In this study, we have identified for the first time that an elevated TNF-α level is correlated to AT1-AA bioactivity in preeclamptic women and provided both in vitro human studies and in vivo mouse evidence that AT1-AA is a novel candidate directly inducing TNF-α production via AT1 receptor activation. Neutralizing AT1-AA-mediated TNF-α induction attenuates the increased placenta apoptosis and sFlt-1 and sEng secretion by cultured human villous explants. Moreover, TNF-α blockade ameliorates the key features associated with PE seen in autoantibody-injected pregnant mice in vivo. Both the mouse and human studies reported here provide strong evidence that AT1 receptor activation by the autoantibody induces TNF-α and that increased TNF-α production may be an underlying mechanism contributing to the pathophysiology of the disease.
While TNF-α is reportedly increased in the circulation of preeclamptic women37-39
, the exact cause of increased cytokine production is unknown, as is its pathogenic role. Multiple in vitro
studies demonstrate that increased inflammatory cytokine production may lead to endothelial dysfunction, increased placenta apoptosis, decreased angiogenesis and kidney abnormalities that are relevant to the pathophysiology of PE40-42
. There are few animal models of PE available and none of them have delineated the cause of increased TNF-α and its pathogenic role. Here, using a novel autoantibody-induced model of PE in pregnant mice, we demonstrate that autoantibody-mediated AT1
receptor activation induces TNF-α, and that its production through this mechanism is pregnancy-dependent. Since IgG purified from normotensive pregnant women did not elicit the same increase, the effect can be attributed to the autoantibody itself and not a non-specific immunologic response.
Next, we found that TNF-α blockade attenuates AT1
-AA-induced preeclamptic features in autoantibody-injected pregnant mice, including hypertension and proteinuria. This finding indicates that anti-TNF-α antibody treatment decreases cytokine induction in autoantibody-injected pregnant mice. We believe that without interference, TNF-α-induced cell damage and inflammation create a detrimental cycle, facilitating further cell damage and inflammation. However, in the presence of an anti-TNF-α antibody which neutralizes TNF-α effects, this damage is decreased, slowing the malicious cycle. Thus, we have revealed that AT1
-AA is a key mediator in inducing the increased TNF-α in PE and blockade of this cytokine can attenuate disease features. In fact, similar to the effects of anti-TNF-α treatment in our AT1
-AA-injected pregnant mice, a soluble TNF-α receptor also attenuates preeclamptic-like features seen in pregnant rats generated by reduced uterine placental perfusion (RUPP)43
. Thus, both of these animal studies provide strong preclinical evidence to support the novel therapeutic possibility of targeting this deleterious cytokine associated with PE.
It is well-established that ANG II can act through the AT1
receptor to increase TNF-α21-27
. In this way, the autoantibody may regulate the secretion of TNF-α resulting in maternal symptoms. Although this potential role of TNF-α in preeclamptic hypertension and proteinuria has been suggested, the pathogenic mechanisms underlying its effects are not clearly identified. Earlier studies have shown that the pro-inflammatory TNF-α is associated with both vascular damage and hypertension44
. Jovinge et al.
have shown that TNF-α-deficient mice have reduced atherosclerotic lesions, suggesting that the cytokine plays a key role in vascular injury45
. Similarly, in salt-sensitive rats, TNF-α blockade has been successful in alleviating both the hypertension and renal damage observed in this model46
. In pregnant rats, TNF-α enhances contraction and inhibits endothelial nitric oxide-cGMP-mediated relaxation in systemic vessels, which could contribute to hypertension47
. Chronic infusion of TNF-α into pregnant rats to achieve two-fold increase in concentration is sufficient to induce hypertension and increase endothelin-1 production, which the authors believe contributes to the vascular damage associated with the maternal symptoms of PE48
. These examples illustrate that the inflammatory properties of TNF-α contribute to vascular damage and high blood pressure, which could therefore do the same in PE. In addition, Muller et al.
report a double transgenic rat model with increased levels of circulating ANG II which exhibits hypertension, renal dysfunction as well as increased TNF-α49
. In this model, the authors believe that increased TNF-α contributes to kidney injury via complement activation and that excess ANG II sensitizes the vasculature to the effects of the cytokine. The induction of TNF-α in the autoantibody-injection model of PE is accompanied with an autoantibody-mediated increases in sFlt-19, 14
. Others have also shown that sFlt-1 and sEng are induced by TNF-α30, 50
. In conjunction with these studies, the results of the PE animal model reported here provide evidence to support the novel concept that autoantibody-mediated AT1
receptor activation induces TNF-α production resulting in the maternal features of PE.
It should not be overlooked that AT1
-AA alone may contribute directly to certain features of PE which are independent of TNF-α. For example, the autoantibody can directly stimulate the AT1
receptors of vascular smooth muscle cells and induce vasoconstriction51-53
. Likewise, the autoantibody could activate AT1
receptors on endothelial cells resulting in the synthesis of endothelin-1, a powerful vasoconstrictive agent54, 55
. The autoantibody may also directly bind to AT1
receptors on renal mesangial cells to induce PAI-1 secretion13
. Therefore, it is not surprising that TNF-α blockade only partially relieves autoantibody-induced features of PE, including the partial attenuation of hypertension and proteinuria observed in the pregnant mice co-injected with the autoantibody and an anti-TNF-α antibody (). However, it is clear through the evidence presented here that reducing TNF-α significantly attenuates the key preeclamptic symptoms initiated by AT1
-AA in pregnant mice, indicating an important role for this cytokine which warrants further investigation.
Decreasing the amount TNF-α circulating in preeclamptic mice may both directly and indirectly alleviate many disease features. In the placenta, decreasing TNF-α production may directly reduce the amount of trophoblast apoptosis and result in a healthier organ. By limiting placental damage, reductions in TNF-α may decrease the release of key anti-angiogenic factors, sFlt-1 and sEng. With little increase in these factors, the subsequent maternal vascular and renal damage may be alleviated, thereby reducing maternal symptoms. As mentioned earlier, TNF-α is capable of inducing vascular injury through the initiation of inflammatory cascades. Should this pathway not be instigated, then the endothelial damage associated with PE may not be as severe, and the symptoms may be lessened. Together, these scenarios indicate that TNF-α may be either directly or indirectly contributing to preeclamptic features and its blockade can reduce their severity.