Unlike during development, blood vessels in the adult are generally thought not to require VEGF for normal function. However, VEGF is a survival factor for many tumor vessels, and there are clues that some normal blood vessels may also depend on VEGF. In this study, we sought to identify which, if any, vascular beds in adult mice depend on VEGF for survival. Mice were treated with a small-molecule VEGF receptor (VEGFR) tyrosine kinase inhibitor or soluble VEGFRs for 1-3 wk. Blood vessels were assessed using immunohistochemistry or scanning or transmission electron microscopy. In a study of 17 normal organs after VEGF inhibition, we found significant capillary regression in pancreatic islets, thyroid, adrenal cortex, pituitary, choroid plexus, small-intestinal villi, and epididymal adipose tissue. The amount of regression was dose dependent and varied from organ to organ, with a maximum of 68% in thyroid, but was less in normal organs than in tumors in RIP-Tag2-transgenic mice or in Lewis lung carcinoma. VEGF-dependent capillaries were fenestrated, expressed high levels of both VEGFR-2 and VEGFR-3, and had normal pericyte coverage. Surviving capillaries in affected organs had fewer fenestrations and less VEGFR expression. All mice appeared healthy, but distinct physiological changes, including more efficient blood glucose handling, accompanied some regimens of VEGF inhibition. Strikingly, most capillaries in the thyroid grew back within 2 wk after cessation of treatment for 1 wk. Our findings of VEGF dependency of normal fenestrated capillaries and rapid regrowth after regression demonstrate the plasticity of the adult microvasculature.

An α 1 -adrenergic receptor (α 1 -AR) antagonist increased heart failure in the Antihypertensive and LipidLowering Treatment to Prevent Heart Attack Trial (ALLHAT), but it is unknown whether this adverse result was due to α 1 -AR inhibition or a nonspecific drug effect. We studied cardiac pressure overload in mice with double KO of the 2 main α 1 -AR subtypes in the heart, α 1A (Adra1a) and α 1B (Adra1b). At 2 weeks after transverse aortic constriction (TAC), KO mouse survival was only 60% of WT, and surviving KO mice had lower ejection fractions and larger end-diastolic volumes than WT mice. Mechanistically, final heart weight and myocyte cross-sectional area were the same after TAC in KO and WT mice. However, KO hearts after TAC had increased interstitial fibrosis, increased apoptosis, and failed induction of the fetal hypertrophic genes. Before TAC, isolated KO myocytes were more susceptible to apoptosis after oxidative and β-AR stimulation, and β-ARs were desensitized. Thus, α 1 -AR deletion worsens dilated cardiomyopathy after pressure overload, by multiple mechanisms, indicating that α 1 -signaling is required for cardiac adaptation. These results suggest that the adverse cardiac effects of α 1 -antagonists in clinical trials are due to loss of α 1 -signaling in myocytes, emphasizing concern about clinical use of α 1 -antagonists, and point to a revised perspective on sympathetic activation in heart failure.

It has been shown that the adeno-associated virus (AAV) vector can deliver the VEGF gene efficiently into the ischemic mouse myocardium. However, the AAV genomes can be found in extracardiac organs after intramyocardial injection. To limit unwanted VEGF expression in organs other than the heart, we tested the use of the cardiac myosin light chain 2v (MLC-2v) promoter and the hypoxiaresponse element to mediate cardiac-specific and hypoxia-inducible VEGF expression. An AAV vector, MLCVEGF, with 250 bp of the MLC-2v promoter and nine copies of the hypoxia-response element driving VEGF expression, was constructed. Gene expression was studied in vitro by infection of rat cardiomyocytes, rat skeletal myocytes, and mouse fibroblasts with the vector and in vivo by direct injection of the vector into normal and ischemic mouse hearts. With MLCVEGF infection, VEGF expression was higher in cardiomyocytes than the other two cell lines and was hypoxiainducible. VEGF expression was also higher in ischemic hearts than in normal hearts. No VEGF expression was detectable in organs with detectable MLCVEGF vectors other than the heart. MLCVEGFinjected ischemic hearts had more capillaries and small vessels around the injection site, smaller infarct size, and better cardiac function than the negative controls. Hence, MLCVEGF can mediate cardiac-specific and hypoxia-inducible VEGF expression, neoangiogenesis, infarct-size reduction, and cardiac functional improvement.cardiac-specific gene expression ͉ hypoxia induction ͉ ischemic heart

To elucidate the time course of sympathovagal balance and its relationship to left ventricular function in heart failure, we serially evaluated left ventricular contractility and relaxation and autonomic tone in 11 conscious dogs with tachycardia-induced heart failure. We determined a dynamic map of sympathetic and parasympathetic modulation by power spectral analysis of heart rate variability. The left ventricular peak +dP/dt substantially fell from 3,364 +/- 338 to 1,959 +/- 318 mmHg/s (P < 0.05) on the third day and declined gradually to 1,783 +/- 312 mmHg/s at 2 wk of rapid ventricular pacing. In contrast, the time constant of left ventricular pressure decay and end-diastolic pressure increased gradually from 25 +/- 4 to 47 +/- 5 ms (P < 0.05) and from 10 +/- 2 to 21 +/- 3 mmHg (P < 0.05), respectively, at 2 wk of pacing. The high-frequency component (0.15-1.0 Hz), a marker of parasympathetic modulation, decreased from 1,928 +/- 1,914 to 62 +/- 68 x 10(3) ms2 (P < 0.05) on the third day and further to 9 +/- 12 x 10(3) ms2 (P < 0.05) at 2 wk. Similar to the time course of left ventricular diastolic dysfunction, plasma norepinephrine levels and the ratio of low (0.05- to 0.15-Hz)- to high-frequency component increased progressively from 135 +/- 50 to 532 +/- 186 pg/ml (P < 0.05) and from 0.06 +/- 0.06 to 1.12 +/- 1.01 (P < 0.05), respectively, at 2 wk of pacing. These cardiac and autonomic dysfunctions recovered gradually toward the normal values at 2 wk after cessation of pacing. Thus a parallel decline in left ventricular contractility with parasympathetic influence and a parallel progression in left ventricular diastolic dysfunction with sympathoexcitation suggest a close relationship between cardiac dysfunction and autonomic dysregulation during development of heart failure.

No abstract