Neutrophils, the Felons of the Brain: Can they be rehabilitated to yield benefit after stroke?

Polymorphonuclear neutrophils (PMNs) are short-lived but powerful immune cells, providing an early and robust inflammatory response following tissue damage. While the defensive role of PMNs in anti-microbial immunity has long been studied, recent attention has focused on the role of these cells in sterile inflammation after tissue injury.

Following stroke, PMNs are rapidly recruited to the injured brain. PMNs have been hypothesized to worsen stroke pathology via several mechanisms, including (1) physical blockade within the microvascular network, further reducing cerebral blood flow and (2) direct entry into the brain parenchyma, followed by the release of granules containing antimicrobial enzymes and chemical species that could further injury brain tissue.^1–3 Based on these assumptions, the prevention of PMN entry into the brain after stroke has been extensively studied as a therapeutic target.^3–5 Although PMN suppressing therapies showed benefit in numerous pre-clinical studies, subsequent clinical trials in stroke patients showed no overall benefit.^{3, 6}

Recently, a growing body of evidence has suggested that PMNs, like other immune cells, may exhibit some level of “functional plasticity”, analogous to Th1/Th2 (for T cells), M1/M2 (for macrophages/microglia; MΦ).⁷ The N1 PMN phenotype refers to PMNs with stronger pro-inflammatory/oxidative properties, which possess effective anti-tumor and anti-microbial actions and may cause greater detrimental effects in the stroke-affected brain. Assuming that the majority of circulating neutrophils are normally in the N1 state, to maximize their ability to effective engage in anti-microbial activities, the PMNs entering the brain early after stroke may be primarily detrimental. Conversely, the N2 PMN phenotype is characterized by reduced pro-inflammatory properties and a higher content of beneficial molecules.^7–11

As such, it is possible that the effect of PNMs on stroke outcome could depend on N1/N2 ratio, which may change with time after stroke, reducing the detrimental effects of infiltrating PMNs – or potentially even conferring benefit to the damaged tissue. Potential mechanisms of PMN-mediated benefit in stroke may involve (1) modification of other cells like macrophages (MΦ) and microglia to an anti-inflammatory “healing” phenotype, (2) PMN self-limitation of pro-inflammatory factors, or (3) direct secretion of beneficial factors. Some recent examples of these PMN behaviors after stroke is outlined below.

Mature segmented neutrophils egress out of bone marrow (BM) to the blood circulation where they have very short half-life (13–19h),¹² meaning that all circulating PMNs are normally replaced within approximately one day of their release. After tissue injury, including stroke, PMNs are recruited to the side of injury within hours, and continue for days, where they release some of their granule contents. Typically, PMNs then die via apoptosis and are consequently removed via MΦ-mediated efferocytosis (phagocytosis-mediated engulfment of apoptotic cells), which is essential to prevent PMN secondary necrosis with subsequent release of cytotoxic and pro-inflammatory PMN content¹³.

The efferocytosis of PMNs has been shown to induce an anti-inflammatory phenotype in the microglia and MΦ that phagocytose them.¹³ Thus, PMN death through apoptosis is believed to act as a signal for MΦ to acquire the M2 phenotype that is essential for efficient phagocytosis and improved healing.^13–15 PMNs have also been shown to phagocytose themselves, a type of “cannibalism” which results in increased production of TGFβ by the engulfing PMN.¹⁶ TGFβ is a cytokine that acts as an inducer of N2 polarization⁷ and has been shown to reduce neuroinflammation and improve recovery after ICH.¹⁷ Thus, the proper clearance of apoptotic PMNs may assist in the resolution of inflammation and the promotion of tissue repair.

Activation of the peroxisome proliferator-activated receptor-γ (PPARγ) promotes the conversion of MΦ to a “healing” (M2) phenotype.¹⁴ Moro et al. recently showed that mice subjected to cerebral ischemia and treated with a PPARγ agonist, rosiglitazone, experienced enhanced PMN infiltration into the brain, and increased proportion of N2 (Ym1+; a prototypic marker of M2 phenotype) anti-inflammatory PMNs and reduced infarct volume.¹⁸ Importantly, Ym1+ PMNs were more effectively phagocytosed by MΦ, and systemic depletion of PMNs prior to stroke abolished the neuroprotective effect of PPARγ agonist treatment.

Under in vitro conditions, PMNs stimulated with pro-inflammatory TNFα or LPS produce and secrete anti-inflammatory (s)IL-1RA (natural inhibitor of the pro-inflammatory IL-1β) at much higher rates than they produce IL-1β,^{19, 20} suggesting a potential mechanism for self-limitation of the pro-inflammatory PMN response. Interestingly, Faraci et al. showed that pre-treatment with LPS reduced ischemic brain injury in mice despite the increased presence of PMNs in the brain.²¹

PMN precursor proliferation and maturation takes place in the BM, where the chemical composition of PMN granules is established by local environmental cues. We have recently shown that IL-27 generated in response to intracerebral hemorrhage (ICH) can modify the production of granule components in maturing BM PMNs.¹¹ Excitingly, we have shown that IL-27 may down-regulate PMN levels of tissue-damaging enzymes (NADPH oxidase, iNOS and MMP-9) and up-regulate the production of potentially beneficial molecules for ICH resolution, including iron-sequestering lactoferrin (PMNs are the primary source of blood lactoferrin) and hemoglobin-neutralizing haptoglobin. Both lactoferrin and haptoglobin have potent protective effects in the ICH-injured brain.^{5, 20} We proposed that PMNs modified by IL-27 may be less damaging or even beneficial in the brain in the later stages of ICH. In agreement with this hypothesis, we found that depletion of PMNs 24 hours after ICH worsened functional outcome in mice.

In conclusion, although PMNs are typically thought of as detrimental in sterile inflammation, we believe that PMNs can assist in the resolution of inflammation under certain conditions. The newly reported plasticity of neutrophils is particularly exciting, as their short-lived and early-responding nature makes them ideal candidates for enhancing tissue repair and regeneration. Future studies should be conducted to determine whether PMNs clearance or function can be beneficially manipulated to enhance the resolution of inflammation and improve functional outcome in sterile inflammatory diseases such as stroke.