Upload
others
View
1
Download
0
Embed Size (px)
Citation preview
Rapid onset of severe septic shock in the pregnant mouse
Julia Zöllner1,2,3, Simon Lambden3,4, Noor Mohd Nasri1,2,3, James Leiper3,5 and Mark R. Johnson1,2
1 Chelsea and Westminster Hospital, 369 Fulham Road, London, SW10 9NH, UK
2 Institute of Reproductive and Developmental Biology, DuCane Road, London W12 0NN, UK
3 Nitric Oxide Signalling Group, MRC Clinical Sciences Centre, Imperial College London, Hammersmith
Hospital Campus, DuCane Road, London W12 0NN, UK
4 Department of Medicine, University of Cambridge, 5th Floor, Addenbrooke’s Hospital, Cambridge,
CB20QQ
5 Institute of Cardiovascular and Medical Sciences, University of Glasgow, University Avenue, Glasgow
G12 8QQ
Correspondence to: Mark Johnson, Imperial College Parturition Research Group, Department of
Obstetrics and Gynaecology, Imperial College School of Medicine, Chelsea and Westminster Hospital,
369 Fulham Road, London, SW10 9NH, UK.
Telephone: 44 20 88 46 78 78 Fax: 44 20 88 46 77 96
Email: mark [email protected]
Word count: 6278.
1
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
Abstract:
Aims:
Globally, sepsis is a major cause of mortality through the combination of cardiovascular collapse and
multi-organ dysfunction. Pregnancy appears to increase the risk of death in sepsis, but the exact
reason for the greater severity is unclear. In this study, we used polymicrobial sepsis induced by
caecal ligation and puncture (CLP) and high-dose intraperitoneal lipopolysaccharide (LPS; 10 or 40mg,
serotype 0111: B4) to test the hypotheses that pregnant mice are more susceptible to sepsis and that
this susceptibility was mediated through an excessive innate response causing a more severe
cardiovascular collapse rather than a reduction in microbe killing.
Methods and Results:
Initial studies found that mortality rates were greater and that death occurred sooner in pregnant
mice exposed to CLP and LPS. In pregnant and non-pregnant CD1 mice monitored with
radiotelemetry probes, cardiovascular collapse occurred sooner in pregnant mice, but once initiated,
occurred over a similar timescale. In a separate study, tissue, serum and peritoneal fluid (for protein,
flow cytometry, nitric oxide and bacterial load studies) were collected. At baseline, there was no
apparent Th1/Th2 bias in pregnant mice. Post CLP, the circulating cytokine response was the same,
but leukocyte marginationinfiltration in the lung was greater in pregnant mice, but only TNFα levels
were greater in lung tissue. The bacterial load in blood and peritoneal fluid was similar in both
groups.
Conclusion:
Sepsis-related mortality was markedly greater in pregnant mice. Cardiovascular collapse and organ
dysfunction occurred sooner in pregnancy, but bacterial killing was similar. Circulating and tissue
cytokine levels were similar, but off-target immune cell extravasation margination into other organs
was greater in pregnant mice. These data suggest that an excessive innate immune system response,
2
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
as shown by the exaggerated lung margination infiltration of leukocytes may be responsible for the
greater mortality. Approaches that reduce off-site trafficking may improve the prognosis of sepsis in
pregnancy.
3
46
47
48
49
50
Introduction
Sepsis in pregnancy is the leading direct cause of maternal mortality in the UK and, together with
haemorrhage and pre-eclampsia, it is a major cause of maternal mortality worldwide [1]. However,
the underlying mechanisms that result in the increased maternal mortality are not well understood.
Despite improvements in the overall management of sepsis, the rate of sepsis has increased over the
last few decades [2]. In the United States, between 1991 and 2003, the rates of maternal sepsis and
of sepsis-related death doubled [3-5]. Furthermore, with the emergence of increasingly virulent
strains and greater anti-microbial resistance, coupled with the trend for women to delay pregnancy
until later life, when obesity and type 2 diabetes are more common, it seems likely that maternal
deaths from sepsis will increase [6].
The immune system during pregnancy is uniquely modified to allow the development and growth of
a healthy pregnancy, while retaining the ability to fight infection. Traditionally, pregnancy was seen
as an immunosuppressed state [7]. However, the evidence indicates that the susceptibility to
infection during pregnancy is not increased, but the ability to fight infection once it has been
acquired is compromised. This suggests that we should actually consider pregnancy as a “modulated
immunologic condition” [8, 9]. However, the exact mechanism(s) responsible for the increase in
sepsis-related mortality during pregnancy are not understood. Indeed, little is known about the
impact of pregnancy on the response to infection. For example, cardiovascular function changes
dramatically during pregnancy with a marked decrease in peripheral resistance [10], which may
exacerbate any sepsis-induced hypotension. Cardiovascular dysfunction is a major driver of mortality
in sepsis and overproduction of nitric oxide (NO) is a mechanism for this [11]. Normally, a septic
insult triggers an inflammatory response involving the activation of inflammatory cells,
predominantly macrophages and neutrophils, which release cytokines and activate complement and
other mediators that include reactive oxygen species [12]. In tandem with the inflammatory
response, a reciprocal anti-inflammatory response occurs, which limits the extent of the pro-
inflammatory response. Dysregulation in this process leads to adverse outcomes in patients with 4
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
sepsis [13]. One previous study suggested that the innate response to LPS was enhanced during
pregnancy, with a greater pro-inflammatory TNFα and anti-inflammatory IL-10 response [14].
However, we recently studied the response in greater detail and found that the increase in
circulating cytokines after LPS treatment was similar in both pregnant and non-pregnant (NP) mice,
but that inflammatory cell tissue infiltration was more marked and that this was associated with a
greater hypotensive response [15].
There are no published basic science studies that have investigated the cardiovascular and immune
response in sepsis during pregnancy. In the current study, we have established for the first time a
reproducible rodent model of polymicrobial sepsis during pregnancy using caecal ligation and
puncture (CLP) in conjunction with radio telemetry to investigate the haemodynamic and innate
immune response to sepsis. In this study we explored the hypothesis that pregnancy is associated
with altered immune and cardiovascular responses to life threatening infection.
5
77
78
79
80
81
82
83
84
85
86
87
88
Methods
Animals: All experiments were performed under UK Home Office License 70/7372 or 70/8237, and in
accordance with the regulations of the UK Animals (Scientific Procedures) Act of 1986. NP female and
male (for breeding purposes only) CD1 outbred mice were obtained from Charles River (Margate, UK)
at 6-8 weeks of age and acclimatised for at least 48 hours before mating and 7 days prior to
operating. All mice were maintained in open cages at 21 ± 1oC, with ad libitum access to food and
water, and a 12:12 light/dark cycle regimen.
Time mating: Female mice were placed overnight into cages with male studs. The following morning
the female mice were inspected for a copulatory plug. The day of the plug would indicate the
gestational day 0 (E0). From previous work in our group E16 compares to about 34 weeks gestation in
human pregnancy.
Cardiovascular telemetry: NP CD1 female mice had a Pa-C10 implantable murine radiotelemetry
pressure catheter (Data Science International, USA) implanted in the left carotid artery. The
operation was conducted under isoflurane anaesthetic (5% for induction and 2% maintenance of
isoflurane, 400mL/min). Pre-operative medication was given prior to induction (Buprenorphine
0.02mg/kg s.c.; Enrofloxacin 150uL of 5% m/v). The left common carotid artery was isolated, a small
incision made, and the catheter tip inserted and advanced distally towards the aortic arch.
Postoperatively the mice were monitored very closely and 7 days recovery was allowed. Following
this, a continuous 24 hour recording was commenced to collect baseline data (mean arterial
pressure, heart rate and body temperature). This allowed the animal to be their own control.
Subsequently (about 12 days post- operatively), female mice were time-mated if appropriate. Data
was recorded at pre-determined time points in each experiment. Implantable temperature
transponders IPTT-300 (BMDS, USA) were placed subcutaneously (s.c) to allow non-invasive
temperature measurements.
6
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
CLP: Mice were subjected to the CLP or sham operation in both pregnant and NP groups. Animals
were prepared and induced as above (see cardiovascular telemetry). A longitudinal midline incision
was made to expose the caecum carefully and 70% ligated. Double caecal puncture was performed
using an 18G needle midway between the ligation and the tip of the caecum. The caecum was then
replaced into the abdominal cavity and the muscle layers and skin closed. In the sham operation the
caecum is identified, presented onto the abdomen and replaced again without intervention.
Following the operation, animals were monitored using telemetric measurements of blood pressure,
heart rate and temperature to determine the onset of sepsis and to enable euthanasia at a pre-
defined and humane endpoint (see Suppl. 1). Euthanasia was performed by cervical dislocation and
cease of circulation was used as confirmation method as per the Animals (Scientific Procedures) Act
1986 (ASPA) and the European Directive 2010/63/EU.
Tissue Collection: Blood and tissue was collected in a separate experiment at 3 and 6 hours following
CLP or sham operation. Blood was obtained by cardiac puncture and spun at 13,000rpm for 20
minutes for serum extraction. Tissue was collected, snap frozen and stored at -80⁰C until further use.
LPS dose response: Lipopolysaccharide (LPS, 0111:B4, Sigma-Aldrich, UK) endotoxin was injected i.p.
to cohort of NP and pregnant animals. The doses used were 10mg, and 40mg per kg.
Flow cytometry: Following tissue collection as described above, the lung tissue was homogenised in
1mL of IC fixation buffer (eBioscience, UK) using gentleMACS M tubes (Miltenyi Biotex, UK) for 1
minute. The cell suspension was filtered using 40μm nylon cell strainer (BD Falcon, UK). This filtered
suspension was centrifuged twice at 2000 rpm for 5 minutes at 4⁰C to extract leukocytes and re-
suspended in facs wash buffer (FWB) before finally re-suspended in 400uL of perm wash buffer
(PWB). The cell suspension was incubated with antibody mix for 30 minutes at 4ᵒC. Fluorophore-
conjugated rat anti-mouse monoclonal antibodies were used (see suppl. Table 1). Cell counts were
determined using AccuCheck counting beads (Invitrogen.UK), 20uL were added per sample. The
7
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
gating strategy used has been described previously [16] and [15]. Samples were run using the BD LSR
II flow cytometer (BD, UK). Data were analysed using FlowJo dongle version V10.
Plasma and peritoneal bacterial load: All samples were collected 8 hours following CLP in NP and
pregnant animals. Peritoneal samples were obtained by filling the peritoneal cavity with 3mls of
sterile PBS. Aliquots of whole blood and peritoneal washouts were serially diluted with PBS and
plated on tryptic soy agar plates (BD, UK). All plates were incubated overnight at 37˚C. The following
day, all plates were inspected and counted for Colony Forming Units (CFU) per ml.
Multiplex assay: To assess cytokine and chemokines levels during sepsis a 23-plex cytokine and
chemokine assay was performed (Bio-rad, UK) using serum samples. The manufacture’s protocol was
strictly followed. Analytes were quantified using a Luminex MAGPIX instrument with xPonent 4.2
software (Luminex Corp., USA). The concentrations of analytes were calculated by comparison to
standard curves.
Nitric Oxide Analyser: To quantify nitric oxide production a Sievers nitric oxide analyser (NOA, GE
Analytical Equipment, UK) was used. To define nitric oxide production, concentrations of both nitrite
and nitrate (NOx) levels were determined. These results were used to measure total NOx
concentration in mouse serum and tissue homogenates. In brief, samples were prepared by protein
precipitation with four volumes of methanol and centrifugation at 16,000 rpm for 30 minutes. To
obtain the total concentrations of NO, nitrites and nitrates in each sample is reduced back to NO.
This occurred in a reaction with VCl2 (Sigma, UK) in 1M HCl (Sigma, UK) at 90⁰C before passing into
the NOA system to react with ozone (O3). The production of a chemiluminescent signal was
quantified by a photo-multiplier system. The concentration of NOx was calculated by comparison to a
calibration curve of known nitrate samples.
Statistical analysis
The Kolmogoroff–Smirnoff test was used to assess the distribution of the data. Unpaired t-test or
Mann Whitney test were used where appropriate, to examine the differences between the groups.
8
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
For repeated measures or >2 groups an analysis of variance (ANOVA) was used followed by
Bonferroni correction. A p-value of <0.05 was considered to be statistically significant. Data was
analysed using GraphPad PRISM software (V.5).
9
163
164
165
166
Results
Mortality following CLP and high dose LPS: All mice were randomly allocated to naïve, sham or CLP
groups. Post-operatively, all animals were monitored closely for a 5 day period or until a humane
end-point was reached. The humane endpoint was defined as a temperature below 25 0C or the
animal demonstrating signs of severe sepsis in their coat/appearance, behaviour, posture/mobility,
breathing and movements (Suppl. Table 1). TAs expected, the mortality rate, assessed over 5 days,
following a septic insultCLP compared to sham operation during pregnancy was 92% (p<0.0001) over
a 5 day period in comparison to 50% (p=0.048) in non-pregnant (NP) mice, with a median survival of
23 hours versus 100 hours, respectively (Figure 1A). These findings established that a polymicrobial
septic insult in the form of CLP of the same magnitude resulted in worse outcomes during pregnancy
as compared to NP mice (p=0.0042). Similarly, mortality rates were observed in pregnant and non-
pregnant mice after 40mg/kg of LPS (serotype 0111:B4; Figure 1B).
Haemodynamic response to CLP: A comparison of all non-pregnant and pregnant mice following CLP
regardless of outcome, demonstrated that there was no statistical difference between mean MAP in
NP CLP 99.56 ± 1.587 vs Pregnant CLP 97.43 ± 1.988 (Suppl. Figure 1A). However, over 24 hours a
significant change difference in heart rate (HR; beats per minute, bpm) was noted in NP CLP 546 ±
14.11 vs Pregnant CLP 421 ± 12.45, p<0.0001 (Suppl. Figure 1B). The CLP NP group was then divided
into non-survivor and survivor groups, whereas the CLP pregnant group was not separated as there
was only one survivor. In the group monitored by telemetry Oforver a 100 hours telemetry period,
the median survival was 30 hours during pregnancy vs. 100 hours in NP mice. A delta analysis of the
mean arterial blood pressure (Figure 2A) and heart rate (Figure 2B) over the 24 hours preceding the
experimental endpoint was conducted. As compared to their sham groups, a statistically significant
decrease was noted in MAP and HR over the 24 hours in the NP non-survivor group and pregnant
CLP, p<0.0001 and p<0.0001, respectively. The maximum increase in ∆MAP was higher in pregnancy,
∆2.02 ± 1.26 mmHg in NP non-survivor mice vs. ∆7.62 ± 1.43 mmHg in pregnant mice as compared to
their respective sham control (NP non-survivor p<0.001, pregnant CLP p<0.05; Figure 2C). The 10
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
maximum decrease in ∆MAP was ∆-34.84 ± 8.97 mmHg in NP non-survivor mice and ∆-25.50 ± 5.16
mmHg in pregnant mice (NP non-survivor p<0.01, pregnant CLP p<0.05; Figure 2D). Comparing the
CLP NP non-survivor and CLP pregnant group there was no statistical difference over 24 hours or at
the experimental endpoint in either MAP or HR.
Organ function: Worsening organ function during septic pregnancy resulted in a The higher base
excess deficit was greater in pregnant mice post CLP (-7.4 ± 1.5) at 6 hours following CLP as
compared to NP mice (-2.3 ± 0.3, p=0.049; Suppl. Figure 2F5A). The bicarbonate levels tended to be
lower in the CLP pregnant group (Pregnant CLP 21.7 ± 1.7 vs. non-pregnant CLP 25.7 ± 0.6; Suppl.
Figure 25EB).
Inflammatory cell trafficking: In separate groups of mice treated in an identical manner, flow
cytometry was performed as previously described [15]. Comparisons were made between pregnant
and NP mice in each group (naïve, sham and CLP). Measurements were made at 3 and 6 hrs after
surgery.
Circulating: At 3 and 6 hours after CLP or sham operations, circulating Ly6CHigh monocytes tended to
increaseddecline, while declined and neutrophils remained unchanged , but both were comparable
in both NP and pregnant mice (Figure 3A-D); similarly, for circulating Ly6CLow monocytes and natural
killer cells (Suppl. Figure 2A-D).
Lung: Lung leukocyte margination infiltrationnumbers increased to a greater extent was more
marked in pregnant CLP mice at both time points (Ly6CHigh monocytes: pregnant CLP vs. NP CLP,
p=0.004 and p=0.028; neutrophils: pregnant CLP vs. NP CLP, p=0.017 and p=0.025; Figure 3E-H). For
Ly6cLow monocytes and natural killer cells, there were no significant differences at 3 hours (Suppl.
11
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
Figure 3A-D), but at 6 hours, lung Ly6cLow monocytes were increased in the pregnant CLP group
(p=0.028; Suppl. Figure 3C).
Peritoneum: NLeukocyte trafficking into peritoneal fluid was assessed 6 hours post CLP (Figure 3I-L).
In pregnant mice, nNeutrophil numbers in the peritoneal fluid were increased higher at 6 hours in
pregnant mice following CLP (p=0.033; Figure 35J).
Circulating and lung cytokine levels: At baseline in the circulation, TNFα and CCL-2 levels were
significantly higher in the naïve pregnant mice (p=0.0004 and 0.045 respectively, Figure 4C&F). At 6
hours post sham surgery, IL-6 levels were marginally higher in the NP mice (p=0.0337, Figure 4B). At
6 hours following CLP, NP mice had higher levels of TNFα, IL-12 p40 and VEGF (p=0.002, p=0.037 and
p=0.025 respectively; Figure 5C and Suppl. Figure 4A&B).
In the lung at 6 hours, cytokine levels were very similar in all groups, with the exception of TNF α,
where the levels were higher in pregnant mice, but this appeared to be due to a decline in levels in
the NP group (p<0.0005, Figure 5C).
Bacterial numbers: The number of colony forming units in whole blood and peritoneal fluid were
similar 8 hours after CLP. In peritoneal fluid, pregnant mice tended to have higher bacterial loads (4.1
x 104 [interquartile range [IQR], 3.8 x103 – 1.4 x 105]) compared to NP mice (1.3 x 103 colony-forming
unit (CFU)/ml [IQR, 283 – 5.9 x 104; Figure 6A]). In whole blood, no significant difference was seen
between pregnant and NP mice (484 [IQR, 225 – 792] vs. 1250 [IQR, 208 – 5075, Figure 6B]).
Organ function: Worsening organ function during septic pregnancy resulted in a higher base excess
deficit (-7.4 ± 1.5) 6 hours following CLP as compared to NP mice (-2.3 ± 0.3, p=0.049; Suppl. Figure
5A). The bicarbonate levels tended to be lower in the CLP pregnant group (Pregnant CLP 21.7 ± 1.7
vs. non-pregnant CLP 25.7 ± 0.6; Suppl. Figure 5B).
12
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
Total nitrates/nitrites: In a separate group of mice treated in an identical manner, no differences in
total nitrates and nitrites (NOx) concentration were found in NP and pregnant mice following CLP as
compared to their respective sham groups 3 or 6 hours following CLP (Suppl. Figure 56).
13
239
240
241
242
Discussion
To date, few studies have compared the response to sepsis in pregnant and NP rodents and those
that have used LPS as a stimulus [14, 15]. Here, we have compared the response to polymicrobial
sepsis induced by CLP in pregnant and NP mice, showing that the rate of onset of sepsis is quicker
and the mortality greater in pregnant mice. The CLP model has significantly contributed to the
understanding of the pathophysiology of sepsis and its associated molecular mechanisms 17. We used
both high-dose LPS and CLP to model sepsis in pregnancy. High-dose LPS resulted in the rapid onset
of septic shock onset underwithin 24 hours and consequently maternal deaths. Here we used
bothThe CLP model and high-dose LPS, finding demonstrated that mortality rates were comparable
however, resulted in maternal deaths over 30 hours. The increased maternal mortality in both
models perhaps suggests that the maternal response perhaps suggesting that it is the maternal
response rather than deficient bacterial killing, which ismay be responsible for the adverse outcome.
Significantly, post CLP, although cardiovascular collapse occurred earlier in pregnant mice, we found
that the degree of hypotension was similar in pregnant and NP non-surviving mice, implying that
although the blood pressure is lower during pregnancy, sepsis does not result in a greater
hypotensive response. Neutrophil migration and bactericidal activity both appeared to be intact, with
similar bacterial numbers in blood and peritoneal fluid, suggesting that the high rates of mortality
with pregnancy were not related to overwhelming infection, in agreement with as suggested by the
LPS data. The key difference we found was that pregnant mice show greater off-target inflammatory
cell margination extravasationnumbers in into off-target (tissues unaffected by the initial insult)other
organs areas, including like the the lungs, which has been implicated in adverse outcomes in other
studies of sepsis 18-21.
In the current study, we observed a more rapid cardiovascular collapse in pregnant mice, but that in
the period prior to reaching the humane end-point, the decline in blood pressure was comparable in
non-surviving pregnant and NP mice, meaning that once septic shock had set in, the rate of
deterioration was similar in pregnant and NP mice. There are 2 key differences: the proportion of 14
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
survivors, pregnant 8% vs. NP 50%, and the speed of deterioration after CLP, pregnant median
survival time 24 hours vs. NP 100 hours. These differences suggest that the response to the same
septic insult is more rapid and severe in pregnancy. Consequently, we opted to focus on the acute
response to sepsis, reasoning that we would distinguish differences in the first 6 hours which would
indicate why the onset of sepsis in pregnancy is more acute and ultimately, more severe. The earliest
deaths in the pregnancy cohort occurred at 6 hours post CLP, consequently, we chose to study
animals at 3 and 6 hours post CLP rather than later time points, in order not to introduce bias into
our data by the loss of some more severely affected mice.
Pregnancy and models of sepsis: There are no comparative studies using CLP in pregnancy. We
previously studied the impact of low-dose LPS on the immune and cardiovascular response in
pregnant and NP mice and found that in pregnancy the hypotensive effect was more marked in
association with greater monocyte margination in the pulmonary vasculature; while markers of
cardiac dysfunction and circulating cytokine levels were similar in pregnant and NP mice suggesting
that the greater monocyte activation/margination could impair vascular function in pregnant mice 15.
The only other study to investigate this used high dose LPS during pregnancy, demonstrated a higher
mortality (76% pregnant vs. 25% NP mice) with increased TNFα and decreased IL-10 immune
response 14. The higher mortality is consistent with our observations, although we found that TNF α
was higher in NP pregnant mice and that IL-10 was similar in both groups. Overall, this paper
supports the hypothesis that sepsis results in increased mortality during pregnancy.
Septic shock: The CLP model shows that the haemodynamic dysfunction in NP and pregnant mice is
closely related to poor outcome and mortality just as in human sepsis 22. It appears that maintaining
a critical threshold of perfusion is important and deteriorating below that provokes terminal decline.
Lower basal blood pressure in pregnancy may contribute to greater susceptibility to septic shock.
Indeed, in our first study, using low-dose LPS, the blood pressure declined markedly in pregnant but
not in the NP mice 15. These data supported the hypothesis that the cardiovascular response to
polymicrobial sepsis would be more marked in pregnancy. Indeed, the hypotensive response in 15
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
pregnant mice appears to occur earlier and result in worse outcomes. However, if only the non-
survivors are considered, then the cardiovascular response is comparable. This suggests that the
ability to maintain BP above a threshold level is the key determinant of survival. Consistent with this,
circulating NO levels were comparable between NP and pregnant mice and VEGF levels were actually
greater in NP mice, perhaps contributing to the lower blood pressure (in the non-survivors) at the
humane endpoint. With regards to the similarity in NO levels, the relatively early time points may
have missed any later increase in NO production or reflect the difficulty to detect changes in plasma
nitrite/nitrates (NOx) associated with specific NOS isoforms 23.
There is evidence in the literature that suggests that immune cell function is impaired during sepsis
24. Further evidence suggests that pregnancy also alters the immune response to infection and it is
likely that pregnancy hormones such as oestrogen or progesterone play a role in these alterations 8.
TheTherefore, the role of the innate immune response following CLP was investigated to evaluate
potential mechanisms to explain the higher mortality noted in this study.
Circulating changes in inflammatory cells were similar; with a decline in Ly6C High monocytes being
apparent in both pregnant and NP mice post CLP. The circulating leukocyte response was also similar
between pregnant and NP mice, but in terms of monocytes at 3 hours, was associated with a decline
in circulating numbers, which tended to be greater for the NP in terms of Ly6C LO and pregnant mice
for the Ly6CHI monocytes. The key differences were found in lung extravasationmarginationnumbers
of Ly6CHigh monocytes and neutrophils, which were greater in pregnant mice at 3 and 6 hours post
procedure.
Indeed, tThe literature describes how non-specific excessive infiltration of leukocytes may be
damaging 18-21. Other studies, using CLP in NP rodents, have been informative and shown changes in
neutrophil function in severe sepsis, which have beenare suggested to contribute to the adverse
outcome. Neutrophils have been shown to migrate less well to infected areas and this dysfunction to
be associated with increased numbers of bacteria in both blood and the peritoneal fluid 25. This
16
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
failure of neutrophil migration is suggested to be NO-mediated 25, 26 and to be associated with greater
off targetinfiltrationnumbers of neutrophils marginalisation in the lung and liver 27. The greater
marginalisation infiltrationinflammatory cell numbers may 27 or may not 28 cause direct damage to
lung or liver, but exposure to a second insult, either LPS or IgG immune complexes, caused greater
lung damage in the context of sepsis 28. In terms of our work, we found that pregnancy was not
associated with impaired peritoneal neutrophil migration, the numbers were in fact greater, or
reduced bacterial killing, where the numbers of bacteria in the blood and peritoneal fluid were the
same.
Where the pregnancy data and these findings are similar is in the greater off-target trafficking of
neutrophils and monocytes into other organs like lung. The “off-target” migration of inflammatory
cells is off target effect has been shown to be CCR-2 mediated, to be prevented in CCR-2 knockout
mice or mice treated with a CCR2 antagonist with an associated improvement in outcome 29. We
have seen a similar greater off-target response in our comparison of the effects of low dose LPS in
pregnant and NP mice, which was associated with a marked hypotensive response 15. Interestingly,
when we studied the effects of RU486, the progesterone and glucocorticoid antagonist, we found
that CCR2 mediated both neutrophil and monocyte myometrial recruitment, suggesting that during
pregnancy neutrophils are CCL2 responsive 16. This may explain why we see an increase in both
monocyte and neutrophil numbers in the lung. Recruiting neutrophils to the site of infection is
important to improve survival outcomes and to initiate a bactericidal immune response 26. Here we
found that the peritoneal neutrophil infiltration was greater and of Ly6C High monocytes similar in
pregnant mice. It is theoretically possible, that the capacity of monocytes to phagocytose and
neutrophil burst activity function in pregnant mice may be reduced, but the similarity in bacterial
colony forming units in the peritoneal fluid and blood suggests that this is not the case.
Using LPS, we observed greater numbers of inflammatory cells in the lung {Edey, 2016 #393}. These
cells were found to be marginalised, to remain in the circulation, but to be adherent to the lung
vasculature prior to potentially transmigrating into the lung tissue {Edey, 2016 #393;Patel, 2015 17
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
#408}. It is possible that CLP induces a greater effect, such that the inflammatory cells actually
migrate into the lung tissue, but this remains to be studied.
In terms of circulating cytokines, the response to CLP was similar in pregnant and NP mice with the
exception of TNFα, which was higher in the NP mice (as mentioned above). Lung cytokine levels
were broadly similar with the exception of TNFα levels, which were much greater in pregnant
animals.
Bacterial Killing: Interestingly, our findings suggest that the increased mortality noted during
pregnancy was unrelated to an increased bacterial load. The similarity in the bacterial numbers
suggests that the pressure exerted by the gravid uterus did not appear to increase or reduce the
release of faecal material into the peritoneal cavity. Studies have shown that sex steroids may
modulate the immune system A study exploring sex differences to Mycobacterium marinum
demonstrated that macrophages from female mice appeared to be more activated and
demonstrated a greater migration to the site of infection 30; this suggests that sex steroid hormones
modulate the immune response. Indeed, they have been suggested to modulate the release of pro-
and anti-inflammatory cytokines, TLR expression and antibody production 31. These data suggest that
the higher sex steroid levels in pregnant mice may be responsible for the increased migration of
neutrophils to the site of injury. Whether they also have a role in the greater off-target effects
remains to be clarified. The fall in base deficit was greater in pregnant mice post CLP, consistent with
the clinical observation that increased lactate and a greater base deficit are associated with increased
mortality and further septic complications 32-34.
This study showed that sepsis is more aggressive in pregnant mice. The markedly greater off-target
inflammatory cell extravasationmargination in the lungs suggests that this off-target effects may
mediate the high rates of mortality and that treatments targeted at reducing off target inflammatory
cell migration may be of benefit. Given the role played by NO in this process 25, 26, and the importance
18
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
of NO in blood pressure regulation, a correctly targeted NO antagonist in combination with the usual
supportive therapy, may be effective.
Funding: This work was supported by a grant from the Chelsea and Westminster Health Charity
Borne.
Acknowledgement: None.
Conflict of Interest: none declared.
19
370
371
372
373
374
375
376
References
1. Knight, M., D. Tuffnell, S. Kenyon, J. Shakespeare, R. Gray, and K.J.E.o.b.o. MBRRACE-UK, Saving Lives, Improving Mothers’ Care - Surveillance of maternal deaths in the UK 2011-13 and lessons learned to inform maternity care from the UK and Ireland Confidential Enquiries into Maternal Deaths and Morbidity 2009-13. 2015, National Perinatal Epidemiology Unit: University of Oxford.
2. Bamfo, J.E.A.K., Managing the risks of sepsis in pregnancy. Best Pract. Res. Clin. Obstet. Gynaecol., 2013. 27(4): p. 583-595.
3. Bauer, M.E., B.T. Bateman, S.T. Bauer, A.M. Shanks, and J.M. Mhyre, Maternal Sepsis Mortality and Morbidity During Hospitalization for Delivery: Temporal Trends and Independent Associations for Severe Sepsis. Anesth. Analg., 2013. 117(4): p. 944-950.
4. Cordioli, R.L., E. Cordioli, R. Negrini, and E. Silva, Sepsis and pregnancy: do we know how to treat this situation? Rev Bras Ter Intensiva, 2013. 25(4): p. 334-344.
5. Acosta, C.D., M. Knight, H.C. Lee, J.J. Kurinczuk, J.B. Gould, and A. Lyndon, The Continuum of Maternal Sepsis Severity: Incidence and Risk Factors in a Population-Based Cohort Study. PLoS One, 2013. 8(7): p. e67175.
6. Holmes, A.H., L.S.P. Moore, A. Sundsfjord, M. Steinbakk, S. Regmi, A. Karkey, P.J. Guerin, and L.J.V. Piddock, Understanding the mechanisms and drivers of antimicrobial resistance. The Lancet. 387(10014): p. 176-187.
7. Medawar, P., Some immunological and endocrinological problems raised by the evolution of viviparity in vertebrates. Symp. Soc. Exp. Biol., 1953. 7: p. 320-338.
8. Kourtis, A.P., J.S. Read, and D.J. Jamieson, Pregnancy and Infection. N. Engl. J. Med., 2014. 370(23): p. 2211-2218.
9. Pazos, M., R. Sperling, T. Moran, and T. Kraus, The influence of pregnancy on systemic immunity. Immunol. Res., 2012. 54(1-3): p. 254-261.
10. Chapman, A.B., W.T. Abraham, S. Zamudio, C. Coffin, A. Merouani, D. Young, A. Johnson, F. Osorio, C. Goldberg, L.G. Moore, T. Dahms, and R.W. Schrier, Temporal relationships between hormonal and hemodynamic changes in early human pregnancy. Kidney Int., 1998. 54(6): p. 2056-2063.
11. Davis, J.S., C.J. Darcy, T.W. Yeo, C. Jones, Y.R. McNeil, D.P. Stephens, D.S. Celermajer, and N.M. Anstey, Asymmetric Dimethylarginine, Endothelial Nitric Oxide Bioavailability and Mortality in Sepsis. PLoS One, 2011. 6(2): p. e17260.
12. Abraham, E. and M. Singer, Mechanisms of sepsis-induced organ dysfunction. Crit. Care Med., 2007. 35(10): p. 2408-2416.
13. Daniel, G.R., Cytokine Therapeutics for the Treatment of Sepsis: Why has Nothing Worked? Curr. Pharm. Des., 2003. 9(1): p. 75-82.
14. Vizi, E., J. Szelenyi, Z. Selmeczy, Z. Papp, Z. Nemeth, and G. Hasko, Enhanced tumor necrosis factor-alpha-specific and decreased interleukin-10-specific immune responses to LPS during the third trimester of pregnancy in mice. J. Endocrinol., 2001. 171(2): p. 355-361.
20
377
378379380381382
383384
385386387
388389
390391392
393394395
396397
398399
400401
402403404405
406407408
409410
411412
413414415
15. Zöllner, J., L.G. Howe, L.F. Edey, K.P. O’Dea, M. Takata, F. Gordon, J. Leiper, and M.R. Johnson, The response of the innate immune and cardiovascular systems to LPS in pregnant and nonpregnant mice†. Biol. Reprod., 2017. 97(2): p. 258-272.
16. Edey, L.F., K.P. O'Dea, B.R. Herbert, R. Hua, S.N. Waddington, D.A. MacIntyre, P.R. Bennett, M. Takata, and M.R. Johnson, The Local and Systemic Immune Response to Intrauterine LPS in the Prepartum Mouse1. Biol. Reprod., 2016. 95(6): p. 125, 1-10-125, 1-10.
17. Buras, J., B. Holzmann, and M. Sitkovsky, Animal Models of Spesis: Setting the Stage. Nat Rev Drug Discov, 2005. 4: p. 854-865.
18. Babayigit, H., C. Kucuk, E. Sozuer, C. Yazici, K. Kose, and H. Akgun, Protective effect of β-glucan on lung injury after cecal ligation and puncture in rats. Intensive Care Med., 2005. 31(6): p. 865-870.
19. Grailer, J.J., M. Kalbitz, F.S. Zetoune, and P.A. Ward, Persistent neutrophil dysfunction and suppression of acute lung injury in mice following cecal ligation and puncture sepsis. J. Innate Immun., 2014. 6(5): p. 695-705.
20. Reutershan, J., A. Basit, E.V. Galkina, and K. Ley, Sequential recruitment of neutrophils into lung and bronchoalveolar lavage fluid in LPS-induced acute lung injury. Am. J. Physiol. Lung Cell Mol. Physiol., 2005. 289(5): p. L807-L815.
21. Asaduzzaman, M., S. Zhang, S. Lavasani, Y. Wang, and H. Thorlacius, LFA-1 and MAC-1 mediate pulmonary recruitment of neutrophils and tissue damage in abdominal sepsis. Shock, 2008. 30(3): p. 254-259.
22. Brun-Buisson, C., F. Doyon, J. Carlet, and et al., Incidence, risk factors, and outcome of severe sepsis and septic shock in adults: A multicenter prospective study in intensive care units. JAMA, 1995. 274(12): p. 968-974.
23. Lauer, T., M. Preik, T. Rassaf, B.E. Strauer, A. Deussen, M. Feelisch, and M. Kelm, Plasma nitrite rather than nitrate reflects regional endothelial nitric oxide synthase activity but lacks intrinsic vasodilator action. Proc. Natl. Acad. Sci. U. S. A., 2001. 98(22): p. 12814-12819.
24. Hotchkiss, R.S., G. Monneret, and D. Payen, Sepsis-induced immunosuppression: from cellular dysfunctions to immunotherapy. Nat. Rev. Immunol., 2013. 13(12): p. 862-874.
25. Alves-Filho, J.C., A. de Freitas, F. Spiller, F.O. Souto, and F.Q. Cunha, The role of neutrophils in severe sepsis. Shock, 2008. 30(7): p. 3-9.
26. Benjamim, C.F., S.H. Ferreira, and F.d.Q. Cunha, Role of Nitric Oxide in the Failure of Neutrophil Migration in Sepsis. J. Infect. Dis., 2000. 182(1): p. 214-223.
27. Mercer-Jones, M.A., M. Heinzelmann, J.C. Peyton, D.J. Wickel, M. Cook, and W.G. Cheadle, The Pulmonary Inflammatory Response to Experimental Fecal Peritonitis: Relative Roles of Tumor Necrosis Factor-α and Endotoxin. Inflammation, 1997. 21(4): p. 401-417.
28. Czermak, B.J., V. Sarma, C.L. Pierson, R.L. Warner, M. Huber-Lang, N.M. Bless, H. Schmal, H.P. Friedl, and P.A. Ward, Protective effects of C5a blockade in sepsis. Nat. Med., 1999. 5(7): p. 788-792.
21
416417418
419420421
422423
424425426
427428429
430431432
433434435
436437438
439440441
442443
444445
446447
448449450
451452453
29. Souto, F.O., J.C. Alves-Filho, W.M. Turato, M. Auxiliadora-Martins, A. Basile-Filho, and F.Q. Cunha, Essential Role of CCR2 in Neutrophil Tissue Infiltration and Multiple Organ Dysfunction in Sepsis. Am. J. Respir. Crit. Care Med., 2011. 183(2): p. 234-242.
30. Yamamoto, Y., H. Saito, T. Setogawa, and H. Tomioka, Sex differences in host resistance to Mycobacterium marinum infection in mice. Infect. Immun., 1991. 59(11): p. 4089-4096.
31. Garcia-Gomez, E., B. Gonzalez-Pedrajo, and I. Camacho-Arroyo, Role of Sex Steroid Hormones in Bacterial-Host Interactions. BioMed Research International, 2013. 2013: p. 10.
32. Duebendorfer, C., A. Billeter, B. Seifert, M. Keel, and M. Turina, Serial lactate and admission SOFA scores in trauma: an analysis of predictive value in 724 patients with and without traumatic brain injury. Eur. J. Trauma Emerg. Surg., 2013. 39(1): p. 25-34.
33. Kellum, J.A., Fluid resuscitation and hyperchloremic acidosis in experimental sepsis: Improved short-term survival and acid-base balance with Hextend compared with saline. Crit. Care Med., 2002. 30(2): p. 300-305.
34. Scheiermann, P., D. Ahluwalia, S. Hoegl, A. Dolfen, M. Revermann, B. Zwissler, H. Muhl, K.A. Boost, and C. Hofstetter, Effects of intravenous and inhaled levosimendan in severe rodent sepsis. Intensive Care Med., 2009. 35(8): p. 1412-1419.
22
454455456
457458
459460
461462463
464465466
467468469
470
471
Figure Legends
Figure 1: Survival in non-pregnant (NP) and Pregnant CD1 mice after CLP or sham operation (CLP, A)
and following administration of i.p. 10 or 40mg/kg of LPS (LPS, B). Data are expressed as mean ± SEM.
A log-rank (Mantel-Cox) test was conducted to analyse survival data. *p<0.05, **p<0.01, ***p<0.001
(Sham Non-pregnant n=6, Sham Pregnant n=9, CLP non-pregnant n=10, CLP Pregnant n=12; NP
10mg/kg n=6, Pregnant 10mg/kg n=6, NP 40mg/kg n=9, Pregnant 40mg/kg n=6).
Figure 2: The effect of CLP on pregnant and non-pregnant mice on delta mean arterial pressure
(∆MAP, A), heart rate (∆HR, B), max increase MAP (∆MAP, C), max decrease MAP (∆MAP, CD), max
increase HR (∆HR, E) and, max decrease HR (∆HR, DF), and organ function following CLP operation (E)
base excess (BEecf) (F) bicarbonate (HCO3-). Serum samples were taken 8 hours following CLP in non-
pregnant and pregnant mice. Data are expressed as mean ± SEM. Recordings were analysed as hourly
averages. One-way ANOVA analysis with Bonferroni’s post-test, and two group analysis using an
unpaired student’s t-test or Mann Whitney U tests depending on the data distribution. ***p<0.001
(Sham NP n=6, CLP NP survivor n=6, CLP NP non-survivor n=6, Sham Pregnant n=6, CLP Pregnant
n=5).
Figure 3: Circulating cell count at 3 hours (A) Ly6Chigh monocytes; (B) Neutrophils and 6 hours (C)
Ly6Chigh monocytes; (D) Neutrophils ; lung leukocyte cell count at 3 hours (E) Ly6Chigh monocytes; (F)
Neutrophils; and 6 hours (G) Ly6Chigh monocytes; (H) Neutrophils; peritoneal leukocyte cell count
densities at 6 hours (I) Ly6Clow monocytes; (J) Ly6Chigh monocytes; (K) natural killer cells; (L) neutrophil
cells following CLP in pregnant (Preg) and non-pregnant (NP) mice. Data are expressed as mean ±
SEM. Two group analysis using an unpaired student’s t-test or Mann Whitney U tests depending on
the data distribution. *p<0.05, (NP Naïve n=6, Preg Naïve n=7, NP Sham n= 7, Preg Sham n=7, NP CLP
n=13, Preg CLP n=8).
Figure 4: Circulating concentrations of pro-inflammatory cytokines (IL-1b, IL-6, TNFa), anti-
inflammatory (IL-4, IL-10) and Chemokines, (CCL2, CCL5, CXCL1, CXCL2) in serum of non-pregnant
23
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
(NP) and pregnant (E16) mice following CLP or sham operation. (A) IL-1b (B) IL-6 (C) TNFa (D) IL-4 (E)
IL-10 (F) CCL2 (G) CCL5 (H) CXCL1 (I) CXCL2. Serum was collected 3 and 6 hours after CLP or sham
operation and in naïve mice. Data shown as mean ± SEM. Two group analysis using an unpaired
student’s t-test or Mann Whitney U tests depending on the data distribution. **p<0.01, (Naïve NP
n=6, Naïve E16 n=6, Sham NP 3HR n=6, Sham E16 6HR n=8, Sham NP 6HR n=7, Sham E16 6HR n=6,
CLP NP 3HR n=9, CLP E16 3HR n=6, CLP NP 6HR n=10, CLP E16 6HR n=6).
Figure 5: Lung concentrations of pro-inflammatory cytokines (IL-1b, IL-6, TNFa), anti-inflammatory
(IL-4, IL-10) and Chemokines, (CCL2, CCL5, CXCL1, CXCL2) in serum of non-pregnant (NP) and
pregnant (E16) mice following CLP or sham operation. (A) IL-1b (B) IL-6 (C) TNFa (D) IL-4 (E) IL-10 (F)
CCL2 (G) CCL5 (H) CXCL1 (I) CXCL2. Lung tissue was collected 6 hours after CLP or sham operation and
in naïve mice. Data shown as mean ± SEM. Two group analysis using an unpaired student’s t-test or
Mann Whitney U tests depending on the data distribution. ***p<0.001, (Naïve NP n=6, Naïve E16
n=6, Sham NP 6HR n=6, Sham E16 6HR n=6, CLP NP 6HR n=6, CLP E16 6HR n=6).
Figure 6: Bacterial count in blood and peritoneal fluid of non-pregnant (NP) and pregnant (E16) mice
following CLP operation. (A) Peritoneal fluid (B) Blood. Bacterial count was assessed in samples taken
8 hours following CLP in non-pregnant and pregnant mice. Data are expressed as median ±
interquartile range, Mann Whitney U test was used to compare NP and E16 samples (NP n=8, E16
n=8). CFU: colony forming unit.
24
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514