Critical role of TNF-α in cerebral aneurysm formation and progression to rupture

Cerebral aneurysm rupture is associated with significant morbidity and mortality [1, 2]. A significant number of patients may be treated with microsurgery or endovascular coiling, but intervention is not without the risk of neurological injury [1, 2]. A large number of patients are followed clinically as they are deemed either at lower risk of hemorrhage or at high risk for treatment. Even in these cohorts, a significant number of patients will go on to have aneurysmal rupture or receive treatment for aneurysm progression despite originally being considered high risk for intervention or low risk of rupture [2, 3]. Medical therapy that stabilizes aneurysmal progression or rupture could be beneficial for a significant number of patients. Currently, there are no pharmacological alternatives for patients with cerebral aneurysms.

Inflammation has been implicated in the pathogenesis of intracranial aneurysm formation and rupture [4]. Alterations in TNF-α have been associated with cerebral aneurysm in humans [5, 6], but a direct role in aneurysm formation or rupture has not been defined. TNF-α is a critical member of the immune system [7] and produces pro-inflammatory alterations in key cells implicated in cerebral aneurysms including macrophages, endothelial and smooth muscle cells [8‐10].

The goals of the present study were to: 1) assess the direct role TNF-α in a model of cerebral aneurysm formation; 2) determine if there are alterations in TNF-α expression in cerebral aneurysm formation and rupture; 3) evaluate if TNF-α inhibition decreases the incidence of aneurysm formation; and 4) test whether TNF-α inhibition after cerebral aneurysm formation may lead to aneurysm stabilization and inhibition of rupture.

This study was carried out in accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Research Council [11]. The protocol was approved by the Institutional Animal Care and Use Committee of Thomas Jefferson University (Philadelphia, PA, USA). Cerebral aneurysms were induced in 8- to 10-week-old male TNF-α gene null (TNF-α −/−) mice (on C57BL/6 J background) or their wild type controls (Jackson Laboratory, Bar Harbor, ME, USA) using previously described methods [9, 12, 13] with alterations as herein described.

To induce hypertension, mice underwent nephrectomy followed by implantation of deoxycorticosterone acetate pellet (Innovative Research of America, Sarasota, FL, USA) 1 week later [14]. On the same day as deoxycorticosterone acetate pellet implantation, animals were started on water containing 1% NaCl (Sigma-Aldrich, St Louis, MO, USA) to induce hypertension [9, 12‐14] and 0.12% beta-aminoproprionitrile (BAPN; Sigma-Aldrich) to reduce collagen cross-linking [15]. Elastase (Sigma-Aldrich) was prepared in sterile PBS (Sigma-Aldrich). Mice underwent a single stereotactic elastase injection (35 mU) into the cerebrospinal fluid at the right basal cistern on the same day as pellet implantation [9, 12, 13]. Sham control mice received a single stereotactic injection of PBS. A single stereotactic injection of dye was performed for every 10 mice to ensure accurate needle placement. Animals were assigned to the sham or aneurysm induction cohorts randomly in an alternating fashion.

Blinded daily neurological examination was carried out using a previously described method [13, 16‐18]. Neurological symptoms were graded: 0, normal; 1, decreased drinking or eating with associated weight loss >2 g of body weight (approximately 10%) over 24 hours; 2, flexion of the torso and forelimbs on lifting of the animal by the tail; 3, circling to one side with a normal posture at rest; 4, leaning or falling to one side at rest; 5, no spontaneous activity. Mice were euthanized when they developed neurological symptoms (score 1 to 5). All asymptomatic mice were euthanized 28 days after aneurysm induction. The brain samples were perfused with PBS followed by a gelatin (Sigma-Aldrich) containing blue dye to visualize cerebral arteries as well as to assess for aneurysm formation and subarachnoid hemorrhage (SAH). Aneurysms were defined as a localized outward bulging of the vascular wall whose diameter was greater than 1.5 times the parent artery diameter by two independent observers blinded to the animal cohort [12, 13]. Animal cohorts were not revealed until all experimental groups had been sacrificed.

Systolic blood pressure was measured by the tail-cuff method with the BP-2000 Blood Pressure Analysis System (Visitech Systems, Apex, NC, USA) after 3 days of training to allow for acclimation and then before aneurysm induction surgery (day 0) and every week until day 28 after surgery [19].

Cerebral aneurysm rupture leads to disability or death in the majority of patients [1, 2]. Treatment is also associated with significant morbidity and mortality, particularly in high risk aneurysms or patients [1, 2]. Although intervention is controversial in select patients with unruptured aneurysms, studies have found that a large number of patients deemed low risk of hemorrhage or high risk for treatment may go on to receive treatment for aneurysm progression or experience SAH [2, 3]. Currently, there are no medical therapies in clinical practice to stabilize aneurysmal progression or prevent rupture. In this study we have found that the expression of the pro-inflammatory cytokine TNF-α is increased in a mouse model of cerebral aneurysms. This expression is increased in unruptured cerebral aneurysms and furthermore in ruptured aneurysms. The incidence of cerebral aneurysm formation and rupture was decreased in TNF-α knockout mice and following pre-treatment with a synthesized TNF-α inhibitor. The TNF-α inhibitor also resulted in aneurysm stabilization and decreased rupture after aneurysm formation.

TNF-α has also been found to be elevated in humans with ruptured cerebral aneurysms [5, 6], but mechanisms behind TNF-α activation in cerebral aneurysms are currently unclear. A number of environmental factors associated with cerebral aneurysm formation have been implicated in TNF-α activation. We have previously found that induced hypertension in conjunction with hemodynamic stress in rats in vivo can increase the expression of TNF-α [8]. We have also found that TNF-α is upregulated following exposure of cultured cerebrovascular smooth muscle cells to cigarette smoke [22, 24], and others have found increased expression of TNF-α in blood vessels following cigarette exposure [25]. Additional risk factors associated with cerebral aneurysms, including aging, gender, and alcohol, have also been associated with TNF-α expression [4‐6].

Environmental factors may activate TNF-α in both macrophages and smooth muscle cells. In this study we have found that TNF-α co-localized to both macrophages and smooth muscle cells. We have previously found that hypertension, hemodynamic stress, and cigarette smoke may activate TNF-α and induce phenotypic modulation in smooth muscle cells [8]. Although the downstream mechanisms by which TNF-α contributes to cerebral aneurysm formation are not completely defined, inflammation is a significant element behind the pathogenesis of cerebral aneurysm formation [4], and TNF-α is a significant pro-inflammatory immune modulator. We have previously found that TNF-α can activate a number of pro-inflammatory and matrix remodeling genes in cerebral vascular smooth muscle cells directly implicated in cerebral aneurysm formation, including: MMP-3, MMP-9, MCP-1, VCAM-1, and IL-1β [8]; and IL-1β, VCAM-1 and MCP-1 [9, 26, 27]. Specifically, increased MCP-1 may attract macrophages leading to increased TNF-α formation, and macrophages have previously been found to be critical in the formation of cerebral aneurysms [9]. Additionally, MCP-1 has been found to be increased in aneurysm walls and MCP-1 knockout mice had lower levels of matrix metalloproteinases (MMPs), and a decreased incidence of aneurysm formation [26]. MMPs degrade the extracellular matrix [28] and have been found to be increased in human cerebral aneurysms [29]. Inhibition of MMPs has also been shown to decrease the incidence of aneurysm formation and progression in animals [12, 13]. Although there are likely many further key mediators, TNF-α may contribute to cerebral aneurysm formation through activation of pro-inflammatory and matrix remodeling genes and recruitment and activation of inflammatory cells.

The role of TNF-α in aneurysm progression and rupture is unknown. Increased expression of TNF-α in human [5, 6] and mouse ruptured cerebral aneurysms may be a reflection of the inflammatory response following rupture rather than a mechanism leading to aneurysm instability and rupture. TNF-α is a key initiator of apoptosis [30, 31] and its pro-apoptotic protein target (Fas-associated death domain) have been increased in human cerebral aneurysms [5]. We have previously found that TNF-α triggers apoptosis in a dose-dependent fashion in cultured cerebrovascular smooth muscle cells [8], and this may contribute to both aneurysm progression and rupture through loss of focal cerebrovascular contractility. This may be explained by the co-localization of TNF-α to smooth muscle cells in cerebral aneurysms found in mice and ultimately the loss of smooth muscle cells in unruptured and furthermore ruptured aneurysms. TNF-α-induced macrophage infiltration and phagocytosis may also contribute to this process [26, 32] as the quantity of apoptotic cells has been associated with cerebral aneurysm rupture [33, 34] and macrophage infiltration was increased furthermore in ruptured versus unruptured aneurysms.

Additional experiments are indicated both in vivo and in humans to assess the role of TNF-α in cerebral aneurysm progression and rupture. Further in vivo experiments following upregulation of TNF-α both in wild type and TNF-α knockout mice would be beneficial to further clarify the molecular mechanisms downstream of TNF-α upregulation in aneurysm formation and rupture. Limitations of this study include detection of aneurysm rupture through assessment of alterations in neurological examination, which may fail to detect subclinical or asymptomatic hemorrhages. Despite this drawback, examination of brains in asymptomatic mice did not reveal signs of significant SAH. Additionally, although DTH has been found to be a specific inhibitor of TNF-α synthesis [20, 21], it may have additional properties that contribute to inhibition of aneurysm formation and rupture.