Articles in PresS. Am J Physiol Regul Integr Comp Physiol (September 6, 2017). doi:10.1152/ajpregu.00271.2017
R-00271-2017 R1
Page 1 – Raff et al.
1
Insulin Sensitivity, Leptin, Adiponectin, Resistin, and Testosterone in Adult Male
2
and Female Rats After Maternal-Neonatal Separation and Environmental Stress.
3 4
Hershel Raffa,b, Brian Hoeyncka, Mack Jablonskia, Cole Leonovicza, Jonathan M.
5
Phillipsa, Ashley L. Gehranda,
6 Endocrine Research Laboratory, Aurora St. Luke’s Medical Center, Aurora Research Institute, Milwaukee, WI 53215
7 8 9 10 11 12 13 14 15
a
16
Keywords: HOMA-IR, adipokines, leptin, androgens, neonatal stress, sex difference
b
Departments of Medicine, Surgery, and Physiology, Medical College of Wisconsin, Milwaukee, WI 53226
Running Head: Adult Insulin Resistance after Neonatal Stress
17 18 19 20 21 22 23 24 25 26
Correspondence: Hershel Raff, PhD Endocrinology Aurora St. Luke’s Medical Center 2801 W KK River Pky Suite 245 Milwaukee WI 53215 USA Phone (414) 649-6411 Fax (414) 649-5747 Email:
[email protected]
Copyright © 2017 by the American Physiological Society.
R-00271-2017 R1 27
Page 2 – Raff et al.
ABSTRACT:
28
Care of premature infants often requires parental and caregiver separation
29
particularly during hypoxic and hypothermic episodes. We have established a neonatal
30
rat model of human prematurity involving maternal-neonatal separation and hypoxia with
31
spontaneous hypothermia prevented with external heat. Adults previously exposed to
32
these neonatal stressors show a sex difference in the insulin and glucose response to
33
arginine stimulation suggesting a state of insulin resistance. The current study used this
34
cohort of adult rats to evaluate insulin resistance (Homeostatic Model Assessment of
35
Insulin Resistance [HOMA-IR]), plasma adipokines (reflecting insulin resistance states),
36
and testosterone. The major findings were that daily maternal-neonatal separation led to
37
an increase in body weight and HOMA-IR in adult male and female rats and increased
38
plasma leptin in adult male rats only; neither prior neonatal hypoxia (without or with
39
body temperature control) nor neonatal hypothermia altered subsequent adult HOMA-IR
40
or plasma adiponectin. Adult male-female differences in plasma leptin were lost with
41
prior exposure to neonatal hypoxia or hypothermia; male-female differences in resistin
42
were lost in the adults with prior neonatal hypoxia allowing spontaneous hypothermia
43
Exposure of neonates to daily hypoxia while preventing spontaneous hypothermia led to
44
a decrease in plasma testosterone in adult male rats. We conclude that neonatal stressors
45
result in subsequent adult sex-dependent increases in insulin resistance and adipokines,
46
and that our rat model of prematurity with hypoxia with the prevention of hypothermia
47
alters adult testosterone dynamics.
48
R-00271-2017 R1 49
Page 3 – Raff et al.
INTRODUCTION:
50
Neonatal stress has dramatic acute effects in the premature infant as well as long-
51
lasting, metabolic effects that are evident in adulthood (4, 30-34, 56, 66, 79). Premature
52
birth often requires separation from direct, tactile parental or caregiver care (i.e. incubator
53
therapy) that can be mimicked in rodents by maternal-neonatal separation (1, 2, 7, 14, 38,
54
41, 45, 46, 57, 64, 85). Furthermore, premature infants often have periods of hypoxia due
55
to apneas or immature lung function that may worsen during “kangaroo care” used to
56
minimize the negative effects of neonatal separation (1-3, 7, 35, 45, 64, 71). In addition,
57
premature infants can experience significant hypothermia even when normoxic (27, 40,
58
41, 55, 76). Finally, hypoxia in the neonate results in profound, endogenous hypothermia
59
(19, 59). The consensus seems to be that isothermia should be maintained during periods
60
of acute hypoxia in the neonate even though therapeutic hypothermia is commonly used
61
to prevent the sequelae of hypoxia at birth, particularly in the central nervous system (73,
62
74).
63
Exploiting the altricial nature of the newborn rat, we have developed a rodent
64
model of prematurity and its treatment. This is accomplished by exposing neonatal rat
65
pups periodically separated from their dams to hypoxia to mimic the apnea of the
66
prematurity, while allowing or preventing the spontaneous hypothermia during hypoxic
67
exposure (9, 10, 23). We have previously shown a significant additional neonatal stress
68
response due to the prevention of hypoxia-induced spontaneous hypothermia in the
69
neonate (9, 10, 23). Furthermore, we have shown long-term alterations in adult
70
physiology subsequent to exposure to maternal-neonatal separation without and with
71
concomitant hypoxia (13, 20, 21).
R-00271-2017 R1 72
Page 4 – Raff et al.
Of relevance to the current study is our previous finding of sex differences in
73
acute adult rat insulin and glucose responses to arginine after neonatal separation and the
74
environmental stresses described above (20). In particular, the adult rats exposed to prior
75
neonatal stress demonstrated hyperinsulinemia without hypoglycemia and greater weight
76
gain reminiscent of the metabolic syndrome and early type 2 diabetes mellitus in humans
77
(48, 62). We now follow up these prior studies by correlating the calculated Homeostatic
78
Model Assessment of Insulin Resistance (HOMA-IR), a validated index of insulin
79
resistance in adult rats (11, 53, 54, 61, 68) with measurement of the adipokines leptin,
80
adiponectin, and resistin which are known to reflect the existence of, and be involved in
81
the adaptation to insulin resistance (5, 16, 39, 42, 50, 51, 69, 72, 77, 87). Furthermore, in
82
order to explore the significantly larger insulin response to arginine in males but not
83
females, we measured adult plasma testosterone levels by LC-MS/MS. We hypothesize
84
that maternal-neonatal separation without (normoxic) or with application of neonatal
85
environment stressors will result in related increases in adult insulin resistance and
86
associated increases in plasma leptin and resistin, and decreases in.adiponectin.
87 88 89
METHODS: This is the second in a series of studies in this cohort of rats exposed to neonatal
90
stress (20). Federal guidelines (http://grants1/nih/gov/grants/olaw/references/phspol.htm)
91
for use and care of laboratory animals were followed and the protocols were approved by
92
the Institutional Animal Care and Use Committee of Aurora Health Care. Timed pregnant
93
Sprague–Dawley rats (N = 10) were obtained at 18 days of gestation and housed in a
94
standardized environment (lights on 0600–1800 h). Rats were provided ad libitum
R-00271-2017 R1
Page 5 – Raff et al.
95
standard diet and water. Dams were delivered normally and cared for their pups until
96
experimentation. Each litter was assigned to a unique treatment group to avoid cross-
97
fostering (47). A total of 110 pups were studied as described below.
98
Neonatal stressors: Male and female rat pups were randomly assigned to different (90
99
min) morning neonatal treatments from postnatal day (PD) 2 to PD6 as described below:
100 101
1) Normoxia-Unseparated – the control group for normoxic separation: These pups (N = 23) were left undisturbed except for weekly cage changes.
102
2) Normoxia-Separated – the normoxic control for the environmental neonatal
103
stressors listed below: These pups (N = 22) were separated from their dams and placed
104
into an environmental chamber with bedding and a variable setting heating pad on the
105
lowest setting required to prevent a decrease in basal body temperature during separation
106
from the dams (23). Pups were allowed to huddle normally.
107
3) Hypoxia allowing spontaneous, endogenous hypothermia: These pups (N = 25)
108
were separated from the dam and placed into a chamber with bedding and a variable
109
setting heating pad on low heat. Hypoxia was induced by decreasing the chamber O2
110
concentration to 8% as described in detail previously (9, 10, 23). This results in a plateau
111
nadir transcutaneous O2 saturation of approximately 80% (10). Body temperature was
112
allowed to spontaneously decrease during hypoxia and was measured in a sentinel pup in
113
each chamber as described previously (23). Sentinel pups were not used for the adult
114
experiments. Body temperature had spontaneously decreased to 23.9±0.5° C (n = 10)
115
after 90 min of hypoxia. After 90 min, the chamber was opened to room air and the pups
116
were warmed to a normal body temperature range of 32–34 °C using a variable setting
117
heating pad set on low before returning the pups to the nest.
R-00271-2017 R1 118
Page 6 – Raff et al.
4) Induced Hypothermia: These pups (N = 15) were separated from the dam and
119
placed in a normoxic chamber with bedding on top of a cold plate (Model
120
#AHP-1200CPV; TECALAB, Chicago, IL) set between 24 and 27 °C and adjusted
121
depending on measured body temperature. Body temperature was measured in a sentinel
122
pup using RET-30-Iso rectal probes and a BAT-12 digital thermometer connected to a
123
SBT-5 switchbox (Physitemp Instruments, Clifton, NJ). Body temperature was
124
decreased to 25 °C over 30 min and held at 25° C by adjusting the temperature of the
125
cold plate. After hypothermia was completed, body temperature was allowed to increase
126
as described for hypoxia above.
127
5) Hypoxia while maintaining isothermia: These pups (N = 25) were treated the
128
same as the hypoxia pups except that body temperature was maintained at 32 °C with a
129
heat plate (Model #AHP-1200CPV; TECALAB) as described previously (23).
130
At PD22, pups were weaned and housed by sex and treatment group. Weaned
131
animals were given a standard diet and water ad libitum and handled only during weekly
132
bedding changes.
133
Adult Measurements: These measurements were obtained after an overnight fast. Rats
134
between PD105 and PD133 had been accustomed to daily handling for 5-10 min for
135
several days before experimentation in order to obtain basal stress hormone levels on the
136
day of experimentation (65). For plasma insulin and glucose measurements, blood was
137
obtained by tail nick as described previously (84). These data have been reported
138
previously and are provided here to support the HOMA-IR calculations (shown below).
139
Rats were exposed to inhaled isoflurane for 5-10 seconds and then killed by rapid
R-00271-2017 R1
Page 7 – Raff et al.
140
decapitation to obtain trunk blood in EDTA-treated tubes to generate plasma for the
141
measurement of hormones described below:
142
Plasma immunoassays: Plasma leptin and adiponectin were measured by ELISA
143
(Crystal Chem #90040 and #80570, respectively, Downer’s Grove, IL)(8). Plasma
144
resistin was measured by ELISA (BioVendor #RD391016200R, Karasek, Czech
145
Republic). The limit of detection is 0.05 ng/ml and intraassay and interassay coefficients
146
of variation are 4.9-5.2% and 4.9-9.3%, respectively.
147
Plasma testosterone: Plasma testosterone was measured by LC-MS/MS. Plasma
148
samples, quality controls or standards (100 µl) were each combined with 100 ng/dl
149
deuterium labeled d3-testosterone internal standard (100 µl) and extracted with 0.4 ml of
150
acetonitrile; the organic phase was evaporated to dryness under nitrogen in a 50 ºC water
151
bath and reconstituted in 100 µl of water/methanol (50:50, v:v). Testosterone
152
measurements were performed using a 1290 Infinity HPLC (Agilent Technologies, Palo
153
Alto, CA, USA) and a Triple Quad LC-MS (Agilent Technologies, Palo Alto, CA, USA)
154
with an ESI ion source in positive mode. The two dimensional LC separation technique
155
was performed with a Zorbax SB-C8 (2.1 x 15 mm, 3.5 µm) loading column (Agilent
156
Technologies, Palo Alto, CA, USA) and a Poroshell 120, EC-C18 (2.1 x 50 mm, 2.7 µm)
157
analytical column (Agilent Technologies, Palo Alto, CA, USA) both maintained at 50º C
158
combined with a 0.3 µm inline filter. The injection volume and flow rate range while
159
injections were on the columns was 5 µl and from 0.3-0.6 ml/min, respectively.
160
Introduction of sample onto the loading column was performed by a gradient elution of
161
mobile phase consisting of 5 mM ammonium formate in water (solvent A) and 5 mM
162
ammonium formate in methanol (solvent B) at 0.5 ml/min A:B 50:50-41:59 (0-1.2). The
R-00271-2017 R1
Page 8 – Raff et al.
163
analytical pump transferred the sample to the analytical column and then introduced it to
164
the mass spectrometer with a gradient elution of A:B 45:55-2:98 (1.1-4.4 min). The flow
165
rate of the analytical pump increased from 0.3 ml/min to 0.6 ml/min (4-4.1 min).
166
Additional, gradients used for eluting waste components and equilibration of columns to
167
initial conditions. MassHunter software (Agilent Technologies, Palo Alto, CA, USA)
168
was used to control the instruments and for data analysis. The mass spectrometer scan
169
type utilized was multiple reaction monitoring with total testosterone quantified and
170
qualified by the ion transition m/z 289.2/97.1 and m/z 289.2/109.1, respectively. d3-
171
testosterone internal standard was analyzed by the ion transition m/z 292.2/97.1. The
172
following source conditions were used: gas temperature, 300 °C; gas flow, 5 l/min;
173
nebulizer pressure, 60 psi; sheath gas temperature, 400 °C; sheath gas flow, 11 l/min;
174
capillary voltage, 3,500 V; nozzle voltage, 0 V; and an electron multiplier voltage of 400
175
V. Fragmentor voltage was 123 V for all compounds. Collision energy was 20 V for
176
internal standard and testosterone quantifier ions and 24 V for testosterone qualifier ions.
177
The functional sensitivity is 0.7 ng/dl, and intraassay and interassay coefficients of
178
variation are 1.5-3.9% and 2.6-7.8%.
179
Calculations and Statistical Analyses: HOMA-IR, a validated index of insulin
180
resistance in rats, was calculated as the product of fasting plasma glucose and fasting
181
plasma insulin divided by a constant (11, 61). Data were analyzed by two factor analysis
182
of variance (ANOVA) followed by the Holm-Sidak multiple comparisons method
183
(Sigmaplot 12.5). Since pups had to be separated from their dams to expose them to the
184
environment stressors, two different null hypotheses were evaluated. For the comparison
185
of separated (normoxia) vs unseparated (control for normoxic separated), the two
R-00271-2017 R1
Page 9 – Raff et al.
186
ANOVA factors were sex and separation. For the comparison of environmental stressors
187
(that required separation) to their control (normoxic separated), the two factors were sex
188
and environmental stressor. Data are presented as median (25%-75% percentile) or mean
189
± standard error of the mean with P