OUP user menu

Platelet-derived growth factor B retention is essential for development of normal structure and function of conduit vessels and capillaries

Henrik C. Nyström, Per Lindblom, Anna Wickman, Irene Andersson, Jenny Norlin, Jenny Fäldt, Per Lindahl, Ole Skøtt, Mattias Bjarnegård, Sharyn M. Fitzgerald, Kenneth Caidahl, Li-ming Gan, Christer Betsholtz, Göran Bergström
DOI: http://dx.doi.org/10.1016/j.cardiores.2006.05.019 557-565 First published online: 1 August 2006


Objective: Extracellular retention of PDGF-B has been proposed to play an important role in PDGF-B signalling. We used the PDGF-B retention motif knockout mouse (RetKO) to study the effects of retention motif deletion on development of micro- and macrovascular structure and function.

Methods Passive and active properties of conduit vessels were studied using myograph techniques and histological examination. Capillary structure and function was studied using measurements of capillary density in skeletal muscle and by assessing aerobic physical performance in a treadmill setup. Cardiac function was assessed using echocardiography.

Results: Myograph experiments revealed an increased diameter and stiffness of the aorta in RetKO. Histological examination showed increased media collagen content and a decreased number of aortic wall layers, however with a similar number of vascular smooth muscle cells. This outward eutrophic remodelling of the aorta was accompanied by endothelial dysfunction. RetKO showed decreased capillary density in skeletal muscle and signs of a defective delivery of capillary oxygen to skeletal muscle, as shown by a decreased physical performance. In RetKO mice, echocardiography revealed an adaptive eccentric cardiac hypertrophy.

Conclusion We conclude that retention of PDGF-B during development is essential for a normal conduit vessel function in the adult mouse. Furthermore, PDGF-B retention is also necessary for the development of an adequate capillary density, and thereby for a normal oxygen delivery to skeletal muscle. The lack of primary effects on cardiac function supports the redundant role of PDGF-B in cardiac development.

  • Platelet-derived growth factor B
  • Extracellular matrix
  • Vascular function
  • Cardiac function
  • In vivo physiology

1. Introduction

Platelet-derived growth factors (PDGFs) play a critical role during development [1–3], however also in the pathogenesis of several diseases, such as atherosclerosis and nephrosclerosis [2]. During development, one of the subtypes, PDGF-B, is excreted mainly by endothelial cells, acting on neighbouring PDGF receptor β (PDGFR-β)-positive vascular smooth muscle cell (VSMC)/pericyte progenitors. A model has been proposed, where endothelial cells in sprouting blood vessels recruit pericyte progenitors in a process dependent on a functional PDGF-B and PDGFR-β [4]. In support of this, knockout of PDGF-B or the PDGFR-β in the mouse leads to prenatal lethality with severe cardiovascular changes [5,6].

Whereas the evidence for an important role of PDGF-B in capillary sprouting appears convincing, the effects of PDGF-B in development of multilayered vessels is disputed [7]. However, some data support the hypothesis that recruitment of VSMC to multilayered vessels is performed using PDGF-B/PDGFR-β signalling in a similar fashion [8]. In support of this, multilayered vessels in PDGF-B as well as PDGFR-β null mutants are thin-walled and dilated [5,6].

Interaction with extracellular matrix is probably of importance for proper function of PDGF-B [9]. In line with this, a retention sequence in the COOH terminus of the PDGF-B molecule, binding to the extracellular matrix and the cell surface, has been described [10]. In in vitro transfection studies, C-terminally truncated PDGF-B accumulated in the culture medium, whereas wild-type PDGF-B was trapped intracellularly [11]. Furthermore, expression of the two different forms resulted in different proliferation patterns in dermal mesenchymal cells [12]. These results point to an important role for the retention motif in limiting the action range of PDGF-B.

Recently, mutant mice have been generated in which the PDGF-B retention motif has been selectively deleted (RetKO) [13]. Morphologically, these mutants show prenatal microvascular, renal and retinal abnormalities similar to the PDGF-B null mutants. In contrast to the null mutants, RetKO mutants also survive into adulthood.

The aim of the present study was to test the hypothesis that PDGF-B retention is important for development of a functional micro- and macrovasculature in the adult mouse. To assess macrovascular structure and function, we used myograph techniques and histological examination of aorta and mesenteric resistance vessels. For assessment of microvascular properties, capillary density was measured in skeletal muscle, whereas oxygen delivery was assessed by testing aerobic physical performance in a treadmill setup. Cardiac function was assessed using echocardiography and electrocardiography.

2 Materials and methods

2.1 RetKO mutants

Knockout of the PDGF-B retention motif was performed as previously described [13]. For the study, male PDGF-Bret/ret were used as retention motif knockouts (RetKO) and male PDGF-Bret/+ and PDGF-B+/+ for the littermate control group. Animals were housed under standard conditions in the Göteborg University Animal Core Facility. All procedures were conducted in accordance with the Guide for the Care and Use of Laboratory Animals (NIH Publication No. 85-23, revised 1996). The study was performed after prior approval from the local ethics committee for animal experimentation at Göteborg University, Sweden.

2.2 Study design

At 10 and 20 weeks of age, tail-cuff systolic blood pressure was assessed and echocardiography performed. Electrocardiographic measurements were performed at 20 weeks of age. Physical performance and tissue oxygen tension were assessed at 15 and 20 weeks respectively. At the end of the study (20 weeks), blood was drawn from the orbital vein for renin measurements. Two days later the animals were anaesthetized and sacrificed. The thoracic aorta and the mesenteric bed were removed and used for morphological examination and studies of vascular function. In a separate group, skeletal muscle was saved for capillary density measurements.

2.3 Myograph assessment of mesenteric and aortic vascular function

The mice were anaesthetized with pentobarbital sodium and the mesentery and an aortic segment (the approx. 3 mm of thoracic aorta directly distal to the aortic arch) were dissected out and kept in physiological salt solution (PSS, composition in mmol l− 1: NaCl 119, NaHCO3 25, glucose 5.5, KCl 4.7, CaCl2 2.5, KH2PO4 1.18, MgSO4 1.17, EDTA 0.026). The mesenteric resistance vessel segments were mounted in a Multimyograph 610M (Danish Myo Technology, Aarhus, Denmark) and the protocol performed as previously described [14]. For the aortic studies, a purpose-built myograph was used. In this device, the aortic segments were mounted on two stainless-steel hooks in PSS-containing glass chambers. Solutions were equilibrated with 95% O2/5% CO2 (pH 7.4) and bath temperature was maintained at 37 °C. Isometric forces were recorded (Grass Instruments Co., Quincy, MA) and collected (Pharmlab 3.0, AstraZeneca, Mölndal, Sweden).

Equilibration of the vessels was performed for 30 min at 3 mN. Thereafter, passive vascular properties were assessed in aortic strips and resistance vessels by exposure to stepwise increases in developed force while recording corresponding diameter. We used the developed force, corresponding diameter and the wall length of the vessel to calculate the luminal cross-sectional area (CSA) and the corresponding pressure in mm Hg [15] for every step: Embedded Image Embedded Image where d is the diameter of the vessel in m, P is the pressure in mm Hg, F is the force in N and l is the wall length in m.

Using linear regression, a CSA–pressure curve could then be constructed for every vessel, describing the increase in luminal area resulting from a given increase in intraluminal pressure.

For assessment of aortic vasoconstrictor properties, aortic vessels were passively stretched to 12 mN, whereas resistance vessels were exposed to a tension corresponding to 100 mm Hg. Vasoconstrictor properties were then assessed in all vessels using potassium chloride (KCl, 100 mmol l− 1, from KEBO Lab, Stockholm, Sweden, all other drugs from Sigma Chemicals, St Louis, MO, USA) and noradrenaline (NA dose–response, range 0.01–10 μmol l− 1). Endothelium-dependent vasodilatation was studied using acetylcholine (ACh dose–response, range 0.03–10 μmol l− 1, with and without l-NNA (N[ω]-nitro-l-arginine, 100 μmol l− 1) and endothelium-independent vasodilatation was assessed using two supramaximal doses of sodium nitroprusside (SNP). At end of experiment, the aorta was fixed in 4% paraformaldehyde for histological analysis. The heart, right lung and adductor musculature were dissected out and fixed in 4% paraformaldehyde. Before fixation, the heart was divided into left and right ventricle and weighed.

2.4 Histological analysis

After fixation, aortic sections, heart, lung and skeletal muscle were paraffin embedded and sectioned at 5 μm. The vascular sections were stained with hematoxylin–eosin and picrosirius red and the heart and lung sections with hematoxylin–eosin only.

On the vascular sections, media thickness, cell count and counting of media layers was performed manually and internal elastica lamina measurements were performed using an automatic wall-tracking tool (Micro Image, Olympus Optical Co., Hamburg, Germany). CSA and media area were calculated from the internal elastica lamina length and the media thickness using ordinary geometry.

Total transectional area, total media area, collagen content area, collagen I area and media collagen area were also measured automatically in picrosirius stained vessels using the Micro Image software [16].

2.4.1 Capillary morphometry

Skeletal muscle sections were stained with isolectin (Bandeiraea simplicifolia, Sigma L-2140) as described [17]. In 10 sections per animal, capillary density was calculated per microscopic field as well as per cross-sectioned muscle fiber and average skeletal muscle cell diameter was measured. The sections were manually examined for changes in capillary and skeletal muscle morphology, such as distorted capillary structure or clear skeletal muscle pathology. Measurements were performed using a Nikon Eclipse E1000 microscope, together with the Easy Image Measurement 2000 software.

2.5 Plasma renin measurements

A blood sample of 200 μl whole blood was collected from the orbital vein during anaesthesia (isoflurane). The plasma was collected and stored at − 20 °C for analysis. Plasma renin activity was measured by radioimmunoassay of angiotensin I using the antibody-trapping technique as previously described [18].

2.6 Electrocardiography and echocardiography

2.6.1 Electrocardiography

The mice were anaesthetized with isoflurane and electrocardiographic leads were placed on the right upper and left lower extremity allowing a lead II bipotential recording. ECG was recorded at 2000 Hz with Pharmlab 3.0 and a mean QRS complex was generated using the Mini Analysis Software (Synaptosoft Inc., Decatur, GA, USA) and electrophysiological conduction times were estimated.

2.6.2 Echocardiography

Cardiac ultrasound imaging was performed under isoflurane anaesthesia using a 15-MHz linear transducer (Sonos 5500, Agilent, Andover, MA) connected to an ultrasound system (HDI 5000, ATL Ultrasound, Bothell, WA, USA) with a frame rate of approx. 230 frames/s when applying HDI zoom. Two-dimensional echocardiographic loops and parasternal long-axis M-mode (motion-mode) tracings were recorded. The M-mode recordings were used for all calculations of cardiac structure and function. Spatial flow profile (velocity–time integral, VTI) in the pulmonary artery were recorded using pulsed-wave Doppler recordings. Together with right ventricular outflow dimensions, these measurements were also used to calculate stroke volume (SV) and cardiac output (CO). Left ventricular outflow tract diameter (LVOT) was also measured.

Off-line measurements were performed using an image analysis system (HDI-Lab, ATL Ultrasound, Bothell, WA, USA). Left ventricular (LV) fractional shortening (FS), SV, CO, left ventricular mass (LVM, using the cubic formula) [19], left ventricular internal diameter in systole and diastole (LVIDs and LVIDd) and end-diastolic volume (EDV, using the Teichholz formula) [20] were calculated according to guidelines from the American Society of Echocardiography.

2.7 Physical performance

The mice were made to run in a treadmill at an inclination of 20° and a speed of 10 m/min according to a specific protocol, where the speed was successively increased (to 13 m/min at 10 min, 18 m/min at 15 min and to 20 m/min at 20 min), finishing after exhaustion or 25 min of running. Gentle tapping on the back was used to affect the willingness of the mice to run. Pre- and post-running lactate and glucose values were measured in a blood sample drawn from the tail by spectrophotometry (ABL 725, Radiometer, Copenhagen, Denmark). The pre-running blood sample was also used for measurements of blood haemoglobin concentration. In a separate experiment, oxygen consumption and carbon dioxide production of the mice were measured at rest and at submaximal exercise (10 m/min at 0, 10° or 20° inclination for 20 min) on the treadmill using an Oxymax system (Columbus Instruments, Columbus, OH, USA).

2.8 Oxygen tension measurements

The animals were anaesthetized with isoflurane and fixed in a supine position on a heated operating table. An incision was made in the groin along the femur, uncovering the adductor muscles. An oxygen electrode (OX-500, Unisense, Aarhus, Denmark, diam 500 μm) was inserted into the muscle using a micromanipulator (no. 11872, Narishige Scientific Instrument Lab., Tokyo, Japan) and oxygen tension was measured at three equidistant locations approx. 3 mm apart.

2.9 Systolic blood pressure measurements

Conscious systolic blood pressure (BP) was measured using a computerized noninvasive tail-cuff system (RTBP Monitor, Harvard Apparatus Inc., South Natick, MA). Measurements were performed on three subsequent days, with at least six recordings for each time point. Final systolic BP was obtained by averaging the mean values for the different time points.

2.10 Statistical analysis

All values are expressed as the mean±S.E.M. Concentration–response relations in the Multimyograph were analyzed with nonlinear regression (Graph Pad Software, Inc. Systems, San Diego, CA). Concentration–response curves were compared using the area under curve method. Statistical significance was assessed by Student's t-test or z-test (physical performance). A value of p<0.05 was considered statistically significant.

3 Results

3.1 Conduit vessel dysfunction in RetKO

3.1.1 Passive vessel function

Fig. 1 shows the increase in CSA in absolute (A) and relative (B) measures, with stepwise increases in pressure in the aorta. CSA was larger in RetKO for every given pressure (P<0.001). However, the relative CSA increase was 53±8% in RetKO aorta and 107±17% in control aorta (P<0.05). Hence, RetKO vessels were significantly stiffer than control vessels.

Fig. 1

Static aorta cross-sectional area (CSA)–pressure curve (A) and relative CSA–pressure curve (% of CSA at 35 mm Hg) (B) during phase of distension from 35 to 60 mm Hg in control mice (solid line, n=7) and RetKO (squares, n=6). Data are expressed as the mean±S.E.M. * denotes P<0.05 vs. controls; *** denotes P<0.001 vs. controls, using the AUC method of comparison between groups.

Resistance vessels showed no difference in absolute diameter between the two groups at any given pressure (data not shown).

3.1.2 Endothelial dysfunction in RetKO aorta

In aortic vessels, we found no significant difference in the response to NA or KCl. However, ACh-induced vasorelaxation was lower in RetKO aorta, compared with control aorta (P<0.05, Table 1).

View this table:
Table 1

Active properties of aorta and mesenteric resistance vessels in control mice and RetKO at 20 weeks of age

AortaMesenteric resistance vessel
Control (n=7)RetKO (n=7)Control (n=7)RetKO (n=7)
K max (N m− 1)2.76±0.412.31±0.301.77±0.182.00±0.03
NA max (N m− 1)2.17±0.262.00±0.340.68±0.180.45±0.18
NA precontraction (N m− 1)1.67±0.181.85±0.370.68±0.150.46±0.12
ACh max dilatation (%)51.5±8.327.2±4.1*42.0±5.344.1±8.2
l-NNA+ACh max dilatation (%)− 36.8±7.3− 30.6±9.316.6±19.421.9±16.8
SNP max dilatation (%)106.9±0.6107.0±1.193.8±3.089.2±4.4
  • Dilatation is expressed as percent of precontraction. K max, maximal contraction on potassium; N, Newton; NA, noradrenaline; ACh, acetylcholine; l-NNA, N[ω]-nitro- l -arginine; SNP, sodium nitroprusside. Data are expressed as the mean±S.E.M.

  • * denotes P<0.05 vs. controls.

Investigations of resistance vessels showed no significant difference in the response to either NA, KCl or ACh (Table 1). In both aorta and resistance vessels, pretreatment with the NO-synthase inhibitor l-NNA abolished dilatation in both RetKO and control vessels. RetKO and control mice showed similar vasorelaxation in response to the NO-donor SNP, indicating preserved endothelium-independent vasorelaxation in aorta and resistance vessels of RetKO mice.

3.2 Outward remodelling and increased media collagen content in RetKO aorta

No significant difference was observed in media thickness (RetKO, 42.6±1.4 μm; control, 47.8±2.4 μm, NS) or cell number (data not shown), but RetKO aorta showed a lower number of VSMC layers compared to control aorta (Figs. 2 and 5). Internal elastica lamina was significantly longer in the RetKO compared to control, creating a correspondingly larger CSA (Fig. 2). No difference was seen in the calculated media area (data not shown).

Fig. 2

Histological properties of conduit vessels (aorta) at 20 weeks of age in control mice (white bars, n=5) and RetKO (black bars, n=7). Panels represent media cell layer count (A), media thickness (B), luminal cross-sectional area (C) and relative media collagen content (D). CSA, luminal cross-sectional area. Data are expressed as the mean±S.E.M. * denotes P<0.05 vs. controls; ** denotes P<0.01 vs. controls.

No difference was observed in any of the parameters calculated using picrosirius-stained sections (data not shown). However, RetKO mice displayed an increase in the relative collagen content of the calculated media area compared to control (Figs. 2 and 5).

3.3 Plasma renin measurements

No difference was observed in plasma renin activity (RetKO, 11.9±3.3 mGU ml− 1; control, 12.0±3.7 mGU ml− 1, NS).

3.4 Echocardiography and electrocardiography reveals cardiac hypertrophy in RetKO

At 10 weeks, RetKO had a larger LVOT, a larger LVIDd and a larger M-mode-derived SV and EDV compared to control mice (Table 2). No difference in any other parameter was observed.

View this table:
Table 2

Cardiac dimensions and function of control mice and RetKO at 10 and 20 weeks of age

10 weeks20 weeks
Control (n=8)RetKO (n=8)Control (n=8)RetKO (n=8)
SV (μl)50.5±5.070.4±4.7*48.1±3.875.0±4.5***
Heart rate (BPM)371±27377±14442±26493±9
CO (ml min− 1)19.3±2.926.6±2.121.0±1.436.9±2.1***
FS (%)35.8±2.937.0±1.936.6±1.732.8±2.0
LVOT (mm)1.81±0.052.11±0.12*1.64±0.072.14±0.09***
IVSTd (mm)0.85±0.051.08±0.140.90±0.030.84±0.07
PWTd (mm)0.68±0.050.78±0.060.85±0.040.82±0.05
LVIDd (mm)4.16±0.134.77±0.14**4.05±0.155.13±0.22**
EDV (μl)77.5±5.5106.7±7.7**73.0±6.4128.2±12.9**
LVM (mg)117.4±7.7202.3±37.4135.0±8.2191.4±21.9*
  • SV, stroke volume; CO, cardiac output; FS, fractional shortening; LVOT, left ventricular outflow tract diameter; IVSTd, interventricular septum thickness in diastole; PWTd, posterior wall thickness in diastole; LVIDd, left ventricular internal diameter in diastole; EDV, end-diastolic volume; LVM, left ventricular mass. Data are expressed as the mean±S.E.M.

  • * denotes P<0.05 vs. controls.

  • ** denotes P<0.01 vs. controls.

  • *** denotes P<0.001 vs. controls.

At 20 weeks of age, LVM, EDV, LVIDd and LVIDs were all significantly larger in RetKO (Table 2, Fig. 3) compared to control. Both M-mode-derived and VTI-derived (data not shown) SV and CO were higher in the RetKO, together with a larger LVOT, when compared to control.

Fig. 3

Echocardiographic estimation of cardiac output (A), left ventricular mass (B), end-diastolic volume (C) and the left ventricular weight-end-diastolic volume ratio (D). Measurements were performed at 20 weeks of age in control mice (white bars, n=8) and RetKO (black bars, n=8). EDV, end-diastolic volume; LVM, left ventricular mass; LV weight, left ventricular weight. Data are expressed as the mean±S.E.M. * denotes P<0.05 vs. controls; ** denotes P<0.01 vs. controls; *** denotes P<0.001 vs. controls.

RetKO displayed a significantly longer QRS time (RetKO, 13.9±0.9 ms; control, 10.2±0.2 ms, P<0.01), QT time (RetKO, 45.2±1.8 ms; control, 35.9±3.3 ms, P<0.05), QTC, a higher R amplitude (RetKO, 1.4±0.2 mV; control, 0.9±0.1 mV, P<0.05), and a longer ventricular activation time compared to control group. PQ time was similar (data not shown).

There was no significant difference in body weight (BW) between RetKO and control mice throughout the study (data not shown). LV weight normalized to BW was significantly higher in the RetKO (RetKO, 44.0±3.96 mg/10 g BW; controls, 32.4±0.49 mg/10 g BW, P<0.05) compared with control mice as was normalized right ventricular weight (RetKO, 10.0±0.74 mg/10 g BW; control, 7.17±0.32 mg/10 g BW, P<0.01). Since femur length was not measured, a cardiac weight/femur length ratio could not be calculated.

3.5 RetKO show reduced capillary density

RetKO exhibited a reduced number of capillaries per microscopic field (850 × 650 μm) as well as per skeletal muscle cell (Fig. 4). Average skeletal muscle cell diameter (RetKO, 41±5 μm; control, 40±4 μm, NS) was similar in RetKO and control, however, the number of skeletal muscle cells per microscopic field (RetKO, 6±0.5; control, 8±0.2, P<0.01) was smaller in RetKO. Capillary morphometry was discretely changed, with small dilatations and irregular capillary diameters (Fig. 5). In heart and lung, no signs of pathology were observed by gross histological examination in the RetKO.

Fig. 5

Representative examples of mouse skeletal muscle (isolectin stain, A–B) and aorta (hematoxylin–eosin, C–D; picrosirius red, E–F) at 20 weeks of age. (A, C, E) Control; (B, D, F) RetKO. In (A and B), scale bar represents 30 μm; in (C–F), scale bar represents 90 μm. By using polarized light, picrosirius-stained sections were also used to measure collagen subtypes.

Fig. 4

Capillary density in skeletal muscle (adductor musculature) at 20 weeks of age in control mice (white bars, n=4) and RetKO (black bars, n=4). Capillary density is shown as number of capillaries per microscopic field (A) and number of capillaries per cross-sectioned skeletal muscle cell (B). Data are expressed as the mean±S.E.M. ** denotes P<0.01 vs. controls.

3.6 RetKO show reduced physical performance

RetKO mice exhibited a significantly shorter treadmill running time (RetKO, 22.0±1.4 min; all controls running for 25 min, P<0.05) compared to control. Basal blood lactate levels and blood glucose levels were similar in RetKO and controls. No increase was seen in blood lactate or glucose levels after running (data not shown) and blood haemoglobin concentration was similar (RetKO, 145±5 g l− 1; control, 151±7 g l− 1, NS).

Oxygen consumption and carbon dioxide production were similar at rest as well as during submaximal exercise in the two groups.

Oxygen tension in resting hindleg skeletal muscle was similar in RetKO (RetKO, 11.1±1.2 kPa; control, 12.4±1.5 kPa, NS) and control.

3.7 Blood pressure

At 10 and 20 weeks of age, systolic BP, measured by tail-cuff, was similar in RetKO and control group (RetKO, 127±1 mm Hg; control, 129±4 mm Hg, NS). At 20 weeks, systolic BP had increased in both RetKO and control (RetKO, 146±2 mm Hg; control, 146±4 mm Hg, NS).

4 Discussion

The main finding in the present study was that PDGF-B retention during development is required for a normal conduit vessel function in the adult mouse. Furthermore, an intact PDGF-B retention appears important for normal oxygen delivery to skeletal muscle, via its role in development of an adequate capillary density. Only adaptive cardiac changes were found, a fact supporting the hypothesis of PDGF-B as redundant in cardiac development.

4.1 PDGF-B retention in development of normal cardiovascular function

Interaction with extracellular matrix (ECM) molecules has been proposed as an important regulator of growth factor action in vivo [9]. In the case of PDGF-B, these ECM-binding properties have been conferred to the retention motif, a sequence of positively charged amino acids in the C-terminal part of the protein [10]. As outlined in Introduction, experimental evidence [10–12] support the importance of the retention sequence in limiting the action range of PDGF-B.

Knockout of the PDGF-B retention motif in the mouse (RetKO) leads to defective investment of pericytes in the vascular wall with severe vascular, glomerular and retinal defects [13]. This is in analogy with the PDGF-B null mouse, displaying similar cardiovascular abnormalities [5].

Based on the present study, we have now for the first time established a role for PDGF-B retention in development of the functional properties of conduit vessels and capillaries. Since PDGF-B is instrumental both in development as well as in several pathophysiological processes in the adult, these findings have implications on processes such as atherogenesis, in which PDGF-B plays an important role [2].

4.2 Conduit vessel dysfunction in RetKO

In RetKO aorta, the diameter was increased with an increased stiffness of the vascular wall, together with endothelial dysfunction. The number of wall layers was decreased without changes in number of VSMC (i.e. eutrophic outward remodelling) and with higher relative media collagen content. However, no changes in resistance vessel function or systolic blood pressure were observed.

The discrepant results between aorta and mesenteric resistance vessels could possibly be attributed to their different ontogenic origin, since the development of central arteries by arteriogenesis differs mechanistically from the sprouting of small resistance vessels in the gut [7]. It could also be attributed to the different location of the vessels in the body.

Furthermore, it is likely that the conduit vessel changes are explained by a primary effect of the mutation on VSMC function. Our hypothesized importance of PDGF-B in proliferation and migration of VSMC in large vessels goes well in line with the observed aortic dilatation and, thereby, the vascular dysfunction observed in the current study.

We also report an increased stiffness of the aorta, which we suggest is secondary to the outward remodelling process. According to the law of Laplace, the increased diameter of the aorta, in combination with a similar blood pressure and a reduced number of VSMC layers, would imply an increased tension in the aortic wall. The most likely explanation for the increased aortic stiffness would then be the observed proportional increase in vascular collagen, constituting the proper physiological response to a higher vascular tension. No difference was observed in vascular contractility, despite the increased collagen expression. This could possibly be due to compensatory changes in the expression of extracellular matrix components.

The increased vascular diameter of the aorta in RetKO was accompanied by impaired vasodilatation to ACh (i.e. endothelial dysfunction). It is possible that the endothelial dysfunction observed in the aorta is secondary to the aortic dilatation. This larger lumen would tend to decrease shear stress, and in turn decrease the expression of endothelial-derived nitric oxide synthase, thereby reducing the capacity to release nitric oxide [21].

4.3 Eccentric cardiac hypertrophy in the RetKO

Left ventricular hypertrophy (LVH) was observed in RetKO at 20 weeks of age, as evident from echocardiography, ECG findings and postmortem cardiac weights. In the heart, increases in cardiac load lead to a highly differentiated LVH. In response to an increased pressure load, a concentric LVH is observed, with an increase in LV weight/LV lumen ratio. Following an increase in volume load, however, both wall thickness and lumen size are proportionally increased, leaving LV weight /LV lumen ratio unchanged, resulting in an eccentric LVH [22].

In the present study, echocardiographic examination of RetKO mice at 20 weeks showed an increased LV mass and LV volume together with an unchanged LV mass/LV volume ratio. These findings, together with an increase in calculated cardiac output and stroke volume and an unchanged blood pressure, point to a hyperkinetic circulation with an adaptive eccentric LVH.

Since PDGF-B is crucial for capillary development, it is highly likely that capillary structure and/or function is changed in the RetKO. We propose a model, where tissue ischaemia arises on the basis of impaired nutrient exchange in the capillaries of RetKO mice. This would lead to an ischaemic vasodilatory response in resistance vessels. The decrease in peripheral resistance would demand an increase in cardiac output to maintain blood pressure at a normal level. The observed eccentric LVH would then be the compensatory response to the reduced peripheral resistance and resultant hyperkinetic circulation.

In support of this physiological hypertrophy we found unchanged plasma renin activity and ECG measurements revealed no signs of electrophysiological disturbances. The changes in cardiac volumes and performance also gradually developed between 10 and 20 weeks and histological examination revealed no obvious signs of cardiac pathology. Since no histological studies on cardiomyocytes were performed, no conclusions could be drawn on hypertrophy on the cellular level.

4.4 Cardiac changes are secondary to capillary rarefaction

In order to elucidate the mechanism behind the observed hyperkinetic circulation, we performed a capillary count in skeletal muscle. Indeed, a decreased capillary density was observed in the RetKO. In order to investigate the functional relevance of these capillary changes, we also performed a physical exhaustion test revealing an impaired physical performance in the RetKO.

It has been shown that the capillary surface area is a limiting factor for the maximal distribution of oxygen to working skeletal muscle [23]. With intact functional mechanisms for oxygen uptake and transport, together with the observed capillary rarefaction, the explanation for the observed defect in oxygen delivery would be a decreased capillary area with a resultant high resistance for oxygen diffusion. The observed morphological changes in skeletal muscle capillaries could also negatively affect gaseous exchange [24]. The decreased number of skeletal muscle cells observed could be secondary to a decreased oxygen delivery.

To further test our hypothesis of a decreased oxygen delivery, direct oxygen tension measurements were performed in resting skeletal muscle. This experiment did not support our hypothesis since we found similar oxygen tension in both RetKO and control mice. However, this could be the result of powerful vasodilatory compensatory mechanisms leaving tissue oxygen tension unchanged at rest. It is likely that a certain level of physical activity is required to unmask this hypothesized ischaemia.

5 Conclusion

We conclude that retention of PDGF-B is essential for development of a normal conduit vessel function. We also conclude that an intact PDGF-B retention is needed for maintenance of an adequate capillary oxygen delivery to skeletal muscle, probably via its role in development of an adequate capillary density. Furthermore, the lack of primary cardiac defects underlines the redundant role of PDGF-B signalling in cardiac development.


This work was supported by the Swedish Medical Research Council (grant no. 12 580), funds at Sahlgrenska University Hospital (LUA/ALF), the Swedish Hypertension Society, the Swedish Heart Lung Foundation and the Emil and Maria Palm Foundation. SWEGENE supports the post-doctoral funding of A.W. (personal support by the King Gustaf V and Queen Victoria Foundation). We are also grateful to the Center for Mouse Physiology and Center for Bio-Imaging at Göteborg University in which parts of these studies were conducted.

The authors wish to thank Mrs. Jing Jia and Mrs. Gunnel Andersson for their excellent technical assistance and Professor Björn Folkow for his invaluable comments on the manuscript.


  • Parts of this work have previously been published in abstract-form at the XIIIth International Symposium on Atherosclerosis, 2003, Kyoto, Japan.

  • Time for primary review 21 days


  1. [1]
  2. [2]
  3. [3]
  4. [4]
  5. [5]
  6. [6]
  7. [7]
  8. [8]
  9. [9]
  10. [10]
  11. [11]
  12. [12]
  13. [13]
  14. [14]
  15. [15]
  16. [16]
  17. [17]
  18. [18]
  19. [19]
  20. [20]
  21. [21]
  22. [22]
  23. [23]
  24. [24]
View Abstract