M Bonnet, I Gourdou, C Leroux, Y Chilliard and J Djianeadipose, epithelial, and myoepithelial cells during pregnancy and lactation.

Leptin expression in the ovine mammary gland: putative sequential involvement of

2002, 80:723-728.J ANIM SCI

http://www.journalofanimalscience.org/content/80/3/723the World Wide Web at:

The online version of this article, along with updated information and services, is located on

www.asas.org

by guest on June 6, 2014www.journalofanimalscience.orgDownloaded from by guest on June 6, 2014www.journalofanimalscience.orgDownloaded from

Leptin expression in the ovine mammary gland: Putative sequentialinvolvement of adipose, epithelial, and myoepithelial cells

during pregnancy and lactation1

M. Bonnet*2, I. Gourdou, C. Leroux, Y. Chilliard*, and J. Djiane

*INRA, Unite de Recherches sur les Herbivores, Equipe Tissu Adipeux et Lipides du Lait,63122 Saint-Gene`s-Champanelle, France and INRA, Laboratoire de Biologie Cellulaire et Moleculaire and

Laboratoire de Genetique biochimique et de Cytogenetique, 78352 Jouy-en-Josas cedex, France

ABSTRACT: We examined the ability of the ovinemammary gland to synthesize leptin throughout preg-nancy and lactation. Leptin gene expression was as-sayed by real-time reverse transcription and polymer-ase chain reaction in mammary gland from ewes at 15,80, 106, 112, or 141 d of pregnancy and at 0 (30 minafter parturition), 3, 48, or 70 d of lactation. LeptinmRNA level was high at the beginning (the first 80 d)and at the end of pregnancy and was lower at mid-pregnancy and throughout lactation. Furthermore, dur-ing these periods of mammary leptin expression, the

Key Words: Lactation, Leptin, Mammary Tissue, Pregnancy, Sheep

2002 American Society of Animal Science. All rights reserved. J. Anim. Sci. 2002. 80:723728

Introduction

Leptin is mainly, but not exclusively, produced byadipose tissue (Zhang et al., 1994; Ahima and Flier,2000) and contributes to the regulation of energy bal-ance by informing the brain about fat store levels, thenregulating food intake and energy expenditure in adultanimals (Houseknetch et al., 1998; Casanueva and Die-guez, 1999). Leptin, via its receptors located in mosttissues, has been implicated in numerous other roles,including modulation of reproduction, endocrine sys-tem, tissue metabolism, blood pressure, hematopoiesis,angiogenesis, brain and bone development, wound heal-ing, and cell differentiation and proliferation (Ahima

1The authors acknowledge A. Gertler for recombinant leptin, L.Belair for total RNA extraction and leptin antibody preparation, S.Taourit for DNA sequencing, B. Vigier, R. Boischard, A. Aubourg,M. Olivier-Bousquet, and M. Guillomot for advice and help in immu-nofluorescence analysis, F. Fort for photographic work, and K. Laudand P. Martin for helpful discussions.

2Correspondence: Centre de Recherches de Clermont-Ferrand/Theix, (phone: 33-4-73-62-47-01; Fax: 33-4-73-62-45-19; E-mail:mbonnet@clermont.inra.fr).

Received March 1, 2001.Accepted August 21, 2001.

723

location of leptin protein, as determined by immunohis-tochemical analysis, changed within mammary tissue.It was located in adipose cells during early stages ofpregnancy, in epithelial cells after full cell differentia-tion just before parturition, and in myoepithelial cellsafter parturition. These data, compared with publisheddata on leptin receptor gene expression, provide evi-dence that leptin could be produced by different celltypes of the mammary gland and could act as a para-crine factor on mammary cell growth and differentia-tion via adipose-epithelial cells and myoepithelial-epi-thelial cell interactions.

and Flier, 2000). The identification of leptin in human(Casabiell et al., 1997; Houseknetch et al., 1997; Smith-Kirwin et al., 1998), rat (Casabiell et al., 1997), murine(Aoki et al., 1999), bovine (Rosi et al., 2000), and porcine(Estienne et al., 2000) milk suggests that this hormonecould also be involved in the physiology of the neonate.However, the presence of leptin in milk opens the ques-tion of the mechanisms by which the epithelial cellstransfer leptin from the blood and(or) synthesize it. Atransfer of leptin from maternal blood to milk throughmammary epithelial cells was suggested by the detec-tion of [125I]-leptin in milk after intraperitoneal injec-tion of [125I]-leptin into lactating rats (Casabiell et al.,1997) and by the characterization of leptin receptormRNA in ovine mammary epithelial cells (Laud et al.,1999). However, the detection of leptin mRNA and(or)protein in human (Smith-Kirwin et al., 1998) and mu-rine (Aoki et al., 1999) mammary tissue suggests alsothat leptin could be produced in the mammary gland.To address the ability of the ovine mammary gland tosynthesize leptin, we quantified leptin mRNA levels byreal-time reverse transcription and polymerase chainreaction (RT-PCR) throughout pregnancy and lacta-tion. In addition, we used immunofluorescence detec-tion to localize the leptin protein among mammarycell types.

by guest on June 6, 2014www.journalofanimalscience.orgDownloaded from

Bonnet et al.724

Materials and Methods

Tissue Samples

Animal care and use procedures were approved bythe French Ministry of Agriculture in agreement withFrench regulations for animal experimentation (guide-line 19/04/1988). Primiparous Prealpes du Sud ewes (n= 26) were allotted in eight groups according to theirpregnancy or lactation stage: 15, 80, 106, 112, or 141d of pregnancy and 0 (30 min after parturition), 3, and48 to 70 d of lactation (three to four animals per group).The diet distributed during the first 3 mo of pregnancyconsisted of 58% ammonia-treated straw, 19% barley,13% dehydrated alfalfa, and 10% peas. The diet distrib-uted from the 3rd mo to the end of pregnancy consistedof 46% ammonia-treated straw, 26% barley, 20% dehy-drated alfalfa, and 8% peas. The diet distributed tolactating ewes with one lamb consisted of 23% wheatand barley straw, 23% pea straw, 9% oats, 18% barley,23% dehydrated alfalfa, and 4% soybean meal. The dietdistributed to lactating ewes with two lambs consistedof 20% wheat and barley straw, 20% pea straw, 8%oats, 20% barley, 24% dehydrated alfalfa, and 8% soy-bean meal. Vitamin-mineral premix was added to thefeed at 15 or 30 g/d for the pregnancy and lactationstages, respectively. The body condition score was 2.4 0.4 (on a 0-to-5 scale). The number of fetuses or lambswas one for 16 ewes and two for 10 ewes. Ewes wereslaughtered by exsanguination and samples of mam-mary tissue were immediately placed either in 2% para-formaldehyde-PBS buffer for immunohistochemicalanalysis or frozen in liquid nitrogen pending gene ex-pression analysis.

Quantification of Leptin mRNA Levelby Real-Time Quantitative RT-PCR Assay

Total RNA was prepared by the guanidium isothiocy-anate/phenol method as described by Puissant andHoudebine (1990). Quantification of leptin mRNA levelwas performed by real-time RT-PCR, using the fluores-cent TaqMan methodology and a 7700 Sequence Detec-tor System (PE Applied Biosystems, Courtaboeuf,France) according to a procedure described previously(Bonnet et al., 2000). Sense (5-TCAGTGGATGGTCCC-TCGA-3) and antisense primers (5-GGGAAACC-CAAGCCTCCTC-3) as well as TaqMan probe (5-CAG-GACCAGCCCCCAGGAGCC-3) (PE Applied Biosys-tems) were chosen (Figure 1) after characterizing 1,076bp of the leptin mRNA 3untranslated region (3UTR).This 1,076-bp cDNA fragment was reverse-transcribedand amplified by PCR with the forward (5-CTTTGTTTCTACTGTGACTGACT-3) and the reverse(5-AGTGCAAGCAGGGTTAGCCTGTG-3) primersand sequenced using an ABI 377A automated se-quencer (PE Applied Biosystems) as described pre-viously (Bonnet et al., 2000). The sequence accessionnumber of this 1,076-bp cDNA fragment is AF310264.

Figure 1. Partial nucleotide sequence of ovine leptincDNA and position of the primers and TaqMan probeused for the real-time reverse transcription and polymer-ase chain reaction assay. We characterized 1,076 bp of theovine leptin cDNA corresponding to a part of the leptinmRNA 3untranslated region (accession numberAF310264). This ovine (o) fragment was aligned with itshuman (h) and pig (p) counterparts (sequence accessionnumbers U43653 and AF026976, respectively). Gaps (.)have been placed to maximize the similarity. Dashes ()correspond to nucleotides that are identical to those ofthe ovine leptin sequence. The primers and TaqMan probeused for quantitative analysis of leptin mRNA level areshaded. Alignment was performed with the Clustalw pro-gram (Version 1.81). This ovine 1,076-bp sequence shows67 and 78% identity with the human and pig se-quences, respectively.

by guest on June 6, 2014www.journalofanimalscience.orgDownloaded from

Leptin expression by the ovine mammary gland 725

The reverse transcription reaction (20 L) of the real-time RT-PCR assay was performed using 4 g of totalRNA, with 100 U of SuperScript reverse transcriptase(Gibco BRL, Life Technologies, Cergy Pontoise, France)and 10 pmol of oligo(dT)18. Amplification reactions (50L) contained diluted (1:500 in water) cDNA sample(10 L), 10 PCR Master Mix (27.5 L, PE AppliedBiosystem), 40 pmol of each primer, and 10 pmol ofTaqMan probe. The cycling conditions included 2 minat 50C and 10 min at 95C. Subsequently, thermalcycling proceeded with 45 cycles at 95C for 15 s andat 60C for 2 min. Each assay was performed in tripli-cate. The concentration of leptin mRNA was deter-mined from a calibration curve prepared by amplifying59,250, 29,625, 7,406, 1,851, and 592 copies of a recom-binant plasmid containing the 1,076-bp fragment de-scribed in Figure 1.

The leptin mRNA copy number was normalized bythe mRNA copy number of the constitutively expressedcyclophilin gene, quantified by real-time RT-PCR asdescribed previously (Bonnet et al., 2000).

Leptin Location by Indirect Immunofluorescence

Mammary fragments obtained after dissection werefixed with 2% paraformaldehyde in PBS buffer, pH 7.2,for 24 h. Fixed tissues were incubated overnight in 40%sucrose in PBS, frozen in liquid nitrogen vapors, andcut in 3-m sections at 35C with a Reichert Cryocut(Leica, Reuil-Malmaison, France).

Leptin antiserum was produced in rabbits by injec-tion of 1 mg of recombinant chicken leptin (Raver etal., 1998) solubilized in saline buffer and emulsified inFreunds complete adjuvant. Three and six weeks laterthe rabbits were reimmunized with 1 mg of recombi-nant leptin solubilized in saline buffer and emulsifiedin Freunds incomplete adjuvant. From the 6th wk afterinitial immunization, antiserum was collected weekly.

Serum and antibody dilutions were made in PBS con-taining 0.2% of fish gelatin.

For labeling of leptin protein, tissue sections (n =2 or 4 for tissues from pregnant and lactating ewes,respectively) were successively incubated with PBS-50mM NH4Cl (20 min), PBS (three times, 15 min eachtime), goat serum (1:10, 1 h) and rabbit anti-chickenleptin antiserum (1:100, 2 h); washed in PBS-0.2% fishgelatin; and then incubated with fluorescein isothiocya-nate (FITC)-conjugated goat anti-rabbit IgG (1:200 for2 h; Sanofi Diagnostics Pasteur, Marnes-La-Coquette,France). Sections were mounted on a drop of Vectas-hield (Vector Laboratories, Burlingame, CA) and ob-served with a Polyvar Reichert microscope (Leica). Con-trol sections were treated similarly with nonimmunerabbit serum or with omission of anti-chicken leptinantiserum. To check the specificity of the staining, sec-tions were incubated with anti-chicken leptin antise-rum fully adsorbed with chicken leptin. This adsorbedantiserum was prepared by incubating, for 2 h, chicken

Figure 2. Expression of the leptin gene in mammarytissue throughout pregnancy and lactation determinedby real-time reverse transcription and polymerase chainreaction assay. Leptin and cyclophilin mRNA levels weremeasured from mammary gland tissue collected fromthree or four ewes at 15, 80, 106, 112, or 141 d of pregnancyand at 0, 3, 48, or 70 d of lactation. Leptin/cyclophilinmRNA ratios were calculated. a,b,c,dMeans ( SEM) withdifferent superscripts differ significantly (P < 0.05).

leptin (Raver et al., 1998) with anti-chicken leptin anti-serum (1:100) to make a final concentration of 1 g/L.

Double-labeling of both leptin protein and F-actinstructures was performed according to the same proce-dure, with the addition of 0.17 mol of tetramethylrho-damine phalloidin (Molecular Probes, Eugene, OR) tothe incubation step with FITC-conjugated goat anti-rabbit IgG.

Statistical Analysis

Data were normalized by log transformation andwere submitted to an analysis of variance by the GLMprocedure of SAS (SAS Inst. Inc., Cary, NC). Since asignificant (P < 0.01) effect of the physiological statewas shown, the differences between two physiologicalstages were tested using Duncans test with a probabil-ity of 0.05.

Results

Temporal Expression of Leptin mRNA in OvineMammary Gland During Pregnancy and Lactation

Leptin mRNA level, normalized by the level ofcyclophilin mRNA, varied significantly (P < 0.01) de-pending on the pregnancy or lactation stage (Figure 2).During pregnancy, the leptin mRNA level decreasedstrongly between d 80 and d 106 or 112 (P < 0.05),before increasing slightly at d 141 (P < 0.05 for d 141vs d 106 of pregnancy) to levels similar to those assayedat d 15 and 80. Throughout lactation, leptin mRNAlevels did not vary significantly but were lower (P