Endometrial expression of the insulin-like growth factor system during uterine involution in the postpartum dairy cow.

Abstract

Rapid uterine involution in the postpartum period of dairy cows is important to achieve a short interval to conception. Expression patterns for members of the insulin-like growth factor (IGF) family were determined by in situ hybridisation at day 14+/-0.4 postpartum (n=12 cows) to investigate a potential role for IGFs in modulating uterine involution. Expression in each uterine tissue region was measured as optical density units and data were analysed according to region and horn. IGF-I mRNA was localized to the sub-epithelial stroma (SES) of inter-caruncular and caruncular endometrium. Both IGF-II and IGF-1R expression was detected in the deep endometrial stroma (DES), the caruncular stroma and myometrium. IGFBP-2, IGFBP-4 and IGFBP-6 mRNAs were all localised to the SES of inter-caruncular and caruncular uterine tissue, and in the DES and caruncular stroma, with IGFBP-4 mRNA additionally expressed in myometrium. IGFBP-3 mRNA was only detectable in luminal epithelium. IGFBP-5 mRNA was found in myometrium, inter-caruncular and caruncular SES and caruncular stroma. These data support a role for IGF-I and IGF-II in the extensive tissue remodelling and repair which the postpartum uterus undergoes to return to its non-pregnant state. The differential expression of binding proteins between tissues (IGFBP-3 in epithelium, IGFBP-2, -4, -5 and -6 in stroma and IGFBP-4 and -5 in myometrium) suggest tight control of IGF activity within each compartment. Differential expression of many members of the IGF family between the significantly larger previously gravid horn and the previously non-gravid horn may relate to differences in their rate of tissue remodelling.

Full text

Page 1

Endometrial expression of the insulin-like growth factor system
during uterine involution in the postpartum dairy cow
S. Llewellyna, R. Fitzpatrickb, D.A. Kennyc, J. Pattond, and D.C. Wathesa⁎
aReproduction, Genes and Development Group, Department of Veterinary Basic Sciences, Royal Veterinary
College, Hawkshead Lane, Hatfield, Herts, London AL9 7TA, UK.
bAnimal Production Research Centre, Mellows Campus, Athenry, Co. Galway, Ireland.
cSchool of Agriculture, Food Science and Veterinary Medicine, University College Dublin, Ireland.
dTeagasc Moorepark, Dairy Production Research Centre, Fermoy, Co. Cork, Ireland.
Abstract
Rapid uterine involution in the postpartum period of dairy cows is important to achieve a short interval
to conception. Expression patterns for members of the insulin-like growth factor (IGF) family were
determined by in situ hybridisation at day 14 ± 0.4 postpartum (n = 12 cows) to investigate a potential
role for IGFs in modulating uterine involution. Expression in each uterine tissue region was measured
as optical density units and data were analysed according to region and horn. IGF-I mRNA was
localized to the sub-epithelial stroma (SES) of inter-caruncular and caruncular endometrium. Both
IGF-II and IGF-1R expression was detected in the deep endometrial stroma (DES), the caruncular
stroma and myometrium. IGFBP-2, IGFBP-4 and IGFBP-6 mRNAs were all localised to the SES of
inter-caruncular and caruncular uterine tissue, and in the DES and caruncular stroma, with IGFBP-4
mRNA additionally expressed in myometrium. IGFBP-3 mRNA was only detectable in luminal
epithelium. IGFBP-5 mRNA was found in myometrium, inter-caruncular and caruncular SES and
caruncular stroma. These data support a role for IGF-I and IGF-II in the extensive tissue remodelling
and repair which the postpartum uterus undergoes to return to its non-pregnant state. The differential
expression of binding proteins between tissues (IGFBP-3 in epithelium, IGFBP-2, -4, -5 and -6 in
stroma and IGFBP-4 and -5 in myometrium) suggest tight control of IGF activity within each
compartment. Differential expression of many members of the IGF family between the significantly
larger previously gravid horn and the previously non-gravid horn may relate to differences in their
rate of tissue remodelling.
Keywords
Bovine; Uterus; IGF; IGFBP; Involution
1  Introduction
In dairy cows, the peri-partum period is critical to future milk production and fertility. Uterine
involution involves extensive restructuring of the extracellular matrix alongside mitogenesis
and apoptosis [1–3]. Initial degeneration of placental cotyledons and maternal caruncles
© 2008 Elsevier Inc.
⁎Corresponding author. Tel.: +44 1707 666553; fax: +44 1707 666371. dcwathes@rvc.ac.uk.
This document was posted here by permission of the publisher. At the time of deposit, it included all changes made during peer review,
copyediting, and publishing. The U.S. National Library of Medicine is responsible for all links within the document and for incorporating
any publisher-supplied amendments or retractions issued subsequently. The published journal article, guaranteed to be such by Elsevier,
is available for free, on ScienceDirect.
Sponsored document from
Domestic Animal Endocrinology
Published as: Domest Anim Endocrinol. 2008 May ; 34(4): 391–402.
Sponsored Document
Sponsored Document
Sponsored Document

Page 2

accumulate as tissue debris in the uterine lumen forming a lochial discharge [4]. Contractions
of the myometrium aid expulsion of lochia, and also restore uterine size, shape and tone to that
of a non-pregnant animal [5,6]. Whilst most of these changes have occurred within 2–3 weeks
postpartum, involution is not considered complete until about 40–50 days postpartum [1]. The
previously non-gravid uterine horn returns to a non-pregnant state 10–15 days earlier than the
previously gravid uterine horn [7]. Histological repair of the endometrium lags physical
involution by 10–20 days [8], completing when caruncles regenerate epithelium [4]. Microbial
contamination of the postpartum uterus is almost universal during the first week postpartum
[9]. When pathogenic bacteria are not cleared the uterus becomes infected and inflamed and
uterine involution is delayed [1,10]. Clinical endometritis is characterised by the continued
presence of a purulent discharge beyond 21 days after calving [1].
Many processes involved in uterine repair are common to those of wound healing in other
tissues (for a review see [11]). Potential mediators of tissue turnover and remodelling in the
uterus include cytokines, matrix-degrading enzymes and growth factors [11,12]. The insulin-
like growth factors (IGF-I and IGF-II) function in such tissue repair processes. In healing-
impaired wounds, the mRNA for IGF-I, IGF-1R, and IGFBP-3 is significantly reduced [13].
The administration of IGF-I to these wounds corrects defective tissue repair [14] and in
combination with other growth factors it increases connective tissue regeneration and
epithelialisation [15]. Components of the IGF system have been described in the uteri of a
variety of species (e.g. humans [16], rodents [17], pigs [18], cattle [19], and sheep [20]). The
proliferative and differentiating effects of IGFs on uterine cells are thought to support the
growth and regression of uterine tissue throughout the estrous cycle and also the regenerative
processes in women following menstruation [16,21]. IGFBP-2 has also been shown to stimulate
endometrial cell mitogenesis directly [22].
An increased rate of uterine involution is associated with earlier resumption of ovarian activity
[23], which is in turn important for increasing pregnancy rate to first service [24]. Conversely,
endometrial damage associated with sub-clinical endometritis leads to prolonged intervals to
conception, with many cows failing to conceive at all [25]. The mechanisms that regulate
uterine involution are not completely understood and, to the best of our knowledge, no previous
studies have investigated the uterine IGF system during involution in lactating dairy cows. We
postulated that changes in IGF bioavailability may be implicated in the rate of postpartum
uterine recovery and thus influence the calving to conception interval and reproductive
efficiency. The objective of the study was to determine patterns of mRNA expression for the
IGF system within the previously gravid (PG) and previously non-gravid (PNG) uterine horns
during the early postpartum period. Samples were obtained at approximately 2 weeks after
calving as we hypothesised that this represents a time by which a delay in the normal recovery
process may predispose cows to the subsequent development of endometritis.
2 
2.1 
 Materials and methods
 Animals and tissue samples
All procedures were carried out under license in accordance with the European Community
Directive, 86-609-EC. Uteri were collected from 12 multiparous Holstein-Friesian dairy cows
(mean parity 4.7) following slaughter at day 14 ± 0.4 postpartum. The diameters of both horns
were measured approximately 5 cm anterior to the bifurcation of the uterus. Samples of inter-
caruncular and caruncular tissue were dissected from the previously gravid and non-gravid
uterine horns approximately 1 cm anterior to the bifurcation of the uterus. A 5 cm square region
was harvested, wrapped in aluminium foil, and frozen in liquid nitrogen-tempered isopentane.
Samples were stored at −80 °C until sectioning.
Llewellyn et al.Page 2
Published as: Domest Anim Endocrinol. 2008 May ; 34(4): 391–402.
Sponsored Document
Sponsored Document
Sponsored Document

Page 3

2.2  In situ hybridisation
The in situ hybridisation procedure was performed as described previously [26]. All chemicals
were purchased from Sigma–Aldrich Company Ltd. (Poole, Dorset, UK) or VWR International
Ltd. (Poole, Dorset, UK) unless otherwise specified. Briefly, sections of 10 μm were cut from
each uterine tissue sample and thaw-mounted onto SuperFrost® Plus or POLYSINE™
microscope slides, fixed in 4% (w/v) paraformaldehyde in 0.01 M PBS, washed in PBS, and
sequentially dehydrated in 70% and 95% ethanol. The oligonucleotide probes for the IGF
system were end-labelled with [35S]dATP (Amersham Biosciences UK Ltd.,
Buckinghamshire, England) using terminal deoxynucleotidyl transferase (Promega UK Ltd.,
Southampton, England). Tissue sections were subsequently treated with
100 000 cpm (100 μl)−1 hybridisation buffer and hybridised overnight at either 42, 45, or 52 °
C (Table 1). Following incubation, slides were washed in a solution of 1 × SSC, 2 g l−1 sodium
thiosulphate at room temperature for 30 min followed by fresh 1 × SSC, 2 g l−1 sodium
thiosulphate at 60 °C for 60 min. Slides were then rinsed in solutions of 1 × SSC, 0.1 × SSC,
75% ethanol and 95% ethanol and air-dried before exposure to β-max hyperfilm (Kodak
BioMax MR Film) for either 4 or 5 days. All uterine sections treated with a particular probe
were hybridized in the same batch. Sense probes, which were identical in sequence to the
respective mRNA targets, were always included as negative controls and any signal from these
was regarded as non-specific. Each batch also contained an appropriate positive control tissue,
based on previous studies. These were cross-sections of uterus from an estrous ewe for IGF-I
and the type 1 IGF receptor [20], IGFBP-1 [27] and IGFBP-6 [28]; ovine placentome for IGF-
II and IGFBPs-2, -3 and -4 [29] and ovine intercotyledonary tissue for IGFBP-5 [30].
2.3  Photographic emulsions
To aid cellular localisation of hybridised probes, slides previously subject to autoradiography
were coated with photographic emulsion LM1 (Amersham Biosciences UK Ltd.,
Buckinghamshire, England) according to the manufacturer's instructions and stored for 28, 30
or 42 days at 4 °C in the dark (Table 1). The slides were developed in 20% phenisol (ILFORD
Imaging UK Ltd., Cheshire, England) fixed in 1.9 M sodium thiosulphate and counterstained
with haematoxylin and eosin. All other slides were also stained with haematoxylin and eosin
to aid identification of tissue region.
2.4  Optical density measurements
Readings were obtained from at least two sections per tissue for each of the antisense (AS) and
sense (S) probes. Autoradiographs were scanned into a computer and optical density (OD)
measurements were recorded from digital images. The relative expression of mRNA for
components of the uterine IGF system was quantified from the autoradiographs using the public
domain NIH ImageJ program (available through the NIH website—http://www.nih.gov),
which calculated the average optical density (OD) over the selected area of film based on a
linear grey scale of 0.01–2.71. The following tissue layers were each assessed separately:
luminal epithelium, sub-epithelial stroma (a layer of dense connective tissue underlying the
luminal epithelium), caruncular stroma (the dense connective tissue forming the caruncles),
deep endometrial stroma (loose connective tissue between the sub-epithelial stroma and the
myometrium) and myometrium. The latter two tissue types were only present in samples
collected from the inter-caruncular region. Each tissue type was measured separately on each
section. The background OD, from a blank area of film, was also measured and subtracted
from both AS and S OD measurements. Finally the S values were subtracted from AS values
to give an average OD value for specific hybridisation [31]. The detection limit was taken as
an OD value of 0.01.
Llewellyn et al.Page 3
Published as: Domest Anim Endocrinol. 2008 May ; 34(4): 391–402.
Sponsored Document
Sponsored Document
Sponsored Document

Page 4

2.5  Statistical analysis
Statistical analyses were performed using Statistical Package for the Social Sciences (SPSS
for Windows, V13.0). Data for uterine diameter measurements at the time of tissue collection
were analysed using Student's t-test. OD measurements were obtained from four samples per
cow, taken from each of the caruncular and inter-caruncular regions of the previously gravid
and non-gravid horns. The effects of uterine horn and tissue region on the level of mRNA
expression for each probe were analysed by general linear model analysis. Cow was entered
as a random effect. For this purpose, data from uteri in which a particular probe showed no
detectable specific hybridisation (OD of <0.01) were given an OD of 0.01, which equated to
the lower limit of detection. Results were considered statistically significant when P < 0.05.
3  Results
At the time of tissue collection, the diameter of the previously gravid uterine horn was larger
than that of the previously non-gravid uterine horn (56 ± 6.9 and 31 ± 3.1 mm, respectively,
mean ± S.E.M., P = 0.005). The spatial distribution of mRNA encoding components of the
uterine IGF system is shown in Figs. 1 and 2. The concentrations of mRNA in OD units are
summarised in Table 2 according to uterine horn and tissue region and their two-way
interactions are illustrated in Figs. 3 and 4. The method used provided a semi quantitative
measure of the intensity of mRNA expression in specific cell types.
3.1  Expression of the IGFs and IGF type 1 receptor
IGF-I mRNA was localized to the sub-epithelial stroma (SES) of inter-caruncular and
caruncular endometrium in both uterine horns (Figs. 1(A) and 2A). Both IGF-II and IGF-1R
expression was detected in the deep endometrial stroma (DES), the caruncular stroma (not
shown) and myometrium (Figs. 1(C), (E) and 2C, E).
Overall expression of IGF-I mRNA was higher in the inter-caruncular than caruncular SES
(P = 0.001, Table 2). There was a significant horn × region interaction (P = 0.032), with lower
levels of IGF-I transcript in the inter-caruncular SES of the PG compared with the PNG horn
(Fig. 3(A)). IGF-II expression was higher in the DES than in the caruncular stroma and
myometrium (P ≤ 0.001, Table 2). When data from tissue regions were pooled the concentration
of IGF-II mRNA did not vary between the PNG and PG horns, but there was a significant
horn × region interaction (P ≤ 0.001). Expression of IGF-II in the DES and caruncular stroma
was lower in the PG than PNG horn, whereas within myometrium expression was higher in
the PG than PNG horn (Fig. 3(B)). For the IGF-1R, expression was highest in myometrium
and similar between DES and caruncular stroma (Table 2 and Fig. 3(C)). Overall, the level of
IGF-1R transcript was higher (P = 0.030) in the PNG than PG horn (Table 2). The horn × region
interaction was not significant for uterine IGF-1R mRNA expression.
3.2  Expression of IGFBPs
IGFBP-1 mRNA could not be detected in any uteri examined, despite expression being
observed in the ovine estrous uterus which was used as positive control tissue (data not shown).
IGFBP-2, IGFBP-4 and IGFBP-6 mRNAs were all localised to the SES of inter-caruncular
and caruncular uterine tissue, and in the DES and caruncular stroma (Figs. 1(G), (K), (O) and
2G, K, O). IGFBP-4 mRNA was additionally expressed in myometrium. In contrast, IGFBP-3
mRNA expression was only detected in the luminal epithelium (LE) of both inter-caruncular
and caruncular samples (Figs. 1(I) and 2I). IGFBP-5 mRNA was found in myometrium, inter-
caruncular and caruncular SES and caruncular stroma (Figs. 1(M) and 2M).
IGFBP-2 mRNA expression in inter-caruncular and caruncular SES was higher than in DES
and caruncular stroma (P ≤ 0.001, Table 2). There was no main effect of horn, but there was
Llewellyn et al. Page 4
Published as: Domest Anim Endocrinol. 2008 May ; 34(4): 391–402.
Sponsored Document
Sponsored Document
Sponsored Document

Page 5

a horn × region interaction (P = 0.034). Within caruncular stroma only, the concentration of
IGFBP-2 mRNA was higher in the PG than the PNG uterine horn (Fig. 4(A)).
For IGFBP-3 mRNA the main effects of uterine horn and tissue region were not significant
but there was an interaction (P ≤ 0.001). Expression in the inter-caruncular LE was higher in
the PNG than PG horn, whereas in the caruncular LE expression was higher in the PG uterine
horn (Fig. 4(B)).
Expression levels of IGFBP-4 mRNA varied between tissue regions, with higher expression
in the caruncular than inter-caruncular SES, lowest expression in myometrium, and
intermediate signal intensity in the DES and caruncular stroma (Table 2). There was no
difference in transcript levels between the PNG and PG uterine horns when regional data were
combined (Table 2). Levels of IGFBP-4 mRNA expression were, however, affected by an
interaction between uterine horn and tissue region (P = 0.024): within DES expression was
lower in the PG than PNG uterine horn (Fig. 4(C)).
Expression of IGFBP-5 mRNA was highest in myometrium, intermediate in the inter-
caruncular and caruncular SES and lowest in caruncular stroma (P ≤ 0.001, Table 2). When
regional expression data were pooled, the PNG uterine horn expressed higher concentrations
of IGFBP-5 mRNA than the PG horn (P ≤ 0.001, Table 2). There was also a significant effect
of the interaction between uterine horn and tissue region (P ≤ 0.001). Expression in both the
inter-caruncular SES and the caruncular stroma was lower in the PG than PNG uterine horn
whereas for the caruncular SES the reverse was true (Fig. 4(D)).
IGFBP-6 mRNA was expressed at higher concentrations in the inter-caruncular and caruncular
SES than in DES and caruncular stroma (P ≤ 0.001, Table 2). Transcript levels were higher
(P ≤ 0.001) in the PNG than PG uterine horn when regional expression data were pooled (Table
2). The interaction between uterine horn and tissue region was significant (P = 0.045).
Expression in each of the inter-caruncular SES, caruncular SES, DES and caruncular stroma
was lower in the PG than PNG horn (Fig. 4(E)).
4  Discussion
The rate of uterine involution is an important factor influencing the subsequent fertility of dairy
cows [24]. In this study we have investigated for the first time a possible role for the IGF family
of proteins in this event in lactating dairy cows. The timing of tissue collection at approximately
day 14 postpartum occurred when the PG uterine horn in our group of multiparous cows was
larger than the PNG horn. At this stage caruncular tissue is expected to have undergone
degeneration and sloughing, but not to have completed re-epithelialisation [2]. In contrast, the
inter-caruncular area does not lose its epithelial layer [32] and recovers from pregnancy more
quickly [2]. The ongoing process of uterine involution at the time of tissue collection was thus
expected to involve tissue regeneration alongside size recovery. An adequate recovery process
may be crucial in preventing the uterus, which is heavily contaminated with bacteria following
calving [1], from developing endometritis. Samples were analysed using in situ hybridisation.
Whilst this approach is considered only semi-quantitative, we have found the technique
described here to be highly repeatable. Furthermore, it enables measurement of mRNA
concentrations in individual cellular types. This is not feasible in a complex organ such as the
uterus using alternative techniques such as RT-PCR, as it is not readily possible to separate
different populations of epithelial and stromal cells for RNA extraction.
IGF-I mRNA was localised to the SES, confirming earlier observations in the cow [33] and
sheep [20,34]. Normal wound healing involves a sequence of inflammation, proliferation, and
maturation or remodelling [12] and local IGF-I production increases as wound healing
progresses [35]. Since IGF-I increases during the late proliferative phase of the human
Llewellyn et al. Page 5
Published as: Domest Anim Endocrinol. 2008 May ; 34(4): 391–402.
Sponsored Document
Sponsored Document
Sponsored Document