Brian Covello: Diabetes Research Paper

1,25-Dihydroxyvitamin D3 and Retinoic Acid in vitro Modulation of PPARγ and Insulin Resistance in

Diabetes Mellitus II  

Brian Covello Florida Southern College

Research: Molecular Biology December 1st, 2012

Covello   1  

Abstract

1,25-Dihydroxyvitamin D3 (D3) and Retinoic Acid (RA) have previously been reported to

down-regulate the master controller of adipogenesis, nuclear receptor PPARγ (Yoshifumi, 1998).

This receptor plays a significant role in understanding diabetes mellitus II, and a class of anti-

diabetic drugs has been found to agonistically bind to PPARγ. For the first time, this study seeks

to combine the aforementioned metabolites together in one treatment scheme. This is a crucial

component for creating a direct relation to in vivo studies, as only concentrations of these

metabolites are found in human serum. We hypothesized that a combined treatment of D3 and

RA will have a synergistic effect upon PPARγ, causing novel protein changes when compared to

individual treatments and leading to a greater down-regulation than previously reported

(Yoshifumi, 1998). Through immunofluorescence microscopy and SDS-PAGE, one is able to see

localization of PPARγ throughout all treatments compared to the positive control, indicating no

mechanism of action for delocalization. Several protein band changes were noted for micromolar

combination and nanomolar combination treatment. Further research through western blot is

warranted to identify these proteins.

Covello   2  

Introduction

Twenty   five   million   people   within   the   United   States   are   currently   inflicted   by  

Diabetes  Mellitus  2  (Powers,  2005).  If  current  statistical  trends  continue,  by  2050,  1  in  3  US  

adults  will   have  diabetes   (Powers,   2005).  Among   adults,   diabetes   is   the   leading   cause   of  

kidney   failure,   new   cases   of   blindness,   and   amputations   of   the   lower   extremities   not  

related   to   accidents   or   injury   (Powers,   2005).   Insulin   is   an   endogenous   hormone  

synthesized  in  the  beta  cells  of   the   islet  of  Langerhans   located  in  the  pancreas.   In  normal  

human  physiology,   an   increase   in   blood   glucose   levels   stimulates   release   of   insulin.   This  

hormone  is  responsible  for  regulating  glucose  metabolism  through  an  increased  uptake  of  

glucose  and  a  decrease  in  gluconeogenesis.  Cellular  uptake  of  glucose  allows  it  to  be  broken  

down   for   energy.   There   are   two   main   types   of   diabetes.   In   type   one,   the   body   fails   to  

produce  insulin.  In  type  two,  the  body  produces  insulin,  yet  cells  have  acquired  a  resistance  

to   insulin   (Powers,   2005).   Type   two   diabetes   has   been   heavily   correlated   with   obesity,  

particularly   in   the   form   of   new   adipocyte   formation   in   the   abdominal   section  

(Wajchenberg,   2000).   Some   of   the  most   promising   research   of   type   two   diabetes   lies   in  

examination  of  insulin  resistance  and  adipocyte  differentiation.    

Peroxisome   proliferated   activated   receptors   (PPAR)   are   a   family   of   nuclear  

receptors   that   act   in   coordination   with   a   ligand   to   become   bound   complexes   that   are  

capable   of   DNA   transcription   (Liang,   2006).   There   are   three   subtypes   of   PPAR:   PPARα,  

PPARβ,  and  PPARγ.  PPARα  expression  is  highest   in  the  liver,  where  it   is  thought  that  this  

receptor   acts   to   regulate   fatty   acid   metabolism   (Liang,   2006).   PPARβ   is   ubiquitously  

expressed   throughout   the   body   and   is   heavily   correlated   with   cancer,   infertility,   and  

dyslipidemia   (Diaradourian,   2005).   Finally,   PPARγ,   the   receptor   of   interest   for   this  

research,  is  predominantly  found  in  tissues  of  the  kidney,  liver  and  small  intestine,  where  it  

is  found  to  modulate  adipocyte  differentiation  and  insulin  sensitivity  (Diradourian,  2005).  

The   physiological   consequences   of   PPARγ   protein   expression  make   it   a   prime   target   for  

studying  type  two  diabetes.  PPARγ  has  three  main  isoforms,  namely,  PPARγ1,  PPARγ2,  and  

PPARγ3.  (Bogazzi,  2007).  These  isoforms  arise  due  to  different  promoter  usage  and  splicing,  

which   may   ultimately   lead   to   transcription   of   different   genes   (Desvergne,   1999).   These  

isoforms,   combined   with   variations   of   ligands,   co-­‐activators,   co-­‐repressors   and   post-­‐

Covello   3  

translational  modifications   through  phosphorylation   lead   to   a  high  degree  of   complexity.  

Upon   ligand   binding,   PPARγ  heterodimerizes  with   retinoid   X   receptors   (RXR),   forming   a  

PPARγ-­‐RXR   complex   that   can   bind   to   the   PPAR   response   element   (PPRE)   and   cause  

transcription   to   take   place,   ultimately   leading   to   transactivation   protein   changes  

(Desvergne,   1999).   Transactivation   activity   includes   up-­‐regulation   of   the   adipocyte   fatty  

acid-­‐binding   protein,   acyl-­‐CoA   synthase,   lipoprotein   lipase,   c-­‐Cbl   associating   protein,  

phosphoenolpyruvate   carboxykinase,   fatty   acid   transport   protein,   and   insulin   receptor  

substrate   2,   and   down-­‐regulation   of   TNFα,   leptin,   plasminogen   activator   inhibitor-­‐1,  

resistin,   IL-­‐6  and  11-­‐beta  hydroxysteroid  dehydrogenase   type  1  (Rangwala,  2005).  These  

proteins  are  responsible  for  lipid  metabolism  and  regulation  of  insulin  resistivity.    

A  class  of  anti-­‐diabetic  drugs  known  as  thiazolidinedione  derivatives  (TZDs),  work  

as   agonistic   ligands   on   PPARγ   (Diradourian,   2005).   Thus,   TZDs   are   thought   to   enhance  

insulin  sensitivity  through  binding  and  subsequent  regulation  of  PPARγ  (Grimaldi,  2007).    

In   addition   to   thiazolidinediones   -­‐   fatty   acids,   hypolipidemic   drugs,   plasticizers,   and  

steroids  have  all  been  shown  to  activate  PPARs  (Kong,  2006).      

RXR  supply  within  a  cell  is  limited,  and  a  series  of  other  receptors,  such  as  vitamin  D  

receptor  (VDR)  and  retinoid  A  receptor  (RAR),  also  heterodimerize  with  RXR  (Yoshifumi,  

1998).   Thus,   competition   abounds   between   PPARγ,   VDR,   and   RAR   for   a   common  

heterodimeric   partner   (Yoshifumi,   1998).   This   fact   has   correlated   VDR,   RAR,   and   their  

corresponding   ligands,   1,25-­‐dihydroxyvitamin   D3   and   retinoic   acid   respectively,   into  

studies  of  PPARγ  and  diabetes  mellitus  type  two.   In   fact,  vitamin  D   insufficiency  has   long  

been   associated   with   obesity   (Martini,   2005).   Earlier   this   year   the   American   Medical  

Association  recommended  an  increase  in  vitamin  D  consumption  for  infants.    

In   response   to   ultraviolet   radiation,   epithelial   cells   produce   fat-­‐soluble   vitamin  D,  

which  is  converted  to  its  metabolite,  1,25-­‐dihydroxyvitamin  D3  by  the  liver  (Kong,  2006).  In  

1988  a  group  of  researchers  hypothesized  and  showed  that  1,25-­‐dihydroxyvitamin  D3  has  

an  inhibitory  effect  on  proliferation  and  differentiation  of  3T3-­‐L1  cells  (Ishida,  1988).  They  

tested  concentrations  of   the  vitamin  D  metabolites  at  10-­‐10,10-­‐9,and  10-­‐8M  (Ishida,  1988).  

This   paper   reported   a   decrease   in   differentiation   and  proliferation  of   adipocytes   of   67%  

less   than   the   control   at   10-­‐8  molar   concentrations.   Even   at   10-­‐10  molar   concentration   of  

1,25-­‐dihydroxyvitamin   D3,   true   human   serum   concentrations,   a   significant   reduction   of  

Covello   4  

adipocyte   differentiation  was   examined,   yet   an   exact   cellular  method  was   still   unknown  

(Ishida,   1988).     Ten   years   later,   a   separate   group   of   researchers   showed   that   1,25-­‐

dihydroxyvitamin   D3   inhibits   PPARγ   expression.   This   research   led   to   the   beginning   of  

understanding   the   underlying   mechanism   responsible   for   inhibition   of   adipocyte  

differentiation  and  proliferation  (Yoshifumi,  1998).    This  research  concludes  that  vitamin  D  

insufficiency  may  cause  obesity  through  repressed  inhibition  of  PPARγ.  The  consequences  

of   an   increased   amount   of   PPARγ  would   correlate   to   increased   adipocyte   differentiation  

and  obesity.  Retinol  or  Vitamin  A,  is  another  fat  soluble  vitamin  (Yoshifumi,  1998).  Retinol  

may  be  taken  up  by  cells  and  oxidized  to  retinoic  acid  by  a  series  of  enzymes.  Retinoic  acid  

also   produced   an   inhibitory   effect   on   the   expression   of   PPARγ.   1μM   concentrations   of  

retinoic   acid   and   1,25(OH)2D3   were   utilized,   and   it   was   found   that   1,25(OH)2D3   had   a  

greater  inhibitory  effect  on  PPARγ  expression  than  retinoic  acid  (Yoshifumi,  1998).    

Several  gaps  exist  between  amongst  previous   research   that  deserves  examination.  

The  effects  of  1,25-­‐dihydroxyvitamin  D3  and  retinoic  acid  have  only  been  tested  on  3T3-­‐L1  

cells   in   the  process  of  differentiation.  No  effects  have  been  tested  on  pre-­‐adipocytes  with  

non-­‐induced  differentiated  or  terminally  differentiated  adipocytes.  3T3-­‐L1  is  a  cell   line  of  

pre-­‐adipocytes  that  express  PPARγ.  Upon  correct  treatment,  these  cells  can  be  induced  to  

differentiate   into   adipocytes,   making   them   a   prime   target   for   adipogenesis   research.  

Additionally,   all   research   conducted   of   the   effects   of   these   metabolites   have   been  

conducted  on  PPARγ  that  is  ligand  bound  to  TZDs.  No  research  exists  which  examines  the  

effects  of  these  metabolites  on  non-­‐ligand  bound  PPARγ.  Mixed  concentrations  of  these  two  

substances  have  also  not  been  tested  on  PPARγ  protein  expression.  Research  into  mixtures  

of   these  metabolites   is   crucial,   as   only  mixtures   of   these  metabolites   are   present   in   the  

human  body.  Research  into  this  area  may  be  able  to  better  predict  in  vivo  results.  There  is  

currently   a   wealth   of   research   into   ligand   binding   activity   of   PPARγ,   however,   ligand  

binding   to   PPARγ   does   not   necessarily   correlate   to   transactivation,   as   cofactors,   co-­‐

repressors  and  phosphorylation  form  a  complex  sequence  of  events  from  DNA  to  mRNA  to  

protein  (Diardourian,  2005).  In  order  to  better  understand  insulin  resistance  with  respect  

to  type  two  diabetes,  transactivation  studies  must  be  conducted.  Surprisingly,  research  into  

transactivation  protein  changes  with  regard  to  PPARγ  modulation  is  lacking.    

Covello   5  

This   proposal   incorporates   three  main   installments   of   research   to   be   carried   out  

over  multiple  semesters.  The  first  part  of  this  proposal  will  specifically  examine  the  mixed  

effects   of   1,25-­‐dihydroxyvitamin   D3   and   retinoic   acid   on   PPARγ   expression   in   chinese  

hamster   ovary   cells.   The   second   portion   of   this   research  will   focus   on   the   effects   of   the  

same  metabolites  on  3T3-­‐L1  cells  the  have  been  induced  to  differentiate.  The  last  portion  of  

this   research   project   will   deal   specifically   with   transactivation   protein   expression   of  

PPARγ,  and  the  study  will  focus  on  a  number  of  different  proteins  that  play  an  integral  role  

in  insulin  resistance,  glucose  metabolism  and  gluconeogenesis.  At  this  stage,  other  variants  

of   PPAR  may  be   studied   in   order   to   gain   a   better   perspective   of   biochemical   action.  The  

goal  of   this   research  project   is   to   fill   the  gaps   in   the  existing  research.  By  utilizing  mixed  

concentrations  of  metabolites,  one  may  begin   to  piece   together  a  more  holistic  picture  of  

PPARγ’s   role   in   insulin   resistance   inside   an   organism.   The   hypothesis   proposed   is   that  

mixed   concentrations   of   1,25-­‐diyhydroxyvitamin   D3   and   retinoic   acid   will   have   an   even  

greater   inhibitory   effect   on   PPARγ   protein   expression   than   1,25-­‐dihydroxyvitamin   D3  

alone,   and   less   1,25-­‐dihydroxyvitamin  D3  will   be   needed   to   produce   the   same   inhibitory  

effect   when   it   is   mixed   with   small   amounts   of   retinoic   acid.   Also,   not   only   will   PPARγ  

protein   expression   be   down-­‐regulated   as   a   result   of   combining   the   metabolites,   but   all  

downstream  consequences  both  during  adipocyte  differentiation  and  transactively  should  

be  affected  accordingly.    

Covello   6  

Fig  1A  portrays  the  intricate  pathway  of  PPARγ  in  adipocytes.  Upon  ligand  binding,  

PPARγ   heterodimerizes   with   RXR   to   act   as   a   transcription   factor,   inducing   a   number   of  

protein  changes  that  ultimately  control  insulin  resistance  and  adipogenesis.  

Fig   1B   shows   PPARγ   (purple)   heterodimerized   with   RXR   (green)   near   the   PPAR  

response  element  (PPRE).  Only  when  bound  to  RXR  can  PPARγ  act  as  a  transcription  factor.  

VDR  and  RAR  compete  with  PPARγ  for  a  limited  amount  of  RXR.  

Materials and Methods Materials

1,25 (OH)2 D3 was purchased from Sigma (St. Louis, MO, USA) and RA from Sigma (St. Louis,

MO, USA). Primary antibody was from BD transduction laboratories and was a polyclonal rabbit

anti-PPAR specification. Secondary antibody was alexa fluora goat anti-rabbit SFX kit. All other

chemical were guaranteed reagent grade or tissue culture grade.

Cell Culture & Treatments

Treatments were conducted for 24 hours each. Treatment numbers for lysate correspond to lane

numbers in Figure 7. Dr. Emily Bradshaw maintained Chinese hamster ovary cells. Cells were

grown until 70% confluence. RA and 1,25 (OH)2 D3 were dissolved in a solution containing

ethanol, whose concentration in each well was less than 0.1% v/v and added at the same time to

incomplete media and ultimately the cells. Cells were investigated before treatment and after

treatment for confluence and survival.

SDS-PAGE

Upon finalization of treatment, a protein lysate was generated, in which nuclei were removed

from the cells. Enumeration of protein took place through use of spectrophotometry. A standard

curve was generated, for use as a comparison for the protein curve. Using the ascertained lysate

concentrations, identical amounts of proteins were pipetted into each well. Separation of proteins

took place through SDS-PAGE. A molecular marker was utilized in the first well, and weight

was measured in kDa. Two trials took place.

Covello   7  

Immunofluorescence

Cells were washed 3x with PBS-MgCl2 quickly at room temperature, and then fixed utilizing 3%

paraformaldehyde. Cells were washed again and then permeabilized with solutions of NH4Cl and

Triton 0.1%x. They were then blocked for 1 hour with PBS-MgCl2-BSA. Cells were placed in a

1:25 dilution with primary antibody. This was followed by an additional wash and staining with

a 1:25 dilution of secondary antibody. Sun exposure was avoided during this process as the

secondary antibody is photosensitive. Glass coverslips were seated onto mounting media. All

pictures were taken with a Canon. Bright light microscopy was taken with 40X objective, while

fluorescent pictures were taken under oil immersion with a 100X objective. Two trials took

place. All photographs were taken by Brian Covello with a Canon on manual settings.

Table 1

Table 1 represents the treatment matrix for lysate generation. Incomplete media was

utilized as the control, as this was the solution that the vitamin metabolites were placed into.

Table 2

Table 2 represents the treatment matrix for immunofluorescence microscopy. Incomplete

media was utilized as a negative control, while 1% EtOH was utilized as a positive control (Powers, 2005).

Covello   8  

Results

Figure 2 (Concentration Curve)

Figure 2A-2H corresponds to immunofluorescence treatments 1-8. Figure 2 above

indicates that none of the treatments had an abrupt effect on cell shape or concentration. Cells

were treated at about 70% confluence, and they retained this confluence level throughout the

treatments. No detrimental effects are shown in these pictures. All pictures were taken under a

40X objective lens by Brian Covello.

Covello   9  

Figures 4A-4L and Figures 5A-5H (Immunofluorescence)

Figures 5A-5H clearly show elongated mammalian cells with no evidence of bacteria or

additional microorganisms. Cells have maintained a confluence between 70-80%. In addition, the

edges of the plasma membranes of these cells are able to be distinguished from the background.

For further evidence of survival, Figure 5C, 5E, and 5H depicts cells mid-cytokinesis. Figure

4A shows clear localization of protein around the nucleus of the cells. The positive control cell in

Figure 4B indicates translocation of protein into the cytosol. Figures 4C-4E indicate

localization of protein similar to that of the negative control. When Figures 4E, 4F, and 4G are

compared to the negative control, slight staining outside of the nucleus may be seen, however,

compared to the positive control these figures fail to produce the same type of drastic

translocation into the cytosol. Figures 4I-4L have been manipulated with adobe photoshop to

Covello   10  

better depict the boundaries of protein staining. Even adjusting saturation for better view of

background staining fails to show staining of the cytosol.

Figure 6A (Standard Curve)

Figure 6A was generated using protein standards with known concentrations and

obtaining absorbance with a spectrophotometer. By utilizing the equation for this line, one is able

to calculate an unknown protein concentration for a treated lysate given the absorbance. In this

manner, the table was created below.

Table 3

Covello   11  

Table 3 shows treatment in the left column and concentration of the lysates generated by

the treatment in the right column. Using these numbers 50 micrograms of protein was measured

in milliliters and placed in one of the five wells for SDS-PAGE.

Figure 7A-B (SDS-PAGE)

Protein concentrates of prepared lysates were determined using a standard curve

generated with an R-squared value of 0.9895. Figure 7A shows disappearance of protein in lanes

4 and 5 in the 28kDa to 33kDa range, while Figure 7B shows disappearance of protein greater

than 155kDa and between 96kDa and 71kDa.

Discussion The present study has demonstrated that RA and 1,25 (OH)2 D3 do not cause any

translocation of PPARγ out of the nucleus and into the cytosol. However, several protein bands

disappeared with SDS-PAGE analysis. It appears that a combined treatment of RA and 1,25

(OH)2 D3 cause protein changes that are not present in solitary metabolite treatments.

A bidirectional approach of immunofluorescence microscopy and SDS-page was

implemented to analyze the effects of RA and 1,25 (OH)2 D3 on nuclear receptor PPARγ. The

concentration of cells did not change throughout any of the treatments, and the bright light

microscopy in figure 5 depict elongated cells with apparent nuclei, indicating that cells survived

the treatment. Additionally, the concentrations of cells were maintained. The negative control of

incomplete media shows localization of PPARγ around the nucleus. This is to be expected, as

Covello   12  

PPARγ is a nuclear receptor. The positive control of 1% EtOH indicates diffusion of PPARγ out

of the nucleus. None of the treatments show such distinct and broad translocation out of the

nucleus when compared to the positive control. Even pictures specifically enhanced to show

staining not originally caught on camera in figures 4I- 4L fail to portray staining outside of the

nucleus. Thus, this data indicates that neither vitamin D treatment, vitamin A treatment, or a

combined mixture of the two translocate PPARγ out of the nucleus. This however, does not

exclude the possibility of PPARγ down-regulation or modulation, as these changes cannot be

detected through immunofluorescence microscopy.

Comparing mixed treatments to individual treatments, mixed RA and 1,25 (OH)2 D3 caused

a disappearance of protein between 33 and 48kDa. Thus, this protein change may take place

through a synergistic mechanism of combined RA and 1,25 (OH)2 D3, ultimately providing some

support to the original hypothesis. Additionally, the nanomolar combined treatment portrays a

disappearance of a protein between 71kDa and 96kDa. It is interesting to note that this change

does not take place with the micromolar combined treatment. This warrants further investigation

as nanomolar concentrations occur in human serum. Ultimately, this data indicates no changes in

positioning of PPARγ as caused by 1,25-dihydroxyvitamin D3 or retinoic acid, but several

protein changes do occur in the combined treatment that are absent in individual treatments.

There are several advancements that must be taken for more conclusive evidence. First,

although a standard curve was generated with an R-squared value of 0.9895, it seems as if

different protein amounts were placed into the wells. This may be due to human or mathematical

error. In order to exclude the possibility that changes in SDS-page occurred through this original

protein distribution, another standard curve must be calculated and a third trial must take place.

Additionally, it is still unknown if PPARγ was up-regulated or down-regulated by the treatments.

In order to analyze this data, western blot analysis must take place. In addition to examination of

PPARγ, western blot may also indicate which proteins disappeared with the combined

treatments.

A transition to 3T3-L1 pre-adipocytes will also prove beneficial in the future, as these

cells are better able to predict consequences as related to diabetes mellitus type II. Insulin,

dexamethasone and 3-isobutyl-1-methylxanthine (IBMX) at concentrations of around 1

microgram/mL, 0.25micromolar, and 0.5mMrespectively have been shown to effectively induce

Covello   13  

differentiation of 3T3-L1 cells (Tang et al., 2012). By inducing differentiation, one may better

predict in vivo results.

Ultimately, this research was successful in beginning to unravel the complexity of

interactions of RA, 1,25 (OH)2 D3, and PPARγ. This study provides a strong foothold for future

studies aimed at bridging basic and translational research for diabetes mellitus type two. The

objectives for this research are widespread. In vitro studies that directly mimic human

physiology may help explain drug interactions. By unraveling these mechanisms, this research

may lead to better dietary suggestions for those patients currently taking TZDs as an anti-diabetic

treatment. While some support was provided for the original hypothesis, more research is

warranted for a better understanding of the PPARγ pathway and its role in adipogenesis and

insulin resistance.

Acknowledgements Thank you Florida Southern College for providing the funding and laboratory space that

made this project possible. Dr. Jung Liu from Mayo Clinic for the 3T3-L1 Cells, Dr. Emily

Bradshaw for material acquisition and cell culture maintenance along with much needed

guidance and direction, and fellow researchers and colleagues for their support.

Covello   14  

Sources Barroso, I. B. (1999). Dominant negative mutations in human ppar gamma associated with severe insulin resistance, diabetes mellitus and hypertension. Letters to Nature, 402(23), 880-889. Bogazzi, F. (2007). Abnormal expression of ppar gamma isoforms in the subcutaneous adipose tissue of patients with cushing's disease. Clinical Endocrinology, 1365(66), 7-12. Desvergne B, Wahli W (1999). Peroxisome proliferator-activated receptors: nuclear control of metabolism. Endocr Rev (20), 649-88. Diradourian C, Girard J, Pegorier JP (2005). Phosphorylation of PPARS: from molecular characterization to physiological relevance. Biochimie (87), 33-8. Grimaldi PA (2007). Lipid Sensing and lipid sensors: peroxisome proliferator activated receptors as sensors of fatty acids and derivatives. Cell Mol Life Sci (64),2459-64. Ishida, Y. (1988). Possible involvement of 1,25-dihydroxyvitamine d3 in proliferation and differentiation of 3t3-l1 cells. Biochemical and Biophysical Research Communications, 151(3), 1122-1127. Kong J, Li YC. Molecular mechanism of 1,25-dihydroxyvitamin D3 inhibition of adipogenesis in 3T3-L1 cells (2006). Am J Physiol Endocrinol Metab (290), E916-924. Liang, G. L. (2006). Peroxisome proliferator activated receptor gamma as a drug target in

the pathogenesis of insulin resistance.Pharmacology & Therapeutics, 111(4), 145-173. Martini LA, Wood RJ (2006). Vitamin D status and the metabolic syndrome. Nutr Rev (64),479-86. Powers, A. C. (2005). Chapter 323. Diabetes mellitus. In D. L. Kasper, A. S. Fauci, D. L. Longo, E. Braunwald, S. L. Hauser, & J. L. Jameson, Harrison’s principles of internal medicine (16th ed.). The McGraw-Hill Companies, Inc. Rangwala, S.M., et al (2003). Genetic modulation of PPARgamma phosphorylation regulates insulin sensitivity. Dev Cell (5),657-63 Tang, Q., & Lane, M. (2012). Adipogenesis: from stem cell to adipocyte. Annual Review Of Biochemistry, 81715-736. Yoshifumi. (1998). Counteraction of retinoic acid and 1,25-dihydroxyvitamin d3 on up- regulation of adipocyte differentiation with ppar ligand, an antidiabetic thiazolidinedione, in 3t3-l1 cells. Pharmacology Letters, 62(14), 205-211. Wajchenberg, B. L. (2000). Subcutaneous and visceral adipose tissue: their relation to the metabolic syndrome. Endocr Rev (21), 697–738.