

[22]. Liver 2-AG levels decreased after 6 and 36 weeks of HFDconsumption

(Fig. 7A). Interestingly, an age-related effect was observed

where 36 weeks of HFD-feeding decreased liver 2-AG levels (Fig. 7A).

The decrease of 2-AG levels on the liver after 36 weeks was accompanied

by a significant increase of AEA (Fig. 7B). The only significant

effect of HFD on liver AEA levels was a significant decrease after

6 weeks on the diet (Fig. 7B). CB1 expression did not change while CB2

expression showed time-dependent increases in HFD-fed mice, although

these were not significant (Fig. 7C).

3.7. Plasma endocannabinoids

Plasma levels of 2-AG and AEA in the LFD-fed mice were quite

stable, apart from a slight increase after 36 weeks of feeding (Fig. 8). In

contrast, HFD-consumption increased plasma levels of 2-AG, but the

effects were bi-phasic where a significant increase was only observed at

6 and 36 weeks, but not after 22 weeks of HFD-consumption (Fig. 8A).

There also appeared to be an effect of age on plasma 2-AG within the

HFD-fed mice as shown by an increase in 2-AG levels following

36 weeks compared to 22 weeks of feeding (Fig. 8A). The effects of

HFD-consumption on the other endocannabinoid, AEA, resembled the

diet's effects on plasma 2-AG, with the only significant effect being an

increase of plasma AEA levels after 36 weeks on HFD (Fig. 8B).

3.8. Plasma esterase activity

The increase in fatty acyls and lysophospholipid species suggests

increased esterase activity. This hypothesis was addressed by assessing

esterase activity at all time points in the plasma. HFD-consumption

increased plasma esterase activity at all time points, with the most

pronounced effect following 6 weeks of feeding (Fig. 9). Although

plasma esterase activity increased following both 22 and 36 weeks, the

effect at 36 weeks was not significant (Fig. 9). It appeared that age

alone affected plasma esterase activity, indicated by heightened esterase

activity after 6 weeks of HFD-consumption followed by lower

levels at later time points (Fig. 9).

3.9. Liver qPCR data of key lipid homeostasis genes

There are many genes that encode for protein regulating energy

balance and lipid metabolism, including peroxisome proliferator-activated

receptors (PPARs) [35,36]. PPARα is best known for its major

role in lipid and lipoprotein metabolism while PPARγ is involved in

adipogenesis and insulin sensitivity [35,37]. HFD-consumption increased

expression of liver mRNA levels of PPARα, PPARγ, and CD36, a

known target of PPARγ [38], significantly at 6 weeks of HFD (Fig. 10A).

Interestingly, PPARα and PPARγ levels increased after 36 weeks of

HFD-consumption, but not after 22 weeks. Liver CD36 mRNA was elevated

at all three time points (Fig. 10A), with the magnitude of elevation

greatest after 36 weeks on HFD.

PPARγ has been shown to increase during lipogenesis, thus we assessed

the expression of fatty acid synthase (FAS), acetyl-CoA carboxylase

(ACC), and stearoyl-CoA desaturase (SCD) (Fig. 10B). HFD-fed

mice demonstrated a numerical trend of time-dependent increases

although not significant, in SCD, ACC, and FAS expression (Fig. 10B).

MGL demonstrated increased expression in mice fed HFD for 6 weeks;

however, this effect was tapered after long- and prolonged-feeding

(Fig. 10B).

3.10. Correlation between lipids, GTT/IST, plasma endocannabinoids, liver

homeostasis

We performed regression analysis to determine if changes in the

blood lipidome correlated to changes in GTT or IST, plasma endocannabinoid

levels, or gene expression of liver homeostasis markers.

Table 1A lists correlations (R2 ≥0.6) between several lipid features and

the expression of CD36, PPARα, and PPARγ mRNA in the liver. There

were only a few lipid species whose abundance correlated to changes in

these genes, including two unidentifiable FA species. Correlations occurred

at all time points with increases in TG (52:4 or 52:5) [PPARγ] in

HFD-mice at 6 weeks, decreases of DG (40:6) [CD36], PE (42:7)

[CD36], and FA (m/z 562.8) [PPARα] at 22 weeks, and an increase in

FA (m/z 438.8) [PPARα] at 36 weeks (Fig. 11). Lipid species indicated

by only their m/z value were unable to be fully characterized by subsequent

MS/MS analysis.

Regression analyses also demonstrated quite a few relationships

between changes in select blood lipids and changes in AUC from glucose

tolerance and insulin sensitivity tests after 6 weeks (Table 1B).

Fewer correlations were identified for changes in the blood lipidome at

22 or 36 weeks (Suppl. Table 3), which is not surprising given the robust

differences observed after 6-weeks of treatment. Interestingly,

several lipids such as PC (44:3) and DG (34:3), which correlated to

glucose tolerance, also demonstrated inverse correlates to insulin sensitivity.

Regarding plasma endocannabinoids, GL species (m/z 734.6, m/z

760.5) and FA species (m/z 353.2) correlated to changes in in 2-AG

exclusively within HFD-mice, whereas AEA did not correlate to any

lipid features (Table 1C).

4. Discussion

The association between increased high-fat consumption and excess

adiposity poses a major global health problem that heightens the risk of

metabolic disorders, diabetes, heart disease, fatty liver, and some forms

of cancer [39]. Growing evidence implicating a role for impaired lipid

metabolism, coupled with the advent of bioinformatics tools has

prompted efforts in characterizing the obese lipidome [40–42]. While

these studies have highlighted alterations in the plasma and/or serum

lipidome, there have been few studies examining the effects of age on

the lipidome and/or the interaction between age and obesity. Further,

the number of studies that address diet-induced obesity within female

models is quite limited, although an increased linear trend of female

obesity among US women within the last decade demonstrates the

significant need [3,43,44]. Here, we used shotgun lipidomics to assess

the effects of dietary fat consumption, age, and their interaction at the

level of the blood lipidome. We correlated changes in the blood lipidome

to changes in metabolic regulation, endocannabinoid levels, and

plasma esterase activity.

One of the most interesting findings of this study is that the effect of

age superseded the effect of HFD with regard to alterations in the blood

lipidome. These data emphasize the need to characterize and stratify

lipidomic alterations not only to diet, but also to age. The accentuated

effect of age on the lipidome between 12-week-old (6-weeks on the

diet) vs. 28-week-old (22 weeks on the diet) and 42-week-old (36 weeks

on the diet) mice indicated distinct shifts in lipid composition and/or

regulation, an interesting note since all ages fall within the mature adult

phase of C57BL/6 mice. Although multivariate analysis indicated a

slight difference between 28- and 42-week-old animals, the separation

was not as robust. With regard to dietary treatment, younger animals

presented striking lipidomic responses to HFD-consumption while older

animals had a more tapered shift, possibly a result of time-dependent

homeostatic mechanisms in response from long-term feeding. In this

regard, as we reported recently, it is interesting to note that in terms of

insulin sensitivity, but not glucose tolerance, age appears to be a major

driver of decreased insulin sensitivity to the point that the effects of

prolonged HFD feeding are overpowered by the effects of age [17].

Given the limitations of the shotgun approach along with the sheer

number of features identified in this study, we report general lipidomic

changes in terms of lipid class with alterations falling into three classes:

glycerolipids (GL), fatty acyls (FA), and phospholipids (PL). MS/MS

analysis was employed to verify these lipid species (Suppl. Table 1).

Features changing at only a few time points showed a class-specific

trend in terms of differential lipid expression; however, most changes

indicated species-specific alterations.

A few studies have demonstrated alterations in lipid classes within

rodent obesity where the majority of reported changes encompass

ceramides, cholesterols, triglycerides, and phospholipids. For example,

changes in lipid species such as lysophosphatidylcholines have been

associated with obesity, insulin resistance, and type 2 diabetes [45–47].

In agreement with some of these studies, we report elevations in PC

(38:5), PC (44:3), and PC (38:3) in the blood of HFD-fed mice, also

reported in Eisinger et al. [40]. Changes in GL and FA reported in the

current study are not in line with another study [48], but it should be

noted that the sex of the mice, feeding durations, and importantly, the

dietary composition in [48] and our study are different. It should also

be pointed out that several studies have demonstrated that HFD typically

increases the levels of TAGs and DAGs [49]. The fact that not

many of these lipids were detected in our own study is most likely a

limitation of the shotgun approach used.

As mentioned above, lipidomic alterations identified by shotgun

analysis revolved around three major lipid classes (GL, FA, and PL),

which also happen to constitute the endocannabinoid system.

Quantification of both 2-AG and AEA demonstrated several correlations

with specific blood lipids. AEA and 2-AG, both derivatives of arachidonic

acid, are signaling lipids that mediate their action via activation

of cannabinoid receptors. Further, changes in plasma levels of 2-AG and

AEA after 6 and 36 weeks associated with decreases in liver 2-AG. This

suggests that increased 2-AG levels in circulation are due to increased

mobilization from the liver, with/without concomitant decreased 2-AG

breakdown [50]. In a study by Caraceni, et al., 2-AG levels were reported

to be higher in the hepatic veins of cirrhosis patients when

compared to peripheral blood, supporting the hypothesis that the liver

contributes to circulating 2-AG levels [51]. This may also suggest that

the source of the increase in plasma 2-AG is non-hepatic, i.e. dietary

[20].

Plasma AEA levels were most affected by the diet after 36 weeks;

however, as opposed to the decreases of liver 2-AG, liver AEA was increased

after 36 weeks irrespective of diet. Together, these data indicate

that in female C57BL/6 mice plasma 2-AG is more sensitive to

HFD-consumption. Further, similar to other endpoints in this study, the

effects of HFD diet are most prominent during early (6 weeks) and late

phases (36 weeks) of the feeding trial. Circulating endocannabinoids

also appear to be more sensitive to non-dietary liver pathology than

their liver levels. For example, in certain conditions, i.e. hepatitis C,

plasma, but not liver, 2-AG was increased [52]. On the other hand,

circulating AEA was significantly higher in cirrhotic patients [51].

A novel finding is the fact that age was associated with decreased 2-

AG and increased AEA in the liver. At the end of the study (36 weeks on

the diet, 42-43-week-old), the mice were middle-aged and close to

becoming reproductively senescent [53]. While liver-specific

endocannabinoid data for female C57BL/6 mice within the context of

age are lacking, it is interesting that a recent study reported decreased

2-AG, but not AEA, levels in the hippocampus of aged mice [54]. In this

study, the decrease in hippocampal 2-AG was attributed to a concomitant

decrease of local 2-AG synthesis and increase of its breakdown

[54]. With that in mind, our data suggest that the age-dependent metabolic

changes in the 2-AG pathway that operate in the brain (hippocampus)

do so in the liver as well. In addition, the increase in liver AEA

levels at the end of the study highlights endocannabinoid metabolitespecific

effects of age. Although Osei-Hyiaman, et al. previously showed

that liver AEA levels increase following HFD, our data indicate lower

levels of hepatic AEA in HFD-fed mice [55,56]. However, it is important

to note that the dietary feeding regimen in Osei-Hyiaman, et al. was

initiated in slightly older mice, which may play a large role given our

data demonstrating the interaction of age and diet. Moreover, the

findings in the above study were based on the use of a combination of

male and female mice and thus, does not reflect the effects within females

alone. This would not be surprising since gender-specific responses

to HFD-intake have been reported in rodents, and it has even

been suggested that females are more susceptible to developing the

secondary effects of HFD-induced obesity [57]. It should also be mentioned

that adipose distribution and function differs across males and

females, so differences are expected to exist in mediators produced by

adipocytes such as endocannabinoids [57–59].

The increased circulating 2-AG levels in the female C57BL/6 mice in

our study are in line with multiple studies in obese human subjects. For

example, obese men, especially those with increased intra-abdominal

adiposity, have increased plasma 2-AG [60]. Interestingly, direct correlation

between plasma endocannabinoids, insulin resistance and

dyslipidemia has been suggested [61]. Moreover, chronic cannabinoid

receptor 1 (CB1) stimulation exacerbates the metabolic dysregulation

caused by HFD-consumption, suggesting key role for the endogenous

endocannabinoids in the process [62]. Circulating endocannabinoids

might contribute to obesity by their central and/or peripheral actions

[20]. The key role of CB1-specific over-activation by excessive endocannabinoids

is further emphasized by the fact that global CB1−/−

mice are not susceptible to HFD and liver-specific CB1−/− mice are

protected from some, but not all, of the adverse effects of HFD intake

[63]. Treatments aimed at reducing plasma endocannabinoids are

beneficial in re-balancing the metabolic dysregulation [64] and obesityrelated

inflammation [65], but not for reducing the body weight in

obese subjects [64]. Interestingly, and in line with our current data,

circulating 2-AG levels were significantly elevated in insulin-resistant

obese women [66].

Peroxisome proliferator-activated receptors (PPARs) are members

of the steroid hormone receptor superfamily of nuclear transcription

factors that are involved in the regulation of various genes encoding

proteins involved in energy balance and lipid metabolism [35,36].

PPARα is best known for its major role in lipid and lipoprotein metabolism

while PPARγ is involved in adipogenesis and insulin sensitivity

[35,37]. It has also been suggested that hepatic PPARγ may mediate the

accumulation of fat via the regulation of genes essential for de novo

lipogenesis, i.e. fatty acid synthase (FAS), acetyl-CoA carboxylase (ACC),

and stearoyl-CoA desaturase (SCD) [67]. Further, it is possible that

glycerolipid alterations can be attributed to changes in lipolysis via

monoacylglycerol lipase (MGL), one of the main lipases involved in the

catabolism of TG.

In our study, we also observed HFD-induced increases in liver

PPARα, PPARγ, and CD36. The increases in PPARα and PPARγ were

biphasic (6 and 22 weeks) and more prominent after 6 weeks of HFD

feeding, indicating that the activation of the PPAR pathways are timedependent.

PPARα regulates fatty acid β-oxidation, is activated by the

AEA analogue oleoylethanolamide (OEA), and pharmacological increases

of OEA are beneficial to diet-induced obese mice [68]. While we

have not measured OEA in our study, it is conceivable that the lack of

significant increase in liver PPARα after 22 weeks of HDF feeding was

due to sensitization-dependent downregulation. Liver PPARγ, which

showed similar kinetics to PPARα, regulates lipid and glucose homeostasis,

is elevated by a HFD [69], and approaches aimed at curbing its

activation are beneficial in obesity and type 2 diabetes [70]. Together,

the changes in liver PPAR levels are in line with the time-dependent

metabolic dysregulation of these mice, especially the sensitivity to insulin

challenge, and reflects the changes in the blood lipidome reported

here [17].

CD36 was the sole lipid homeostasis/inflammation molecule whose

expression was increased by HFD throughout the study. Increases in

CD36 were greatest at the end (36 weeks) of the experiment. CD36 is

associated with obesity, diabetes, and liver dysfunction; hence, our

findings are not surprising. Liver CD36 was previously shown to increase

due to aging [71]. Thus, the marked increase in liver CD36 at the

end of the feeding duration could be a sum of the effects of age and HFD

on its expression or, due to increased demand for hepatic lipid uptake in

the face of continued HFD-consumption. In this regard, CD36 plays a

major role on hepatic fat uptake [71].

Circulating esterases are predominantly investigated for their role in

the metabolism of drugs and toxicants, though they also metabolize

endogenous, i.e. dietary, and exogenous lipids [72]. The early robust

increase in plasma esterase that we observed in the current study might

be the result of host's attempt to maintain lipid homeostasis in the face

of excessive dietary fat. Over time, the increases in esterase activity

were still present, but less robust. This may reflect saturation of this

likely protective mechanism. In support of this hypothesis, esterasedeficient

mice are not only more susceptible to pesticides that are detoxified

by it, but also to diet-induced metabolic dysregulation and

atherosclerosis [73]. Our data indicating an age effect, i.e. decreased

plasma esterase due to age, further supports the notion that this mechanism

of metabolizing excessive dietary fat is less robust in older

mice.

In conclusion, we demonstrated an interaction between dietary fat

consumption and aging with widespread effects on the blood lipidome

in female mice. This study indicates that the effects of HFD feeding

occur in an age-dependent manner with robust responses at a younger

age. Further, we identified several associations between lipids and

metabolic and liver regulation, providing a basis for female-specific

obesity- and age-related lipid biomarkers. These findings highlight the

need for additional age-dependent tracking studies, prior to sexual

maturity into advanced age, to obtain comprehensive understanding of

the evolving lipidome with regard to dietary changes.

Based on the article below:

Title: High fat diet and Endocannabiods.

Question: Please write a summary of the article, with a deeper understanding of the important informations on the high fat diet and endcannabiods. And list the Advantage and disavantage of a high fat diet and endocannabiods.

Article:

1. Introduction

Obesity is a growing public health concern that increases the risk of

inflammatory and metabolic disorders such as type 2 diabetes, fatty

liver, and pulmonary inflammation [1,2]. The incidence of obesity has

drastically increased over the past few decades. In a nationally representative

survey (National Health and Nutrition Examination

Survey, 2014) of adults in the US, the prevalence of obesity was 35%

among men and 40% among women, where linear trends significantly

increased for women between 2005 and 2014 [3]. The prolonged and

excessive inflammation associated with obesity has also been associated

with increases in certain cancers, cardiovascular disease, and Alzheimer's

disease [4,5]. While the mechanisms linking obesity and metabolic

disorders are not fully understood, several studies suggest that

alterations in lipid-mediated metabolism play a significant role [1,2,6].

These studies have led to the hypothesis that change in the blood lipidome

can be exploited to identify lipid markers as prognostic indicators

for obesity and type 2 diabetes [7].

Lipids are a diverse subset of biomolecules that are not only responsible

for energy storage and structural regulation, but also participate in complex signaling networks whose disruption results in

the pathogenesis of obesity and other ailments. A few studies have

identified several lipid and lipoprotein abnormalities among obese

patients [8,9]. For example, Hu, et al. reported decreases in HDL cholesterol

along with altered triglyceride levels in nondiabetic obese patients

[9]. Additionally, the role of dietary fat in obesity and influence

of fatty food intake on inflammatory responses are well-established

[4,10]. Obesity-associated inflammation is not restricted to impaired

lipid metabolism, but is also strongly linked with type 2 diabetes, as

obesity is associated with insulin resistance, which heightens the risk

for metabolic syndrome [11,12].

The higher prevalence of type 2 diabetes among adults supports the

assertion that aging is the precursor to insulin resistance [13–16]. While

insulin resistance, type 2 diabetes, and metabolic syndrome have been

studied within the context of age to some extent, including by us [17],

the effect of age on the lipidome and/or age-obesity interactions have

not received significant attention. However, two more recent studies

demonstrate that age exerts appreciable, lipid species-specific effects on

the brain lipidome [18] and, importantly, that age has profound effects

on the female reproductive system (oocytes) lipidome [19]. These

studies support the hypothesis that age-obesity interactions alter the

lipidome. In comparison to the lack of knowledge on the effect of

obesity and/or age on the plasma lipidome as a whole, it is known that

obesity can alter the levels of select lipid species. For example, increased

circulating endocannabinoids, especially 2-arachidonoylglycerol

(2-AG), have been associated with obesity in both humans and

laboratory animals, i.e [20,21]; however, it is not known how the endocannabinoid

system is altered within the context of age in the face of

high-fat diet consumption. The endocannabinoid (EC) system participates

in the control of lipid and glucose metabolism and dysregulation

of this system can occur following unbalanced energy intake [22]. Such

dysregulation often results in overactivity across various organs involved

in energy homeostasis such as intra-abdominal adipose tissue

[23]. Over-activation of the endocannabinoid system has been shown to

promote insulin resistance [6]. Additionally, the essential role of the EC

system in adipogenesis and lipogenesis has been reviewed in detail by

Silvestri and Marzo, et al. [22,23].

In this study, we further investigated the effects of dietary fat consumption,

age, and their interaction at the level of the lipidome using

shotgun lipidomics with electrospray ionization-mass spectrometry

(ESI-MS). Because of the paucity of data and the linear increase of female

obesity among US women in the most recent decade [3], we assessed

the blood lipid profiles of female C57BL/6 mice following HFDconsumption

for short (6 weeks), long (22 weeks), and prolonged

(36 weeks) periods to evaluate the persisting effects of feeding. To

compare lipid alterations with metabolic and liver regulation, markers

of liver homeostasis were assessed and correlations between them and

indices of glucose tolerance and insulin sensitivity with the blood lipidome

were determined. Circulating and liver levels of the two major

endocannabinoids, 2-arachidonoylglycerol (2-AG) and anandamide

(AEA), were also measured to determine the effects of HFD-consumption

and age on the endocannabinoid system.

2. Materials and methods

2.1. Animals

Experiments were performed with female C57BL/6 mice (Harlan,

Indianapolis, IN). The mice were housed (4–5/cage) and maintained at

22–24 °C with food and water available ad libitum on a 12 h light/dark

cycle in an AAALAC accredited facility throughout the study. All experimental

procedures were in accord with the latest National Institutes

of Health (NIH) guidelines and the study was approved prior to initiation

by the Institutional Animal Care and Use Committee (IACUC) of

the University of Georgia.

2.2. Dietary treatment

The diets and dietary treatment are described in detail in our recent

publication [17], where the body weight changes, metabolic and behavioral

effects of the same experimental paradigm are reported.

Briefly, 6–7 weeks old female mice weighing 16.0 ± 0.20 g

(mean ± SEM) were randomly divided into two groups (n = 8/group/

time point) and placed on either a low-fat diet (LFD; 10% kcal from fat,

D12450J, Research Diets, Inc., New Brunswick, NJ) or a high-fat diet

(HFD; 60% kcal from fat, D12492, Research Diets) for either 6, 22, or

36 weeks. The LFD diet (3.85 kcal/g) consisted of 70% carbohydrate,

20% protein, 10% fat, of which 23.5% were saturated fatty acids [SFA],

29.7% monounsaturated fatty acids [MUFA], and 46.8% polyunsaturated

fatty acids [PUFA]) (Suppl. Table 4). The HFD diet

(5.24 kcal/g) consisted of 20% carbohydrate, 20% protein, 60% fat, of

which 32.2% were SFA, 35.9% MUFA, 31.9% PUFA (Suppl. Table 4).

2.3. Blood, plasma, and liver tissue collection

Mice were sacrificed at three time points (6, 22 and 36 weeks);

considerations for the selection of these time points are explained in

detailed in Krishna, et al., body weights (BW) were recorded and liver

was collected and quickly frozen at −80 °C [17]. Blood (1 ml) was

collected via cardiac puncture and immediately split into two aliquots:

500 μl was placed in Na citrate-containing tubes, mixed thoroughly,

and the plasma was separated by centrifugation (3500 Å~g; 10 min;

4 °C). Harvested plasma was then aliquoted and placed at −80 °C until

its use for endocannabinoid and esterase activity analyses as described

in detail below. The other 500 μl of blood was immediately mixed, by

vortexing, with 1 ml of methanol:water (1.0:0.4 v/v) and then placed at

−80 °C until lipid extraction as described below. Liver (6, 22, and

36 weeks) samples were used for qPCR and endocannabinoid analyses.

2.4. Glucose tolerance test (GTT) and insulin sensitivity test (IST) areas

under the curve (AUCs)

Glucose tolerance (GTT) and insulin sensitivity (IST) tests were

performed after 5, 20 and 33 weeks on respective diets as described in

our recent study [17]. We used the blood glucose integrated areas

under the curve (AUC) in the GTT and IST tests, as calculated using the

trapezoidal method [24], to determine if mice's response to oral glucose

challenge or to insulin correlates with specific lipid metabolites (described

below).

2.5. Bligh-dyer blood lipid extraction

Phospholipids were extracted using chloroform and methanol according

to the method of Bligh and Dyer [25]. Briefly, blood samples

designated for lipidomics analysis were suspended in 1.25 ml of methanol

and 1.25 ml of chloroform. Tubes were vortexed for 30 s, allowed

to sit for 10 min on ice, centrifuged (213 Å~g; 5 min), and the

bottom chloroform layer was transferred to a new test tube. The extraction

steps were repeated a second time and the chloroform layers

combined. The collected chloroform layers were dried under nitrogen,

reconstituted with 50 μl of methanol: chloroform (3:1 v/v), and stored

at −80 °C until analysis.

2.6. Lipid phosphorus assay

Lipid phosphorus was quantified using the phosphorus assay [26].

400 μl of sulfuric acid (5 M) was added to lipid extracts (10 μl) in a glass

test tube, and heated at 180–200 °C for 1 h. 100 μl of 30% H2O2 was

then added to the tube while vortexing, and heated at 180–200 °C for

1.5 h. 4.6 ml of reagent (1.1 g ammonium molybdate tetrahydrate in

12.5 ml sulfuric acid in 500 ml ddH20) was added and vortexed,

followed by 100 μl of 15% ascorbic acid and vortexing. The solution

was heated for 7–10 min at 100 °C, and a 150 μl aliquot was used to

measure the absorbance at 830 nm.

2.7. Phospholipid characterization with electrospray ionization-mass

spectrometry (ESI-MS)

Lipid extract samples (500 pmol/μl) were prepared by reconstitution

in chloroform: methanol (2:1, v/v). ESI-MS was performed as described

previously [27–29] using a Trap XCT ion-trap mass spectrometer

(Agilent Technologies, Santa Clara, CA) with a nitrogen drying

gas flow-rate of 8 l/min at 350 °C and a nebulizer pressure of 30 psi.

The scanning range was from 200 to 1000 m/z on 5 μl of the sample

scanned in positive and negative ion mode for 2.5 min with a mobile

phase of acetonitrile: methanol: water (2:3:1) in 0.1% ammonium formate.

As described previously [30], qualitative identification of individual

phospholipid molecular species was based on their calculated

theoretical monoisotopic mass values, subsequent MS/MS analysis, and

their level normalized to either the total ion count (TIC) or the most

abundant phospholipid.

MSnth fragmentation was performed on an Agilent Trap XCT iontrap

mass spectrometer equipped with an ESI source. Direct injection

from the HPLC system was used to introduce the analyte. The nitrogen

drying gas flow-rate was 8.0 l/min at 350 °C. The ion source and ion

optic parameters were optimized with respect to the positive molecular

ion of interest. Initial identification was typically based on the loss of

the parent head group followed by subsequent analysis of the lysophospholipid.

In the event that neutral loss scanning could not confirm

the species, the tentative ID was assigned based on the m/z value and

the LIPIDMAPS database (http://www.lipidmaps.org).

2.8. Multivariate statistical analysis of blood lipids

Multivariate principal component analysis (PCA) was performed

using MetaboAnalyst 3.0 (http://www.metaboanalyst.ca/). Automatic

peak detection and spectrum deconvolution was performed using a

peak width set to 0.5. Analysis parameters consisted of interquartile

range filtering and sum normalization with no removal of outliers from

the dataset. Features were selected based on volcano plot analysis and

were further identified using MS/MS analysis. Significance for volcano

plot analysis was determined based on a fold change threshold of 2.00

and p ≤0.05. Following identification, total ion count was used to

normalize each parent lipid level, and the change in the relative

abundance of that phospholipid species as compared to its control was

determined. This method is standard for lipidomic analysis as reported

in our previous studies [27,29].

2.9. Liver endocannabinoid (2-AG and AEA) levels

2-AG and AEA were extracted from liver using a modification of the

method of Kingsley and Marnett (2007) [74]. In brief, ~0.05–0.1 g of

frozen liver tissue (exact weight recorded) was Dounce homogenized in

2:2:1 v/v/v ethyl acetate:hexane:0.1 M potassium phosphate (pH 7.0)

[total volume 5 ml; supplemented with butylated hydroxytoluene and

triphenylphosphine, 0.05% w/v each (antioxidants)] containing deuterated

standards for 2-AG and AEA (5.6 pmol d8-AEA and 518 pmol d8-

2-AG). The mixture was vortexed (1 min) and centrifuged to separate

organic and aqueous phases (1400 Å~g, 10 min). The organic layer was

removed, dried under a stream of N2 and residues dissolved in 2:2:1 v/

v/v water:methanol:isopropanol (200 μl). After filtration (0.1 μm),

10 μl of the resolubilized lipid was injected onto a Acquity UPLC BEH

C18 column (2.1Å~ 50 mm, 1.7 μm) equipped with VanGuard precolumn

(2.1 Å~ 5 mm, 1.7 μm). The mobile phase was a blend of solvent

A (2 mM ammonium acetate/0.1% acetic acid in water) and solvent B

(2 mM ammonium acetate/0.1% acetic acid in methanol). Analytes are

eluted with the following gradient program: 0 min (95% A, 5% B),

0.5 min (95% A, 5% B), 5 min (5% A, 95% B), 6 min (5% A, 95% B),

7 min (95% A, 5% B), 8 min (95% A, 5% B). The flow rate was 0.4 ml/

min and the entire column eluate was directed into a Thermo Quantum

Access triple quadrupole mass spectrometer (heated electrospray ionization

in positive ion mode). Single reaction monitoring (SRM) of each

analyte was as follows: 2-AG, [M +NH4]+ m/z 396.3 > 287.3; 2-AGd8,

[M+NH4]+ m/z 404.3 > 295.3; AEA, [M+ H]+ m/z 348 > 62;

AEA-d8, [M+H]+ m/z 356 > 63. Endocannabinoids were quantified

by measuring the area under each peak in comparison to the deuterated

standards and normalized on tissue weight.

2.10. Plasma endocannabinoid (2-AG and AEA) levels

Plasma levels of the two endocannabinoids 2-arachidonoylglycerol

(2-AG) and anandamide (N-arachidonoylethanolamine; AEA) were determined

using mass spectrometry. First, 50 μl of mouse plasma was

placed into a glass vial. Deuterated standards, 6 pmol AEA-d8 and

52 pmol 2-AG-d8 were added to each sample, followed by 2 ml of ethyl

acetate for extraction. The mixture was vortexed (1 min) and centrifuged

at 1400 g for 10 min. The organic layer (~1.5 ml) was transferred

into a clean glass vial and was dried under a stream of N2. The

residues were reconstituted in 1:1 v/v water: methanol (100 μl). After

filtration (0.1 μm), 10 μl of samples was injected onto an Acquity UPLC

system (Waters, Milford, MA) coupled to a TSQ Quantum Ultra tandem

mass spectrometer equipped with a heated electrospray (H-ESI) source

(Thermo Fisher). Chromatographic separation was carried out using an

Acquity UPLC BEH C18 column (2.1 mmÅ~ 100 mm, 1.7 μm) equipped

with a VanGuard precolumn (2.1 mmÅ~ 5 mm, 1.7 μm) at 40 °C using

column oven. The mobile phases used were water containing 0.1%

acetic acid (A) and methanol containing 0.1% acetic acid (B). Mobile

phase gradient conditions were as follows: hold at 15% A and 85% B for

0.5 min, linear increase of B to 95% in 2 min, hold at 95% B for 4 min,

decrease of B to 5% in 1 min and re-equilibrate for 2 min at the starting

conditions. The overall run time was 10 min and flow rate was 0.2 ml/

min. Eluate from the LC was directly electrosprayed into the mass

spectrometer using an electrospray ionization interface in the positive

mode. MS conditions were set as follows: spray voltage = 3500 V, vaporizer

temperature =350 °C, sheath gas= 25 units, auxiliary gas =2

and capillary temperature = 350 °C. Samples were run in positive

single reaction monitoring (SRM) mode and the following precursor-toproduct

ion transitions were used for quantification: 2-AG, [M+ H]+

m/z 379.2 > 287.1; AEA, [M+ H]+ m/z 348.2 > 287.2; 2-AG-d8,

[M + H]+ m/z 387.2 > 292.3; AEA-d8, [M+H]+ m/z 356.2 > 294.2. Scan time was

0.2 s per SRM, and the scan width was

m/z 0.01. Optimum collision energy and S-lenses conditions were determined

for each compound by using autotune software for each

analyte by post-column infusion of the individual compounds into a

50% A/50% B blend of the mobile phase being pumped at a flow rate of

0.2 ml/min. Xcalibur software was employed for data acquisition and

processing. For quantification, each calibration standard was prepared

ranging from 1 to 1000 nM by fortifying phosphate-buffered saline with

stock standards of 2-AG or AEA prepared in methanol. Quality control

samples were prepared at a concentration of 50 nM for each endocannabinoid

in triplicate. Weighted calibration curves were constructed

using 1/x as a weighting factor for 2-AG and AEA, respectively.

2.11. Plasma esterase activity

Plasma esterase activity was determined using the substrate paranitrophenyl

valerate, as previously described [31]. Production of p-nitrophenol

liberated from pNPV was monitored at 405 nm on a spectrophotometer

[32]. An extinction coefficient of 13 cm−1 mM−1 [33]

was used to convert the slopes of each activity curve to specific enzyme

activities. All enzymatic reaction rates were corrected for non-enzymatic

hydrolysis rates as we have described previously [31].

2.12. Real-time quantitative PCR (qPCR)

qPCR was performed on liver samples as described in [17,34].

Briefly, total liver RNA (20 mg tissue) was isolated using a GeneJET ™

RNA Purification Kit (Thermo Fisher Scientific, Pittsburgh, PA) and

quantified using a Take 3 plate and an Epoch microplate spectrophotometer

(Bio-Tek, Winooski, VT). RNA was converted to cDNA

using qScript cDNA SuperMix (Quanta Bioscience, Gaithersburg, MD)

and a Peltier thermal cycler (Bio-Rad, Hercules, CA). Using 3 ng of

cDNA per sample (with each sample run in duplicate), expression of

peroxisome proliferator-activated receptor alpha (PPARα), peroxisome

proliferator-activated receptor gamma (PPARγ), hepatic fatty acid

transporter (CD36), fatty acid synthase (FAS), acetyl-CoA carboxylase

(ACC), stearoyl-CoA desaturase (SCD), monoacylglycerol lipase (MGL),

cannabinoid receptor type 1 (CB1), cannabinoid receptor type 2 (CB2)

and 18S, was determined by qPCR using mouse-specific primers (Real

Time Primers, Elkins Park, PA) and SYBR Green-based master mix

(Qiagen, Valencia, CA). Amplifications were performed on a Mx3005P

qPCR machine (Stratagene) and treatment differences were calculated

as a fold change by the ΔΔ Ct method with 18S used as a house-keeping

gene (HKG) as previously reported [17,34]. Correlation between the

results from the liver qPCR and the blood lipidome was investigated as

described below.

2.13. Regression analysis between lipid features, GTT/IST, plasma

endocannabinoids, and liver homeostasis

GraphPad Prism for windows version 5.04 (GraphPad Software,

Inc., La Jolla, CA) was used for all correlation analyses comparing the

differential lipid expression via the relative abundance of features to

multiple endpoints used in the study such as liver mRNA expression

values, plasma endocannabinoid levels, and AUC values from GTT/IST.

Reported correlations meet the fairly stringent cutoff correlation coefficient

criteria of R2 ≥0.6.

2.14. Statistical analysis

All statistical analyses were compiled using GraphPad Prism for

windows version 5.04 (GraphPad Software, Inc., La Jolla, CA). For all

analysis, the experimental unit was individual animals and samples

from a total of 6–8 animals/diet/time point were assessed. For all

analyses, significance was set at p ≤ 0.05 where data are expressed as

mean ± SEM based on t-test for pairwise analysis and/or ANOVA

analysis (two-factor repeated-measures with Bonferroni post hoc test).

3. Results

3.1. Morphometric, GTT, IST, and intestinal permeability data

The lipid changes reported below were correlated to glucose tolerance

(GTT) and insulin sensitivity (IST) outcomes, which were recently

reported [17]. Briefly, HFD-fed mice were significantly heavier than

LFD mice at all three time points, and had impaired glucose tolerance.

Interestingly, HFD-feeding had the greatest effect on glucose tolerance

rather than insulin sensitivity [17]. On the other hand, while HFD decreased

IST at the earlier time points, this trend was not seen at the end

of the study due to age-dependent decreases in IST in LFD-fed mice

[17]. HFD-consumption by the female C57BL/6 mice also increased the

gastrointestinal permeability, more so after longer feeding durations

[17].

3.2. Multivariate analysis of lipidome

Multivariate, unsupervised principal component analysis (PCA) of

spectral data comparing high-fat diet (HFD) and low-fat diet (LFD)

consumption showed distinct clustering within the blood lipidome

where diet and age were the major effectors (Fig. 1, Fig. 2). Scores plots

of all groups, for both positive (Fig. 1A) and negative ion mode

(Fig. 2A), demonstrated a striking separation between 6-week vs. 22-

and 36-week treatment groups. The separation of populations occurred

regardless of dietary treatment indicating a significant role for age on

the lipidome. However, HFD-consumption altered lipid profiles within

each respective time point where 6-week treatment (Fig. 1C, Fig. 2C)

had the most pronounced effect. Urine lipid profiles demonstrated a

similar trend (Suppl. Fig. 1). The baseline lipidome was also assessed

with PCA analysis comparing blood lipid profiles of ~5–6-week-old

mice to 6-week HFD/LFD-fed mice, indicating group differences along

with variability in the baseline lipidome, which diminished following

6 weeks of dietary treatment (Suppl. Fig. 4).

Volcano plots identified and ranked potentially important features

based on fold change and statistical significance level for age- (Fig. 3,

Suppl. Fig. 2) and diet-related (Fig. 4, Suppl. Fig. 3) effects. Age-related

pairwise comparisons within dietary treatment groups for short vs.

long/prolonged (6-week vs. 22- and 36-week) (Fig. 3A-3B, Suppl.

Fig. 2A–B) yielded the greatest number of features while 22-week vs.

36-week (Fig. 3C, Suppl. Fig. 2C) resulted in very few. Based on the

number of features altered, diet-related pairwise comparisons of LFDvs.

HFD-consumption indicated that HFD-induced alterations were

most robust following a short consumption period (Fig. 4A) with fewer

alterations following longer periods of exposure (Fig. 4B, C).

3.3. Phospholipid species

Diet- and age-dependent alterations elevated the relative abundances

of phospholipid (PL) species in blood lipid profiles of HFD-fed

mice after 6 weeks of consumption (Fig. 5A), and presented the most

changes following 22 weeks of treatment where HFD-fed mice had

decreased relative abundances of various PL species, differing from

those species affected after 6 weeks on the diet (Fig. 5A). None of the

age−/diet- changes persisted after 36 weeks of HFD feeding (Fig. 5A).

Significance analysis of lipid features was performed for all time

points within each dietary group (i.e. 6 weeks vs. 22 weeks of LFD

feeding) to identify age−/diet- alterations. These age−/diet- features

were subsequently excluded during analysis of LFD vs. HFD treatment

to characterize the effects of HFD feeding alone. Interestingly, the effects

of HFD-consumption alone did not appear following 6 weeks of

feeding (Fig. 6A). 22-week consumption demonstrated the greatest effect,

much like that observed in features altered due to diet−/age-,

although differences in PL relative abundances were bidirectional and

did not show a class-specific uniform trend (Fig. 6A). Following

36 weeks of feeding, the relative abundance of only one feature (m/z

578.3) changed due to HFD-consumption alone (Fig. 6A). MS/MS

analysis and neutral loss scanning was performed to validate phospholipid

class identities (Suppl. Table 1–2).

3.4. Fatty acyl species

Fatty acyl (FA) species altered based on diet- and age-related effects

demonstrated bidirectional effects at all three time points where the

majority of features identified were in blood profiles of mice following

22 weeks of treatment (Fig. 5B). One feature (m/z 562.8) persisted

between short- and long-term feeding, demonstrating an increase in

relative abundance in HFD-fed animals after 6-weeks followed by a

decrease after 22-weeks (Fig. 5B).

Diet alone altered the blood lipidome of the 6-week treatment group

in which HFD-consumption decreased the relative abundances of FAs

by ~2-fold (Fig. 6B). This was also observed in urine lipid profiles of

the 6-week treatment group where three features (m/z 337, m/z 385,

m/z 381) detected and altered in blood were also detected in urine with

comparable magnitudes of differences between LFD- and HFD-fed mice

(Fig. 6B). The effects of HFD-consumption alone did not persist after 22-

weeks and 36-weeks (Fig. 6B). MS/MS analysis and neutral loss scanning

was performed to validate fatty acyl class identities (Suppl.

Table 1–2).

3.5. Glycerolipid species

Most diet- and age-dependent feature alterations in glycerolipids

occurred after short- (6 weeks) and long-term feeding (22 weeks) where

changes were primarily bidirectional across groups (Fig. 5C). 22-week

feeding displayed a trend of general decreases in glycerolipid (GL) relative

abundances in HFD-fed mice (Fig. 5C). We identified one feature

that was altered after 36-weeks of dietary treatment, (m/z 708.6),

which was decreased in HFD-mice (Fig. 5C).

Resembling the pattern observed across FA species (Fig. 6B), the

majority of features altered due to HFD-consumption alone were

identified in the 6-week treatment group. Further, GL features demonstrated

species-specific increases and decreases in both blood and

urine profiles (Fig. 6C). Long-term feeding affected a few GL features

with net decreases in HFD-mice, which were no longer present after

36 weeks (Fig. 6C). MS/MS analysis and neutral loss scanning was

performed to validate glycerolipid class identities (Suppl. Table 1–2).

3.6. Liver endocannabinoids

Given that many of the species altered in HFD-fed mice were

phospholipids containing polyunsaturated fatty acids (PUFAs), and

because many of the fatty acyls correlated to derivatives of fatty acids,

we focused on arachidonic acid-containing metabolites, including 2-

arachidonoylglycerol (2-AG) and N-arachidonoylethanolamide (AEA).

Upon release, these endocannabinoids target cannabinoid receptors

(CB1 and CB2) and work together to play a role in energy homeostasis

[22]. Liver 2-AG levels decreased after 6 and 36 weeks of HFDconsumption

(Fig. 7A). Interestingly, an age-related effect was observed

where 36 weeks of HFD-feeding decreased liver 2-AG levels (Fig. 7A).

The decrease of 2-AG levels on the liver after 36 weeks was accompanied

by a significant increase of AEA (Fig. 7B). The only significant

effect of HFD on liver AEA levels was a significant decrease after

6 weeks on the diet (Fig. 7B). CB1 expression did not change while CB2

expression showed time-dependent increases in HFD-fed mice, although

these were not significant (Fig. 7C).

3.7. Plasma endocannabinoids

Plasma levels of 2-AG and AEA in the LFD-fed mice were quite

stable, apart from a slight increase after 36 weeks of feeding (Fig. 8). In

contrast, HFD-consumption increased plasma levels of 2-AG, but the

effects were bi-phasic where a significant increase was only observed at

6 and 36 weeks, but not after 22 weeks of HFD-consumption (Fig. 8A).

There also appeared to be an effect of age on plasma 2-AG within the

HFD-fed mice as shown by an increase in 2-AG levels following

36 weeks compared to 22 weeks of feeding (Fig. 8A). The effects of

HFD-consumption on the other endocannabinoid, AEA, resembled the

diet's effects on plasma 2-AG, with the only significant effect being an

increase of plasma AEA levels after 36 weeks on HFD (Fig. 8B).

3.8. Plasma esterase activity

The increase in fatty acyls and lysophospholipid species suggests

increased esterase activity. This hypothesis was addressed by assessing

esterase activity at all time points in the plasma. HFD-consumption

increased plasma esterase activity at all time points, with the most

pronounced effect following 6 weeks of feeding (Fig. 9). Although

plasma esterase activity increased following both 22 and 36 weeks, the

effect at 36 weeks was not significant (Fig. 9). It appeared that age

alone affected plasma esterase activity, indicated by heightened esterase

activity after 6 weeks of HFD-consumption followed by lower

levels at later time points (Fig. 9).

3.9. Liver qPCR data of key lipid homeostasis genes

There are many genes that encode for protein regulating energy

balance and lipid metabolism, including peroxisome proliferator-activated

receptors (PPARs) [35,36]. PPARα is best known for its major

role in lipid and lipoprotein metabolism while PPARγ is involved in

adipogenesis and insulin sensitivity [35,37]. HFD-consumption increased

expression of liver mRNA levels of PPARα, PPARγ, and CD36, a

known target of PPARγ [38], significantly at 6 weeks of HFD (Fig. 10A).

Interestingly, PPARα and PPARγ levels increased after 36 weeks of

HFD-consumption, but not after 22 weeks. Liver CD36 mRNA was elevated

at all three time points (Fig. 10A), with the magnitude of elevation

greatest after 36 weeks on HFD.

PPARγ has been shown to increase during lipogenesis, thus we assessed

the expression of fatty acid synthase (FAS), acetyl-CoA carboxylase

(ACC), and stearoyl-CoA desaturase (SCD) (Fig. 10B). HFD-fed

mice demonstrated a numerical trend of time-dependent increases

although not significant, in SCD, ACC, and FAS expression (Fig. 10B).

MGL demonstrated increased expression in mice fed HFD for 6 weeks;

however, this effect was tapered after long- and prolonged-feeding

(Fig. 10B).

3.10. Correlation between lipids, GTT/IST, plasma endocannabinoids, liver

homeostasis

We performed regression analysis to determine if changes in the

blood lipidome correlated to changes in GTT or IST, plasma endocannabinoid

levels, or gene expression of liver homeostasis markers.

Table 1A lists correlations (R2 ≥0.6) between several lipid features and

the expression of CD36, PPARα, and PPARγ mRNA in the liver. There

were only a few lipid species whose abundance correlated to changes in

these genes, including two unidentifiable FA species. Correlations occurred

at all time points with increases in TG (52:4 or 52:5) [PPARγ] in

HFD-mice at 6 weeks, decreases of DG (40:6) [CD36], PE (42:7)

[CD36], and FA (m/z 562.8) [PPARα] at 22 weeks, and an increase in

FA (m/z 438.8) [PPARα] at 36 weeks (Fig. 11). Lipid species indicated

by only their m/z value were unable to be fully characterized by subsequent

MS/MS analysis.

Regression analyses also demonstrated quite a few relationships

between changes in select blood lipids and changes in AUC from glucose

tolerance and insulin sensitivity tests after 6 weeks (Table 1B).

Fewer correlations were identified for changes in the blood lipidome at

22 or 36 weeks (Suppl. Table 3), which is not surprising given the robust

differences observed after 6-weeks of treatment. Interestingly,

several lipids such as PC (44:3) and DG (34:3), which correlated to

glucose tolerance, also demonstrated inverse correlates to insulin sensitivity.

Regarding plasma endocannabinoids, GL species (m/z 734.6, m/z

760.5) and FA species (m/z 353.2) correlated to changes in in 2-AG

exclusively within HFD-mice, whereas AEA did not correlate to any

lipid features (Table 1C).

4. Discussion

The association between increased high-fat consumption and excess

adiposity poses a major global health problem that heightens the risk of

metabolic disorders, diabetes, heart disease, fatty liver, and some forms

of cancer [39]. Growing evidence implicating a role for impaired lipid

metabolism, coupled with the advent of bioinformatics tools has

prompted efforts in characterizing the obese lipidome [40–42]. While

these studies have highlighted alterations in the plasma and/or serum

lipidome, there have been few studies examining the effects of age on

the lipidome and/or the interaction between age and obesity. Further,

the number of studies that address diet-induced obesity within female

models is quite limited, although an increased linear trend of female

obesity among US women within the last decade demonstrates the

significant need [3,43,44]. Here, we used shotgun lipidomics to assess

the effects of dietary fat consumption, age, and their interaction at the

level of the blood lipidome. We correlated changes in the blood lipidome

to changes in metabolic regulation, endocannabinoid levels, and

plasma esterase activity.

One of the most interesting findings of this study is that the effect of

age superseded the effect of HFD with regard to alterations in the blood

lipidome. These data emphasize the need to characterize and stratify

lipidomic alterations not only to diet, but also to age. The accentuated

effect of age on the lipidome between 12-week-old (6-weeks on the

diet) vs. 28-week-old (22 weeks on the diet) and 42-week-old (36 weeks

on the diet) mice indicated distinct shifts in lipid composition and/or

regulation, an interesting note since all ages fall within the mature adult

phase of C57BL/6 mice. Although multivariate analysis indicated a

slight difference between 28- and 42-week-old animals, the separation

was not as robust. With regard to dietary treatment, younger animals

presented striking lipidomic responses to HFD-consumption while older

animals had a more tapered shift, possibly a result of time-dependent

homeostatic mechanisms in response from long-term feeding. In this

regard, as we reported recently, it is interesting to note that in terms of

insulin sensitivity, but not glucose tolerance, age appears to be a major

driver of decreased insulin sensitivity to the point that the effects of

prolonged HFD feeding are overpowered by the effects of age [17].

Given the limitations of the shotgun approach along with the sheer

number of features identified in this study, we report general lipidomic

changes in terms of lipid class with alterations falling into three classes:

glycerolipids (GL), fatty acyls (FA), and phospholipids (PL). MS/MS

analysis was employed to verify these lipid species (Suppl. Table 1).

Features changing at only a few time points showed a class-specific

trend in terms of differential lipid expression; however, most changes

indicated species-specific alterations.

A few studies have demonstrated alterations in lipid classes within

rodent obesity where the majority of reported changes encompass

ceramides, cholesterols, triglycerides, and phospholipids. For example,

changes in lipid species such as lysophosphatidylcholines have been

associated with obesity, insulin resistance, and type 2 diabetes [45–47].

In agreement with some of these studies, we report elevations in PC

(38:5), PC (44:3), and PC (38:3) in the blood of HFD-fed mice, also

reported in Eisinger et al. [40]. Changes in GL and FA reported in the

current study are not in line with another study [48], but it should be

noted that the sex of the mice, feeding durations, and importantly, the

dietary composition in [48] and our study are different. It should also

be pointed out that several studies have demonstrated that HFD typically

increases the levels of TAGs and DAGs [49]. The fact that not

many of these lipids were detected in our own study is most likely a

limitation of the shotgun approach used.

As mentioned above, lipidomic alterations identified by shotgun

analysis revolved around three major lipid classes (GL, FA, and PL),

which also happen to constitute the endocannabinoid system.

Quantification of both 2-AG and AEA demonstrated several correlations

with specific blood lipids. AEA and 2-AG, both derivatives of arachidonic

acid, are signaling lipids that mediate their action via activation

of cannabinoid receptors. Further, changes in plasma levels of 2-AG and

AEA after 6 and 36 weeks associated with decreases in liver 2-AG. This

suggests that increased 2-AG levels in circulation are due to increased

mobilization from the liver, with/without concomitant decreased 2-AG

breakdown [50]. In a study by Caraceni, et al., 2-AG levels were reported

to be higher in the hepatic veins of cirrhosis patients when

compared to peripheral blood, supporting the hypothesis that the liver

contributes to circulating 2-AG levels [51]. This may also suggest that

the source of the increase in plasma 2-AG is non-hepatic, i.e. dietary

[20].

Plasma AEA levels were most affected by the diet after 36 weeks;

however, as opposed to the decreases of liver 2-AG, liver AEA was increased

after 36 weeks irrespective of diet. Together, these data indicate

that in female C57BL/6 mice plasma 2-AG is more sensitive to

HFD-consumption. Further, similar to other endpoints in this study, the

effects of HFD diet are most prominent during early (6 weeks) and late

phases (36 weeks) of the feeding trial. Circulating endocannabinoids

also appear to be more sensitive to non-dietary liver pathology than

their liver levels. For example, in certain conditions, i.e. hepatitis C,

plasma, but not liver, 2-AG was increased [52]. On the other hand,

circulating AEA was significantly higher in cirrhotic patients [51].

A novel finding is the fact that age was associated with decreased 2-

AG and increased AEA in the liver. At the end of the study (36 weeks on

the diet, 42-43-week-old), the mice were middle-aged and close to

becoming reproductively senescent [53]. While liver-specific

endocannabinoid data for female C57BL/6 mice within the context of

age are lacking, it is interesting that a recent study reported decreased

2-AG, but not AEA, levels in the hippocampus of aged mice [54]. In this

study, the decrease in hippocampal 2-AG was attributed to a concomitant

decrease of local 2-AG synthesis and increase of its breakdown

[54]. With that in mind, our data suggest that the age-dependent metabolic

changes in the 2-AG pathway that operate in the brain (hippocampus)

do so in the liver as well. In addition, the increase in liver AEA

levels at the end of the study highlights endocannabinoid metabolitespecific

effects of age. Although Osei-Hyiaman, et al. previously showed

that liver AEA levels increase following HFD, our data indicate lower

levels of hepatic AEA in HFD-fed mice [55,56]. However, it is important

to note that the dietary feeding regimen in Osei-Hyiaman, et al. was

initiated in slightly older mice, which may play a large role given our

data demonstrating the interaction of age and diet. Moreover, the

findings in the above study were based on the use of a combination of

male and female mice and thus, does not reflect the effects within females

alone. This would not be surprising since gender-specific responses

to HFD-intake have been reported in rodents, and it has even

been suggested that females are more susceptible to developing the

secondary effects of HFD-induced obesity [57]. It should also be mentioned

that adipose distribution and function differs across males and

females, so differences are expected to exist in mediators produced by

adipocytes such as endocannabinoids [57–59].

The increased circulating 2-AG levels in the female C57BL/6 mice in

our study are in line with multiple studies in obese human subjects. For

example, obese men, especially those with increased intra-abdominal

adiposity, have increased plasma 2-AG [60]. Interestingly, direct correlation

between plasma endocannabinoids, insulin resistance and

dyslipidemia has been suggested [61]. Moreover, chronic cannabinoid

receptor 1 (CB1) stimulation exacerbates the metabolic dysregulation

caused by HFD-consumption, suggesting key role for the endogenous

endocannabinoids in the process [62]. Circulating endocannabinoids

might contribute to obesity by their central and/or peripheral actions

[20]. The key role of CB1-specific over-activation by excessive endocannabinoids

is further emphasized by the fact that global CB1−/−

mice are not susceptible to HFD and liver-specific CB1−/− mice are

protected from some, but not all, of the adverse effects of HFD intake

[63]. Treatments aimed at reducing plasma endocannabinoids are

beneficial in re-balancing the metabolic dysregulation [64] and obesityrelated

inflammation [65], but not for reducing the body weight in

obese subjects [64]. Interestingly, and in line with our current data,

circulating 2-AG levels were significantly elevated in insulin-resistant

obese women [66].

Peroxisome proliferator-activated receptors (PPARs) are members

of the steroid hormone receptor superfamily of nuclear transcription

factors that are involved in the regulation of various genes encoding

proteins involved in energy balance and lipid metabolism [35,36].

PPARα is best known for its major role in lipid and lipoprotein metabolism

while PPARγ is involved in adipogenesis and insulin sensitivity

[35,37]. It has also been suggested that hepatic PPARγ may mediate the

accumulation of fat via the regulation of genes essential for de novo

lipogenesis, i.e. fatty acid synthase (FAS), acetyl-CoA carboxylase (ACC),

and stearoyl-CoA desaturase (SCD) [67]. Further, it is possible that

glycerolipid alterations can be attributed to changes in lipolysis via

monoacylglycerol lipase (MGL), one of the main lipases involved in the

catabolism of TG.

In our study, we also observed HFD-induced increases in liver

PPARα, PPARγ, and CD36. The increases in PPARα and PPARγ were

biphasic (6 and 22 weeks) and more prominent after 6 weeks of HFD

feeding, indicating that the activation of the PPAR pathways are timedependent.

PPARα regulates fatty acid β-oxidation, is activated by the

AEA analogue oleoylethanolamide (OEA), and pharmacological increases

of OEA are beneficial to diet-induced obese mice [68]. While we

have not measured OEA in our study, it is conceivable that the lack of

significant increase in liver PPARα after 22 weeks of HDF feeding was

due to sensitization-dependent downregulation. Liver PPARγ, which

showed similar kinetics to PPARα, regulates lipid and glucose homeostasis,

is elevated by a HFD [69], and approaches aimed at curbing its

activation are beneficial in obesity and type 2 diabetes [70]. Together,

the changes in liver PPAR levels are in line with the time-dependent

metabolic dysregulation of these mice, especially the sensitivity to insulin

challenge, and reflects the changes in the blood lipidome reported

here [17].

CD36 was the sole lipid homeostasis/inflammation molecule whose

expression was increased by HFD throughout the study. Increases in

CD36 were greatest at the end (36 weeks) of the experiment. CD36 is

associated with obesity, diabetes, and liver dysfunction; hence, our

findings are not surprising. Liver CD36 was previously shown to increase

due to aging [71]. Thus, the marked increase in liver CD36 at the

end of the feeding duration could be a sum of the effects of age and HFD

on its expression or, due to increased demand for hepatic lipid uptake in

the face of continued HFD-consumption. In this regard, CD36 plays a

major role on hepatic fat uptake [71].

Circulating esterases are predominantly investigated for their role in

the metabolism of drugs and toxicants, though they also metabolize

endogenous, i.e. dietary, and exogenous lipids [72]. The early robust

increase in plasma esterase that we observed in the current study might

be the result of host's attempt to maintain lipid homeostasis in the face

of excessive dietary fat. Over time, the increases in esterase activity

were still present, but less robust. This may reflect saturation of this

likely protective mechanism. In support of this hypothesis, esterasedeficient

mice are not only more susceptible to pesticides that are detoxified

by it, but also to diet-induced metabolic dysregulation and

atherosclerosis [73]. Our data indicating an age effect, i.e. decreased

plasma esterase due to age, further supports the notion that this mechanism

of metabolizing excessive dietary fat is less robust in older

mice.

In conclusion, we demonstrated an interaction between dietary fat

consumption and aging with widespread effects on the blood lipidome

in female mice. This study indicates that the effects of HFD feeding

occur in an age-dependent manner with robust responses at a younger

age. Further, we identified several associations between lipids and

metabolic and liver regulation, providing a basis for female-specific

obesity- and age-related lipid biomarkers. These findings highlight the

need for additional age-dependent tracking studies, prior to sexual

maturity into advanced age, to obtain comprehensive understanding of

the evolving lipidome with regard to dietary changes.

Please read the article and answer about questions.

International Strategies

When you are struggling to get through that first year of business, international sales are about the last thing on your mind. The U.S. Department of Commerce, however, indicates that large compa- nies account for only about 4 percent of all exporters, meaning the other 98 percent of the exporters in 2010 were small businesses.22

Entrepreneurs typically fall into three categories. There are those who realistically will never go international (for example, a restaurant owner or dry cleaner working from a single site). There are those who intentionally start international businesses23 (for example, import-export businesses), such as Peter P., the director of procurement for a Russian trading company, who saw a trading opportunity with the opening up of Eastern Europe and the former Soviet Union. Educated in the United States, he is of Ukrainian descent and speaks both Russian and Ukrainian.24 Last, there are those who think international business might be something they’ll do someday way off in the future. This section is primarily for the last two categories.

Thanks to the Internet, once a company has a website, it is essentially an international business, a whole new breed of firms known as born internationals.25 Potential foreign customers see the website and before you know it—or before you are prepared—the first international order rolls in. Even “website–free” companies aren’t exempt. A foreign visitor comes across your products and sees a need for it in his or her country, and here comes that order.

Some international orders aren’t all that difficult to handle. If the order is small enough, if the product or service is not highly regulated domestically, and if the country is one with which the United States has rather liberal trade such as Canada, the order processing may offer few or no head- aches. The customer may use a credit card or international money order, and the product ships in the mail without much more effort than figuring the extra postage. That’s okay for the occasional order, but more complex situations will require more time and effort on the part of the entrepreneur. The ideal situation is to consider and prepare an international strategy before it becomes a hit-and-miss method that is too cumbersome or before serious and costly mistakes are made.

Entrepreneurs have available to them the same options as large companies including wholly owned subsidiaries, joint ventures, licensing, franchising, and exporting. For most, though, an ex- port strategy is sufficient and is all that is covered in this section. It’s usually inexpensive, quick to start, easy to change, and less risky than other ventures. It has the additional advantage of allowing the entrepreneur the opportunity to learn about doing business abroad in case the company reaches the point of moving further. For U.S. entrepreneurs, the U.S. government offers detailed and useful help for exporters, including seminars and other training, export assistance, websites and reports, financing, insurance, and legal and collection assistance.

Putting together an export strategy involves answering three questions:

1. Are we ready? 2. Where should we go? 3. Whom do we contact over there?

There are many sources for assistance in answering these questions, and many good ones are free or almost free. One excellent resource is the U.S. Commerce Department’s report entitled, “A Basic Guide to Exporting” that can be found at www.unzco.com/basicguide.

Question 1: Are you ready to export? Exporting requires a different kind of thinking and prepa- ration from selling locally or even nationally. Are you going to target one country, a region, or the whole world? Do you know what customers want? Do you know what the import requirements are? What aspects will you handle, and which ones will you contract out? Are you ready for the costs and headaches of exporting? To see how you are coming along, you can check your readiness online at the U.S. government’s exporting site www.export.gov/, which provides extensive exporting basics, including a “readiness test” at www.fas.usda.gov/agexport/exporttest.asp.

Consider your product as well. Will your U.S. designed product fit an international lifestyle or needs? Clothing sizes are different—both in how they are numbered and what the sizes mean. A woman’s medium in the United States is an XXL in mainland China. Electrical currents are

different, as are various other safety and product standards, and the United States is one of only two countries that’s not on the metric system.27

There are several ways you can export. One is to use online services such as eBay. Approximately one-fifth of eBay’s sales are out of country.28 If you’re handling your international business this way, a lot of the rest of this section isn’t really for you until you want or need to change methods. Another is to work from personal contacts gained through school, travel, or family. Most exporting small businesses start with countries where they have had personal experience or support.29 These two methods are called direct exporting, since you are selling directly to foreign buyers or distributors.

If you want to use outside experts, there are three intermediaries who can help. With indirect exporting, you use agents, export management companies, or export trading companies as inter- mediaries to handle most of the exporting process. Direct exporters can also get help from freight forwarders. Freight forwarders are specialists in export-related activities including tariff sched- ules, shipping, insurance, packing, transportation arrangements, customs clearing, and other export details. (By the way, many agents, export management companies, export trading companies, and freight forwarders are themselves small businesses. They know exactly what problems you’ve faced and are much easier to approach than some megacompany.) The Small Business Administration’s Export Assistance Center can help you find one.

Question 2: Where should we go? The United Nations has 193 member countries in the world; chances are not all of them are right for your product. Even if your product should have wide appeal, it makes good sense to pick one or two as first markets. One of the safest bets is to consider countries that are similar to the United States—Canada, United Kingdom, Australia, for example. In those countries, you have few language issues, the culture is pretty close, the governments and economies are stable, and the people there are likely to want or need about the same kinds of products as people in the United States do. Should you decide to go further afield, those are the same sorts of things you want to look for—language and culture issues, government and legal situations, economic situ- ations, and peoples’ wants and needs. Here’s a good time to use those personal contacts mentioned earlier; if they live there, they are likely to be able to tell you if the product makes sense or not.

International marketing research isn’t cheap and can be difficult to do. Contacts are a valuable resource. Additionally, the U.S. government and world trade centers can give a lot of free or low-cost assistance. See Table 11.1 for a list of some of the major ones.

TABLE 11.1   Sources of Export Assistanc Question 3: Whom do we contact over there? You may already have international contacts through school, friends, travel, or other methods. If so, you’re ahead of the game. Even if they cannot help you with specific questions, they probably know someone who can. On the other hand, if you do not have any contacts, a lot of the government services you have already used can provide lists of poten- tial intermediaries or end users. In addition to the free services available, U.S. Commercial Services (www.export.gov) provides a number of levels of fee-based customized services. For $500, they offer their International Partner Search service, which will identify up to five potential businesses to work with you as licensees, agents, distributors, or strategic partners, and prequalify them based on your criteria. The government’s www.export.gov site offers a database of sales leads that can be searched for free by industry, region, or country. They also are the point of contact for catalog exhi- bitions which can get your product or service catalogs into the hands of potential buyers in specific markets (or at specific trade shows) overseas.

Other good ways to make international contacts are to participate in trade shows and trade mis- sions. In a trade mission, a U.S. government official takes a small group of business owners to dif- ferent foreign countries in order to help establish relationships and promote exporting. There are not a lot of these missions, and they are usually specific to a particular type of business and region of the world, so they are not always appropriate. At an international trade fair, similar to domestic trade fairs, you have a booth displaying your products or services and the opportunity for exposure to thousands of potential clients. Again, some fairs are industry-specific, while others are more general. The U.S. government often has a U.S. pavilion featuring export-oriented companies. These companies often have the opportunity to tie into other U.S. government services such as meeting with local U.S. embassy officials, prearranged meetings with qualified customers, market research information, trade barrier information, transportation and customs information, and assistance and access to U.S. trade show experts. Even if you can’t exhibit in the fair, attending the fair may give you a chance to meet the sort of people you need to know.

The U.S. government through www.export.gov also provides such services as printed and video catalogs, online databases, and personalized (fee-based) contact services. The U.S. Commercial Service will also assist a company in arranging private promotional activities, including exhibitions, press releases, and receptions when appropriate.

Still another way is to look for foreign companies with a resident representative in the United States, a type of private importing agent. Often these representatives are interested in bringing U.S. products back to their home countries and will already have a good idea if your product is right, and how to promote and distribute it.30 To find these resident representatives, try a Google search with the terms “resident representative” US, importing-site:.gov.

The next step is to export your products. But there are a few other things to consider first. Pricing becomes complicated as you need to cover transportation, the additional documents you may need,

possible tariffs (taxes on incoming goods), potential currency valuation changes, the cost of convert- ing currencies, and the additional packaging necessary to ship abroad. The importer usually covers foreign taxes, tariffs, additional shipping charges, port handling fees, and the like, but this must be carefully spelled out in your contracts in order to avoid potential differences of opinion.

Shipping documentation and other paperwork are very specific to the product and the country to which it is going. The International Trade Administration (www.ita.doc.gov) provides extensive information about tariffs, taxes, specific country information, and other general exporting informa- tion. The U.S. Country Commercial guides also provide some assistance in this area, as do some country government websites. The Bureau of Export Administration (www.bxa.doc.gov) provides information about when export licensing is necessary and also information on exporting of politi- cally sensitive products. In addition, companies such as NetShip (www.netship.net) have arisen specifically to handle shipping and documentation issues for e-commerce.31

There are a variety of payment procedures available. The easiest for you is to require up-front cash payment prior to shipment (or credit card if appropriate). This eliminates your risk, but puts the customer at risk. Providing credit to your customers reverses the risk, and puts it all on you. Both of these are possible methods of receiving payments, but less often used. More typical meth- ods include letters of credit or documentary drafts. In both cases, the payment procedure now includes four parties—you, your customer, and both of your banks—and payments are made upon proper presentation of certain documents, including the letter of credit or draft, bills of lading, and other paperwork. Although the system is somewhat complex, it provides a lower level of risk for all parties than cash in advance or an open account. You can find assistance about these methods at your current bank.

Financing and insurance become important because of the length of time it may take for interna- tional payments to be processed and the risk of default, as well as the difficulty of recovery in case of default in international transactions. The Small Business Administration (www.sba.gov), the Ex-Im Bank (www.exim.gov), and the Overseas Private Investment Corporation (www.opic.gov) provide loans and insurance to cover exporting. In some cases, these loans may also be used to finance trade show participation, to translate brochures and catalogs for international distribution, to renovate or expand existing facilities necessary to produce products for export, to set up lines of credit for potential customers, to provide export working capital, and to provide funding for developing an export program.

Last is the consideration of conflict resolution. Although the possibility exists for pirating, prod- uct misuse, and other unfortunate occurrences, the primary areas for conflict resolution include nonpayment and contract default issues. There is no universal court of law that can handle these situations. The U.S. Department of Commerce can provide advice and offer reputable local coun- sel, but only for sizable losses, typically several thousand dollars or more. The U.S. Council of the International Chamber of Commerce (www.iccwbo.org) provides international arbitration services and offers some other suggestions, but arbitration, too, is costly and probably not worthwhile un- less the loss is significant. This difficulty in international dispute resolution underscores the need to carefully select partners and to do a thorough job of prescreening. This is an area in which various government agencies can help you. The U.S. Commerce Department, for example, prequalifies potential customers in many cases prior to recommending them; you should check the particular program specifics to verify. Ex-Im Bank provides credit information on potential customers and, as mentioned earlier, many agencies provide insurance for export payments.

Importing

Importing strategy is similar to exporting, but with the buyers and sellers reversed. Instead of cus- tomers to buy your products, you are looking for sources to sell products to you (which, of course, you’ll eventually resell). If you have the opportunity to travel abroad, look for products that are selling well in the country you’re visiting and aren’t available in the United States or products that are considerably cheaper than similar ones found in the United States (labor and manufacturing costs are often cheaper in other countries than the United States). Trade mission and domestic and international trade shows are also good sources. If you can’t travel, ask your international contacts for this information. Next, find out who manufactures them and write the manufacturer a letter, introducing yourself and your company and the potential you see in your market for its product. You’re selling yourself, so be sure to tell the producer why you are the best person or company to be representing the product (i.e., experience in that product or in importing, contacts and distribution systems already in place, familiarity with the market, etc.). International mail can be painfully slow, so a fax or e-mail letter is probably best. Also, avoid slang terms (e.g. “your product is da bomb!”) and idiomatic expressions (like “break the ice”) that are likely to be misunderstood. Since English is rapidly becoming the language of business, a translation is usually not necessary. Follow up with a phone call or visit in which you can pitch the specifics of your marketing plan for the product.32 One way to conduct international calls for free is to register for Skype, an Internet service which lets you use a broadband connected computer to call other Skype users for free (www.skype.com). If you and the overseas company both use Skype, having long conversations to get an understanding of each other will not pose a financial problem. Along the same lines, it is worthwhile these days to check to find out whether an overseas company has video capabilities. Video cameras for PCs are inexpensive, and videoconferencing services are often available on campuses or at commercial locations such as FedEx Kinko’s for low costs.

With importing, many of the paperwork and insurance details will be your source’s responsibil- ity. Import buying works the same way as export selling, that is, the same sorts of paperwork and procedures are followed only in reverse.

Concluding Thoughts on International Business

One of the major mistakes commonly made by U.S. businesspeople (entrepreneurs or major compa- nies) is being insensitive to cultural differences. You’re likely to make some mistakes, but take time to learn at least the basics about the culture you’re dealing with to avoid the biggest errors. Travel guides and U.S. government country reports often offer brief cultural assistance as do books such as Kiss, Bow or Shake Hands and a plethora of “doing business in ———” guides.33

Although international business might seem a little daunting with all the paperwork and regula- tions, small businesses just like yours do it every day. There’s a lot of free or very inexpensive help out there; make use of it.

Location

When you ask real estate agents the best three things to look for in a house, they will tell you, “Loca- tion, location, location.” The same holds true in your business. What location—in particular, good location—means for your business is highly dependent on what your business is, the amount of money you can afford to budget for it, your particular business philosophy, and the marketing niche you are seeking. Let’s start with some general information about location, then move onto specific issues for services (including retailing) and manufacturing businesses. We then discuss some spe- cific choices such as site selection and layout and the buy, build, or lease option.

The first choice, and often only choice, for many entrepreneurs is their hometown because it offers convenience and a familiar setting, and it eliminates a lot of possible family issues. There may also be valid business reasons for this choice: The local banker knows you and is more likely to loan you money; you know your market—the potential customers in the area—and understand their wants and needs; you have seen an unmet need that you can fill; and, for many entrepreneurs, friends and family (usually local) are often the first customers and are great at spreading the word about your business. (Remember that word-of-mouth is often the first method of getting to your customers.)

There may also be some compelling reasons to consider a different location. What are the busi- ness laws like in your area? Local zoning ordinances specify what sorts of businesses are allowed and not allowed in specific locations.34 Certain types of businesses—usually those deemed hazard- ous or that produce foul odors—may be banned or severely restricted. State and local pollution standards, worker’s compensation, wage rates, and other such legislation might increase the cost of doing business to the point that other locations become much more favorable. State and local taxes in particular vary considerably from state to state. For example, Wyoming has no personal or corpo- rate income tax, while California has relatively high rates. On the other hand, certain locations often offer attractive incentives for new businesses ranging from tax credits to low-interest loans, from favorable business laws to business incubators (discussed later). Most of this information can be found on the Internet. Try the state or city business development office (a good place to start is the Federation of Tax Administrators’ state list at www.taxadmin .org/fta/rate/tax_stru.html) or the local chambers of commerce (look in the phone book or at www.uschamber.com/chambers/directory/default to find your local Chamber affiliate). There is also information by state available for your state at business.USA.gov, and the Small Business Ad- ministration offers links to state-based resources at www.sba.gov/category/navigation-structure/ counseling-training. Site Selection Magazine’s website (www.siteselection.com) has a number of tools that can help you find the right location. Many of these require being a registered user, but registration is free.

Other reasons to consider other locations are tied to your customer. Your hometown may not be the best place for you to find your target market customers. Are you close to the people who will use your product or service? Other considerations include population growth or decline (especially in your target sector), income levels, and predicted increases or decreases in income. Is the location expanding economically or slowly dying? Perhaps the best source for this infor- mation is the US Census Bureau. State and local municipality business development offices may also carry such information, but they are likely to be slanted toward attracting new businesses. Being positioned to benefit your customer can also be key. Zappos’s primary distribution hub was placed in Louisville, Kentucky, to be close to a major UPS air cargo hub in order to speed delivery.

Also consider the type of business you are planning. Do you need skilled labor? If so, what areas will provide you with the necessary employees? Do you need to be near raw materials or particular methods of transportation? These issues will help determine your choices. Where are your competitors? Certain industries tend to be clustered in certain regions where they can make efficient use of services and employees. Think of California’s Silicon Valley or the financial district of New York City.

Doing business in your hometown may be perfectly appropriate; however, the cost of moving a company—whether across town or across the country—can be very expensive. It pays to plan ahead.

Service Firms

There are three typical locations for services: at the client’s location, at a mutually accessible loca- tion, and at your firm’s location. Traditionally, services may have been tied to one or another of these, but marketing niches have been carved out by people daring to be different. Typically, dry cleaning and restaurant dining are services provided at a place accessible to both parties, but some dry cleaners now offer pick up and delivery from the client’s home, and not only pizza restaurants offer delivery these days. Thanks to the Internet, video rental like Netflix.com and other services are handled electronically, and the customer and service provider may never meet face to face. Whatever innovative niche you select, there are a few things to keep in mind.

At the Client’s Location

Typically, these services include things such as house or office cleaning, pest control, remodeling, lawn and gardening services, carpet cleaning, and similar services which must be performed at the client’s location. Business headquarters can be a home office with enough room to store and maintain any necessary equipment used in the service. Reliable transportation, preferably modified to organize and store tools efficiently, is imperative. More importantly, the range of your client’s locations must be planned to prevent transportation times from being unmanageable. For example, facing a one-hour drive to a client’s location might mean you have tied up two hours in commuting. If you cannot charge for travel and do not have other clients nearby, it means you have two hours in which you cannot make any money that day.

If you’ve done your homework carefully, you already know the geographic area(s) most likely to use your service. Certain services may be organized into a rotating schedule. For example, a house cleaning service may clean a certain set of neighborhoods on Monday, different set on Tuesday, and so on. In other cases, more remote clients may be charged a transportation fee. In some cases, a mile- age fee may be appropriate for your business (delivery services, for example).

As the firm grows, it may outgrow its home-based headquarters. As your clients seldom, if ever, visit you, you have more latitude in where you can be located and the ability to seek out low-cost space (see site selection section below). Reasonable distance to the clients and adequate storage room for your expanded fleet and equipment are key to choosing a site.

Mutually Accessible Location

Services using this approach often have too much specialized equipment to be readily transported and a need for at least some client involvement. Barbershops, dentist offices, video rental stores, and restaurants are services typically located at a site that is extremely convenient for the client and reasonably so for the owner and employees.

Even though your service may be traditionally located in a mutually accessible area, consider what you might do to make it home-based (see Chapter 5). Your watch repair shop might generate clientele by being located in a shopping center, but will the added sales be offset by the high cost of rent, utilities, insurance, and other payments? Can you offer pickup and delivery and do the work at

home? Your restaurant idea might work as a catering service. Instead of a specialized clothing shop, why not try mail order or Internet-based sales?

Remote Location

In this type of service, face-to-face meetings with the client are infrequent. Typical services that meet this criterion include medical transcription, data processing, fulfillment centers, and some consulting work. These services generally are ideal for home-based businesses. Certain services, for example, fulfillment centers, generally take more space—at the minimum, an attic, garage, or basement. The biggest advantage of these sorts of businesses is that they can be located anywhere in the world. U.S. hospitals, for example, use medical transcription services located in India. One such company, Infoflow/TSVI, operates from a U.S. sales base (which makes handling calls from U.S. hospitals easier) with transcription being done in India, managed there by a co-owner, who is a cousin and long-time friend of one of the two American owners (www.tsviinc.com).35 Other than perhaps some initial sales meetings, all business is transacted via phone, fax, electronic exchanges, or mail.

Manufacturers

What if you are selling a product and not a service? What are your considerations about location now? Where you make the product is really dependent on the product. Some products that do not require a lot of specialized or bulky equipment can be produced at home unless zoning ordinances forbid it. In addition to whatever office space is needed, adequate work space is also required. The basement, a garage, or a home workshop may be adequate for some time. As business expands and as you add employees, it will become awkward if not illegal to continue production at home (see Chapter 5).

Some products require bulky and specialized equipment, utility demands atypical of homes, or sizable warehousing requirements and are never suitable for home businesses. Certain production characteristics—for example, use of hazardous materials and materials with strong odors or noisy operations—may make a home-based business undesirable. Many cities have zoning ordinances prohibiting manufacturing in residential areas. When you start to hire employees, providing the amenities they will expect or that are required by law will usually require moving the business from your home.

Contract manufacturing might be a better option, at least for awhile. In this case, a firm with the capabilities to produce your product is contracted to manufacture it for you, usually for a flat per unit fee and a possible setup charge. Some firms will also assist in marketing and sales as well. Trade magazines in your field often list ads for contract manufacturers. An interesting possibility here is to use sheltered workshops to perform light manufacturing or assembly sorts of businesses. These workshops exist in nearly every state and offer very competitive pricing, often including tax benefits for the business.

Site Selection

Once you have determined the general location of your business, you need to determine the exact location for your operation. What you should look for falls into three main categories: home-based businesses (covered in Chapter 5), high customer contact (e.g., retail), and low customer contact (e.g., manufacturing). Each has certain criteria to be considered.

High Customer Contact Businesses

Businesses with high customer contact include such diverse operations as medical or legal offices, restaurants, retail establishments, dry cleaners, and other businesses that are highly dependent on being convenient to the customer. For these operations, there are three critical site selection consid- erations: traffic, customer ease, and competition.36

First of all, you want a site that is convenient to your target market and to enough of the customers to make you profitable. If you are considering a franchise, many will offer site criteria to help you make your selection. If you are going it alone, consider the population density of the area and how many of the people in the area meet your target market criteria. The U.S. Census Bureau website

and a number of free nongovernment sites like www.zipwho.com and www.city-data.com can be a good starting place for free information. See Skill Module 11.2 and the Online Learning Center. If plowing through the Census Bureau website doesn’t get you what you want, several commercial services mentioned in Chapter 10 including Prizm and ESRI will sell you data about the population in a particular zip code for several hundred dollars. More detailed and specific commercial infor- mation is also available and can range in price from several thousand dollars to over $100,000 and is probably not an option for most entrepreneurs.37 Again, the website of Site Selection Magazine (www.siteselection.com) mentioned above has tools and additional articles that can help. Another consideration is the presence of traffic generators in the area. These are other busi- nesses that draw customers to the area and may include supermarkets, office complexes, schools, and malls. If the customers are drawn there, for example, to grocery shop, might they not stop at your video store next door? Reflect on the type of customer you are seeking and the likelihood of these businesses in attracting them. If you want a teen customer, a location near a high school works well. If you are looking for evening clientele, offices that close at five aren’t the right traffic genera- tors for you. Drive around likely areas and locate possible sites. Visit during the hours you anticipate to be peak times for your business and evaluate foot and car traffic.38 Look at the crowds or lines in similar businesses and decide whether there is room for you. If most businesses seem empty, you probably will be too.

Intersections of major streets offer high automobile traffic, but because of divided roads and other barriers, they may not make it easy for your customers to get to you. Businesses along inter- states have high visibility, but the frontage roads can be so convoluted that the clients seek easier- to-get-to competitors. Even some malls and shopping centers have such tortuous access problems that customers avoid them when possible. Sometimes entry is easy, but getting out is difficult. For example, no signals for left turns when most of the traffic needs to head in that direction can turn off customers.

Parking is also an issue. Is it conveniently located to your place of business? Do customers need to cross busy streets to get to you? Is parking free or paid? Are parking areas well lit and safe? Are there wheelchair ramps or other accommodations for disabled customers?

Customers have strong ideas about how far they should have to drive for things. These vary some- what from major metropolitan areas to more rural towns and from one region of the United States to another. Generally convenience stores, fast-food restaurants, and gas stations need to be close to con- sumers. Grocery stores and banks can be somewhat farther away, but not much. Discount stores and midscale restaurants can be even farther away, while specialty stores, upscale restaurants, and malls can be relatively remote. Where does your business fit? How far are customers willing to drive to get to you?

Malls are great traffic generators, but space at malls is costly. If it is appropriate for your product, you might consider a kiosk or cart in the mall as a way of testing the market and location without making a large investment.39 Neighborhood shopping centers (those anchored by a supermarket, drugstore, or major retailer) or strip centers (shopping centers without major anchors) are more modestly priced, but lack the drawing power of malls.4 Generally, competition in the area can draw away valuable clients, but this is not true in every case. Many cities have a restaurant row, an antique district, or an automobile mile (as well as other business types) where many competitors cluster. Clients wish to comparison shop or have choices and are drawn to areas where they can see several similar businesses at one time. Locating far from these will mean that you are free from competition, but this benefit may easily be offset by the cost of at- tracting customers to a different place. Additionally, you can capitalize on competitive advertisements that bring potential customers to the area. Your competitor’s high-budget TV ad might get customers to the neighborhood, but the “sale” sign in your window may get them to stop at your place instead.

Another instance when you want to be near competitors is when your business provides a strong contrast to the competition in the area. Do you offer better assistance, additional services, unique advantages, or other benefits that can easily be seen by customers? They may be drawn to the area by a well-known competitor’s name, but they may select your establishment instead because of what you offer that differentiates you.

Low Customer Contact Businesses

Generally low customer contact businesses are manufacturing businesses, the headquarters of client location-based services, or remote location services. Customer access is relatively unimportant. More critical are access for your employees, reasonable cost, and the space necessary to do your business. Certain manufacturing operations will need adequate utilities and specialized transportation too. Unless you plan to use some of this space as a high-traffic showroom, commercial space in a business park or light industrial park might be appropriate. These parks are located near major transportation routes, often have rail spurs, and are designed for industrial utilities; that is, they have adequate electricity, gas, water, waste water treatment, and the like. Frequently, support businesses will be located in or near the park such as warehousing, shipping firms, copy centers, and office supply stores. Industrial or business park space tends to be cheaper in smaller cities and rural towns than in major metropolitan areas. If distribution to customers can be arranged, these locations are certainly cost-effective.

Some major metropolitan areas offer empowerment zones. These zones, often in economically depressed areas, offer businesses low-cost space and tax advantages for locating there. For more information, see www.rurdev.usda.gov/BCP-EZEC-Home.html or www.siteselection.com.

A third possibility is a business incubator. The National Business Incubator Association (www .nbia.org) shows over 1,400 business incubators in North America sponsored by government, uni- versities, or private investment groups. These business incubators are specifically designed for the entrepreneur, and, in addition to relatively low cost space, they offer a multitude of small business support services. These services range from copy machines, faxes, and conference rooms to ac- counting, finance, and consulting services. Since the building is populated by other entrepreneurs, it’s a great place to talk to others who might have had some of the same problems or to brainstorm new ideas. Most incubators require a stake in your company in exchange for their assistance— maybe as much as 50 percent—and often have quite a bit to say about how you run your business. Opinion is mixed on how much real help a company can get; just like all businesses, there are better and worse incubators, so do your homework.42

General Comments on Site Selection

How do you go about finding potential sites? Looking for “for sale” and “for rent” signs is a start, but not all space will be advertised that way. Just as a good real estate agent can warn you about the pro- posed freeway project going through the backyard of the house you are considering or let you know about houses not yet listed but likely to be, an experienced real estate broker will also be able to assist you in your search for your business location. Many have relationships with landlords that can work to your benefit. They are also likely to have at least some of the market statistics you may need to help you decide if the location is right for your business.43 Level with them about what you can spend. You have your business plan and know the cost per square foot you can afford and be profitable. If you are looking at property more expensive than that, you’ll need to cut corners elsewhere.44

Leasing

It is rare that a small business start-up buys its first location. The reason is financial. It takes a lot of money to buy a place, and beyond that a long-term commitment to pay for it. For most small busi- nesses, it is not a worthwhile risk. It makes more sense to rent or lease your facility to leave more money for other aspects of the business. But leasing is one of the most complex of the issues an entrepreneur faces when starting a business.

In reality, most landlords (especially those from big national commercial real estate and mall companies) have fairly standard contracts which they don’t like to change. These typically start out as very pro-landlord. That said, in many cases they are also likely to accept certain standard clauses that are more tenant-friendly. However, it is unlikely they will offer those. You will need to ask for them. In this section you will learn about the major types of tenant-friendly clauses you might want to seek. However, it makes tremendous sense to get a real estate lawyer involved to help you. They should be able to tell you what kinds of clauses are typical, and who else offers them in your area, and if there are any other likely traps in the lease. You can learn how to choose a lawyer in Chapter 18.

You should start the leasing process by looking for locations. You can start using the local news- paper’s or business journal’s classified ads for commercial real estate. If you know of a great loca- tion, but there is no “for rent” or “for lease” on it, consider asking the owner or current renter about subleasing a portion of the location. If your product or service complements the current tenant, you could find a home. Local real estate websites may also have listings, and there are national websites like LoopNet.com which compile listings from a variety of sources. There is a how-to video for

using LoopNet on the Online Learning Center. You may contact a real estate agent with commercial experience to help you, but make sure you know how the agent is making his or her money. You want the agent to work in your best interests.

It helps to have two or three possibilities identified before you begin negotiating leases. This is a use of the idea of the power of rivalry from Chapter 7. This gives you a basis for comparison, and an alternative for leasing when negotiating. But note that the more alike the properties, the greater your power at the negotiating table. Also, if you are opening a franchised operation, you will want to contact your franchisor before you start looking for locations. Most franchisors have specifications for locations, and advice on costs and other features. They often have a lease review department to help you with this process.45

The best way to start thinking about the clauses is to separate them into those clauses related to choosing a property, day-to-day operations, and endings. In reality, though, all of these clauses will get negotiated when you and the landlord discuss the lease agreement.

There are several issues that could pop up as you are narrowing your choices and trying to decide which location and deal is the best for you. These are:46

?          “As is” versus compliant property: If the location has problems, who should fix them? The landlord would like to have you do it, and will try to push you to accept the property “as is.” You, of course, want the landlord to fix it before you move in, so you would ask the property to be “in compliance with all applicable laws, rules, and regulations.” Realize that the landlord will get back the money paid for repairs eventually, through fees or higher rents, but it can save you money on the front end.

?          HVAC: This is the commercial jargon for “heating, ventilation, and air conditioning.” It can be the most expensive type of repair, and since it is mechanical, one of the types most likely to go wrong. The landlord wants it to be your responsibility. You want it to be the landlord’s. This is particularly important if the location has a central air system so everyone shares the same air conditioner and heating equipment. This type of equipment needs to be the landlord’s responsibility. For any type of equipment, you want the landlord to at least guarantee the first year of operation.

?          Signs: Called “signage” in the business, it can be on the street, on the building, or around the door. You probably have ideas for your signs. If you are a franchise, you face signage require- ments from the franchisor. You want a landlord who will work with you on the size, place- ment, and visibility of signs. Make sure you have written

agreement on the signs and, if possible, a clause that says approval cannot be unreasonably denied for future changes.

There are other benefits possible if you know to ask for them. Often these are called concessions. Examples include “leasehold improvements,” which are permanent changes made to the loca- tion to fit your business’s needs. You cover these by seeking a “tenant improvement allowance” or “construction allowance” which are rent dollars (typically $5 to $25 a square foot) they agree to let you put into improving your location.47 This amount should be based on a firm estimate from a construction profes- sional. Another concession is a “rent-free use period” which cov- ers the time while you prepare your location prior to opening.

As you start thinking about how you would operate day-to- day, there will be several different issues you will face. These include:

?          Hidden charges: Many leases include charges that do not have to be listed in the term sheet given you for the prop- erty. An example is a monthly operating expense. This may be justified. You may be leasing a thousand feet of space, but there are also common areas, restrooms, parking and the like that the landlord keeps up for everyone. Ask spe- cifically for a list of all expenses or charges for which you will be liable, and compare to other locations. Also make sure to learn the conditions under

which you can lose all or part of your security deposit. ?     Use of premises: You specify in the lease what you will sell or do at the leased location, but

too exact a description could prevent you from expanding the products or services you offer. Try to add the clause “and related goods and services” to any description you give to provide reasonable flexibility.

?          Noncompete: If you have a pet store in a strip mall, you’d like to be the only one there. For many types of businesses, you negotiate a clause limiting the landlord’s ability to lease to a competing business. This can be just for your facility or for a radius. You should expect to pay for the exclusivity and the farther you want it, the more it will cost. Note that competition in terms of different types of restaurants, or another store selling some of your products, is still likely.

?          Hours of operation: Mall landlords want stores open the same hours and days, and landlords of other types of properties may have some of the same desires. You need to negotiate times that fit your business model. Look for stores in the landlord’s properties that match your hours. Precedence helps here.

?          Rent default: When you are late paying rent, all sorts of penalties and problems emerge. It also hurts your credit rating. Some leases require the renter to keep track. Ask to change the lease to specify the landlord needs to alert you immediately on the rent due date in written or telephonic (usually fax) form, and get the 5–10 day default period for paying rent before default starts from that notice.

?          Moves and remodels: There may be a clause that gives the landlord the right to move your business elsewhere, at their discretion. If this is to update or repair an area, fine, but what if it is to get a higher-paying tenant in your space? Set time limits and return rights on any forced move. Similarly, if the landlord decides to remodel, you should not have to pay for it.

As an entrepreneur negotiating a lease, you need to prepare for the good and the bad as time moves along. The good is the prospect your business grows and you need more space. The bad is that your business doesn’t do well, and you need to get out of your lease before the end of the term. Dealing with these issues is like worrying about a prenuptial agreement while you are taken with the romance of getting married. It might be painful to imagine, but it is important to keeping what is yours.

If your business falters, you are obligated to continue paying your monthly rent and fees for the duration of the lease. A landlord has the power to let you out of a lease, but he or she is only likely to do this if a better tenant is lined up. Once you tell your landlord you may need to vacate the prem- ises, she or he is supposed to look for a replacement tenant, but not all do, or do a good job of it. If you can find a replacement tenant, it can help this process along, but you need to make sure there is a clause that lets you sublease the property, and further, that the landlord can’t unreasonably deny the sublease. If your pet store is closing, finding a dress shop is probably reasonable (as long as it doesn’t violate some other tenant’s noncompete clause), but finding an adult book store is probably not a reasonable replacement.

Three other ways to handle an early termination are to set up a short-term (e.g., 6 months) lease initially, ask for a bailout clause, or for a “cap” or limit on how long you need to continue paying rent. The bailout clause lets you out of the lease if sales do not meet an agreed-to level. You negotiate this with the landlord up front. To understand a cap, think about a three-year lease. If you close down after only six months, you are still obligated to pay 30 more months’ rent. With a one-year cap you are only obligated to pay 6 months’ more rent. This is like a type of insurance, and like insurance policies, you will probably have to pay a slightly higher rent from the start to cover this possibility.

Realize there can be problems you face caused by the property itself. What if you move into a mall with a major chain like Sears, Penny’s or Macy’s, or a strip mall with a major supermarket or discount store. Part of what you are paying a premium for is the traffic and reputation these anchor stores bring. What if they leave? Your location’s quality could dramatically drop. To get out of your lease under these unfortunate circumstances, you want to ask for a cotenancy clause.

Although we’ve segmented a renter’s concerns by stage of the leasing process, all of these issues need to be negotiated at the start when crafting a lease. Although landlords often start with a lease they call “standard,” nearly everything about it can be negotiated. But be fair; the space may mean a lot to you, but it is a drop in the bucket to large commercial realtors. You can learn more about negotiating in general in Chapter 18, but there are some special considerations for lease negotia- tions. Because so many aspects are potentially negotiable and areas have different norms for what are typical tenant-friendly clauses, work with a real estate lawyer of your own to help you in the negotiation and phrasing of the lease terms.

Layout

Since so much of this is particular to the type of business you are in, what you’ll read below is a general guide. Check out competitors or similar types of businesses to see what you like and don’t like, what seems to work well, and what seems to cause a lot of problems. In addition, certain categories—restaurants and retailing, for example—have numerous books from college textbooks to do-it-yourself books, like the “For Dummies” series. Try your local library or bookstore.

The layout of a potential site must be considered carefully. Is the building setup appropriate for your use? A restaurant will have different needs than a retail area or a manufacturing plant. The amount of area allocated to the “front room” (e.g., eating or retail areas) versus “back room” (storage, kitchen, warehouse, and office areas) needs to be adequate for the purpose of your business. If you are operating a restaurant or retail operation, how important is space in the front room? A coffee shop or fast-food restaurant squeezes in more customers per square foot than a gourmet restaurant. Do you need specialized areas, such as a kitchen or laboratory that are expensive to retrofit into existing buildings? Is there adequate storage area? How much dock space is appropriate for your business? A manufacturing firm usually needs more dock space and storage than a retailer or a restaurant, while a service company may need very little of either. Retail operations need display windows, while manufacturers do not. For restaurants or other services, this need varies. Is there room for expansion should the business grow? Remember: Moving can be expensive. A good strategy is to rough out the desired layout of your operation on graph paper to get a basic idea of the square footage needed and how it should look. What exactly you want may not be out there, but you’ll be able to see what’s close and what’s impos- sible to live with.

Consider the amenities that are already there. Carpeting may be appropriate for a retail area and perhaps the office or dining area of a restaurant, but not appropriate for manufacturing or cooking areas. What about the walls? What sort of ceilings and lighting is appropriate? Again, a visit to the competition will help you decide what works and what doesn’t, as well as where you want to be different.

Check the exterior, too. Is the building attractive and inviting? Are the sidewalks and landscaped areas in decent shape? Is parking adequate, well lit, and safe? Is employee parking separate from customer parking? What about handicap accessibility? The 1990 Americans with Disabilities Act (ADA) specifies that businesses (with few exceptions) must accommodate persons with disabilities. Many buildings have been brought up to code, and all new construction should meet the require- ments of this act, but keep your eyes open.

Once the building has been selected, how you lay out the interior also needs to be considered. While retailers, restaurants, offices, and manufacturers all have different layout needs and consider- ations, two facts hold true: (1) layouts need to be designed so as to eliminate unnecessary and exces- sive employee movement, and (2) the layout says something about who you are to your customers, employees, and visitors. In retailing, this last factor is called atmospherics. While the opportunities for variation are limitless, let’s consider the major types of retail and manufacturing layouts as well as what atmospherics might mean to a business.

Traditionally, manufacturing processes are laid out in one of two formats: production line layout and process layout displayed in Figure 11.2. In the production line layout, material flows in on one side of the operation and continues to the other end of the operation. Most assembly manufacturing is done this way, often with conveyors moving subassemblies from one station to another. Although a rather rigid flow, it works well for mass production and high-volume manufacturing. The second method, process layout, groups similar machines/or functions together, not unlike a typical machine shop. This format is much more appropriate for lower-volume, flexible manufacturing.

There are also two traditional layouts for retail operations, which are shown in Figure 11.3. The first one, the grid layout, has aisles running from the front of the store to the back like the typical grocery, discount, or convenience store. It’s a very efficient and organized layout although it lacks some visual impact. The second layout, the free-form layout, alleviates this problem. In this layout, the store is laid out in sections with aisles that angle or meander through the store. This is the layout more typically found in upscale department stores, clothing stores, and the like.

Atmospherics include all the ambiance items that might be considered in your business. An up- scale women’s clothing store is likely to have wider aisles, deep carpeting, soft “elevator music,” indirect lighting, and, perhaps, a lightly perfumed aroma. These are appropriate atmospherics for the target market. A shop catering to edgy teen fashions may be done in black and chrome with loud rock or alternative music and strobe or black lights. Both of these send a message about whom the store is likely to appeal to. Restaurants use atmospherics, too. Compare a family restaurant to a gourmet restaurant to an ethnic restaurant. Services and the office and public areas of manufacturing firms do this as well in their choice of colors, furniture styles, and background music.48

Build, Buy, or Lease49

Ultimately, there are three choices available to the business: build, buy, or lease. Building has the advantage of having the perfect layout in the perfect location and the street appeal of a new build- ing, but it is costly and slow. Buying something already in existence shortens the time and may be

somewhat cheaper, but any remodeling or retrofitting that needs to be done may overshadow any time or money savings. In both cases, though, business owners have an asset that they can leverage, as well as the depreciation tax advantage. They have the flexibility to make the changes they need and know what the long-term costs will be.

Leasing, which we detailed earlier in the chapter, is an option with a considerably lower initial cash outlay, and it is often the only feasible choice for new businesses. Lease expenses are deduct- ible business expenses. One of the main downsides of leasing is that you are usually limited in the renovations you can do. Another one is that leases tend to get higher with each renewal contract, and your landlord may choose not to extend a lease, forcing you to move before you are ready to do so.

The issues of location and distribution are decisions that business owners make only occasion- ally. Many businesses operate from the same location for their entire existence. Distribution deci- sions may come up more often. For example, a business started on eBay develops its own website, and then grows into a store in the city’s commercial center. Regardless of how often these decisions are made, they are central to the success of the small business, because placing a business in the right location and equipping it with the right channels of distribution are essential to finding and connecting with customers. Done right, managing the issues of location and distribution can turn an average firm into a major success.

1. According to this chapter, what are three key considerations in determining the location of your business?

2. According to this chapter, what are the typical locations for service businesses?

3. What are the advantages and disadvantages of buying, building or leasing your business’ facility?