bqae048-html.html

© The Author(s)  2024. Published  by Oxford University  Press on behalf  of the Endocrine  Society. All rights
reserved.  For commercial  re-use,  please contact reprints@oup.com  for reprints  and translation  rights  for
reprints.  All other permissions  can be obtained  through  our RightsLink  service via the Permissions  link on the
article page on our site

—

for further information please contact journals.permissions@oup.com. This article

is published and distributed under the terms of the Oxford University Press, Standard Journals Publication
Model (https://academic.oup.com/pages/standard-publication-reuse-rights)

Impact of ghrelin on islet size in non-pregnant and pregnant female mice

Authors and affiliations

Deepali Gupta

Avi W. Burstein

, Kripa Shankar

, Salil Varshney

, Omprakash Singh

, Sherri

Osborne-Lawrence

, Corine P. Richard

, Jeffrey M. Zigman

1,2,3

Center for Hypothalamic Research, Department of Internal Medicine, UT Southwestern Medical

Center, Dallas Texas, USA

Division of Endocrinology & Metabolism, Department of Internal Medicine, UT Southwestern

Medical Center, Dallas Texas, USA

Department of Psychiatry, UT Southwestern Medical Center, Dallas Texas, USA

Keywords

Islet, pregnancy, ghrelin, LEAP2, GHSR

Corresponding author

Jeffrey M. Zigman

Center for Hypothalamic Research, Department of Internal Medicine,

UT Southwestern Medical Center, 75390-9077, Dallas Texas, USA

Phone: +1 214-648-6422

Email: jeffrey.zigman@utsouthwestern.edu

ORCID - https://orcid.org/0000-0003-3477-1295

Disclosure summary

J.M.Z. owns stock in Eli Lilly, Novo Nordisk, and Medtronic.

ACCEPTED MANUSCRIPT

Downloaded from https://academic.oup.com/endo/advance-article/doi/10.1210/endocr/bqae048/7646761 by UCSD-Philosophy user on 23 April 2024

Abstract:

Reducing ghrelin by ghrelin gene knockout (GKO), ghrelin-cell ablation, or high-fat

diet feeding increases islet size and β-cell mass in male mice. Here, we determined if reducing

ghrelin also enlarges islets in females, and if pregnancy-associated changes in islet size are

related to reduced ghrelin. Islet size and

β-cell mass were larger (

=0.057 for β-cell mass) in

female GKO mice. Pregnancy was associated with reduced ghrelin and increased LEAP2 [a

ghrelin receptor (GHSR) antagonist] in WT mice. Ghrelin deletion and pregnancy each

increased islet size (by ~19.9-30.2% and ~34.9-46.4%, respectively), percentage of large islets

(>25 µm

x 10

, by ~21.8-42% and ~21.2-41.2%, respectively)

and β-cell mass (by ~15.7-23.8%

and ~65.2-76.8%, respectively). Neither islet cross-

sectional area, β-cell cross-sectional area,

nor β-cell mass correlated with plasma ghrelin, although all positively correlated with LEAP2

(

=0.081 for islet cross-sectional area). In

ad lib

-fed mice, there was an effect of pregnancy, but

not ghrelin deletion, to change (raise) plasma insulin without impacting blood glucose. Similarly,

there was an effect of pregnancy, but not ghrelin deletion, to change (lower) blood glucose area

under the curve during a glucose tolerance test. Thus, genetic deletion of ghrelin increases islet

size and β-cell cross-sectional area in female mice, similar to males. Yet, despite pregnancy-

associated reductions in ghrelin, other factors appear to govern islet enlargement and changes

to insulin sensitivity and glucose tolerance in the setting of pregnancy. In the case of islet size

and β-cell mass, one of those factors may be the pregnancy-associated increase in LEAP2.

ACCEPTED MANUSCRIPT

Downloaded from https://academic.oup.com/endo/advance-article/doi/10.1210/endocr/bqae048/7646761 by UCSD-Philosophy user on 23 April 2024

Introduction

Ghrelin is an acylated peptide hormone secreted primarily by ghrelin cells in the stomach and

duodenum (1). Ghrelin, which acts by binding to and activating GHSRs (growth hormone

secretagogue receptors) located in the brain, pituitary, pancreatic islets of Langerhans, and

elsewhere, is best known for its actions to increase food intake (including reward-based eating),

body weight, and GH secretion (1-5). Ghrelin also has several essential glucoregulatory actions

(6). The related liver- and intestinal tract-derived hormone LEAP2 (liver-expressed antimicrobial

peptide-2) also binds GHSRs, although it serves as a GHSR antagonist/inverse agonist that

potently blocks ghrelin action and decreases constitutive signaling (7,8). LEAP2 and/or LEAP2

analogs reduce food intake, binge eating, and body weight and block ghrelin-induced food

intake (9,10). Further, pre-prandial plasma LEAP2 levels are negatively correlated with hunger

sensation in human subjects while pre-prandial plasma ghrelin levels are positively correlated

(11).

Regarding regulation of blood glucose, ghrelin raises blood glucose when injected and prevents

life-threatening hypoglycemia during prolonged caloric restriction modeling starvation (12,13).

Ghrelin permits the normal counterregulatory response to insulin-induced hypoglycemia in

healthy and diabetic mice (14,15). In contrast, LEAP2 infusions reduce post-prandial glucose

excursions following a standardized liquid mixed meal in humans

– a finding that is also

observed in mice (10). These effects are thought to involve both direct and indirect actions on

the brain, pituitary, and/or islets (4,6,16-19). Within islets, GHSR expression is found in all 4

traditional islet endocrine cell-types, including glucagon-

secreting α-cells, insulin-secreting β-

cells, pancreatic polypeptide-secreting



-cells, and somatostatin-

secreting δ-cells, the latter of

which express the most islet GHSR (4,6,16-19).

ACCEPTED MANUSCRIPT

Downloaded from https://academic.oup.com/endo/advance-article/doi/10.1210/endocr/bqae048/7646761 by UCSD-Philosophy user on 23 April 2024

Ghrelin’s glucoregulatory actions and islet GHSR expression previously led us to begin

investigating effects of reducing ghrelin on islet size. In a recent study performed using male

mice, we reported that ghrelin deletion [as occurs in ghrelin-KO (GKO) mice] increases mean

islet size (as assessed by measuring islet cross-

sectional area and Ferret’s diameter, which is

the longest diameter within an islet), percentage of the largest islets, and β-cell cross-sectional

area in juvenile (4 wk-old) and adult (10-12 wk-old) mice (20)

. Higher β-cell numbers from

decreased β-cell apoptosis drove the increase in β-cell cross-sectional area (20). To confirm

that the changed islet morphology in GKO mice was not simply from a developmental change

resulting from germline deletion of the ghrelin gene, we inducibly ablated ghrelin-cells from mice

at 6 wks-of-age and then examined islets 4 wks post-ablation (20). This inducible ghrelin cell

ablation, which was achieved using a ghrelin-cell selective diphtheria toxin-mediated approach,

also increased islet size and β-cell mass (20). Further, as high-fat diet-fed, diet-induced obese

mice have low circulating plasma ghrelin and increased islet size, we investigated if the

physiologically reduced ghrelin mediates the islet enlargement (20). A negative correlation

between islet size and plasma ghrelin in high-fat diet-fed plus standard chow-fed wild-type (WT)

mice together with even higher islet size in high-fat diet-fed GKO mice than in high-fat diet-fed

WT mice suggested that reduced ghrelin contributes to, but is not solely responsible for, diet-

induced obesity-associated islet enlargement (20).

Importantly, the above study (20) focused only on male mice. Yet, evidence exists regarding

sexually dimorphic effects of manipulating elements of the ghrelin system, for instance on body

weight and body composition, and estrogen-dependent changes in both ghrelin levels & ghrelin

orexigenic efficacy have been described (21,22). Thus, it is unclear if the effects of reducing

ghrelin on islet size observed in male mice extend to female mice. Moreover, it is as yet

unknown if a reduction in ghrelin also contributes to the enlargement in islet size and β-cell

mass occurring during pregnancy. Notably, while some inconsistencies regarding pregnancy -

ACCEPTED MANUSCRIPT

Downloaded from https://academic.oup.com/endo/advance-article/doi/10.1210/endocr/bqae048/7646761 by UCSD-Philosophy user on 23 April 2024

associated changes in plasma ghrelin exist in the literature, many studies performed in humans,

mice, rats, and pigs demonstrate lower plasma ghrelin, desacyl-ghrelin and/or total ghrelin

(which comprises ghrelin and desacyl-ghrelin) in later pregnancy stages as compared to non-

pregnant or postpartum levels [see Discussion for details; (23-36)]. Regarding the effect of

pregnancy on islet size and β-cell mass, Banerjee et al demonstrated a higher percentage of Ki-

67-

positive (proliferating) β-cells in pregnant mice at gestational day 15.5 (GD15.5) than in non-

pregnant females

– a phenomenon which the investigators demonstrated was dependent on β-

cell-expressed prolactin receptors (37). Kim et al showed that relative

β-cell mass and

percentage of BrdU-

positive (proliferating) β-cells were higher in mice at GD13 vs. non-pregnant

mice

– a phenomenon which the investigators demonstrated was dependent on β-cell-

expressed 5-HTR2b receptors (38). In rats, the numbers of BrdU-labeled cells/islet

progressively increased over the course of pregnancy beginning between GD6 (similar to non-

pregnant controls) and GD10 (3-fold increase), peaking at GD14 (10-fold increase), at which

point they again fell down to non-pregnant levels by GD18 (39). When Rieck et al examined

GD14.5 islets in mice, they observed a 3.8X increase in β-cell mass with an associated increase

in BrdU-labeled, Ki-67-labeled

β-cells compared to non-pregnant controls, as well as β-cell

hypertrophy and changed expression of a number of genes, including a 5-fold increase in

expression of the gene encoding the apoptosis inhibitor

Birc5

(40).

An increase in β-cell mass is also recognized during pregnancy in humans. For instance, Butler

et al reported an ~1.4-fold increase in pancreatic fractional

β-cell area during pregnancy in

women as compared to non-pregnant women (41). The higher

β-cell area in that autopsy series

of 18 pregnant women, who averaged 24.6



2.6 weeks of gestation (range: 10-40 weeks) at

the time of death, was associated with increases in the percentage of small islets, islet density,

frequency of isolated β-cells, and percentage of insulin-positive pancreatic duct cells (41). The

cross-sectional area of the largest islet identified per pancreatic section was not increased by

ACCEPTED MANUSCRIPT

Downloaded from https://academic.oup.com/endo/advance-article/doi/10.1210/endocr/bqae048/7646761 by UCSD-Philosophy user on 23 April 2024

pregnancy, nor was mean

β-cell size increased (41). Similar findings were reported in another

autopsy series by Moin et al of 14 pregnant women (at 24.5



2.7 weeks of gestation; range: 10-

40 weeks) and 9 age-matched non-pregnant controls (42). Specifically, pregnant women had a

higher islet density and increased number of smaller islets (42). Also, pregnancy was

associated with an

increase in the percentage of β-cells within clusters of 3 or fewer

chromogranin A-positive cells, which are thought to represent newly-formed islets (42). Thus,

although β-cell mass is increased in pregnancy in both mice and humans, this results from an

increase in the percentage of large islets which drives an increase in mean islet size in mice vs.

higher islet density and increased number of small islets in human.

These pregnancy-

associated increases in islet size and β-cell mass observed in humans and

rodents are widely considered to be adaptations to the increased demand for insulin that

develops during pregnancy (39)

. Insufficiency of this adaptive expansion in β-cell mass is

thought to be a main contributor to gestational diabetes (43,44). The current study was

performed in order to establish the effects of ghrelin levels on islet size and β-cell cross-

sectional area in female mice and to determine if pregnancy-associated reductions in ghrelin

contribute to pregnancy-associated islet enlargement.

Methods

Study animals and study approval

All experiments were performed using female mice and were approved by the UT Southwestern

Medical Center Institutional Animal Care and Use Committee. Except as noted, mice had

libitum

(ad lib

) access to standard chow diet [2916 Teklad Global 16% Protein Rodent Diet

(Envigo, Indianapolis, IN)] and water and were individually housed (beginning one week prior to

the studies) at room temperature (21.5

–22.5°C) in a 12-hr light-dark cycle. GKO and WT

ACCEPTED MANUSCRIPT

Downloaded from https://academic.oup.com/endo/advance-article/doi/10.1210/endocr/bqae048/7646761 by UCSD-Philosophy user on 23 April 2024

littermates were generated by crossing GKO (line GKO1) heterozygotes on a C57BL/6N

background, as previously reported (14,15,45). Mice were weaned at 3-4 wks-of-age. Non-

pregnant adult female WT and GKO mice were individually housed and handled daily.

Assessments were performed on mice aged 8-15 weeks. Except as indicated, the same

assessments described for pregnant mice were performed in non-pregnant mice.

Induction of timed pregnancies and routine metabolic analyses on pregnant mice

As schematized in Figure 1A, timed pregnancies were achieved by synchronization of estrus

cycles and mating. Adult GKO and WT littermate females were group housed (4-5 mice per

cage) for 2 wks. Simultaneously, adult C57BL/6N males were individually housed for 2 wks.

Next, one of the above female mice was placed together with one of the above males for mating

between 10:00 to 11:00 am, followed by checking for vaginal copulation plugs at 4:00 pm on the

same day and again at 7:00 am and 3:00 pm for the next 3 consecutive days. Once the vaginal

plug was visualized, the male was removed. The day on which the vaginal plug was detected

was set as gestational day 0.5 (GD0.5). We observed vaginal copulation plugs in 93.1% of

females paired with males over the 3-day observation period. Of the females in which vaginal

copulation plugs were observed, 74.1% became pregnant, as evidenced by presence of fetuses

on GD13.5. As mice in which plugs were not observed over the 3-day period were excluded

from further analysis, we do not know the percentage of mice without observed plugs who

nonetheless became pregnant.

Pregnant mice were weighed daily from GD0.5 to GD13.5. An assessment of 24 h food intake

was performed from 9:00 am on GD9.5 to 9:00 am on GD10.5. On GD10.5, body composition

was measured using an EchoMRI-100 apparatus (EchoMRI LLC, Houston, TX). On GD11.5, an

oral glucose tolerance test (GTT) was performed (see below for details). On GD13.5, between

9:00 am

– 10:00 am, tails were nicked in the

ad lib

-fed condition to measure blood glucose with

ACCEPTED MANUSCRIPT

Downloaded from https://academic.oup.com/endo/advance-article/doi/10.1210/endocr/bqae048/7646761 by UCSD-Philosophy user on 23 April 2024

a Contour Next EZ monitoring system (Bayer, Parsippany, NJ) and to collect tail blood for

hormone analysis, followed by perfusion, fixation and organ collection (see below for details).

Oral glucose tolerance test (GTT)

Mice were fasted beginning at 7:00 a.m. for 6 h, after which blood glucose was measured from

nicked tails (t = 0 min). D-glucose (Sigma-Aldrich, St. Louis, MO; Cat# 50-99-7) 100 mg/mL was

prepared in distilled water and 20 µL/g of body weight (corresponding to 2 g D-glucose/kg body

weight) was administered orally at t = 0 min after blood glucose was measured. Blood glucose

was measured again from nicked tails at 15, 30, 60, 90 and 120 min post glucose

administration. To assess the trajectory of plasma insulin during the GTT, extra blood samples

(~15 µL each) were taken from nicked tails at t = 0, 15, and 30 min.

Blood collection and hormone analysis

Blood samples were collected into EDTA-coated microtubes kept on ice. Collection tubes

contained either the protease inhibitor p-hydroxymercuribenzoic acid (Sigma Aldrich; final

concentration 1 mM; for ghrelin measurement), the protease inhibitor aprotinin (final

concentration 250 KIU/mL; Sigma Aldrich; for LEAP2 measurement), or no protease inhibitor

(for insulin measurement). Samples were immediately centrifuged at 4°C at 1,500 g for 15 min.

For stabilization of the acyl-group on ghrelin, 1N HCl was added to the p-hydroxymercuribenzoic

acid-treated plasma to achieve a final concentration of 0.1 N. Processed samples were stored at

−80°C. Ghrelin, LEAP2, and insulin were measured using ELISA kits (Cat # EZRGRA-90K,

Millipore-Merck, Burlington, MA, RRID:AB_2889905; Cat# EK-075-40, Phoenix

Pharmaceuticals, Inc. Burlingame, CA, RRID:AB_2909436 and Cat # 90080, Crystal Chem,

Downers Grove, IL, RRID:AB_2783626, respectively), with the aid of a BioTek PowerWave XS

Microplate spectrophotometer (BioTek, Winooski, VT) and BioTek KC4 junior software, as

described previously (4,46).

ACCEPTED MANUSCRIPT

Downloaded from https://academic.oup.com/endo/advance-article/doi/10.1210/endocr/bqae048/7646761 by UCSD-Philosophy user on 23 April 2024

Perfusion, tissue collection and pancreas processing

Non-pregnant WT and GKO mice and pregnant (at GD13.5) WT and GKO mice were

anesthetized with chloral hydrate (700 mg/kg BW, i.p.) and then perfused transcardially with

PBS, pH 7.0 followed by 10% neutral buffered formalin. Pancreata with attached spleens and

uteri with attached fetuses were removed and weighed immediately. Average weights of

individual fetuses within a single uterus were estimated by dividing the total weight of each

isolated uterus with attached fetuses by the number of fetuses within the uterus; the average

fetus weight per dam was calculated. Pancreata with attached spleens were stored in formalin

overnight at 4°C and then immersed in graded (5%, 10%, 18% and 30%) sucrose solutions in

PBS for 24 h each at 4°C. Eight µm-

thick pancreatic sections were cut on a cryostat at 50 μm

intervals. Sections were mounted on SuperFrost Plus glass slides, air dried, and stored at

−80°C until further processing.

Immunohistochemistry

Immunohistochemistry was performed as described (4,47). Briefly, slides were washed three

times with PBS, incubated with 3% normal donkey serum for 1 h at room temperature, and

incubated overnight with guinea pig anti-Insulin (DakoCytomation, Carpinteria, CA; A0564;

RRID: AB_2617169; diluted 1:300) + rabbit anti-Glucagon (Cell Signaling; 2760; RRID:

AB_210732; diluted 1:200) antisera. Then, the slides were washed three times with PBS and

incubated for 1 hr with secondary antibodies diluted 1:500: Alexa Fluor 594

goat anti-guinea

pig (A11076; RRID: AB_141930) + Alexa Fluor 488

donkey anti-rabbit (A32790; RRID:

AB_2762833) (Invitrogen, Life Technology Corporations, Eugene, OR). Next, slides were

washed with PBS and coverslipped with Vectashield-DAPI mounting medium (Vector

Laboratories, Burlingame, CA).

ACCEPTED MANUSCRIPT

Downloaded from https://academic.oup.com/endo/advance-article/doi/10.1210/endocr/bqae048/7646761 by UCSD-Philosophy user on 23 April 2024

Islet morphology

Islet morphology was performed as described (20). Briefly, all islets within a series of four 8 µm-

thick head-to-tail pancreas sections, separated from each other by at least 50 µm, were studied

by first imaging each section in its entirety using the 20X objective of a ZEISS Axio Scan.Z1

Slide Scanner coupled with Zen Lite 2.3 (ZEISS Research Microscopy, White Plains, NY) and

then by extracting individual 8-bit RGB images of each islet. These images were processed and

islet morphology was characterized by an investigator blinded to genotype and treatment by

analyzing each islet image with a set of programs (20) which interface with Fiji is Just ImageJ

(http:/rsbweb.nih.gov/ij/; https://imagej.net/Fiji/Downloads) software. Islet cross-sectional area,

Ferret’s Diameter, circularity (4



a/p

;

where a = area and p = length of perimeter),

β-cell cross-

sectional area per islet, and α-cell cross-sectional area per islet were measured. β-cell mass

was calculated by multiplying the total β cell cross-sectional area within 4 pancreatic

sections/total pancreas area of those 4 sections by the pancreas plus spleen weight. Of note,

we weighed the pancreas together with the spleen to help orient the pancreas for histology, and

thus do not have the pancreas weight alone. Hence, instead of multiplying the relative

β-cell

cross-

sectional area by pancreas weight to calculate the β-cell mass, it was calculated as stated

above). Total islet number within 4 pancreatic sections per mouse and total islet number per

total pancreas area (in µm

x 10

) of those 4 sections also was calculated.

Review of litter sizes

Average litter sizes from C57BL/6N x C57BL/6N crosses and GKO heterozygote x GKO

heterozygote crosses were determined using data from the last 100 litters of each cross that

were set up in our mouse colony. Notably, these numbers reflect the numbers of mice in the

litters that reached weaning age (3-4 weeks); we did not record pups that may have died prior to

weaning.

ACCEPTED MANUSCRIPT

Downloaded from https://academic.oup.com/endo/advance-article/doi/10.1210/endocr/bqae048/7646761 by UCSD-Philosophy user on 23 April 2024

Statistical analysis

Data are presented as mean ± SEM. Two-tailed statistical analysis and graph preparations were

performed using GraphPad Prism 10.1.2. An unpaired Student's

-test, two-way ANOVA or

repeated measures two-way ANOVA were performed. Post hoc analysis -- in particular, a

Tukey’s multiple comparisons test -- was performed and shown in the figure panels, when the

interaction of the factors (column factor and row factor) was found to be statistically significant

(48)

. Data with unequal variance as determined by “F test to compare variances” for datasets of

two groups or “Spearman's test for heteroscedasticity” (a test for homogeneity of variances) for

datasets including more than two groups (e.g. those analyzed by two-way ANOVA) were log-

transformed prior to performing analyses. Notably, even though unequal variance was noted in

the data set for Figure 1B, we were unable to log transform those data due to the values being 0

for two of the four groups. For datasets analyzed by repeated measures two-way ANOVA, the

Geisser-Greenhouse correction was used instead to correct unequal variances. Tests for

residuals were performed on all datasets including more than two groups, and showed normal

distribution in each case. The strength of the linear relationship between 2 sets of variables was

compared by Pearson’s correlation coefficient. Outliers, if any, were removed using Grubb’s

test

. ‘ns’ represents no statistical significance.

values < 0.05 were considered statistically

significant and

values ≥ 0.05 and < 0.1 were considered evidence of a statistical trend.

Percent change values included in the Results section represent averages and are statistically -

significant; if instead, the % change is associated with a

value ≥ 0.05 and < 0.1, that

value

is listed alongside the % change.

ACCEPTED MANUSCRIPT

Downloaded from https://academic.oup.com/endo/advance-article/doi/10.1210/endocr/bqae048/7646761 by UCSD-Philosophy user on 23 April 2024

Results

Plasma ghrelin and LEAP2 levels in pregnant and non-pregnant female ghrelin-KO mice

and wild-type littermates

We started by determining

ad lib

-fed plasma ghrelin and LEAP2 levels in pregnant (GD13.5)

GKO mice and WT littermates. Ghrelin was on average ~59% lower in pregnant WT mice than

in non-pregnant WT mice, whereas it was undetectable in both pregnant and non-pregnant

GKO mice (Fig. 1B). In contrast, pregnancy increased LEAP2 (by ~51% in WT mice and by

~68% in GKO mice), without an effect of genotype (Fig. 1C). The plasma LEAP2/ghrelin molar

ratio increased significantly (by ~131%) in pregnant vs. non-pregnant WT mice (Fig. 1D).

Effects of ghrelin deletion on metabolic and biophysical adaptations in pregnant and

non-pregnant female mice

Body weight curves for both WT and ghrelin-KO mice progressively increased over the

pregnancy period examined (from GD0.5 to GD13.5), with that of ghrelin-KO mice being slightly

lower (

=0.089; Fig. 2A). Body weight gain during these 13 days was genotype-independent

(WT: 6.78 g; GKO: 6.14 g; Fig. 2B). Pregnancy significantly raised GD13.5 body weights (by

~23.7-34.5%) and GD9.5-GD10.5 24 h food intake (by ~31.6-41.7%), in a genotype-

independent manner (Fig. 2C-D). Mean body weights of pregnant mice at GD13.5 after

subtracting the weights of the uteri with fetuses were genotype-independent (data not shown).

While fat mass was independent of both genotype and pregnancy, lean mass was independent

of genotype but reduced by pregnancy (by 2.4-3.5%; albeit the measures determined for the

pregnant mice do not account for contributions by fetal tissues) (Fig. 2E-F).

Ad lib

-fed plasma

insulin levels were increased as a result of pregnancy (by ~319-535% in WT and GKO mice) but

not ghrelin deletion, and there were no effects of pregnancy or genotype on corresponding

ad-

lib

-fed blood glucose (Fig. 2G-H). Pancreas + spleen weight, number of fetuses per mouse, and

average fetus weights per mouse were similar in pregnant WT and GKO mice (Fig. 2I-K).

ACCEPTED MANUSCRIPT

Downloaded from https://academic.oup.com/endo/advance-article/doi/10.1210/endocr/bqae048/7646761 by UCSD-Philosophy user on 23 April 2024

Effects of ghrelin deletion on islet morphology in pregnant and non-pregnant mice

Nearly all islets within four 8 µm-thick head-to-tail pancreas sections (each section separated

from the next by





m) from each of 6-8 mice per group were examined by an investigator

blinded to genotype. This amounted to 3,073 islets from non-pregnant WT mice (n = 8 mice;

208-620 islets per mouse), 3,292 islets in pregnant WT mice (n = 6 mice; 361-793 islets per

mouse), 2,572 islets in non-pregnant GKO mice (n = 7 mice; 262-500 islets per mouse), and

3,664 islets in pregnant GKO mice (n = 7 mice; 261-813 islets per mouse). The overall

organization of the islet (centrally-distributed insulin-

immunoreactive β-cells and peripheral

glucagon-

immunoreactive α-cells; Fig. 3A) and consistent pattern of islet distribution throughout

the pancreas were similar in both WT and GKO mice and in both non-pregnant and pregnant

mice. Pregnancy and ghrelin deletion both independently increased islet cross-sectional area

(by ~34.9-46.4% and ~19.9-30.2%, respectively) (Fig. 3B). Similarly, pregnancy and ghrelin

deletion both independently increased the % of very large islets (islet cross-sectional area >25

µm

x 10

; pregnancy: by ~21.2-41.2%; ghrelin deletion: by ~21.8-42%) while decreasing the %

of very small islets (islet cross-sectional area <5 µm

x 10

; pregnancy: by ~3.9-4.9% (

=0.063);

ghrelin deletion: by ~4.2-4.9%) (Fig. 3C-D). Ghrelin deletion and pregnancy both independently

increased Ferret’s diameter but did not impact islet circularity (Fig. 3E-F). β-cell cross-sectional

area was increased ~43.3-56.1% by pregnancy and ~19.7-30.3% by ghrelin deletion (Fig. 3G).

β-cell mass was increased ~65.2-76.8% by pregnancy and ~15.7-23.8% by ghrelin deletion

(

=0.057) (Fig. 3H). Neither pregnancy nor ghrelin deletion impacted α-cell cross-sectional area

(Fig. 3I). Pregnancy but not ghrelin deletion impacted both mean number of islets observed per

mouse (raised;

=0.008) and islet density (total islet number/total pancreatic area; decreased;

=0.088) from four head-to-tail pancreas sections (Fig. 3I-J).

ACCEPTED MANUSCRIPT

Downloaded from https://academic.oup.com/endo/advance-article/doi/10.1210/endocr/bqae048/7646761 by UCSD-Philosophy user on 23 April 2024

Correlations of plasma ghrelin, LEAP2, and LEAP2/ghrelin molar ratio with islet size

No correlations between

ad lib

-fed plasma ghrelin levels and islet cross-

sectional area, Ferret’s

diameter, β-cell cross-sectional area, β-cell mass, α-cell cross-sectional area, islet number, or

ad lib

-fed plasma insulin levels of pregnant and non-pregnant WT mice were observed

(Supplementary Fig. 1A-G) (49). Nor did correlations become apparent when data from

pregnant and non-pregnant GKO mice were included (Supplementary Fig. 1H-N) (49). Neither

were correlations observed between

ad lib

-fed plasma LEAP2 levels and those same

parameters in pregnant and non-pregnant WT mice (Supplementary Fig. 2A-G) (49). However,

positive correlations between plasma LEAP2 and those parameters did become apparent for

islet cross-sectional area (

=0.081), Ferret’s diameter, β-cell cross-sectional area and β-cell

mass, but not for the other parameters, when data from pregnant and non-pregnant GKO mice

were included (Fig. 4A-G and Supplementary Fig. 2A-G) (49). LEAP2/ghrelin molar ratio was

not correlated with any of these parameters in pregnant and non-pregnant WT mice

(Supplementary Fig. 3A-G) (49).

Effect of ghrelin deletion on glucose tolerance in pregnant and non-pregnant female mice

Oral glucose tolerance tests were performed on 6 h fasted pregnant (GD11.5) and non-pregnant

GKO mice and WT mice. The trajectories of the blood glucose curves over the 120 min

following glucose gavage differed on the basis of pregnancy rather than genotype (Fig. 5A).

This was most apparent subsequent to the spike in blood glucose at 15 min, at which time

curves of pregnant WT and GKO littermates were lower than those of non-pregnant mice (Fig.

5A). These lower blood glucose curves were also reflected in the blood glucose area under the

curves data which indicated an effect of pregnancy (lower by ~72.9-74.2%) but not of genotype

(Fig. 5B). The spike in plasma insulin 15 min after glucose gavage also was impacted by

pregnancy but not genotype (Fig. 5C), leading to ~112.6-123.8% higher plasma insulin area

under the curve in pregnant mice (Fig. 5D).

ACCEPTED MANUSCRIPT

Downloaded from https://academic.oup.com/endo/advance-article/doi/10.1210/endocr/bqae048/7646761 by UCSD-Philosophy user on 23 April 2024

Discussion

We demonstrate that in female mice, reducing ghrelin as a consequence of germline genetic

deletion of the gene encoding ghrelin is associated with increases in mean islet size,

percentage of very large islets, and β-cell cross-sectional area, just as had been described

previously in males. Also, ghrelin-

KO females exhibited increased β-cell mass. Furthermore,

although we confirmed both lower plasma ghrelin and enlarged islets (plus greater

β-cell cross-

sectional area) in GD13.5 pregnant mice than in non-pregnant mice, the two were not

correlated. However, plasma LEAP2 was positively correlated with islet size and

β-cell cross-

sectional area. Furthermore, pregnancy but not ghrelin deletion affected

ad lib

-fed insulin levels

and glucose plus insulin responses to oral glucose administration. Nor did ghrelin deletion

impact pregnancy-associated body weight gain, GD9.5- GD10.5 24 h food intake, fetus number

per mouse, or average fetus weight per mouse.

These results raise several topics worthy of discussion. The first topic concerns how well the

new plasma ghrelin and LEAP2 data match prior studies. Indeed, many prior studies that also

added protease inhibitor to and acidified plasma samples to prevent degradation of ghrelin

reported similar trajectories of lowered ghrelin during pregnancy. These include Tham et al.,

which showed markedly lower ghrelin and total ghrelin in 2

and 3

trimester pregnant women

than 9 wks postpartum (23). Baykus et al. also showed lower ghrelin and desacyl-ghrelin in

pregnant women than in the postpartum period (24). Gibson et al. found lower fasting desacyl-

ghrelin in 3

trimester diabetic women compared to 2

trimester values (25). Johnson et al.

reported that fasted plasma total ghrelin fell over the course of pregnancy in rats, recovering to

control values at parturition (26). Kaur et al. showed that lights-on plasma ghrelin was lower at

GD12.5 than in non-pregnant diestrus mice or earlier in pregnancy, although it rebounded to

early pregnancy levels by the end of pregnancy (27).

ACCEPTED MANUSCRIPT

Downloaded from https://academic.oup.com/endo/advance-article/doi/10.1210/endocr/bqae048/7646761 by UCSD-Philosophy user on 23 April 2024

Several prior studies that did not include the above-described procedural steps to best stabilize

ghrelin similarly showed lower total ghrelin in later stages of pregnancy. Makino et al. reported

lower ghrelin in 3

trimester women than in postpartum or non-pregnant women (28,29).

Fuglsang et al. reported a gradual decrease in fasting serum total ghrelin over the course of

pregnancy from the 2

to the end of the 3

trimester, rising again by 5-9 wks postpartum (30).

Riedl et al. also reported lower total ghrelin in pregnant women than postpartum women (29).

Garces et al. reported that 3

trimester fasting serum total ghrelin was lower than the 1

and 2

trimester levels (which were equivalent) (31)

. Damjanovic et al. reported lower “ghrelin” in 2

and 3

trimester women than in 1

trimester women (32). Further, Palik et al. reported fasting

serum ghrelin that was lower in 3

trimester women as compared to non-pregnant women

(although higher in in the 2

trimester women) (33). Shibata et al. demonstrated a gradual fall in

plasma total ghrelin in rats from diestrus to GD5 to GD10 to GD15 (although with a rebound by

GD20) (34). Govoni et al. reported that over the course of pregnancy in pigs, ghrelin was

highest at GD30 and lower at GD60 and GD90 (35).

However, not all studies included results that aligned with those findings. For instance, three

studies that optimally processed plasma reported either unchanged plasma ghrelin as

pregnancy progressed (in diabetic women), higher total ghrelin but unchanged ghrelin (in

GD17.5 mice at lights-off), or higher desacyl-ghrelin but unchanged ghrelin (in GD19 and GD21

Wistar rats) (25,27,36). Additionally, other studies that did not take measures to best stabilize

ghrelin reported unchanged desacyl-ghrelin over the course of pregnancy in humans (50),

higher ghrelin in 2

and 3

trimesters compared to 1

trimester in humans (51), unchanged

total ghrelin during pregnancy in rats (52), unchanged total ghrelin over the course of pregnancy

ad lib

-fed rats but higher total ghrelin over the course of pregnancy in calorie-restricted rats

(53), increased ghrelin in GD13 and GD18

ad lib

-fed rats compared to postpartum (54), or a

ACCEPTED MANUSCRIPT

Downloaded from https://academic.oup.com/endo/advance-article/doi/10.1210/endocr/bqae048/7646761 by UCSD-Philosophy user on 23 April 2024

progressive increase in serum total ghrelin from GD12 to GD16 to GD21 in rats (all higher than

levels in virgin rats) (31).

To our knowledge, Garces et al was the only study besides the current one to have measured

LEAP2 levels in pregnancy (31). They reported a progressive reduction in serum LEAP2 levels

from the 1

, to the 2

, and then to the 3

trimester in humans (31). A similar finding was

observed in rats, specifically with serum LEAP2 progressively falling from GD12 to GD16 to

GD21 (all of which were lower than levels in virgin rats) (31). The exact reasons for the

discrepancies between those results and our current finding of increased plasma LEAP2 in

pregnant mice, as well as for the discrepancies in general among the many studies that

measured plasma ghrelin in pregnancy, are unclear although we presume may be related to

species differences, differential sample processing (especially for the ghrelin levels), body

weights, time-of-day, or time of blood draw in relation to meals. It is also important to note that

the current study did not specifically investigate whether the observed plasma LEAP2 and

ghrelin levels result from pregnancy per se or the body weight increase associated with

pregnancy.

The next discussion topic concerns the effects of ghrelin reduction on islet size and

β-cell cross-

sectional area in females and as a result of pregnancy. Previously, we demonstrated that

juvenile and adult male GKO mice exhibited increases in mean islet size, percentage of very

large islets, and β-cell cross-sectional area. The same is observed here in adult female mice,

suggesting that the effect of reducing ghrelin to increase these parameters is sex-independent.

However, the effect of genetic ghrelin deletion to increase islet and β-cell cross-sectional areas

did not extend to physiologically reducing ghrelin as a result of pregnancy. Specifically, although

pregnancy was associated with higher mean islet size, higher percentage of very large islets,

and higher β-cell cross-sectional area, as well as with reduced plasma ghrelin, no correlation

ACCEPTED MANUSCRIPT

Downloaded from https://academic.oup.com/endo/advance-article/doi/10.1210/endocr/bqae048/7646761 by UCSD-Philosophy user on 23 April 2024

was observed between islet cross-

sectional area or β-cell cross-sectional area and plasma

ghrelin levels when the data from non-pregnant and pregnant WT mice were pooled. Nor did the

addition of data from non-pregnant and pregnant GKO mice change this result. Additionally,

although pregnant GKO mice also developed increased mean islet cross-sectional area,

percentage of very large islets, and mean β-cell cross-sectional area as compared to non-

pregnant GKO mice, there was not an interaction between pregnant state and genotype. In

other words, pregnancy led to a similar degree of islet enlargement in GKO mice as in WT mice.

These data suggest that the mechanisms responsible for pregnancy-associated islet reductions

do not include the observed reductions in plasma ghrelin. This observation is opposite that

which we suggested for diet-induced obesity (in males), which is also associated with both

reduced plasma ghrelin and islet enlargement. In particular, a negative correlation between islet

size and plasma ghrelin in high-fat diet-fed plus chow-fed WT mice, together with even larger

islet sizes in high-fat diet-fed GKO mice than in high-fat diet-fed WT mice, suggested that

reduced ghrelin contributed to, but was not wholly responsible for diet-induced obesity

–

associated islet enlargement (20).

Perhaps, the differential effect of physiological ghrelin reduction on islet size in pregnancy vs.

diet-induced obesity is related to magnitude of the reduction. However, such is unlikely as

plasma ghrelin was ~59% lower in pregnant WT mice than non-pregnant WT mice but only

~38% lower in diet-induced obese WT mice than lean mice (20). Another possibility is that a

differential exposure time to reduced ghrelin is responsible. Particularly, here we examined

islets at ~2 weeks of the usual 3-week gestation whereas the islets from diet-induced obese

male WT mice were examined after 10 weeks on high-fat diet. Then again, an effect of reduced

ghrelin on islet enlargement in diet-induced obesity but not pregnancy is not necessarily

unexpected. Indeed, Rieck et al demonstrated distinct β-cell gene transcriptional changes in

pregnant mice vs. diabetes-resistant leptin-deficient mice vs. the PANIC-ATTAC model of

ACCEPTED MANUSCRIPT

Downloaded from https://academic.oup.com/endo/advance-article/doi/10.1210/endocr/bqae048/7646761 by UCSD-Philosophy user on 23 April 2024

inducible and reversible

β-cell ablation, all of which are associated with expanded β-cell mass

(40). Based on those findings, the authors of that study suggested that diverse signaling

mechanisms are engaged to achieve increased β-cell mass depending on the physiological

setting (40).

Of great interest, although plasma ghrelin was not correlated with islet cross-sectional area,

Ferret’s diameter, or β-cell cross-sectional area in pregnant and non-pregnant WT mice, plasma

LEAP2 was positively correlated with those parameters when pregnant and non-pregnant WT

and GKO mice were considered. In other words, more plasma LEAP2 was associated with

larger islets, including in the setting of pregnancy in which plasma LEAP2 levels are higher. This

result suggests that reduced GHSR signaling remains an important determinant of enlarged islet

size in pregnancy (just as it is in diet-induced obesity). However, in pregnancy, instead of

reduced ghrelin-dependent GHSR signaling resulting from reduced plasma ghrelin being the

culprit, reduced ghrelin-dependent and ghrelin-independent (constitutive) GHSR signaling

resulting from higher plasma LEAP2 may be to blame. Further studies are needed to test that

hypothesis.

As a final Discussion topic, it is also worthwhile to briefly discuss the effect of ghrelin deletion, or

rather, the lack thereof, on pregnancy outcomes and metabolic changes in pregnancy. As

summarized above, the current data suggest that the presence or absence of maternally -

produced ghrelin is immaterial to

ad lib

-fed insulin levels and glucose or insulin responses to

oral glucose administration during pregnancy or to pregnancy-associated body weight gain, 24 h

food intake when measured on GD9.5-GD10.5, fetus number per mouse, or average fetus

weight per mouse. Notably, as the GKO females in our study were mated with WT males, all of

their fetuses were GKO heterozygotes (carrying one chromosome with the WT ghrelin gene and

one chromosome with a deleted ghrelin gene). These GKO heterozygote fetuses undoubtedly

were producing some ghrelin especially by GD10.5 [as had previously been reported in (55)],

ACCEPTED MANUSCRIPT

Downloaded from https://academic.oup.com/endo/advance-article/doi/10.1210/endocr/bqae048/7646761 by UCSD-Philosophy user on 23 April 2024

although not enough to be detectable in plasma samples from the dams. While uncertain, it is

possible that this fetal source of ghrelin might have protected the fetuses against a decrease in

fetal body weight, which is otherwise observed in pregnant Wistar rats that have received

passive immunization against ghrelin (36). As our pregnancy experiments in the current study

were focused on islet morphology in the dams at GD13.5, we did not follow the mice to delivery

to assess the effects of reducing ghrelin on litter size. That said, the fetus numbers per mouse

were equivalent in the WT and GKO dams, suggesting that litter sizes likely would be the same.

Similarly, a retrospective review of the last 100 litters derived in our mouse colony from crosses

of GKO heterozygote females with GKO heterozygote males (on a C57BL/6N background) to

generate WT and GKO (and GKO heterozygote) littermate mice, and the last 100 litters derived

in our colony from crosses of C57BL/6N females X C57BL/6N males to generate C57BL/6N

mice, does not reveal a statistically-significant difference in litter size (GKO heterozygote

crosses: 6.8



0.2; C57BL/6N crosses: 6.4



0.3). Nonetheless, another group has shown that

exposure of pups to reduced ghrelin

in utero

impairs their fertility later in life (56). Specifically,

GKO heterozygotes first were shown to have reduced 24-hr fasted plasma ghrelin levels as

compared to WT mice [at least in the non-pregnant state (plasma ghrelin was not measured in

pregnant mice in that study)] (56). Adult WT female progeny of GKO heterozygote females had

significantly reduced litter sizes as compared to adult WT female progeny of WT (“ghrelin-

replete”) females (56). Although details were not included, that study also reported that female

GKO mice (which were on a B6D2F1 background) "infrequently conceived and rarely produced

viable offspring” (56). Clearly, more studies are needed to further define the impacts of ghrelin

and LEAP2 on issues of fertility and fetal development.

In conclusion, we newly demonstrate that reducing ghrelin as a result of genetic deletion

increases islet size and

β-cell cross-sectional area in female mice, similar to our recent report in

male mice. Yet, despite demonstrating reduced plasma ghrelin in pregnant mice and confirming

ACCEPTED MANUSCRIPT

Downloaded from https://academic.oup.com/endo/advance-article/doi/10.1210/endocr/bqae048/7646761 by UCSD-Philosophy user on 23 April 2024

increased islet size and

β-cell cross-sectional area in pregnant mice, ghrelin deletion did not

impact islet morphology in pregnancy nor was plasma ghrelin level correlated with islet size and

β-cell cross-sectional area in pregnant and non-pregnant mice. In contrast, plasma LEAP2,

which was elevated in pregnant mice, was positively correlated with those parameters when

pregnant and non-pregnant WT and GKO mice were considered. This result suggests that

reduced GHSR signaling as a result of higher plasma LEAP2 is a potentially important

contributor to enlarged islet size in pregnancy, just as reduced GHSR signaling as a result of

reduced plasma ghrelin was previously shown to be a potentially important determinant of islet

size in diet-induced obesity.

Acknowledgments

The authors thank the UT Southwestern Histo Pathology Core for assistance with tissue

sectioning and the UT Southwestern NORC Metabolic Phenotyping Core for access to the

EchoMRI.

Funding

This work was supported by the David and Teresa Disiere Foundation (to JMZ); the Diana and

Richard C. Strauss Professorship in Biomedical Research (to JMZ); the Mr. and Mrs. Bruce G.

Brookshire Professorship in Medicine (to JMZ); the Kent and Jodi Foster Distinguished Chair in

Endocrinology, in Honor of Daniel Foster, MD (to JMZ); UT Southwestern NORC

(P30DK127984, to JMZ) and the NIH (R01DK103884 and R01DK119341, to JMZ).

Author Contributions

DG co-designed the study, performed the experiments, helped process the pancreas images,

analyzed data, constructed figures, and co-wrote the manuscript. AWB designed and wrote the

codes to assess islet morphology, helped process the pancreas images, performed the blinded

ACCEPTED MANUSCRIPT

Downloaded from https://academic.oup.com/endo/advance-article/doi/10.1210/endocr/bqae048/7646761 by UCSD-Philosophy user on 23 April 2024

References

Kojima M, Hosoda H, Date Y, Nakazato M, Matsuo H, Kangawa K. Ghrelin is a growth-

hormone-releasing acylated peptide from stomach.

Nature

1999;402(6762):656-660.

Muller TD, Nogueiras R, Andermann ML, et al. Ghrelin.

Mol Metab

2015;4(6):437-460.

Tschop M, Smiley DL, Heiman ML. Ghrelin induces adiposity in rodents.

Nature

2000;407(6806):908-913.

Gupta D, Dowsett GKC, Mani BK, et al. High coexpression of the ghrelin and LEAP2

receptor GHSR with pancreatic polypeptide in mouse and human islets.

Endocrinology

2021;162(10).

Zigman JM, Jones JE, Lee CE, Saper CB, Elmquist JK. Expression of ghrelin receptor

mRNA in the rat and the mouse brain.

J Comp Neurol

2006;494(3):528-548.

Mani BK, Shankar K, Zigman JM. Ghrelin's relationship to blood glucose.

Endocrinology

2019.

Ge X, Yang H, Bednarek MA, et al. LEAP2 is an endogenous antagonist of the ghrelin

receptor.

Cell Metab

2018;27(2):461-469.

Mani BK, Puzziferri N, He Z, et al. LEAP2 changes with body mass and food intake in

humans and mice.

J Clin Invest

2019;129(9):3909-3923.

Cornejo MP, Castrogiovanni D, Schioth HB, et al. Growth hormone secretagogue

receptor signalling affects high-fat intake independently of plasma levels of ghrelin and

LEAP2, in a 4-day binge eating model.

J Neuroendocrinol

2019;31(10):e12785.

10.

Hagemann CA, Jensen MS, Holm S, et al. LEAP2 reduces postprandial glucose

excursions and ad libitum food intake in healthy men.

Cell Rep Med

2022;3(4):100582.

11.

Andreoli MF, Fittipaldi AS, Castrogiovanni D, et al. Pre-prandial plasma liver-expressed

antimicrobial peptide 2 (LEAP2) concentration in humans is inversely associated with

hunger sensation in a ghrelin independent manner.

Eur J Nutr

2023.

12.

Zhao TJ, Liang G, Li RL, et al. Ghrelin O-acyltransferase (GOAT) is essential for growth

hormone-mediated survival of calorie-restricted mice.

Proc Natl Acad Sci U S A

2010;107(16):7467-7472.

13.

Li RL, Sherbet DP, Elsbernd BL, Goldstein JL, Brown MS, Zhao TJ. Profound

hypoglycemia in starved, ghrelin-deficient mice is caused by decreased

gluconeogenesis and reversed by lactate or fatty acids.

J Biol Chem

2012;287(22):17942-17950.

14.

Shankar K, Gupta D, Mani BK, et al. Acyl-ghrelin is permissive for the normal

counterregulatory response to insulin-induced hypoglycemia.

Diabetes

2020;69(2):228-

237.

15.

Shankar K, Gupta D, Mani BK, et al. Ghrelin protects against insulin-induced

hypoglycemia in a mouse model of type 1 diabetes mellitus.

Front Endocrinol

(Lausanne)

2020;11:606.

16.

Adriaenssens AE, Svendsen B, Lam BY, et al. Transcriptomic profiling of pancreatic

alpha, beta and delta cell populations identifies delta cells as a principal target for ghrelin

in mouse islets.

Diabetologia

2016;59(10):2156-2165.

17.

DiGruccio MR, Mawla AM, Donaldson CJ, et al. Comprehensive alpha, beta and delta

cell transcriptomes reveal that ghrelin selectively activates delta cells and promotes

somatostatin release from pancreatic islets.

Mol Metab

2016;5(7):449-458.

18.

Mani BK, Zigman JM. A strong stomach for somatostatin.

Endocrinology

2015;156(11):3876-3879.

19.

Pradhan G, Wu CS, Villarreal D, et al. beta Cell GHS-R Regulates Insulin Secretion and

Sensitivity.

Int J Mol Sci

2021;22(8).

20.

Gupta D, Burstein AW, Schwalbe DC, et al. Ghrelin deletion and conditional ghrelin cell

ablation increase pancreatic islet size in mice.

J Clin Invest

2023;133(24).

ACCEPTED MANUSCRIPT

Downloaded from https://academic.oup.com/endo/advance-article/doi/10.1210/endocr/bqae048/7646761 by UCSD-Philosophy user on 23 April 2024

21.

Zigman JM, Nakano Y, Coppari R, et al. Mice lacking ghrelin receptors resist the

development of diet-induced obesity.

J Clin Invest

2005;115(12):3564-3572.

22.

Clegg DJ, Brown LM, Zigman JM, et al. Estradiol-dependent decrease in the orexigenic

potency of ghrelin in female rats.

Diabetes

2007;56(4):1051-1058.

23.

Tham E, Liu J, Innis S, et al. Acylated ghrelin concentrations are markedly decreased

during pregnancy in mothers with and without gestational diabetes: relationship with

cholinesterase.

Am J Physiol Endocrinol Metab

2009;296(5):E1093-1100.

24.

Baykus Y, Gurates B, Aydin S, et al. Changes in serum obestatin, preptin and ghrelins in

patients with Gestational Diabetes Mellitus.

Clin Biochem

2012;45(3):198-202.

25.

Gibson W, Liu J, Gaylinn B, et al. Effects of glucose and insulin on acyl ghrelin and

desacyl ghrelin, leptin, and adiponectin in pregnant women with diabetes.

Metabolism

2010;59(6):841-847.

26.

Johnson ML, Saffrey MJ, Taylor VJ. Gastrointestinal capacity, gut hormones and

appetite change during rat pregnancy and lactation.

Reproduction

2019;157(5):431-443.

27.

Kaur H, Muhlhausler BS, Sim PS, et al. Pregnancy, but not dietary octanoic acid

supplementation, stimulates the ghrelin-pituitary growth hormone axis in mice.

Endocrinol

2020;245(2):327-342.

28.

Makino Y, Hosoda H, Shibata K, et al. Alteration of plasma ghrelin levels associated with

the blood pressure in pregnancy.

Hypertension

2002;39(3):781-784.

29.

Riedl M, Maier C, Handisurya A, Luger A, Kautzky-Willer A. Insulin resistance has no

impact on ghrelin suppression in pregnancy.

J Intern Med

2007;262(4):458-465.

30.

Fuglsang J, Skjaerbaek C, Espelund U, et al. Ghrelin and its relationship to growth

hormones during normal pregnancy.

Clinical endocrinology

2005;62(5):554-559.

31.

Garces MF, Buell-Acosta JD, Angel-Muller E, et al. Study of the ghrelin/LEAP-2 ratio in

humans and rats during different phases of pregnancy.

Int J Mol Sci

2022;23(17).

32.

Damjanovic SS, Stojic RV, Lalic NM, et al. Relationship between basal metabolic rate

and cortisol secretion throughout pregnancy.

Endocrine

2009;35(2):262-268.

33.

Palik E, Baranyi E, Melczer Z, et al. Elevated serum acylated (biologically active) ghrelin

and resistin levels associate with pregnancy-induced weight gain and insulin resistance.

Diabetes Res Clin Pract

2007;76(3):351-357.

34.

Shibata K, Hosoda H, Kojima M, et al. Regulation of ghrelin secretion during pregnancy

and lactation in the rat: possible involvement of hypothalamus.

Peptides

2004;25(2):279-287.

35.

Govoni N, Parmeggiani A, Galeati G, et al. Acyl ghrelin and metabolic hormones in

pregnant and lactating sows.

Reprod Domest Anim

2007;42(1):39-43.

36.

Nakahara K, Nakagawa M, Baba Y, et al. Maternal ghrelin plays an important role in rat

fetal development during pregnancy.

Endocrinology

2006;147(3):1333-1342.

37.

Banerjee RR, Cyphert HA, Walker EM, et al. Gestational diabetes mellitus from

inactivation of prolactin receptor and MafB in islet beta-cells.

Diabetes

2016;65(8):2331-

2341.

38.

Kim H, Toyofuku Y, Lynn FC, et al. Serotonin regulates pancreatic beta cell mass during

pregnancy.

Nat Med

2010;16(7):804-808.

39.

Parsons JA, Brelje TC, Sorenson RL. Adaptation of islets of Langerhans to pregnancy:

increased islet cell proliferation and insulin secretion correlates with the onset of

placental lactogen secretion.

Endocrinology

1992;130(3):1459-1466.

40.

Rieck S, White P, Schug J, et al. The transcriptional response of the islet to pregnancy in

mice.

Mol Endocrinol

2009;23(10):1702-1712.

41.

Butler AE, Cao-Minh L, Galasso R, et al. Adaptive changes in pancreatic beta cell

fractional area and beta cell turnover in human pregnancy.

Diabetologia

2010;53(10):2167-2176.

ACCEPTED MANUSCRIPT

Downloaded from https://academic.oup.com/endo/advance-article/doi/10.1210/endocr/bqae048/7646761 by UCSD-Philosophy user on 23 April 2024

42.

Moin ASM, Zeng K, Rizza RA, Dhawan S, Butler AE. Chromogranin A-positive hormone-

negative endocrine cells in pancreas in human pregnancy.

Endocrinol Diabetes Metab

2021;4(2):e00223.

43.

Sorenson RL, Brelje TC. Adaptation of islets of Langerhans to pregnancy: beta-cell

growth, enhanced insulin secretion and the role of lactogenic hormones.

Horm Metab

Res

1997;29(6):301-307.

44.

Halban PA, Polonsky KS, Bowden DW, et al. beta-cell failure in type 2 diabetes:

postulated mechanisms and prospects for prevention and treatment.

J Clin Endocrinol

Metab

2014;99(6):1983-1992.

45.

Rodriguez JA, Bruggeman EC, Mani BK, et al. Ghrelin receptor agonist rescues excess

neonatal mortality in a prader-willi syndrome mouse model.

Endocrinology

2018;159(12):4006-4022.

46.

Shankar K, Takemi S, Gupta D, et al. Ghrelin cell-expressed insulin receptors mediate

meal- and obesity-induced declines in plasma ghrelin.

JCI Insight

2021;6(18).

47.

Mani BK, Uchida A, Lee Y, et al. Hypoglycemic effect of combined ghrelin and glucagon

receptor blockade.

Diabetes

2017;66(7):1847-1857.

48.

Steel RG, Torrie JH, Dickey DA. Principles and procedures of statistics: a biometrical

approach

49.

Gupta D. Impact of ghrelin on islet size in non-pregnant and pregnant female mice. V1

ed: Texas Data Repository; 2024.

50.

Saylan F, Koken G, Cosar E, et al. Maternal and fetal leptin and ghrelin levels:

relationship with fetal growth.

Arch Gynecol Obstet

2011;284(2):327-329.

51.

Valsamakis G, Papatheodorou DC, Naoum A, et al. Neonatal birth waist is positively

predicted by second trimester maternal active ghrelin, a pro-appetite hormone, and

negatively associated with third trimester maternal leptin, a pro-satiety hormone.

Early

Hum Dev

2014;90(9):487-492.

52.

Taylor VJ, Patterson M, Ghatei MA, Bloom SR, Wilson CA. Ghrelin and peptide YY

(PYY) profiles in gastrointestinal tissues and the circulation of the rat during pregnancy

and lactation.

Peptides

2009;30(12):2213-2220.

53.

Gualillo O, Caminos JE, Nogueiras R, et al. Effect of food restriction on ghrelin in

normal-cycling female rats and in pregnancy.

Obes Res

2002;10(7):682-687.

54.

Szczepankiewicz D, Skrzypski M, Pruszynska-Oszmalek E, et al. Importance of ghrelin

in hypothalamus-pituitary axis on growth hormone release during normal pregnancy in

the rat.

J Physiol Pharmacol

2010;61(4):443-449.

55.

Wierup N, Sundler F, Heller RS. The islet ghrelin cell.

J Mol Endocrinol

2014;52(1):R35-

49.

56.

Martin JR, Lieber SB, McGrath J, Shanabrough M, Horvath TL, Taylor HS. Maternal

ghrelin deficiency compromises reproduction in female progeny through altered uterine

developmental programming.

Endocrinology

2011;152(5):2060-2066.

ACCEPTED MANUSCRIPT

Downloaded from https://academic.oup.com/endo/advance-article/doi/10.1210/endocr/bqae048/7646761 by UCSD-Philosophy user on 23 April 2024

Figure Legends

Figure 1.

Plasma ghrelin and plasma LEAP2 levels in pregnant WT and GKO mice.

(

)

Schematic showing the design and timeline of inducing timed pregnancy and the following

experiments. (

) Plasma ghrelin, (

) Plasma LEAP2 and (

) LEAP2:Ghrelin molar ratio in

pregnant (P) and non-pregnant (NP) WT and GKO mice. n = 8-10 mice. Data were analyzed by

way ANOVA followed by Tukey’s multiple comparisons test (

) or 2-way ANOVA (

) or

Student's unpaired

-test (

values are indicated in the figure panels (

B-C

), *

< 0.05; **

0.01; ***

< 0.001; actual

values of 0.05 or greater and of less than 0.1 are indicated

;

ns = not

significant.

Figure 2.

Metabolic changes and biophysical parameters in pregnant WT and GKO mice

(

) Body weight and (

) Body weight gain curves of WT and GKO

mice from GD0.5 to GD13.5.

(

) Body weight, (

)

Ad lib

-fed 24 h food intake between GD9.5-GD10.5, (

) Percent fat mass,

(

) Percent lean mass, (

)

Ad lib

-fed plasma insulin and (

)

Ad lib

-fed blood glucose in

pregnant (P) and non-pregnant (NP) WT and GKO

mice. (

) Fetus number, (

) Fetus + Uterus

weight/Body weight and (

) Pancreas + Spleen weight/Body weight in

ad lib

-fed WT-P and

GKO-P mice. n = 10 mice in (

A-B

), 7-10 mice in (

), 4-7 (

D-F

), 8-16 mice in (

G-H

), 10 mice in (

I-

). Data were analyzed by repeated measures two-way ANOVA (

A-B

) or two-way ANOVA (

C-

) or Student's

-test (

I-K

values are indicated in the figure panels; ns = not significant.

Figure 3.

Islet morphologic changes in pregnant WT and GKO mice

. (

) Representative

islet images from

ad lib

-fed 8-9 week-old non-pregnant (NP) and 13-15 week-old pregnant (P)

WT and GKO littermates. (

), Islet cross-sectional area. (

), Percentage of islets in islet cross-

sectional area range > 25 µm

x 10

and

(

), percentage of islets in islet cross-sectional area

range <5 µm

x 10

from pregnant (P) and non-pregnant (NP) WT and GKO mice.

(

), Ferret’s

diameter of islets. (

), Circularity of islets. (

), β-cell cross-sectional area per islet, (

), β-cell

mass, (

), α-cell cross-sectional area per islet. (

), Islet number (as counted in 4 head-to-tail

ACCEPTED MANUSCRIPT

Downloaded from https://academic.oup.com/endo/advance-article/doi/10.1210/endocr/bqae048/7646761 by UCSD-Philosophy user on 23 April 2024

pancreatic sections) per mouse, (

), Total number of islets/total pancreatic area from 4 head-to-

tail pancreatic sections per mouse. n = 6-8 mice. Data were analyzed by two-way ANOVA.

values are indicated in the figure panels.

Figure 4.

Correlations of plasma LEAP2 with islet size or plasma insulin in pregnant (P)

and non-pregnant (NP), WT and GKO mice.

Correlations of (

ad lib

-fed plasma LEAP2 with

islet cross-sectional area, (

ad lib

fed plasma LEAP2 with Ferret’s diameter, (

ad lib-

fed

plasma LEAP2 with β-cell cross-sectional area, (

ad lib-

fed plasma LEAP2 with β-cell mass,

(

ad lib-

fed plasma LEAP2 with α-cell cross-sectional area, (

ad lib

-fed plasma LEAP2 with

islet number per mouse, (

ad lib

-fed plasma LEAP2 with

ad lib-

fed plasma insulin, in

pregnant and non-

pregnant WT and GKO mice. n = 28 mice. Data were analyzed by Pearson’s

correlation and simple linear regression analysis. Pearson’s correlation coefficient (r) and

value are indicated in the figure panels.

Figure 5.

Glucose tolerance and insulin secretion in WT and GKO mice when challenged

with oral glucose.

(

) Blood glucose, (

) Blood glucose-area under curve (AUC), (

) Plasma

insulin and (

) Plasma insulin-AUC in pregnant (P; GD11.5) and non-pregnant (NP), WT and

GKO mice in oral glucose tolerance test conducted after 6 h fast. n = 4-7 mice. Data were

analyzed by 2-way repeated measures ANOVA (

A,C

) or 2-way ANOVA (

B,D

values are

indicated in the figure panels.

ACCEPTED MANUSCRIPT

Downloaded from https://academic.oup.com/endo/advance-article/doi/10.1210/endocr/bqae048/7646761 by UCSD-Philosophy user on 23 April 2024