Download Articles in PresS. Am J Physiol Heart Circ Physiol (September 25

Articles in PresS. Am J Physiol Heart Circ Physiol (September 25, 2015). doi:10.1152/ajpheart.00397.2015
1
2
3
4
5
Afterload-induced diastolic dysfunction contributes to high filing pressures in experimental heart
failure with preserved ejection fraction
6
Sara Leite1,2*, Sara Rodrigues1,2*, Marta Tavares-Silva1,2,3, José Oliveira-Pinto1,2,4, Mohamed Alaa1,2,5,
Mahmoud Abdellatif1,2, Dulce Fontoura1,2, Inês Falcão-Pires1,2, Thierry C. Gillebert6, Adelino F. LeiteMoreira1,2,7, André P. Lourenço1,2,8
7
8
9
10
1
11
12
13
14
15
16
17
Department of Physiology and Cardiothoracic Surgery, Faculty of Medicine, University of Porto,
Porto, Portugal; 2Cardiovascular Research Centre, Faculty of Medicine, University of Porto, Porto,
Portugal; 3Department of Cardiology, Hospital São João, Porto, Portugal; 4Department of Vascular
Surgery, Hospital São João, Porto, Portugal; 5 Department of Cardiothoracic Surgery, Suez Canal
University, Egypt; 6Department of Cardiology, Ghent University, Ghent, Belgium; 7Department of
Cardiothoracic Surgery, Hospital São João, Porto, Portugal; 8Department of Anesthesiology, Hospital
São João, Porto, Portugal
18
*these authors equally contributed to the manuscript.
19
20
21
Running Head: Afterload induced diastolic dysfunction in HFpEF
22
23
24
Corresponding author:
25
André P. Lourenço
26
Department of Physiology and Cardiothoracic Surgery, Faculty of Medicine, University of Porto
27
Medical Research Centre
28
Alameda Professor Hernâni Monteiro
29
4200-319 Porto, Portugal
30
Tel.: +351 225 513 644; fax: +351 225 513 646
31
E-mail address: [email protected]
1
Copyright © 2015 by the American Physiological Society.
32
Abstract
33
Myocardial stiffness and upward-shifted end-diastolic (ED) pressure-volume (P-V) relationship
34
(EDPVR) are the key to high filling pressures in heart failure with preserved ejection fraction (HFpEF).
35
Nevertheless, many patients may remain asymptomatic unless hemodynamic stress is imposed on
36
the myocardium. Whether delayed relaxation induced by pressure challenge may contribute to high
37
end-diastolic pressure (EDP) remains unsettled. Our aim was to assess the effect of suddenly
38
imposed isovolumic afterload on relaxation and EDP exploiting a highly controlled P-V experimental
39
evaluation setup in the ZSF1 obese rat model of HFpEF. Twenty-week old ZSF1 obese (ZSF1 Ob,
40
n=12), healthy Wistar-Kyoto (WKY, n=11) and hypertensive ZSF1 lean control rats (ZSF1 Ln, n=10)
41
underwent open-thorax left ventricular (LV) P-V hemodynamic evaluation under anesthesia with
42
sevoflurane. EDPVR was obtained by inferior vena cava occlusions to assess LV ED chamber stiffness
43
constant β and single-beat isovolumic afterload acquisitions were obtained by swift occlusions of the
44
ascending aorta. ZSF1 Ob showed increased ED stiffness, delayed relaxation, as assessed by time
45
constant of isovolumic relaxation (τ), and elevated EDP with normal ejection fraction. Isovolumic
46
afterload increased EDP without concomitant changes in ED volume or heart rate. In isovolumic
47
beats relaxation was delayed to the extent that time for complete relaxation as predicted by 3.5*τexp
48
exceeded effective filling time. EDP elevation correlated with reduced time available to relax which
49
was the only independent predictor of EDP rise in multiple linear regression. Our results suggest
50
delayed relaxation during pressure challenge to be an important contributor to lung congestion and
51
effort intolerance in HFpEF.
52
Keywords: afterload; relaxation; diastolic function; heart failure with preserved ejection fraction
53
2
54
New and Noteworthy
55
In a highly controlled hemodynamic evaluation set up in experimental heart failure with preserved
56
ejection fraction we demonstrate that delayed relaxation independently explains end-diastolic
57
pressure elevation during suddenly imposed afterload. This is an important contribution to
58
understand the response to exercise or hypertensive stress in preserved ejection fraction heart
59
failure.
60
3
61
Glossary
62
BSA, body surface area
63
CO, cardiac output
64
dP/dtmax, maximum rate of pressure rise
65
dP/dtmin, maximum rate of pressure fall
66
67
E/E’, ratio of peak early transmitral pulsed-wave Doppler flow velocity (E) to peak early diastolic
tissue Doppler velocity (E’)
68
EDP, end-diastolic pressure
69
EDPVR, end-diastolic pressure volume relationship
70
EDV and EDVi, end-diastolic volume and indexed end-diastolic volume
71
EES, end-systolic elastance
72
EF, ejection fraction
73
HFpEF, heart failure with preserved ejection fraction
74
HR, heart rate
75
LV, left ventricle or left ventricular
76
LVMi, left ventricular mass index
77
NO, nitric oxide
78
Pmax, maximum developed pressure
79
PRSW, preload recruitable stroke work
80
S’, peak systolic tissue Doppler motion velocity
81
TAR, time available to relax
82
τ, time constant of isovolumic relaxation
83
4
84
Introduction
85
Heart failure with preserved ejection fraction (HFpEF) remains a major unsolved health issue and a
86
complex disease in which the contribution of various aspects is still incompletely understood (22,
87
24). Although HFpEF pathophysiology is multifarious and HFpEF patients constitute a heterogeneous
88
group comprising several risk factors (26) most experts would agree upon abnormalities of left
89
ventricular (LV) relaxation and high passive stiffness as underlying hemodynamic mechanisms (31) as
90
well as upon comorbidity and systemic inflammation-induced microvascular coronary endothelium
91
disturbances, decreased cardiomyocyte nitric oxide (NO) availability and protein kinase G activity as
92
key
93
hypophosphorylation, and interstitial fibrosis that ultimately lead to LV stiffening (23). On top of
94
these essential aspects, vascular stiffening and the effects of vascular reflection on myocardial load
95
and function are often underscored. Both magnitude and timing of load play a role in diastolic
96
function (5, 18). We have recently characterized a new cardiometabolic risk model of HFpEF that
97
recapitulates this paradigm. Despite preserved systolic function, hypertensive, obese and diabetic
98
ZSF1 obese rats show high end-diastolic pressure (EDP) and increased cardiomyocyte stiffness which
99
can be ascribed mostly to titin hypophosphorylation (9). Moreover, these animals show impaired
100
max and decreased anaerobic threshold in effort testing and low diastolic function tolerance to
101
both afterload and preload closely mimicking clinical HFpEF (20). Indeed, many patients who have
102
high passive stiffness may not show symptoms at rest. Underlying disturbances are only revealed
103
during exercise or hemodynamic stress suggesting that further diastolic function abnormalities come
104
into play (2). Our group has pioneered the investigation of the effects of sudden single-beat
105
afterload elevations in diastolic function. This in situ experimental preparation allows a full and
106
unique understanding of the interplay between afterload, relaxation and filling pressures without
107
the confounding of neurohumoral and reflex responses, overall hemodynamic status, ventricular
108
interdependence and pericardial constraint (16-19). In dogs and rabbits we have demonstrated a
109
biphasic response between afterload and relaxation. Small afterload elevations accelerate relaxation
determinants
to
molecular
remodeling
(namely
myocardial
hypertrophy)
titin
5
110
whereas large elevations delay relaxation (18) to the point where it becomes incomplete at end-
111
diastole enabling the build-up of residual EDP (17) defying the rough dichotomy between
112
independent active relaxation and passive elastance determinants of diastolic function and the view
113
that no matter how delayed relaxation is it may never impart on EDP (8). We have proposed that
114
delayed and incomplete relaxation might contribute to residual active force development at end-
115
diastole and thus constitute a pathophysiological mechanism in acute hypertensive lung edema (7)
116
which we labelled afterload-induced diastolic dysfunction (16, 17). Indeed, we have documented
117
that the diseased myocardium tolerates afterload elevations poorly both in patients with impaired
118
ejection fraction (19) and in a rat disease model (6). In healthy rats however likely owing to
119
particularities in myofilament composition and Ca2+ kinetics (21) only isovolumic afterload
120
significantly impaired relaxation with faint consequences in EDP (6). Most importantly, although it
121
may provide important mechanistic views on myocardial response to sudden afterload this
122
experimental set-up was never applied to the stiff HFpEF myocardium.
123
The aim of this study was to explore diastolic function response to sudden single-beat isovolumic
124
afterload elevation in ZSF1 obese rats and to ascertain the potential role of afterload-induced
125
diastolic dysfunction in experimental HFpEF.
6
126
Methods
127
Animal model
128
ZSF1 obese (ZSF1 Ob, n=12), ZSF1 lean (ZSF1 Ln, n=10) and Wistar-Kyoto rats (WKY, n=11; Charles
129
River Laboratories, Barcelona, Spain) were fed with standard diet (Purina diet 5008) to achieve
130
metabolic syndrome and HFpEF, systemic arterial hypertension or absence of any cardiovascular risk
131
factor, respectively, as previously reported (9). Animals were kept in individually ventilated
132
chambers, in groups of 2-per cage under controlled environment with a 12-h-light-dark cycle at 22°C
133
room temperature. All animals received humane care. Experimental procedures were approved by
134
the ethical committee of the Faculty of Medicine of Porto and were performed in accordance with
135
Portuguese law on animal welfare, EU Directive 2010/63/EU for animal experiments and the
136
National Institutes of Health Guide for the Care and Use of Laboratory Animals (NIH publication no.
137
85-23, revised 2011). Although we have formerly reported echocardiography data for this
138
experimental model (9), we performed confirmation echocardiography in 6 animals per group, as
139
previously described (9).
140
Hemodynamic studies
141
Briefly, animals were anaesthetized with sevoflurane (8 and 2.5–3% sevoflurane for induction and
142
maintenance, respectively; Penlon Sigma Delta) after sedation and analgesia with 100 µg.kg-1 and 5
143
mg.kg-1 intraperitoneal midazolam and fentanyl, respectively. Mechanical ventilation with positive
144
end-expiratory pressure held at 5 cmH2O, tidal volume set at 6 mL.Kg-1 and respiratory rate adjusted
145
to achieve normocapnia (MouseVent™ - Automatic Ventilator, Physiosuite, Kent Scientific) was
146
instituted upon endotracheal intubation (14G). Electrocardiogram (Animal Bio Amp, FE136,
147
ADInstruments), peripheral oximetry (MouseSTAT™ - Pulse Oximeter & Heart Rate Monitor,
148
Physiosuite, Kent Scientific), capnography, minute ventilation (CapnoScan™ - End-Tidal CO2 Monitor,
149
Physiosuite, Kent Scientific) and body temperature were recorded throughout. Temperature was
150
kept at 38oC on a heating pad (RightTemp™ - Temperature Monitor & Homeothermic Controller,
151
Physiosuite, Kent Scientific). Fluid replacement with heated Ringer’s lactate at 32 mL.kg-1.h-1 (NE-
7
152
1000, New Era Pump Systems) was instituted through a peripheral dorsal foot vein catheter (24G).
153
The heart was exposed by a left thoracotomy and the pericardium was widely opened. A flowmeter
154
probe was transiently placed in the ascending aorta for cardiac output (CO) calibration (2.5PS,
155
Transonic), a pressure-volume catheter (SPR-847 Millar Instruments) was inserted through the apex
156
and positioned along the LV long axis and 3-0 silk threads were passed around the inferior vena cava
157
and through a plastic tube to enable transient occlusions. Parallel conductance was determined by
158
40 μL 10% hypertonic saline injection and slope factor α was derived by simultaneous measurement
159
of CO by a flowmeter (TS420, Transonic). After a stabilization period of 30 mins, recordings were
160
obtained at suspended end-expiration. Transient 5-7 cycle occlusions of the inferior vena cava
161
occlusions were performed to obtain load independent indexes of LV contractility and chamber
162
stiffness by fitting the end-systolic and end-diastolic pressure-volume relationships (EDPVR) to a
163
linear and an exponential function, respectively, as previously described (9). After careful dissection,
164
single beat occlusions of the ascending aorta were performed with a non-traumatic vascular clamp
165
to assess diastolic response to isovolumic afterload. Clamping was performed during diastole and
166
maintained throughout systole, as reported (16). Three separate acquisitions were obtained and
167
averaged in every animal. Data with arrhythmia, heart rate (HR) changes higher than 2% or evidence
168
of incomplete afterload elevations were excluded. When several heartbeats were spanned in the
169
occlusion only the first beat after clamping was analyzed. Resting periods were allowed between
170
each intervention. Signals were continuously acquired (MPVS 300, Millar Instruments), digitized at
171
1000 Hz (ML880 PowerLab 16/30, ADinstruments), and analyzed offline (PVAN 3.5, Millar
172
Instruments). Time constant of isovolumic relaxation (τ) was computed by the logistic (τ log) and the
173
monoexponential (τ exp) methods. Prediction of time to complete relaxation was derived as 3.5* τ exp
174
based on Weisfeldt, et al.(29) and effective diastolic filling time was established as the time interval
175
from the maximum rate of pressure fall (dP/dtmin) to end-diastole (peak R wave of the ECG). To
176
account for large differences in body weight between groups, volumes were indexed for body
8
177
surface area (BSA) as estimated by 9.1*body weight in grams¾. Upon completion of experiments,
178
animals were euthanized by exsanguination under anesthesia.
179
Statistical analysis
180
Groups were compared regarding baseline hemodynamic parameters with one-way ANOVA or
181
Kruskal-Wallis according to normal or non-normal distribution, respectively, as verified by Shapiro-
182
Wilk’s test. Homogeneity of variances in ANOVA was checked by Levene’s test and corrected by
183
Welch’s F-ratio with adjusted degrees of freedom whenever violations occurred. Post hoc
184
comparisons were performed with Kramer’s modification for unequal group sizes of Tukey’s
185
method. The effects in isovolumic beats were assessed with mixed general linear models with group
186
and change from baseline to isovolumic as between-effect and within effect predictors, respectively.
187
Homogeneity of variances was verified with Bartlett’s test. Multivariate general linear models with
188
overall model estimation by Pillai’s trace followed by univariate test were used to compare groups
189
regarding joint changes in end-diastolic pressure and volume and to assess the influence of τ and
190
time available to relax on EDP. Deviations from sphericity were corrected with Huynh-Feldt’s
191
method. The influence of time available to relax and LV chamber stiffness constant β on EDP rise
192
during isovolumic beats was assessed by best subsets multiple linear regression with group recoded
193
as dummy variables. Standardized residual were carefully checked for normal distribution. Statistical
194
significance level was set at two-tailed 0.05. Data are presented as mean±standard error of mean.
195
9
196
Results
197
Baseline hemodynamic and single beat isovolumic parameters are summarized in Table 1 and
198
illustrated in Figures 1 and 2. Compared with WKY, ZSF 1 Ln showed higher maximum pressure
199
reached during systole (Pmax) which was further increased in ZSF 1 Ob. Chronic afterload was
200
accompanied by higher LV mass indexed for BSA (LVMi) which was significantly increased only in
201
ZSF1 Ob compared with WKY (Table 2). LV hypertrophy was confirmed on morphometric evaluation,
202
the ratio between LV plus interventricular septum mass to tibial length was increased (P=0.02) in
203
ZSF1 Ob (29.8 ± 2.1 mg.mm-1) compared with WKY (23.0 ± 0.7 mg.mm-1) but not with ZSF1 Ln (24.9 ±
204
1.2 mg.mm-1). ZSF1 Ob also showed baseline elevation in τexp, delayed onset of diastole as assessed
205
by dP/dtmin, higher EDP and an upward shift in EDPVR (Figure 2) as assessed by the LV chamber
206
stiffness constant for indexed volumes βi (Table 1). Cardiac index, ejection fraction (EF) and indexed
207
end-diastolic volume (EDVi) did not differ between groups (Table 1). Baseline maximum velocity of
208
pressure rise (dP/dtmax) was increased in ZSF1 Ob compared with WKY whereas end-systolic
209
elastance (EES) was increased in both ZSF1 groups compared with WKY, but the chamber size-
210
independent preload recruitable stroke work (PRSW) was unchanged (Table 1). Preserved EF and
211
unaltered peak systolic tissue Doppler motion velocity (S’) were confirmed on echocardiography,
212
along with an increase in the ratio of peak early transmitral pulsed-wave Doppler flow velocity to
213
peak early diastolic tissue Doppler velocity (E/E’), a surrogate of higher filling pressures, in ZSF1 Ob
214
compared with both ZSF1 Ln and WKY (Table 2). Changes in monoexponentially-derived τ were
215
confirmed by the logistic method. Because τexp is commonly used to predict the length of relaxation
216
and because τlog behaved like τexp we present results simply for τexp. No group differences were
217
observed in the maximum systolic pressure developed during isovolumic single beats Pmax (ISO) or
218
active developed pressures. Nevertheless, Pmax rose by 88 ± 11 % in WKY in the isovolumic beat
219
which was significantly higher (P<0.005) compared with both ZSF1 Ln and Ob (50 ± 4 and 33 ± 5 %,
220
respectively), denoting baseline work at a lower pressure level and better contractile reserve when
221
subjected to pressure challenge (represented by arrow in panel A of Figure 1). Indeed, if EES is
10
222
calculated by a single beat method that linearly fits the end-systolic pressure-volume of the baseline
223
beat and Pmax (ISO) at the corresponding volume (dotted lines in the lower panels of figure 1) ZSF1 Ob
224
has a slope of 0.16 ± 0.02 mmHg.μL-1.cm-2 which is significantly lower (P=0.002) than WKY (0.33 ±
225
0.04 mmHg.μL-1.cm-2) while ZSF1 Ln present intermediate values (0.25 ± 0.03 mmHg.μL-1.cm-2)
226
suggesting impaired contractility in the high pressure range in ZSF1 Ob. Regarding filling pressures,
227
the rise in EDP at the completion of the isovolumic beat however was higher in ZSF1 Ob compared
228
with both ZSF1 Ln and WKY (Table 1 and figure 1) independently of EDVi as confirmed by
229
multivariate analysis. Indeed, EDVi was virtually unchanged (Figure 3). To evaluate the determinants
230
of EDP elevation in ZSF1 Ob we conducted multivariate analysis of τexp and time available to relax
231
(TAR). Since prolongation of relaxation (τexp) can only impart on EDP in as much as its completeness
232
is delayed beyond the point of end-diastole, the variable TAR was defined as the difference between
233
effective diastolic filling time and time predicted to complete relaxation. Though time to dP/dtmin
234
was similarly increased by the isovolumic intervention in all groups (Table 1, Figure 4), τexp and TAR
235
were significantly prolonged and curtailed, respectively, in ZSF1 Ob compared with ZSF1 Ln and WKY
236
(Figure 4). Indeed, 7 out of 12 ZSF1 Ob animals presented negative TAR values, denoting incomplete
237
predicted relaxation at end-diastole whereas all of ZSF1 Ln and WKY animals showed positive values.
238
To further clarify the determinants of EDP rise in isovolumic beats we explored correlations with the
239
independent variables TAR and βi. Because individual subjects had a single observation for stiffness
240
constant but two for TAR the influence of TAR was assessed by expressing it as percentage change
241
from baseline to isovolumic beat (Figure 5). Although both correlations were significant, a stronger
242
correlation coefficient was observed for percent drop in TAR. Indeed, accounting for group as a
243
potential confounder, and after recoding as dummy variables, in multiple linear regression analysis
244
change in TAR and not βi was the only independent predictor of EDP rise (Table 3).
11
245
Discussion
246
We demonstrate that sudden single-beat isovolumic elevations of afterload delay relaxation and
247
shorten the time available to relax leading to residual high EDP independently from end-diastolic
248
stiffness in ZSF1 Ob rats with metabolic syndrome and HFpEF compared with hypertensive ZSF1 Ln
249
and healthy normotensive WKY controls.
250
Based on computer model simulations of the human cardiovascular system Hay et al. have predicted
251
that isolated impairment of active myocardial relaxation could generate elevated filling pressures in
252
the setting of rapid HR and increased systolic intervals (11). Our previous works with this
253
experimental set-up have been mainly conducted on healthy rabbits and dogs. The magnitude of
254
afterload that leads to incomplete relaxation and residual EDP was more obvious in the dog than in
255
the rabbit (17) and unremarkable in the healthy rat (6). Compared with previous works which were
256
carried out using volume estimation by sonomicrometry we used gold standard volume assessment
257
by conductance catheters clearly demonstrating that any changes in EDP cannot be attributed to
258
slight shifts in EDV. With this work we also further extend the role of afterload-induced diastolic
259
dysfunction to the setting of HFpEF. Contrarily to healthy WKY controls and hypertensive ZSF1 Ln
260
controls, ZSF1 Ob showed delayed and prolonged relaxation as assessed by τ and time to dP/dtmin,
261
respectively, that altogether led to insufficient time to relax, incomplete relaxation and high residual
262
EDP. Independently of LV myocardial diastolic stiffness, in this open thorax preparation a single
263
isovolumic beat raised EDP on average from 13 to 20mmHg. Moreover, as aforementioned rats are
264
particularly resistant to afterload induced diastolic dysfunction thus it is foreseeable that the effect
265
may be more relevant in closed-chest unanesthetized large animal and human physiology. This
266
finding addresses an important unsolved issue in HFpEF pathophysiology whether delayed relaxation
267
during exercise or afterload elevation can lead to EDP elevation per se. Westermann et al. compared
268
the response to handgrip exercise in HFpEF patients and controls and reported that despite similar
269
rises in systolic blood pressure up to 179 mmHg EDP rose from 15 to 24 mmHg in HFpEF patients
12
270
while no change were observed from the 6 mmHg baseline in controls (30). Curiously, and though no
271
significant changes were reported for EDV the authors attributed the rise in EDP to the steeper
272
EDPVR in HFpEF. In a similar study on another small subset of HFpEF patients with pressure-volume
273
evaluation during handgrip exercise Kawaguchi et al. on the other hand reported prolonged
274
relaxation and filling pressure elevation and suggested that ventricular-arterial stiffening
275
exacerbates hypertensive stress responses by delaying relaxation, limiting filling and raising diastolic
276
pressures (15). Our experimental results with isovolumic beats clearly indicate that the rise in EDP
277
can neither be attributed to a steeper EDPVR nor to increased end-diastolic volume. Indeed, in order
278
to achieve end-diastolic pressures of 20 mmHg, ZSF1 Ob would require an important increase in end
279
diastolic volume as represented in Figure 2 which was clearly ruled out by the conductance
280
methodology (Figures 1 and 3). Indeed, according to our multivariate analysis relaxation
281
disturbances clearly influenced EDP to a larger extent than chamber stiffness itself. Finally, we must
282
emphasize that our results in isovolumic beats were obtained under constant HR, increasing HR
283
during exercise will predictably enhance the effects of afterload–induced diastolic dysfunction.
284
Nevertheless, we must acknowledge that handgrip exercise is not a good representative of activities
285
of daily living. Borlaug et al. examined changes with supine dynamic exercise in HFpEF patients and
286
found enhanced relaxation. Although this was inadequate to compensate for tachycardia-induced
287
shortening of diastolic filling period, relaxation was still estimated to be complete by 50% of the
288
filling period (1). This can be partly explained by the distinct hemodynamic responses elicited by
289
isotonic aerobic exercise and isometric exercise. Indeed, isometric exercise usually evokes higher
290
blood pressure rise (15). Though isovolumic pressure challenge has no clinical counterpart, the
291
results from our proof-of-concept work underscore the role of sudden pressure challenge, during
292
effort or psychological, environmental, pharmacological and other forms of acute cardiovascular
293
stress, as a mechanism of acute decompensation in a stiff high-gain and poor-reserve ventricular-
294
arterial system as HFpEF. The pathophysiology of effort intolerance in HFpEF is multifarious but
13
295
delayed relaxation elicited by sudden pressure challenge may importantly contribute to elevate EDP
296
and cause an HFpEF patient to become symptomatic, as proposed by Borlaug et al. (4).
297
Beyond load and ventricular stiffening, many cellular and molecular determinants likely contribute
298
to impaired relaxation in HFpEF. Changes in protein kinase G activity (28), changes in myofilament
299
composition namely regarding titin a giant sarcomeric protein that mediates restoring forces and
300
length-dependent deactivation during the early phase of diastole (12) and that we have shown to be
301
hypophosphorylated in this experimental model (9) as well as disturbances in calcium kinetics (13)
302
may justify disturbed cross-bridge detachment and relaxation kinetics. Relaxation is both dependent
303
on cross-bridge detachment and calcium kinetics and impaired protein kinase G signaling which has
304
been reported in HFpEF patients (28) may be involved not only in titin hypophosphorylation but also
305
in increased calcium sensitivity of the myofilaments and slow cross-bridge detachment (14).
306
Recently, in old hypertensive dogs with HFpEF decreased troponin I, Myosin binding protein C and
307
myosin light-chain 2 hypophosphorylation along with increased calcium sensitivity were also
308
described and related to altered protein kinase activity (10) further supporting changes in regulatory
309
myofilament proteins as a likely molecular explanation for delayed relaxation in HFpEF. Moreover,
310
impaired myocardial bioenergetics may also be involved in delayed relaxation, as observed in HFpEF
311
patients (25). All of the aforementioned mechanisms that are already disturbed at baseline
312
explaining delayed relaxation in HFpEF will likely exacerbate load-induced slowing of relaxation in
313
HFpEF although their relative contribution and that of other potential molecular determinants
314
remains to be established.
315
Finally, though contractility indexes such as EES and dP/dtmax suggest hypercontractility in ZSF1 Ob,
316
PRSW which is not confounded by the size and geometry of the LV was unaltered. In fact, isovolumic
317
beats unmasked an impaired response to pressure challenge in ZSF1 Ob. Accordingly, we have
318
previously demonstrated that relaxation and contractility are closely related in sudden pressure
319
challenge (18) and subtle impairment of systolic function has been shown in HFpEF patients
14
320
particularly during effort (27). As proposed by Borlaug, et al. the inability to enhance contractility
321
upon imposed afterload may limit cardiac output reserve during effort and contribute to effort
322
intolerance (3).
323
In conclusion, in a highly controlled experimental setup that assesses the interplay between
324
afterload, relaxation and filling in the intact heart we have demonstrated in the ZSF1 obese rat
325
model of metabolic syndrome and HFpEF that acutely imposed afterload delays relaxation to the
326
extent that it imparts on end-diastole. Based on our experimental observations, incomplete
327
relaxation and build-up of residual EDP may contribute to high filling pressures and lung congestion
328
during hypertensive stress in HFpEF.
329
Grants
330
This work was supported by grants from the Portuguese Foundation for Science and Technology to
331
the Cardiovascular Research Centre of the Faculty of Medicine of Porto (EXCL/BIM-MEC/0055/2012
332
and PEST-C/SAU/UI 0051/2014) and from the European Commission (FP7-Health-2010; MEDIA-
333
261409).
334
Disclosures
335
The authors have no conflict of interest do disclose.
336
337
15
338
References
339
1.
340
relaxation and compliance reserve during dynamic exercise in heart failure with preserved ejection
341
fraction. Heart (British Cardiac Society) 97: 964-969, 2011.
342
2.
343
enhance diagnosis of early heart failure with preserved ejection fraction. Circulation Heart failure 3:
344
588-595, 2010.
345
3.
346
cardiovascular reserve dysfunction in heart failure with preserved ejection fraction. J Am Coll Cardiol
347
56: 845-854, 2010.
348
4.
349
diagnosis, and treatment. Eur Heart J 32: 670-679, 2011.
350
5.
351
St John Sutton M, and Gillebert TC. Early and late systolic wall stress differentially relate to
352
myocardial contraction and relaxation in middle-aged adults: the Asklepios study. Hypertension 61:
353
296-303, 2013.
354
6.
355
G, Vasques-Novoa F, Gillebert TC, and Leite-Moreira AF. Time course and mechanisms of left
356
ventricular systolic and diastolic dysfunction in monocrotaline-induced pulmonary hypertension.
357
Basic Res Cardiol 104: 535-545. Epub 2009 Mar 2014., 2009.
358
7.
359
pathogenesis of acute pulmonary edema associated with hypertension. N Engl J Med 344: 17-22,
360
2001.
361
8.
362
pressure-volume relation. Circ Res 64: 827-852, 1989.
Borlaug BA, Jaber WA, Ommen SR, Lam CS, Redfield MM, and Nishimura RA. Diastolic
Borlaug BA, Nishimura RA, Sorajja P, Lam CS, and Redfield MM. Exercise hemodynamics
Borlaug BA, Olson TP, Lam CS, Flood KS, Lerman A, Johnson BD, and Redfield MM. Global
Borlaug BA, and Paulus WJ. Heart failure with preserved ejection fraction: pathophysiology,
Chirinos JA, Segers P, Rietzschel ER, De Buyzere ML, Raja MW, Claessens T, De Bacquer D,
Correia-Pinto J, Henriques-Coelho T, Roncon-Albuquerque R, Jr., Lourenco AP, Melo-Rocha
Gandhi SK, Powers JC, Nomeir AM, Fowle K, Kitzman DW, Rankin KM, and Little WC. The
Gilbert JC, and Glantz SA. Determinants of left ventricular filling and of the diastolic
16
363
9.
Hamdani N, Franssen C, Lourenco A, Falcao-Pires I, Fontoura D, Leite S, Plettig L, Lopez B,
364
Ottenheijm CA, Becher PM, Gonzalez A, Tschope C, Diez J, Linke WA, Leite-Moreira AF, and Paulus
365
WJ. Myocardial titin hypophosphorylation importantly contributes to heart failure with preserved
366
ejection fraction in a rat metabolic risk model. Circulation Heart failure 6: 1239-1249, 2013.
367
10.
368
Myofilament Phosphorylation and Function in Experimental Heart Failure with Preserved Ejection
369
Fraction. Cardiovasc Res 97: 464-471, 2013.
370
11.
371
in the production of elevated left ventricular filling pressure. American journal of physiology Heart
372
and circulatory physiology 288: H1203-1208, 2005.
373
12.
374
Starling relation in early diastole. The Journal of general physiology 121: 97-110, 2003.
375
13.
376
Platzer D, Sokolow S, Herchuelz A, Antoons G, Sipido K, Pieske B, and Heinzel FR. Intracellular
377
dyssynchrony of diastolic cytosolic [Ca(2)(+)] decay in ventricular cardiomyocytes in cardiac
378
remodeling and human heart failure. Circulation research 113: 527-538, 2013.
379
14.
380
Kim KH, Kim YJ, Kim SJ, and Zhang YH. Myofilament Ca2+ desensitization mediates positive
381
lusitropic effect of neuronal nitric oxide synthase in left ventricular myocytes from murine
382
hypertensive heart. Journal of molecular and cellular cardiology 60: 107-115, 2013.
383
15.
384
stiffening in patients with heart failure and preserved ejection fraction: implications for systolic and
385
diastolic reserve limitations. Circulation 107: 714-720, 2003.
386
16.
387
pressure-volume relation. American journal of physiology Heart and circulatory physiology 280: H51-
388
59, 2001.
Hamdani N, Bishu KG, von Frieling-Salewsky M, Redfield MM, and Linke WA. Deranged
Hay I, Rich J, Ferber P, Burkhoff D, and Maurer MS. Role of impaired myocardial relaxation
Helmes M, Lim CC, Liao R, Bharti A, Cui L, and Sawyer DB. Titin determines the Frank-
Hohendanner F, Ljubojevic S, MacQuaide N, Sacherer M, Sedej S, Biesmans L, Wakula P,
Jin CZ, Jang JH, Kim HJ, Wang Y, Hwang IC, Sadayappan S, Park BM, Kim SH, Jin ZH, Seo EY,
Kawaguchi M, Hay I, Fetics B, and Kass DA. Combined ventricular systolic and arterial
Leite-Moreira AF, and Correia-Pinto J. Load as an acute determinant of end-diastolic
17
389
17.
Leite-Moreira AF, Correia-Pinto J, and Gillebert TC. Afterload induced changes in myocardial
390
relaxation: a mechanism for diastolic dysfunction. Cardiovascular research 43: 344-353, 1999.
391
18.
392
its regulation by load and contractile state. Circulation 90: 2481-2491, 1994.
393
19.
394
MJ, Almeida J, Pinho P, and Gillebert TC. Diastolic tolerance to systolic pressures closely reflects
395
systolic performance in patients with coronary heart disease. Basic research in cardiology 107: 251,
396
2012.
397
20.
398
Lourenco AP. Echocardiography and invasive hemodynamics during stress testing for diagnosis of
399
heart failure with preserved ejection fraction: an experimental study. American journal of physiology
400
Heart and circulatory physiology ajpheart 00076 02015, 2015.
401
21.
402
research: advantages and disadvantages. Pharmacology & therapeutics 141: 235-249, 2014.
403
22.
404
prevalence and outcome of heart failure with preserved ejection fraction. The New England journal
405
of medicine 355: 251-259, 2006.
406
23.
407
fraction: comorbidities drive myocardial dysfunction and remodeling through coronary
408
microvascular endothelial inflammation. J Am Coll Cardiol 62: 263-271, 2013.
409
24.
410
inconvenient truth! Journal of the American College of Cardiology 55: 526-537, 2010.
411
25.
412
Steendijk P, Ashrafian H, Henning A, and Frenneaux M. Heart Failure With Preserved Ejection
413
Fraction Is Characterized by Dynamic Impairment of Active Relaxation and Contraction of the Left
Leite-Moreira AF, and Gillebert TC. Nonuniform course of left ventricular pressure fall and
Leite-Moreira AF, Lourenco AP, Roncon-Albuquerque R, Jr., Henriques-Coelho T, Amorim
Leite S, Pinto JO, Silva MT, Abdellatif M, Fontoura D, Falcao-Pires I, Leite-Moreira AF, and
Milani-Nejad N, and Janssen PM. Small and large animal models in cardiac contraction
Owan TE, Hodge DO, Herges RM, Jacobsen SJ, Roger VL, and Redfield MM. Trends in
Paulus WJ, and Tschope C. A novel paradigm for heart failure with preserved ejection
Paulus WJ, and van Ballegoij JJ. Treatment of heart failure with normal ejection fraction: an
Phan TT, Abozguia K, Shivu GN, Mahadevan G, Ahmed I, Williams L, Dwivedi G, Patel K,
18
414
Ventricle on Exercise and Associated With Myocardial Energy Deficiency. J Am Coll Cardiol 54: 402-
415
409, 2009.
416
26.
417
and Deo RC. Phenomapping for novel classification of heart failure with preserved ejection fraction.
418
Circulation 131: 269-279, 2015.
419
27.
420
The pathophysiology of heart failure with normal ejection fraction: exercise echocardiography
421
reveals complex abnormalities of both systolic and diastolic ventricular function involving torsion,
422
untwist, and longitudinal motion. J Am Coll Cardiol 54: 36-46, 2009.
423
28.
424
JG, van der Velden J, Stienen GJ, Laarman GJ, Somsen A, Verheugt FW, Niessen HW, and Paulus
425
WJ. Low myocardial protein kinase G activity in heart failure with preserved ejection fraction.
426
Circulation 126: 830-839, 2012.
427
29.
428
relaxation in the dog: prediction from the time constant for isovolumic pressure fall. The Journal of
429
clinical investigation 62: 1296-1302, 1978.
430
30.
431
Poller W, Pauschinger M, Schultheiss HP, and Tschope C. Role of left ventricular stiffness in heart
432
failure with normal ejection fraction. Circulation 117: 2051-2060, 2008.
433
31.
434
and passive stiffness of the left ventricle. N Engl J Med 350: 1953-1959, 2004.
Shah SJ, Katz DH, Selvaraj S, Burke MA, Yancy CW, Gheorghiade M, Bonow RO, Huang CC,
Tan YT, Wenzelburger F, Lee E, Heatlie G, Leyva F, Patel K, Frenneaux M, and Sanderson JE.
van Heerebeek L, Hamdani N, Falcao-Pires I, Leite-Moreira AF, Begieneman MP, Bronzwaer
Weisfeldt ML, Frederiksen JW, Yin FC, and Weiss JL. Evidence of incomplete left ventricular
Westermann D, Kasner M, Steendijk P, Spillmann F, Riad A, Weitmann K, Hoffmann W,
Zile MR, Baicu CF, and Gaasch WH. Diastolic heart failure--abnormalities in active relaxation
435
19
436
Figure Legends
437
Figure 1. Superimposed representative baseline (dashed line) and isovolumic (continuous line)
438
pressure (panels A-C) and pressure-volume tracings (panels D-F) from Wistar-Kyoto (WKY, black
439
lines, panels A and D), ZSF1 lean (ZSF1 Ln, light gray lines, panels B and E) and ZSF1 obese rats
440
(ZSF1 Ob, dark gray lines, panels c and f). Tracings were obtained by averaging 20 evenly spaced
441
isophasic determinations in each of the animals from every group. Notice the conspicuous rise in
442
end-diastolic pressure (EDP) in the isovolumic beats from ZSF1 Ob compared with ZSF1 Ln and WKY.
443
Not only maximum pressure (Pmax) rises in the isovolumic beat (Pmax (ISO)) but also EDP (EDP (ISO)). In
444
the upper panels dotted lines are added to improve visualization of the differences between Pmax and
445
Pmax (ISO) between groups. In the lower panels dotted lines represent end-systolic elastance (EES) as
446
derived from the slope of single beat end-systolic pressure-volume relationships. In panel A the
447
arrow represents the higher contractile reserve in WKY (see text for detailed results).
448
Figure 2. End-diastolic pressure volume relationships (EDPVR) derived from transient inferior vena
449
cava (IVC) occlusions in Wistar-Kyoto (WKY, black lines), ZSF1 lean (ZSF1 Ln, light gray lines) and
450
ZSF1 obese rats (ZSF1 Ob, dark gray lines). Baseline (continuous) and sequential loops after IVC
451
occlusion (progressive shading) are represented. Tracings were obtained by averaging 20 evenly
452
spaced isophasic determinations in each of the animals from every group. A thick continuous line
453
represents the exponential fitting of the EDPVR. Notice the upward shift in ZSF1 Ob compared with
454
ZSF1 Ln and WKY.
455
Figure 3. End-diastolic pressure (EDP) and indexed volume (EDVi) at baseline (circles) and
456
isovolumic beats (triangles) in Wistar-Kyoto (WKY, white symbols), ZSF1 lean (ZSF1 Ln, light gray
457
symbols) and ZSF1 obese (ZSF1 Ob, dark grey symbols). Results for main effects of group (G),
458
intervention (I) and interactions (G*I) in multivariate (Multiv) and univariate (Univ) analyses are
459
provided. P<0.05 in post hoc tests: *vs WKY, †vs ZSF1 Ln, ‡vs rise of EDP in WKY and ZSF1 Ln.
20
460
Figure 4. Time available to relax (TAR) as a function of isovolumic relaxation time constant
461
calculated by the monoexponential method (τexp) at baseline (circles) and isovolumic beats
462
(triangles) in Wistar-Kyoto (WKY, white symbols), ZSF1 lean (ZSF1 Ln, light gray symbols) and ZSF1
463
obese (ZSF1 Ob, dark gray symbols). TAR was calculated as the difference between predicted
464
relaxation time as estimated by 3.5*τexp and effective diastolic filling time as assessed by the time
465
interval from the maximum rate of pressure fall (dP/dtmin) to end-diastole. Results for main effects of
466
group (G), isovolumic beat intervention (I) and interactions (G*I) in multivariate (Multiv) and
467
univariate (Univ) analyses are provided. ZSF1 Ob showed longer τexp compared with ZSF1 Ln
468
(†P<0.001) and WKY (*P<0.001) and shorter TAR compared with WKY (*P<0.01). Changes in τexp and
469
TAR during the isovolumic beat were more pronounced in ZSF1Ob compared with the other groups
470
(‡P<0.001).
471
Figure 5. Correlations between upward shift in end-diastolic pressure (ΔEDP) and curtailing of time
472
available to relax as percentage (panel A) and chamber stiffness constant for indexed volumes (βi,
473
panel B). Individual data from Wistar-Kyoto (WKY, white circles), ZSF1 lean (ZSF1 Ln, light gray
474
squares) and ZSF1 obese rats (ZSF1 Ob, dark gray triangles) are differentiated. Correlation
475
coefficients (R) and significance levels are presented.
476
477
478
479
480
21
Table 1. Hemodynamics.
WKY
ZSF1 Ln
ZSF1 Ob
HR, min-1
330 ± 12
374 ± 8
319 ± 16†
BSA, cm2
454 ± 6
504 ± 6
659 ± 4*†
CI, μL.min-1.cm-2
125 ± 9
128 ± 10
136 ± 6
EF, %
60 ± 3
55 ± 3
61 ± 4
EDVi, μL.cm-2
0.65 ± 0.03
0.68 ± 0.07
0.76 ± 0.07
dP/dtmax, mmHg.s-1
7860 ± 571
9340 ± 386
10240 ± 582*
0.077 ± 0.007
0.197 ± 0.031*
0.156 ± 0.020*
67 ± 3
64 ± 12
70 ± 9
0.016 ± 0.004
0.022 ± 0.006
0.027 ± 0.009†
109 ± 4
135 ± 2*
170 ± 6*†
204 ± 10‡
204 ± 5‡
223 ± 3‡
EDP, mmHg
6±1
5±1
13 ± 1*†
EDP (ISO), mmHg
9 ± 1‡
9 ± 1‡
20 ± 2*†‡ §
8.7 ± 0.4
8.2 ± 0.4
12.5 ± 0.5*†
12.3 ± 0.8‡
11.9 ± 1.4‡
25.6 ± 1.6*†‡§
Time at dP/dtmin, ms
80 ± 4
81 ± 3
109 ± 8*†
Time at dP/dtmin (ISO), ms
99 ± 5‡
100 ± 6‡
130 ± 6*†‡
ESPVR EES i, mmHg.μL-1.cm-2
PRSW, mmHg
EDPVR βi, μL-1.cm2
Pmax, mmHg
Pmax (ISO), mmHg
τexp, ms
τexp (ISO), ms
Summary of hemodynamic variables for Wistar-Kyoto (WKY, n=11), ZSF1 lean (ZSF1 Ln, n=10) and ZF1 obese rats (ZSF1 Ob,
n=12). HR, heart rate; BSA, body surface area, CI, cardiac index; EF, ejection fraction; EDVi, end-diastolic volume indexed for
BSA; dP/dtmax, maximum rate of pressure rise; ESPVR EES i, end-systolic elastance derived from end-systolic pressure-volume
relationship; PRSW, preload recruitable stroke work; EDPVR βi, left ventricular chamber stiffness constant derived from enddiastolic pressure-volume relationships; Pmax, maximum pressure; EDP, end-diastolic pressure; τexp, time constant of isovolumic
relaxation derived by the monoexponential method; dP/dtmin, maximum rate of pressure fall. Data: mean ± SEM; P<0.05: * vs
WKY and †vs ZSF1 Ln in post hoc group comparisons, ‡vs control beat as main effect, §rise vs other groups, denoting an
interaction term.
1
Table 2. Echocardiography.
WKY
ZSF1 Ln
ZSF1 Ob
HR, min-1
286 ± 18
336 ± 10*
345 ± 13*
LVMi, g.cm-2
1.0 ± 0.1
1.2 ± 0.1
1.3 ± 0.2*
74 ± 3
73 ± 4
74 ± 3
S’, cm.s-1
4.3 ± 0.9
5.7 ± 0.9
6.1 ± 1.2
E/E’
11.4 ± 0.9
12.4 ± 0.7
18.0 ± 0.5*†
EF, %
Summary of echocardiography data for Wistar-Kyoto (WKY), ZSF1 lean (ZSF1 Ln) and ZF1 obese rats (ZSF1 Ob). HR, heart rate;
LVMi, left ventricular mass indexed for body surface area; EF, ejection fraction; S’, maximum systolic tissue Doppler velocity at
the lateral mitral annulus; E/E’, ratio between peak velocity of early filling in transmitral flow pulsed-wave Doppler and
maximum velocity of early diastolic myocardial motion at the lateral mitral annulus. Data: mean ± SEM; P<0.05: * vs WKY and †vs
ZSF1 Ln in post hoc group comparisons; n=6 per group.
Table 3. Multiple linear regression of end-diastolic pressure elevation determinants.
Standardized β ± SE
95% CI of β
T (28)
P value
ΔTAR
0.549 ± 0.14
0.29 to 0.88
3.85
<0.001
CSC βi
0.05 ± 0.17
-0.38 to 0.42
0.26
0.80
Group 1
-0.03 ± 0.17
-0.40 to 0.36
-0.15
0.87
Group 2
0.16 ± 0.22
-0.38 to 0.66
0.71
0.48
Intercept
1.10 ± 0.90*
-1.84 to 3.98
1.22
0.22
Overall model assessment: adjusted R2=0.30, F(1,31)=14.844, P<0.001. Standardized coefficients (β), standard errors (SE) and
confidence intervals (CI) are provided as well as T test results for the appropriate degrees of freedom. Group 1 and 2 are
dummy variables that recode the 3 groups in 2 binary variables. ΔTAR, percentage change from baseline to isovolumic beat of
time available to relax; CSC βi, chamber stiffness constant for indexed volumes.