Kristina Kozusko,1 Venessa H.M. Tsang,2 William Bottomley,3 Yoon-Hi Cho,4,5 Sheetal Gandotra,6

Michael Mimmack,1 Koini Lim,1 Iona Isaac,1 Satish Patel,1 Vladimir Saudek,1 Stephen O’Rahilly,1

Shubha Srinivasan,4 Jerry R. Greenfield,2,7 Ines Barroso,3 Lesley V. Campbell,2,7 and David B. Savage1

Clinical and MolecularCharacterization of a NovelPLIN1 Frameshift MutationIdentified in Patients WithFamilial Partial LipodystrophyDiabetes 2015;64:299–310 | DOI: 10.2337/db14-0104

Perilipin 1 is a lipid droplet coat protein predominantlyexpressed in adipocytes, where it inhibits basal andfacilitates stimulated lipolysis. Loss-of-function muta-tions in the PLIN1 gene were recently reported inpatients with a novel subtype of familial partial lipodys-trophy, designated as FPLD4. We now report theidentification and characterization of a novel heterozy-gous frameshift mutation affecting the carboxy-terminus(439fs) of perilipin 1 in two unrelated families. Themutation cosegregated with a similar phenotype inclu-ding partial lipodystrophy, severe insulin resistanceand type 2 diabetes, extreme hypertriglyceridemia, andnonalcoholic fatty liver disease in both families. Poormetabolic control despite maximal medical therapyprompted two patients to undergo bariatric surgery,with remarkably beneficial consequences. Functionalstudies indicated that expression levels of the mutantprotein were lower than wild-type protein, and in stablytransfected preadipocytes the mutant protein was asso-ciated with smaller lipid droplets. Interestingly, unlike thepreviously reported 398 and 404 frameshift mutants,this variant binds and stabilizes ABHD5 expression but

still fails to inhibit basal lipolysis as effectively as wild-type perilipin 1. Collectively, these findings highlight thephysiological need for exquisite regulation of neutrallipid storage within adipocyte lipid droplets, as well asthe possible metabolic benefits of bariatric surgery inthis serious disease.

Adipocytes within white adipose tissue are uniquelyadapted to storing large quantities of triglyceride in asingle unilocular lipid droplet, providing a highly econom-ical mechanism for surplus energy storage. Perilipin 1 isthe most abundant phosphoprotein in adipocytes, whereit constitutively associates with the phospholipid surfacemonolayer of the lipid droplet (1). Here it regulateslipases, particularly adipose tissue triglyceride lipase(ATGL) and hormone sensitive lipase (HSL). ATGL andHSL catalyze the sequential hydrolysis of tri- and thendiacylglycerol, releasing fatty acids and monoacylglycerol,which undergoes a final hydrolytic cleavage step by mono-acylglycerol lipase (MGL) (2). Precise regulation of this

1University of Cambridge Metabolic Research Laboratories, Wellcome Trust–Medical Research Council Institute of Metabolic Science, University of Cambridge,Cambridge, U.K.2Diabetes Centre and Department of Endocrinology, St Vincent’s Hospital, Sydney,Australia3Wellcome Trust Sanger Institute, University of Cambridge, Hinxton, U.K.4Institute of Endocrinology and Diabetes, The Children’s Hospital at Westmead,Sydney, Australia5Discipline of Paediatrics and Child Health, The Children’s Hospital at WestmeadClinical School, University of Sydney, Sydney, Australia6Council of Scientific and Industrial Research-Institute of Genomics & IntegrativeBiology, New Delhi, India7Diabetes and Obesity Research Program, Garvan Institute of Medical Research,Sydney, Australia

Corresponding author: David B. Savage, dbs23@medschl.cam.ac.uk, or Lesley V.Campbell, l.campbell@garvan.org.au.

Received 20 January 2014 and accepted 24 July 2014.

This article contains Supplementary Data online at http://diabetes.diabetesjournals.org/lookup/suppl/doi:10.2337/db14-0104/-/DC1.

K.K. and V.H.M.T. contributed equally.

L.V.C. and D.B.S. contributed equally.

V.H.M.T. is a current PhD student at Kolling Institute of Medical Research,University of Sydney, Australia.

© 2015 by the American Diabetes Association. Readers may use this article aslong as the work is properly cited, the use is educational and not for profit, andthe work is not altered.

Diabetes Volume 64, January 2015 299

GENETIC

S/G

ENOMES/P

ROTEOMIC

S/M

ETABOLOMIC

S

process is a major factor in determining the efficacy ofadipose tissue as a lipid buffer in the fed state and thensubsequently as a source of lipid fuel in the fasting stateor during exercise.

The importance of perilipin 1 function in humanmetabolism was very recently highlighted by the discov-ery of two heterozygous PLIN1 frameshift mutationsaffecting the COOH-terminus of the protein in patientswith FPLD4 (3). Both mutations cosegregated with par-tial lipodystrophy, severe insulin resistance, hepaticsteatosis, and severe dyslipidemia in three French fam-ilies. Molecular characterization of these mutationsdemonstrated the inability of both mutants to inhibitbasal lipolysis. In each case, this was attributable to thefailure of the proteins to effectively prevent ABHD5(ab-hydrolase domain containing 5), a coactivator ofATGL, from activating ATGL and thus increasing basallipolysis (4). Here, we report the identification, clinicalcharacterization, and functional analysis of a third PLIN1heterozygous mutation.

EXPERIMENTAL PROCEDURES

Genetic AnalysisThe study was conducted in accordance with the princi-ples of the Declaration of Helsinki and was approved bythe U.K. National Health Service Research Ethics Com-mittee. Each participant, or a parent in the case of minors,provided written informed consent; minors provided oralconsent.

Genomic DNA was isolated from peripheral bloodleukocytes. After excluding mutations in the coding regionsand splice junctions of LMNA and PPARG, the codingregions and splice junctions of PLIN1 were amplified byPCR and sequenced as described previously (3).

Clinical and Biochemical StudiesBlood samples for biochemical assays were taken afteran overnight fast. Patients underwent skin-fold mea-surements using calipers; results were expressed as themean of two independent measurements and comparedwith previously reported reference ranges (5). Dual-energy X-ray absorptiometry (DEXA) (Lunar Prodigy,software version 12.20) was used to evaluate bodycomposition.

Cloning StrategySite-directed mutagenesis was performed on pCR-BluntII-TOPO PLIN1 WT (3) using QuikChange II XL Site-Directed Mutagenesis Kit (Agilent Technologies) to generatethe PLIN1 p.Pro439ValfsX125 (from here onwards re-ferred to as PLIN1 439fs) mutant using the followingset of primers 59-CGGAGCGCAGAGCGTCGGGCGCCGTCCGCCGG-39 and 59-CCGGCGGACGGCGCCCGACGCTCTGCGCTCCG-39. For the generation of stable cell lines, a N-terminallyMyc-tagged PLIN1 439fs mutant sequence was subclonedusing SalI restriction enzyme sites into the retroviral ex-pression vector pBABE-puro (Clontech) harboring a puro-mycin resistance cassette. For transient transfections, the

same sequence was subcloned into pcDNA3.1 vector usingEcoRV-XbaI restriction sites. For bimolecular fluorescencecomplementation (BiFC) assays, pcDNA3.1 mycPLIN1-439fs-Yc was cloned in frame with the COOH-terminalYc fragment in previously generated pcDNA3.1-Yc usingEcoRV-BlpI (4). The following constructs pcDNA3.1-Yc,pcDNA3.1-Yn, pcDNA3.1-Yn-hABHD5, pcDNA3.1-hATGL(S47A)-Yc, and pcDNA3.1-MycPLIN1-398fs-Yc werepreviously generated as described by Gandotra andcolleagues (3,4).

Cell CultureStable cell lines were generated and maintained asdescribed by Gandotra et al. (3). In order to study theeffects of PLIN1 439fs in the presence of WT perilipin 1,a previously described 3T3-L1 preadipocyte stable cellline expressing pLXSN-Flag-PLIN1-WT (3) was subjectedto a second round of retrovirus transduction withpBABE-puro myc-PLIN1 WT or mutants.

Quantitative real-time PCR mRNA was harvested andextracted using a RNeasy Mini Kit (Qiagen). cDNAsynthesis was performed using the SuperScript II ReverseTranscriptase enzyme (Invitrogen). Quantitative real-timePCR was performed using an Applied Biosystems TaqManMaster Mix according to the manufacturer’s protocols.The following primer sets were used: hPLIN1 (forward,59-CCCCCTGAAAAGATTGCTTCT-39; reverse, 59-GGAACGCTGATGCTGTTTCTG-39; probe, 59FAM-CATCTCCACCCGCCTCCGCA TAMRA-39) and PPIA (forward, 59TTCCTCCTTTCACAGAATTATTCCA-39; reverse, 59-CCGCCAGTGCCATTATGG-39; probe, 59FAM-6ATTCATGTGCCAGGGTGGTGACTTTACAC-TAMRA-39). Results were analyzed on anABI PRISM 7900HT (Applied Biosystems) and normalizedto housekeeping gene (Cyclophilin A) expression levels.

Western BlottingFor patient adipose tissue protein expression analysis,tissue was homogenized in liquid nitrogen and directlylysed in RIPA buffer (Sigma-Aldrich) containing pro-tease and phosphatase inhibitors (Roche). Protein wasquantified using Bio-Rad DC protein quantification kitand 10 mg of protein lysate was diluted in NuPAGE 4XLDS sample buffer (Invitrogen) containing 0.05%b-mercaptoethanol and subjected to SDS-PAGE, follow-ing transfer onto a nitrocellulose membrane. Followingtransfer, membranes were washed in Tris-buffered sa-line containing 0.1% TWEEN-20 (Sigma-Aldrich), thenblocked for 1 h at room temperature in 3% bovine se-rum albumin (BSA) or 5% powdered skimmed milk di-luted in TWEEN-20. Membranes were further incubatedwith appropriate primary antibody diluted in blockingbuffer for 16 h at 4°C. The following antibodies wereused: perilipin 1 N-terminal antibody (GP29, Progen),perilipin 1 COOH-terminal antibody (Abcam), perilipin2 (GP40, Progen), Myc (Millipore), Flag (Sigma-Aldrich),ABHD5 (Abnova), Calnexin (Abcam), or b-actin (Abcam).For cellular protein extracts, 50 mg of lysate was subjectedto SDS-PAGE.

300 PLIN1 Mutation With Lipodystrophy Diabetes Volume 64, January 2015

Estimation of Protein Degradation3T3-L1 preadipocytes stably expressing either WT or439fs perilipin 1 were grown until confluency and treatedwith either DMSO as a control or 100 mg/mL cyclohexi-mide, 10 mmol/L MG132, and 25 mmol/L NH4Cl for in-dicated times before lysis.

Microscopy StudiesFor microscopy studies, cells were seeded into 12-wellplates containing slides previously rinsed in 70% ethanol.Cells were harvested by rinsing three times in PBS andfixing with 4% formaldehyde diluted in PBS for 15 min atroom temperature. Membrane permeabilization wasachieved with 0.5% Saponin for 10 min, followed byblocking in 3% BSA, 0.05% TWEEN-20 PBS for 1 h.Afterward, cells were incubated with primary antibodydiluted in blocking buffer for 16 h at 4°C. Then cells werewashed three times in 0.1% BSA-PBS and incubated withfluorescent probe conjugated secondary antibody for 1 hat room temperature wrapped in foil. Where necessary,cells were stained for neutral lipid by rinsing three moretimes with 0.1% BSA-PBS and adding LipidTOX Deep Redreagent (Invitrogen) diluted in PBS (1:1,000) for 20 minat room temperature. Afterward, cells were rinsed in PBSand slides mounted using ProLong Gold antifade mountingreagent with DAPI (Invitrogen) for nuclear staining.

Lipid Droplet Volume Measurements3T3-L1 preadipocytes were treated with 400 mmol/L oleicacid (Sigma-Aldrich) for 48 h and harvested for micros-copy studies as described above. LipidTOX Deep Red wasused to stain neutral lipids within lipid droplets, and anti-Myc primary antibody (1:500) dilution in 3% BSA, fol-lowed by goat anti-mouse Alexa Fluor 488 fluorescentsecondary antibody, was used for perilipin 1 staining.Cell imaging was performed on a Zeiss LSM 510 META

confocal microscope using the Zen software package(Carl Zeiss Microscopy GmbH). On average 10 Z-stackimages were taken per sample using the 363 objective,and lipid droplet volume was measured using Volocity 5software (PerkinElmer).

LipolysisLipolysis experiments were performed as described byGandotra et al. (4) in the presence of 6 mmol/L Triacsin Cin order to prevent fatty acid re-esterification. Lipolysiswas expressed as a ratio of the amount of radioactivityreleased into the medium over the total amount of in-corporated radioactivity.

BiFCBiFC experiments and signal quantification were per-formed in COS-7 cells as described by Gandotra et al. (4)and Patel et al. (6). In cells expressing perilipin 1 (WT ormutants), ATGL (S47A)-Yc, and ABHD5-Yn, total yellowfluorescent protein (YFP) signal was normalized to controlcells only expressing ATGL(S47A)-Yc and ABHD5-Yn. Allcomponents were normalized for equal protein expressionin these experiments.

Statistical AnalysisQuantitative data are presented as mean 6 SEM. One-way or two-way ANOVA with post hoc Bonferroni analy-ses were performed on data at a minimum P , 0.05.

RESULTS

Case Studies

Proband AProband A, a 41-year-old Caucasian woman (Fig. 1A, I.1),was referred to specialist physicians in Sydney in 1998 witha 10-year history of type 2 diabetes, complicated by dia-betic retinopathy, extreme hypertriglyceridemia leading to

Figure 1—A heterozygous PLIN1 439fs mutation cosegregates with insulin-resistant diabetes, dyslipidemia, and partial lipodystrophy. A:Family pedigrees of probands A and B showing cosegregation of p.Pro439ValfsX125 with insulin resistance, dyslipidemia, lipodystrophy,and, to a lesser extent, hypertension. M1 denotes PLIN1 p.Pro439ValfsX125 mutation; N, WT; and ND, not determined. Probands aredenoted by an arrow. B: Sequence chromatogram confirming deletion of a TG base pair within exon nine (nucleotide 1298_1299) and thepredicted consequences of this mutation on the amino acid sequence of perilipin 1.

diabetes.diabetesjournals.org Kozusko and Associates 301

recurrent pancreatitis, nonalcoholic fatty liver disease(NAFLD; visualized on ultrasound and computerized to-mography), and hypertension. She also had new-onsetexercise intolerance (exertional dyspnea without chestpain) and muscle cramping. Her diabetes was subopti-mally controlled despite high-dose insulin therapy(.500 units daily). Insulin resistance was confirmed bya hyperinsulinemic-euglycemic clamp (glucose infusionrate 10.6 mm/min/kg fat-free mass). The patient had nohistory of polycystic ovarian syndrome (PCOS) and nodifficulty conceiving as is apparent from four successfulpregnancies. Her first pregnancy was uncomplicated, buther second child was delivered early because of large sizefor gestational age. The diagnosis of diabetes was madeafter her third pregnancy.

Her BMI was 31.2 kg/m2 and waist circumference93 cm. Lipoatrophy was most notable in the femoroglu-teal depot and lower limbs, with abdominal fat largelypreserved (Supplementary Fig. 1). Skin-fold measure-ments (Supplementary Fig. 2) confirmed excess subcuta-neous truncal fat and reduced peripheral fat. In addition,DEXA scan confirmed disproportionate fat loss on thelimbs (Table 1). Her cardiovascular and respiratoryexaminations were normal. There were no dysmorphicfeatures.

Biochemical investigations revealed severe hypertri-glyceridemia and suboptimal glycemic control (Table 1).Her lipoprotein-a level was elevated at 680 mg/L (,300).and her leptin level was low (Table 1). An echocardiogramrevealed a reduced ejection fraction of 30%, and a coro-nary angiogram showed an occluded right coronary artery.Nerve conduction studies and electromyography werereported as normal, with only occasional small atrophicfibers. A deltoid muscle biopsy was normal.

The patient’s eldest daughter, aged 38 years (Fig. 1A,II.1), was diagnosed with PCOS, dyslipidemia, fatty liverdisease (ultrasound), hypertension, and type 2 diabetesat age 24 years. She was initially managed with metfor-min but required .500 units/day of insulin during thelater stages of her first pregnancy, during which sheweighed up to 120 kg (BMI 40.1 kg/m2). She has hadfour pregnancies, all of which were complicated by hy-pertension, necessitating hospital admission in two preg-nancies. All four children are well and have not beentested for diabetes.

The proband’s oldest son, aged 37 years (Fig. 1A, II.2),has a lipodystrophic phenotype and a BMI of 29.1 kg/m2

but has not been genetically tested or assessed for diabe-tes or dyslipidemia. Her middle son (Fig. 1A, II.3), aged 33years, is centrally obese with a BMI of 39 kg/m2, diabetes,and mild hypertriglyceridemia. Her youngest son, aged 18years (Fig. 1A, II.4), has type 2 diabetes managed with oralmedication and a BMI of 29.1 kg/m2. The proband’sbrother, aged 55 years (Fig. 1A, I.2), has a BMI of 25.3kg/m2, diabetes, a fat mass ratio (% fat trunk to % fatlower limb ratio) of 2.2, and skin-fold measurements con-sistent with a lipodystrophic phenotype.

In 2005, proband A underwent Roux-en-Y gastricbypass surgery in an effort to improve her glycemiccontrol and hypertriglyceridemia. This resulted in a 20-kgweight loss, substantial reduction in insulin requirements(from 500 to 100 units/day), improved lipid profile, andreversal of cardiomyopathy on echocardiography. Thepatient’s daughter also underwent Roux-en-Y bypass sur-gery, resulting in a 55-kg loss of weight (BMI 22.4 kg/m2),cessation of insulin, and a subsequent twin birth and twofurther successful uncomplicated pregnancies. In the past2 years, the proband has been diagnosed with a poplitealartery aneurysm, requiring surgical repair and vascularstenting. Her diabetes is currently treated with a lowerdose of insulin and a very low2fat diet. Hypertriglyceridemiais controlled with gemfibrozil and maxepa. Her echocardio-gram is now within normal limits, with an ejection fractionincreasing from 30 to 59%.

Proband BProband B, an unrelated 15-year-old boy of Caucasianorigin, presented with acanthosis nigricans, hepatomeg-aly, and NAFLD (Fig. 1A). His BMI was 25.1 kg/m2 (1.51SD score) with central adiposity (waist to height ratio0.54, normal ,0.5) and a striking paucity of limb fat. Bio-chemical investigations showed severe hypertriglyceridemiawith normal total cholesterol and low HDL cholesterol levels(0.4 mmol/L). An oral glucose tolerance test confirmed thepresence of extreme insulin resistance. He had impairedglucose tolerance with a 2-h glucose level of 10.7 mmol/Land an astonishing peak insulin level of 12,396 pmol/L. Hisleptin level was low (Table 1). Liver biopsy confirmed signif-icant hepatic steatosis.

The patient’s mother had been diagnosed with PCOS,treated with metformin since age 25 years, and required invitro fertilization for her first pregnancy at age 28 years.She was subsequently diagnosed with type 2 diabetes re-quiring .200 units of insulin daily during pregnancy. Shecontinued to have poor glycemic control, microalbuminu-ria, and NAFLD complicated by cirrhosis (on liver biopsy).Lipoatrophy was first recognized clinically in her early 40saffecting her limbs and buttocks, with mild central adipos-ity and prominent deltoid, gluteal, and calf muscles. Shehad coarse facial features and androgenic alopecia. Shewas lean (BMI 21.8 kg/m2) and her leptin level was low(Table 1). The patient’s father was obese (BMI 37 kg/m2)with associated dyslipidemia but did not have features ofsevere insulin resistance (Fig. 1A).

Identification of a Novel PLIN1 p439fs MutationBoth probands were wild type (WT) for PPARG andLMNA gene sequencing. Candidate gene sequencing of allexons and splicing regions of the PLIN1 gene revealed aheterozygous deletion of two adjacent nucleotides (thymi-dine and guanine) within exon 9 (c.1298_1299delTG),resulting in a translational frameshift (Fig. 1B). The muta-tion is predicted to lead to the incorporation of 125 aber-rant amino acids from amino acid position 439 onwards,thus producing an elongated protein (563 amino acids

302 PLIN1 Mutation With Lipodystrophy Diabetes Volume 64, January 2015

Tab

le1—Clinicaland

biochem

icalcharacterizationofproband

sand

relativeswith

thePLIN

1439fs

mutatio

n

Prob

andA

I.2II.1

II.4Prob

andB

Mother

Referencerange

PLIN

1mutation

439fs439fs

439fs439fs

439fs439fs

Age

(years)56

5538

1815

48

Sex

Female

Male

Female

Male

Male

Female

BMI(kg/m

2)31.2

25.430.4

29.125.1

21.8

Fatdistrib

utionLim

band

femorogluteal

lipod

ystrophy

Limband

femorogluteal

lipod

ystrophy

Limband

femorogluteal

lipod

ystrophy

Limband

femorogluteal

lipod

ystrophy

Limband

femorogluteal

lipod

ystrophy

Limband

femorogluteal

lipod

ystrophy

PCOS

-NA

+NA

NA

+

Fattyliver

+ND

+ND

++

Pancreatitis

+-

--

-+

Card

iomyop

athy+

--

--

+

DEXAFM

R1998,

1.4;2005,

1.9;2010,

1.9

2.2ND

ND

ND

ND

Lipod

ystrophy

ifFM

R.1.2

Totalcholesterol(mmol/L)

10.54.6

5.54.1

4.43.0

,5.2

Triglyceride(m

mol/L)

56.12

6.12.3

26.32.2

,1.7

HbA1c[m

mol/m

ol(%)]

63(7.9)

79(9.4)

59(7.5)

51(6.8)

32(5.1)

97(11.0)

,53.0

(7.0)

Glucose

(mmol/L)

11.36.4

ND

4.85.1

14.4,6.1

Insulin(pmol/L)

ND

17ND

ND

285169

,60

Leptin

(ng/mL)

2.92.8

3.8ND

2.13.6

3.7–11.1

FMR,fat

mass

ratio(%

fattrunk

to%

fatlow

erlim

bratio);

NA,not

available;

ND,not

determ

ined.

diabetes.diabetesjournals.org Kozusko and Associates 303

rather than 522) (Fig. 2A). The same mutation was pres-ent in both probands, although they are not known to berelated. The mutation cosegregated with partial lipodys-trophy, severe insulin resistance, severe dyslipidemia, andNAFLD in both kindreds (Fig. 1A). Gestational hyperten-sion was prominent in affected women.

Detection of the PLIN1 439fs Mutant Protein in PatientAdipose TissueIn order to determine whether the frameshift perilipin 1mutant protein is expressed in vivo, samples of both visceralabdominal and subcutaneous adipose tissue were obtainedfrom proband A (at the time of a medically indicatedsurgical procedure) and subjected to SDS-PAGE alongsidesubcutaneous fat samples from sex-matched control sub-jects (Fig. 2B). Both the WT and 439fs copy of perilipin 1 aredetectable with an N-terminal epitope-targeted antibody;however, the higher molecular weight band correspondingto the 439fs protein was not detected by an antibody thattargets a COOH-terminal epitope. Notably, the levels of WTand particularly the 439fs protein were reduced in the pa-tient compared with control subjects.

In addition, in both murine and cell models, a decreasein perilipin 1 expression levels in adipocytes is typicallyassociated with upregulation of perilipin 2 expression (7).

In accordance with this, proband A manifested an in-crease in perilipin 2 expression in both visceral and sub-cutaneous adipose tissue when compared with controlsubjects (Fig. 2B).

Functional Characterization of the PLIN1 439fsMutationIn order to functionally characterize the effects of thePLIN1 439fs mutation, 3T3-L1 preadipocytes were stablytransfected with retroviral vectors expressing pBABE-puro PLIN1 WT or the PLIN1 398fs or 439fs mutants,all bearing a N-terminal Myc tag. 3T3-L1 preadipocytes donot express endogenous perilipin 1 enabling us to directlycompare their function in the absence of the potentiallyconfounding influence of endogenous perilipin 1. ThePLIN1 398fs mutation was characterized in previous workand served as a negative control in these studies (3). All ofthe above mutant proteins are predicted to have a disor-dered COOH-terminus following the frameshift. In keepingwith the patient biopsy findings, expression levels of the439fs mutant were lower than those of the WT proteindespite equivalent mRNA expression levels (Fig. 3A and B).Expression of the 398fs mutant was even lower than thatof the 439fs. In the presence of cycloheximide, the half-lifeof the 439fs protein was reduced compared with WTperilipin, which is known to be particularly stable whenassociated with lipid droplets (Fig. 3C and D) (8–10). Thefact that coincubating the stable cell lines exposed to cy-cloheximide with a proteasomal inhibitor (MG132) mod-estly increased 439fs expression whereas the addition ofa lysosomal inhibitor (NH4CL) had no discernible affect(Fig. 3E) suggests that the 439fs perilipin 1 mutant isprobably degraded via the proteasome. We also notedthat the addition of MG132 increased endogenous perilipin2 expression, which has been previously shown to be de-graded via the proteasome (Fig. 3E), providing useful cor-roborative evidence for the efficacy and specificity ofMG132 as a proteasomal inhibitor in these cells.

PLIN1 439fs Mutant Is Associated With Smaller LipidDropletsAs perilipin 1 is critical for lipid accumulation in adipocytes(11), we assessed the impact of expression of the 439fsmutant on lipid droplet size in 3T3-L1 fibroblasts. After48 h of exposure to medium containing oleic acid, perilipin1 WT2overexpressing cells accumulated several large lipiddroplets, whereas cells expressing the perilipin 1 439fsmutant contained a more heterogeneous population ofsignificantly smaller lipid droplets (Fig. 4A and B). Therewas no difference in the ability of COOH-terminal peril-ipin 1 mutants to target to the lipid droplets (Fig. 4A).Perilipin 1 has previously been shown to exert its effecton lipid accumulation primarily through the inhibition ofbasal lipolysis rather than an increase in triglyceride syn-thesis (12). We thus hypothesized that the 439fs mutantmay facilitate higher rates of basal lipolysis. In order to testthis hypothesis, cells were incubated with oleic acid over-night, promoting lipid droplet accumulation, followed by

Figure 2—Predicted consequences of the perilipin 1 439fs mutantand documentation of its expression in adipose tissue. A: A sche-matic representation of the expected effects of the perilipin 1 439fsmutation on the protein domain organization. B: Immunoblots show-ing perilipin 1 and 2 expression in adipose tissue samples from pro-band A (visceral fat and subcutaneous leg fat sample) compared withexpression in subcutaneous abdominal fat samples from two con-trol subject samples. The N-terminal perilipin 1 antibody (GP29)detects both mutant (arrow) and WT perilipin 1, whereas theCOOH-terminal antibody only detects WT perilipin 1. Calnexinwas used as a loading control.

304 PLIN1 Mutation With Lipodystrophy Diabetes Volume 64, January 2015

a 4-h chase period in the presence of Triacsin C to inhibitfree fatty acid (FFA) re-esterification. WT perilipin 1 signif-icantly suppressed FFA release, whereas the 439fs perilipin1 mutant did so less effectively, albeit to a greater extentthan the 398fs perilipin 1 protein (Fig. 4C).

Perilipin 1 439fs Mutant Is Able to Bind ABHD5In an effort to reveal the molecular mechanisms un-derpinning the 439fs mutant’s failure to fully suppress

basal lipolysis, we used BiFC assays to assess the interac-tion of the perilipin 1 439fs with ABHD5 (Fig. 5). As boththe 439fs and 398fs mutants are expressed at lower levelsthan the WT protein, expression plasmids were titrated inorder to achieve equal protein expression. YFP signal wasreconstituted when ABHD5-Yn was expressed with eitherthe WT or 439fs mutant, but not in the presence of the398fs mutant, implying that 439fs is still able to interactwith ABHD5. This is in agreement with previous studies

Figure 3—Expression of the perilipin 1 439fs mutation in stably transduced 3T3-L1 preadipocyte cell lines. 3T3-L1 preadipocytes stablyexpressing either empty vector (EV) or N-terminally Myc-tagged perilipin 1 WT, perilipin 1 439fs, and perilipin 1 398fs mutants, which weregenerated as described in RESEARCH DESIGN AND METHODS. A: Immunoblot analysis of perilipin 1 expression using an anti-Myc antibody; b-actinwas used as a loading control. B: mRNA levels of hPLIN1 normalized to mCyclophilin A expression. C and D: Stably transfected 3T3-L1preadipocytes expressing WT and 439fs perilipin 1 were grown to confluency and treated with 100 mg/mL cycloheximide (CHX) for theindicated times. Control (CT) cells were not treated with CHX. Cells were lysed and perilipin 1 expression levels were detected by immuno-blotting with anti-Myc antibody and b-actin to assess loading. C: A representative image is shown. D: Quantification of perilipin 1 WT and439fs expression levels from three independent experiments. Data represented as mean 6 SEM. E: Stably transfected 3T3-L1 preadipocytesexpressing WT and 439fs perilipin 1 were grown to confluency and treated with 100 mg/mL CHX and/or 10 mmol/L MG132 or 25 mmol/LNH4Cl for 24 h. CT cells were not exposed to CHX. Cells were lysed and perilipin 1 and 2 expression levels detected by immunoblotting;b-actin was used to assess loading. A representative image is shown. *Nonspecific band; arrow denotes corresponding band.

diabetes.diabetesjournals.org Kozusko and Associates 305

where amino acids 382–429, equivalent to amino acids380–427 in the human sequence of perilipin 1, wereshown to be crucial for ABHD5 binding (13).

One could hypothesize that the conformational changeinduced by the COOH-terminal frameshift would allowthe 439fs mutant to bind ABHD5 but might notnecessarily prevent the subsequent interaction betweenABHD5 and ATGL, as occurs with WT perilipin 1. Wetested this hypothesis by assessing the ability of theperilipin 1 439fs mutant to inhibit direct interactionbetween ABHD5 and ATGL. In this assay, expression ofATGL (S47A)-Yc and ABHD5-Yn results in bright YFPfluorescence, which is reduced by ;70% when either WT

or the 439fs perilipin 1 mutant is expressed (Fig. 6). Incontrast, expression of the 398fs mutant has no effect onthe interaction between ATGL and ABHD5.

Expression of WT Perilipin 1 Does Not Rescue 439fsMutant PhenotypeFinally, in an effort to mimic the in vivo consequences ofthe heterozygous mutation, the 439fs mutant was coex-pressed with flag-tagged WT perilipin 1. 439fs proteinlevels were stabilized to some extent by the expression ofa WT copy of perilipin 1 (Fig. 7A). The results suggest thatincreasing the amount of WT perilipin 1 protein increaseslipid droplet volume (Fig. 7C and D), whereas coexpression

Figure 4—Characterization of the impact of the perilipin 1 439fs mutant on lipid droplet morphology and lipolysis. A: Stable cell lines weretreated with 400 mmol/L oleic acid for 48 h and fixed and stained with anti-Myc and anti-mouse Alexa Fluor 488 for perilipin 1 (green), LipidTOXDeep Red for lipid droplets (red), and DAPI for nuclei (blue); scale bar 10 mm. B: Lipid droplet (LD) volume was measured by taking an averageof 10 Z-stacks at 363 magnification with 2–3 cells per focal plane, followed by analysis using the Volocity 5 software (PerkinElmer). Resultsfrom three independent experiments are expressed as mean 6 SEM. C: Stable cell lines were loaded with 14C oleic acid for 16 h to label theintracellular triglyceride pool. Release of incorporated radioactivity was measured over 4 h in the presence of 6 mmol/L Triacsin C to preventfatty acid re-esterification. Lipolysis is expressed as the percentage of released radioactivity in the medium over total radioactivity incorpo-rated. Results from three independent experiments are expressed as mean 6 SEM. In B and C, post hoc pairwise comparisons are shownversus empty vector (EV) and between pairs, indicated with an overlying bar. ***P < 0.001, **P < 0.01.

306 PLIN1 Mutation With Lipodystrophy Diabetes Volume 64, January 2015

of either the 439fs or 398fs mutants fails to further in-crease lipid droplet volume. In accordance with these data,measurement of basal lipolysis in these cell lines showedless inhibition of basal lipolysis by the 439fs perilipin 1mutant than WT perilipin 1 even in the presence of WTperilipin 1 (Fig. 7E).

DISCUSSION

Lipodystophic syndromes are a rare but fascinating causeof insulin-resistant type 2 diabetes. The archetypal featureof all lipodystrophies is a lack of adipose tissue, which caneither be partial or generalized. Nevertheless, the conse-quences of lipodystrophy, which include dyslipidemia,NAFLD, hyperandrogenism, PCOS in women, and acceler-ated cardiovascular disease, are remarkably similar to thoseusually associated with obesity as part of the metabolicsyndrome (14). This observation in itself constitutes a keypiece of evidence underpinning the so-called lipid overflowhypothesis, which posits that the capacity of mammalian

adipose tissue to accommodate surplus lipid is finite, andthat when this capacity is exceeded, lipids accumulate inectopic sites, such as the liver and skeletal muscle, wherethey are instrumental in causing insulin resistance (15).

Within the last 15 years, several monogenic defectshave been causally linked to human lipodystrophicsyndromes, collectively advancing the understanding ofhuman adipogenesis and adipocyte function and thesystemic metabolic consequences of adipocyte dysfunc-tion (16,17). More recently, our laboratory has reportedthe consequences of loss-of-function mutations in pro-teins implicated in the formation and functional regula-tion of adipocyte lipid droplets (3,18). Interestingly,CIDEC and PLIN1 encode lipid droplet proteins almostexclusively expressed in white adipocytes, emphasizingfirst their importance in determining the unique propertiesof adipocyte lipid droplets and second the whole-organismconsequences of dysfunction of proteins primarily ex-pressed in adipocytes (perilipin 1 is also expressed in

Figure 5—The perilipin 1 439fs mutant interacts with ABHD5 on the surface of lipid droplets. Interaction between perilipin 1 (WT, 439fs, or398fs) and ABHD5 was assessed using BiFC in COS-7 cells after normalizing for equal expression of each protein. A: ABHD5-Yn and Myc-PLIN1-Yc (WT, 439fs, or 398fs mutants) were cotransfected in COS-7 cells and stained with anti-Myc and anti-mouse Alexa Fluor 594 forperilipin 1 (green) and DAPI for nuclei (blue). The presence of YFP BiFC signal indicates direct interaction between perilipin 1 and ABHD5; scalebar 10 mm. B: The YFP BiFC signal was quantified as described in RESEARCH DESIGN AND METHODS. Results from three independent experimentsare expressed as mean 6 SEM, where post hoc comparison is shown versus WT. A.U., arbitrary unit; ns, not significant. **P < 0.001. C: Theamount of Myc-perilipin 1 and ABHD5 expressed was assessed by immunoblotting. b-Actin was used as a loading control.

diabetes.diabetesjournals.org Kozusko and Associates 307

adrenal cells, and both proteins can be expressed in steatotichepatocytes).

The original description of patients with FPLD4 due toPLIN1 mutations was based on a description of sixaffected subjects. In order to better understand this novelphenotype, we have attempted to identify more affectedsubjects. These efforts enabled us to identify another twoprobands and four affected relatives with a differentPLIN1 mutation (439fs) to those previously reported. Thekey features of patients with the 439fs mutant includelack of peripheral fat depots, severe insulin resistance,and extreme hypertriglyceridemia. These features areconsistent with what was observed in patients with eitherthe PLIN1 398fs or 404fs mutations. The effect ofbariatric surgery on the metabolic phenotype of probandA was particularly striking. She experienced a 20-kgweight loss, with an 80% reduction in insulin require-ments, improved lipid profile, reversal of cardiomyopathy,and improvement in muscle strength. Her daughterexhibited a similar improvement in metabolic profile.These metabolic improvements are consistent with casereports describing the response to bariatric surgery inother forms of partial lipodystrophy (19–21) and high-light the importance of relieving the energetic stress im-posed upon their dysfunctional adipocytes. The beneficialsymptomatic and functional effect of bariatric surgery oncardiac function in obese patients with cardiomyopathy

has been described in a limited number of reports but notyet in patients with lipodystrophy (22).

Functional studies of the 439fs mutant suggest that itis associated with smaller lipid droplets and higher basallipolytic rates. These data are consistent with previouslyreported studies by Garcia et al. (11), who observed lesstriglyceride accumulation in cells overexpressing an arti-ficially generated perilipin 1 truncation mutant that trun-cates the protein at an amino acid very close to the site ofour 439fs mutant, namely E427 stop. However, under-standing why the 439fs mutant is associated with higherbasal lipolytic rates is more complex than was the case forthe previously reported human mutants (398fs and 404fs)as those fail to bind and stabilize ABHD5, whereas ourstudies suggest that the 439fs mutant does bind ABHD5,stabilize its expression, and sequester it from ATGL. Onepossible explanation might be the strikingly reduced ex-pression of the 439fs mutant, which was observed in tissuesamples from proband A and in transfected cells. The ab-errant COOH-terminus of 439fs mutant protein is likely tobe disordered, which would account for its propensity todegradation as occurs with many mutant proteins (23).Interestingly, even partial loss of perilipin 1 is sufficientto reduce fat mass as observed in Plin1 heterozygousknockout mice (7). Absent perilipin 1 expression in Plin1null mice was also associated with concomitant upregula-tion of perilipin 2 (7,24). Thus, the observed upregulation

Figure 6—The perilipin 1 439fs mutant inhibits the interaction between ATGL and ABHD5. Interaction between ATGL and ABHD5 wasassessed using BiFC in COS-7 cells after normalizing for equal expression of each protein. A: ABHD5-Yn, ATGL (S47A)-Yc, and myc-PLIN1(WT, 439fs, or 398fs mutants) were cotransfected in COS-7 cells and stained with anti-Myc and anti-mouse Alexa Fluor 594 for perilipin 1(green) and DAPI for nuclei (blue). The presence of YFP BiFC signal indicates a direct interaction between ATGL and ABHD5; scale bar 10mm. B: The YFP BiFC signal was quantified as described in RESEARCH DESIGN AND METHODS. Results from three independent experiments areexpressed as mean 6 SEM. Post hoc pairwise comparisons are shown versus empty vector (EV). ns, not significant. **P < 0.001.C: Expression of Myc-perilipin 1, ABHD5, and ATGL was assessed by immunoblotting. b-Actin was used as a loading control.

308 PLIN1 Mutation With Lipodystrophy Diabetes Volume 64, January 2015

Figure 7—Characterization of the consequences of the PLIN1 439fs mutation in preadipocytes coexpressing WT perilipin 1. 3T3-L1preadipocytes stably expressing pLXSN-Flag-PLIN1-WT were infected with pBABE-puro empty vector (EV) or Myc-PLIN1 (WT, 439fs,or 398fs mutants) as described in RESEARCH DESIGN AND METHODS. A: Immunoblot analysis of perilipin 1 expression using an anti-Myc or anti-Flag antibody. b-Actin was assessed as a loading control. B: mRNA levels of total hPLIN1 normalized to mCyclophilin A expression. C:Stable cell lines were treated with 400 mmol/L oleic acid for 48 h and fixed and stained with anti-Myc and anti-mouse Alexa Fluor 488 forperilipin 1 (green), LipidTOX Deep Red for lipid droplets (red), and DAPI for nuclei (blue); scale bar 10 mm. D: Lipid droplet (LD) volume wasmeasured by taking the average of 10 Z-stacks at 363 magnification with 2–3 cells per focal plane, followed by analysis using Volocity 5software. Results from three independent experiments are expressed as mean 6 SEM. E: Stable cell lines were loaded with 14C oleic acidfor 16 h to label the intracellular triglyceride pool. Release of incorporated radioactivity was measured over 4 h in the presence of 6 mmol/LTriacsin C to prevent fatty acid re-esterification. Lipolysis is expressed as the percentage of released radioactivity in the media over the totalradioactivity incorporated. Results from three independent experiments are expressed as mean 6 SEM. In D and E, post hoc pairwisecomparisons are shown versus EV, and between pairs, indicated with an overlying bar. *P < 0.05; **P < 0.01; ***P < 0.001.

diabetes.diabetesjournals.org Kozusko and Associates 309

of perilipin 2 expression in fat biopsies from proband Aprovides further evidence for reduced total perilipin 1 expres-sion in vivo. We cautiously conclude that reduced perilipin 1expression is likely to be an important factor in the pheno-type observed in patients with the 439fs PLIN1 mutation. Itis also very likely that the COOH-terminus of perilipin 1 hasadditional, as yet incompletely understood, biological func-tions that are likely to be impaired by this mutation.

In summary, we report the clinical and molecularcharacterization of a novel PLIN1 frameshift mutation intwo kindreds with FPLD4. The severity of metabolic diseaseseen in affected patients further highlights the importanceof precise lipolytic regulation in adipose tissue in humanhealth. Finally, the highly beneficial response of twopatients with this disorder to bariatric surgery emphasizesthe importance of weight loss in these patients. Surgerylessens ectopic fat deposition and offers at least one usefulmanagement option for an otherwise devastating metabolicillness.

Acknowledgments. The authors are very grateful to the patients whoparticipated in these studies and to the medical staff who assisted in theirmanagement. This work was supported by grants from the Wellcome Trust(D.B.S., S.O., I.B.), the U.K. National Institute for Health Research CambridgeBiomedical Research Centre, the Medical Research Council Centre for Obesityand Related Metabolic Diseases, and a Biotechnology and Biological SciencesResearch Council CASE studentship (K.K.).Duality of Interest. No potential conflicts of interest relevant to this articlewere reported.Author Contributions. K.K. performed functional studies. K.K., V.H.M.T.,S.P., J.R.G., L.V.C., and D.B.S. contributed to data analysis and discussion andwrote the manuscript. V.H.M.T., Y.-H.C., S.S., and L.V.C. undertook clinical studies.K.K., W.B., and I.I. performed genotyping. S.G., M.M., K.L., and S.P. assisted withfunctional studies. V.S. provided functional insight into the expected consequencesof the mutation. D.B.S. designed the study in collaboration with S.O. and I.B. Allauthors reviewed and edited the manuscript. D.B.S. and L.V.C. are the guarantorsof this work and, as such, had full access to all the data in the study and takeresponsibility for the integrity of the data and the accuracy of the data analysis.

References1. Bickel PE, Tansey JT, Welte MA. PAT proteins, an ancient family of lipiddroplet proteins that regulate cellular lipid stores. Biochim Biophys Acta 2009;1791:419–4402. Zechner R, Zimmermann R, Eichmann TO, et al. FAT SIGNALS—lipases andlipolysis in lipid metabolism and signaling. Cell Metab 2012;15:279–2913. Gandotra S, Le Dour C, Bottomley W, et al. Perilipin deficiency and auto-somal dominant partial lipodystrophy. N Engl J Med 2011;364:740–7484. Gandotra S, Lim K, Girousse A, Saudek V, O’Rahilly S, Savage DB.Human frame shift mutations affecting the carboxyl terminus of perilipin in-crease lipolysis by failing to sequester the adipose triglyceride lipase (ATGL)coactivator AB-hydrolase-containing 5 (ABHD5). J Biol Chem 2011;286:34998–35006

5. Donadille B, Lascols O, Capeau J, Vigouroux C. Etiological investigations inapparent type 2 diabetes: when to search for lamin A/C mutations? DiabetesMetab 2005;31:527–5326. Patel S, Yang W, Kozusko K, Saudek V, Savage DB. Perilipins 2 and 3 lacka carboxy-terminal domain present in perilipin 1 involved in sequesteringABHD5 and suppressing basal lipolysis. Proc Natl Acad Sci U S A 2014;111:9163291687. Tansey JT, Sztalryd C, Gruia-Gray J, et al. Perilipin ablation results in a leanmouse with aberrant adipocyte lipolysis, enhanced leptin production, and re-sistance to diet-induced obesity. Proc Natl Acad Sci U S A 2001;98:6494–64998. Xu G, Sztalryd C, Londos C. Degradation of perilipin is mediated throughubiquitination-proteasome pathway. Biochim Biophys Acta 2006;1761:83–909. Brasaemle DL, Barber T, Kimmel AR, Londos C. Post-translational regulationof perilipin expression. Stabilization by stored intracellular neutral lipids. J BiolChem 1997;272:9378–938710. Kovsan J, Ben-Romano R, Souza SC, Greenberg AS, Rudich A. Regulation ofadipocyte lipolysis by degradation of the perilipin protein: nelfinavir enhanceslysosome-mediated perilipin proteolysis. J Biol Chem 2007;282:21704–2171111. Garcia A, Subramanian V, Sekowski A, Bhattacharyya S, Love MW,Brasaemle DL. The amino and carboxyl termini of perilipin a facilitate the storageof triacylglycerols. J Biol Chem 2004;279:8409–841612. Brasaemle DL, Rubin B, Harten IA, Gruia-Gray J, Kimmel AR, Londos C.Perilipin A increases triacylglycerol storage by decreasing the rate of tri-acylglycerol hydrolysis. J Biol Chem 2000;275:38486–3849313. Subramanian V, Rothenberg A, Gomez C, et al. Perilipin A mediates thereversible binding of CGI-58 to lipid droplets in 3T3-L1 adipocytes. J Biol Chem2004;279:42062–4207114. Garg A, Agarwal AK. Lipodystrophies: disorders of adipose tissue biology.Biochim Biophys Acta 2009;1791:507–51315. Savage DB, Petersen KF, Shulman GI. Disordered lipid metabolism and thepathogenesis of insulin resistance. Physiol Rev 2007;87:507–52016. Vigouroux C, Caron-Debarle M, Le Dour C, Magré J, Capeau J. Molecularmechanisms of human lipodystrophies: from adipocyte lipid droplet to oxidativestress and lipotoxicity. Int J Biochem Cell Biol 2011;43:862–87617. Huang-Doran I, Sleigh A, Rochford JJ, O’Rahilly S, Savage DB. Lipodys-trophy: metabolic insights from a rare disorder. J Endocrinol 2010;207:245–25518. Rubio-Cabezas O, Puri V, Murano I, et al.; LD Screening Consortium. Partiallipodystrophy and insulin resistant diabetes in a patient with a homozygousnonsense mutation in CIDEC. EMBO Mol Med 2009;1:280–28719. Ciudin A, Baena-Fustegueras JA, Fort JM, Encabo G, Mesa J, Lecube A.Successful treatment for the Dunnigan-type familial partial lipodystrophy withRoux-en-Y gastric bypass. Clin Endocrinol (Oxf) 2011;75:403–40420. McGrath NM, Krishna G. Gastric bypass for insulin resistance due tolipodystrophy. Obes Surg 2006;16:1542–154421. Utzschneider KM, Trence DL. Effectiveness of gastric bypass surgery ina patient with familial partial lipodystrophy. Diabetes Care 2006;29:1380–138222. Grapsa J, Tan TC, Paschou SA, et al. The effect of bariatric surgery onechocardiographic indices: a review of the literature. Eur J Clin Invest 2013;43:1224–123023. Yue P, Li Z, Moult J. Loss of protein structure stability as a major causativefactor in monogenic disease. J Mol Biol 2005;353:459–47324. Martinez-Botas J, Anderson JB, Tessier D, et al. Absence of perilipin resultsin leanness and reverses obesity in Lepr(db/db) mice. Nat Genet 2000;26:474–479

310 PLIN1 Mutation With Lipodystrophy Diabetes Volume 64, January 2015