Targeting Brain Endothelial Gasdermin D: A Shortcut to Remodel the Blood-Brain Barrier

The blood-brain barrier (BBB), an intricate interface of the neurovascular unit, plays a vital role in maintaining the function of the central nervous system (CNS). The BBB interacts with various components, such as glial cells, circulating cells, exosomes, and hormones, to regulate communication between the CNS and the periphery [1].

BBB injury accompanies brain-damaging events, such as infections, cerebrovascular diseases, and neurodegenerative disorders [2]. However, previous research has focused its spotlight only on the BBB alterations in cerebral injuries, including septic encephalopathy, stroke, and cognitive decline. In these scenarios, the BBB always shows dissolution of tight junctions, increase of permeability, and infiltration of immune cells, all difficult to reverse. However, our understanding of how BBB injury is initiated, or what “licenses” the injury, remains vague. This knowledge gap hampers the development of preemptive interventions to mitigate BBB injury, especially that with a high propensity to irreversibility.

Brain endothelial cells (bECs), critical building blocks of the BBB, are sensitive to external stimuli [3, 4]. The BBB can be broken down during pathogen infections and systemic inflammation, especially in septic encephalopathy caused by bacteria. Lipopolysaccharide (LPS), a component of the outer wall of gram-negative bacteria, can be recognized by the host cell using its pattern recognition receptor TLR4/MD2 complex, which activates the downstream transcription of genes encoding cytokines and chemokines via NF-κB and IRF3. It has been established that LPS induces the secretion of pro-inflammatory factors to dissolve tight junctions and disrupt BBB integrity. However, BBB disruption has not been reported in mice receiving retro-orbital injection of LPS-free tumor necrosis factor or interleukin-6. Pyroptosis, a form of programmed lytic cell death proposed in recent studies, is closely implicated in innate immunity, inflammation, and infection. Gasdermin D (GSDMD), as the primary executive protein, mediates pyroptosis under inflammatory and infectious conditions. As confirmed in numerous studies, GSDMD can be activated by mouse caspase-11 or human caspase-4 and caspase-5, a mechanism possibly contributing to LPS-induced mortality in mice and bacterial infection-induced sepsis in clinical settings [5, 6].

In the present article contributed by Wei et al. [7], an endotoxic shock model was established based on mice induced with a single, high dose of LPS, in conjunction with single-gene knockout (Tlr4^-/-, Caspase-11^-/-, or Gsdmd^-/-). The experimental results demonstrated that TLR4, caspase-11, and GSDMD in bECs are required for BBB breakdown in mice. However, Tlr4^-/- mice become sensitive to BBB disruption by LPS under pretreatment with the TLR3 agonist polyinosinic-polycytidylic acid (poly I:C), which has primed cytosolic caspase-11 expression prior to LPS challenge. Interestingly, the BBB in Caspase-11^-/- or Gsdmd^-/- mice remains resistant after LPS induction in poly (I:C)-primed mice, reminding us that intracellular LPS disrupts the BBB via caspase-11-mediated activation of GSDMD, while this process relies on TLR4-mediated caspase-11 priming. However, TLR4-mediated caspase-11 priming is irrelevant to BBB disruption in humans, because human caspase-4 (a homolog of mouse caspase-11) is constitutively expressed.

LPS-binding protein (LBP) and the CD14 receptor facilitate the presentation and endocytosis of intracellular LPS. Notably, LBP and CD14 knockout mice show no response to LPS-induced BBB disruption in poly(I:C)-primed mice, suggesting that LBP-CD14-mediated transport/endocytosis may be necessary for the activation of the intracellular caspase-11-GSDMD pathway by mediating LPS endocytosis. Moreover, single-cell sequencing revealed that during systemic inflammation, both caspase-11 and GSDMD are highly expressed in bECs, rather than in other cell types that constitute the BBB. Electron microscopy further captures the ultrastructural damage in the LPS-disrupted BBB, such as pyroptosis of endothelial cells, detachment of the endothelial layer, enlargement of the perivascular space, and parenchymal edema. Selective knockout of GSDMD in bECs, as well as comprehensive AAV-based rescue of GSDMD deficiency, underscores that brain endothelial GSDMD activation allows the LPS-caused BBB breakdown. This discovery is groundbreaking in that it demonstrates that GSDMD-mediated membrane perforation and the subsequent bEC pyroptosis contribute to BBB disruption during gram-negative bacterial infection. Recent findings [8], along with this study, confirm the involvement of pyroptosis in various neurological diseases, underscoring the urgency to elucidate its mechanisms.

GSDMD is primarily known for its involvement in pyroptosis [9]. New chapters remain to be written about its effects on pore formation in the BBB, particularly in cerebral vascular endothelial cells. In this study, the authors propose that pores form in the BBB upon activation of GSDMD alone, without the need for LPS stimulation, suggesting that GSDMD can perforate the BBB in a direct and cytokine-independent manner. These pores compromise the integrity of the endothelial cells in the BBB, thus making it more permeable. This finding expands the gallery of BBB breakdown mechanisms, in which disrupted tight junctions and enhanced endocytosis/transcytosis have already been well portrayed. A higher BBB permeability allows harmful substances to enter and evoke inflammation in the brain. On the other hand, in scenarios where the BBB is intact, GSDMD agonists can temporarily “open” the BBB, facilitating the entry of therapeutic agents into the diseased brain. This study stands out by shifting its focus from cytokine-regulated immune responses, long considered mainstream BBB breakdown mechanisms, to a new one involving cellular disruption directly mediated by GSDMD.

This study fathomed the potential of GSDMD as a therapeutic target for BBB-related pathologies. VHH_GSDMD-1, an inhibitory nanobody that targets the human GSDMD-N domain to neutralize its pore-forming activity, is delivered by AAV-BI30 virus to bECs. This approach effectively blocks the BBB breakdown induced by LPS and inhibits the BBB destruction caused by Klebsiella pneumoniae infection in CASP4^Tg mice, providing a strategy for protecting the BBB during septic and bacteremic conditions (Fig. 1). The translational importance of this research is further accentuated by the efficacy of a GSDMD-neutralizing nanobody in preventing the BBB disruption triggered by LPS and by infection. In addition, GSDMD-neutralizing nanobodies may also be applied to manage CNS diseases complicated by BBB breakdown.

Fig. 1.

Brain endothelial-derived GSDMD regulates BBB homeostasis. Under normal conditions, endothelial cells, the main components constituting the BBB, maintain extremely low permeability and homeostasis. However, during infections and systemic inflammation, LPS secreted by bacteria promotes intracellular LPS presentation and endocytosis through LBP and the CD14 receptor, thus activating the caspase-4/11 GSDMD pathway in bECs. Through GSDMD-mediated pore formation and subsequent pyroptosis, the BBB exhibits a higher permeability, and its function is disrupted. However, with the use of drugs targeting the GSDMD-N domain, these changes are effectively blocked and the BBB is remodeled. The dotted lines indicate the potential role of CD14 in LPS endocytosis. Solid lines indicate the cleavage of the N-terminal and C-terminal domains of the GSDMD protein by caspase-4/11, release of the membrane-binding N-terminal domain, and pore formation, thereby disrupting the cell membrane and triggering pyroptosis.

However, it is important to note that neuroinflammation has been recognized as a double-edged sword for brain injuries and rehabilitation. In the early stages of acute brain injury, inflammatory cells such as microglia are recruited to the injury site, releasing various inflammatory mediators, such as cytokines and chemokines, which help to limit the extent of the injury [10]. Therefore, although in an inflammatory context, GSDMD signaling-mediated pyroptosis has a pathogenic role in disrupting the BBB and mediating brain damage, conversely, its mediated natural immune response may play a key role in anti-infection as well. However, excessive or persistent inflammation may lead to secondary brain injury, further aggravating neuronal death and dysfunction. For this reason, inhibiting the activity of GSDMD in some pathological processes may help to protect the BBB and reduce the severity of brain injury [7]. Conversely, moderate activation of GSDMD may help clear damaged tissue and facilitate the repair process. Hence, the role of GSDMD in brain endothelial cells is not limited to causing BBB disruption but also involves complex inflammatory responses and pyroptotic responses in various pathological contexts. Therefore, future studies are needed to further understand the dual role of GSDMD-mediated pyroptosis in different pathological processes, with the hope that precise therapeutic strategies can be developed to improve the prognosis of patients with various brain injuries.

In conclusion, this research provides a new lens through which to look into the molecular mechanisms underlying BBB breakdown during inflammation. These findings may be exploited to design targeted therapies protecting against BBB disruption and related consequences in various CNS diseases. These findings also lay a solid foundation for future studies investigating the molecular and cellular mechanisms of BBB disruption during inflammatory events and the development of targeted therapies.