Increased level of compleasomes in cerebrospinal fluid of patients with herpes simplex encephalitis

Herpes simplex encephalitis (HSE), most often caused by herpes simplex virus type 1 (HSV-1), induces an acute focal, necrotizing inflammation with a predilection for the frontotemporal regions of the brain. Although acyclovir treatment has been reported to reduce mortality (Raschilas et al. 2002; Sili et al. 2014; Skoldenberg et al. 1984), a large proportion of surviving patients suffer from permanent neurological sequelae, including cognitive, memory and behavioural impairments (Raschilas et al. 2002), as well as a high risk of developing subsequent epilepsy (Hjalmarsson et al. 2007). The acute intrathecal inflammatory response may be followed by long-term inflammation (Aurelius et al. 1994; Aurelius et al. 1993; Lind et al. 2017; Skoldenberg et al. 2006), which may be part of the destructive process leading to neurological damage in HSE survivors. Recently, the complement system, which is a part of the innate immune response, was found to be activated in HSE. The complement factors C3a, C3b, C5 and C5a were increased in the cerebrospinal fluid (CSF) from HSE patients compared with healthy controls, and C3a and C5a concentrations remained increased at long-term follow-up (Eriksson et al. 2016). The innate immune response aims to suppress viral replication and cell death and also enhances adaptive immunity (van Beek et al. 2003; Veerhuis et al. 2011). In addition, the proteasome system has emerged as regulator of innate immune responses and different types of proteasomes have been detected in the CNS (Ramachandran and Margolis 2017). In CSF, extracellular proteasomes have been suggested to play a role in neurodegenerative inflammatory diseases of the CNS (Mueller et al. 2012). One component of the regulatory part of the proteasome is antisecretory factor 1 (AF1, also named S5a/rpn10/PSMD4) (Tomko Jr and Hochstrasser 2013). The 43-kD ubiquitously expressed AF1 protein has been shown to exert a potent antisecretory and anti-inflammatory effect (Davidson and Hickey 2004; Johansson et al. 1995; Lange and Lonnroth 2001). The 16-amino acid-long sequence 36–51 has been designated AF-16 and on account of its stability and potency is used in experimental work (Johansson et al. 1997; Matson Dzebo et al. 2014; Nicolas and Lievin-Le Moal 2015). Intranasally administered AF-16 in rats experimentally infected with HSV-1 has been demonstrated to abolish sickness and death during time course of infection (Jennische et al. 2008), most likely accomplished by an AF-16-mediated absence of increased intracranial pressure (ICP) ubiquitously found in HSE-affected rats. Decreasing of a high ICP in infected rats has also been achieved by a diet-induced rise in endogenous AF1 (Johansson et al. 2013). The role of a high ICP is scarcely investigated in human HSE, but since a high ICP leads to decreased consciousness which is associated to poor prognosis in HSE (Singh et al. 2016), a high ICP might be a bad prognostic factor (Jouan et al. 2015; Barnett et al. 1988) and a target for therapy.

We have previously described that during the initial phase of infection, a complex is formed between circulating proteasomes and complement factors, which we have named the “compleasome” (Lonnroth et al. 2016). During formation of compleasomes, the C3 complement factor is split into C3c and C3d. The proteolysis of C3 into C3c has been described as deactivation of complement activity (Gros et al. 2008). This split changes the proteasome conformation, resulting in an exposure of previously hidden antisecretory epitopes. Thus, rats infected with HSV-1 for experimental encephalitis studies were shown to have induced compleasomes in CSF (Lange et al. 2016). However, no hydrolyses of C3 or exposure of AF occurred in that complex. This was probably because, for ethical reasons, the rats were sacrificed before clinical symptoms developed (Lange et al. 2016). Since AF-16 abolished HSE in HSV-1 infected rats, it is of interest to study the benefit formation of AF compleasomes in human CSF during HSV-1 infection.

The aim of the present study was to investigate the formation of compleasomes in human CSF in response to HSE, and to study this response over time.

In the present work, we demonstrate an increased CSF compleasome concentration in response to HSE, especially during the acute phase of the disease. The early formation of compleasomes was followed by declining compleasome levels over time. This kind of kinetic response suggests an important role for this specific protein complex in the intrathecal innate host defence against HSE infection.

This early formation of compleasomes in human CSF is similar to the one we found in our previous investigation of HSV-1 rat encephalitis. In that study, we demonstrated an early response of proteasome subunit 20Sα6 compleasomes in CSF, which did not respond to AF1 antibodies (Lange et al. 2016). In the rat model, the response after the acute phase was not investigated, since the rat encephalitis was lethal. By contrast, the present study of human CSF revealed a significant AF1 compleasome response during or after the HSV-1 acute phase of the disease. This result suggests that a total conformational change had occurred in the compleasome as a result of the HSV-1 infection. Our previous in vitro experiments suggest that AF1-reactive compleasomes are formed in a later phase of the disease, in close correlation to a C3 complement degradation (Lonnroth et al. 2016). This theory was confirmed in the present study. Thus, the HSV-1 infection induced a considerably stronger compleasome formation in the acute phase compared with a later stage of the disease. This fast reactivity against the infection can be interpreted as a most essential part of the innate immune response with the purpose of counteracting infections in the brain.

The compleasome levels were also enhanced in CSF collected from PCs compared with HCs. The PCs had undergone lumbar puncture for a suspected clinical CNS infection. However, CNS infection was ruled out because of lack of CSF pleocytosis. These PCs all had headache and other unspecific symptoms, which might have been elicited by an infection of unknown origin, without specific involvement of the meninges or the neuronal brain tissue, and these symptoms with unknown aetiology had induced compleasome formation. However, the compleasome formation was lower in the PC group compared with the HSE group.

The formation of compleasomes declined over time. Ten days after the onset of neurological symptoms was arbitrarily chosen, but the time point generally corresponds well to the disappearance of acute symptoms. Levels of HSV DNA in CSF usually start declining at 10–21 days (Schloss et al. 2009). There is probably a continuous development of the immunological processes in the brain, and the time points at which lumbar puncture was performed may therefore reflect different stages of the disease. It is well known that prodromal symptoms, sometimes lasting for up to a week, are present in approximately half of HSE cases, while the onset of disease is more abrupt in others (Kennedy et al. 1988). We can, however, conclude that compleasome formation significantly and continuously decreases after the initial month of HSE disease.

The acute CSF compleasome levels were higher in patients with good recovery (GOS 5) compared with patients with worse clinical outcome (GOS 1–4). This suggests that a rapid and strong formation of the innate compleasome complex may be important in the control of brain HSV-1 infection. The retrospective design of this study and analysis of frozen CSF samples imply limitations, i.e. lack of standardised timing of lumbar punctures and selected bias when using stored samples. However, the acute and late CSF samples in this study were from patients with both good and bad outcome, since the patients had previously participated in studies with long-term follow-up. The clinical data with outcome measurements (GOS) after 6 months were accurately documented in the patient’s records. Intrathecal complement activity has been examined in other CNS infections such as cryptococcal meningitis, although no test of correlation with the clinical outcome was performed (Shen et al. 2017). In a previous study, the pro-inflammatory anaphylatoxins C3a and C5a were significantly increased in CSF samples from HSE patients in the acute phase. The concentrations of C3a, C3b, C5 and C5a were, however, also increased in the late phase of infection. Thus, complement activity contributes to inflammation during both the acute and the later phase of HSV-1 infection (Eriksson et al. 2016). An earlier reduction in C3a activity was associated with better clinical outcome, measured by GOS, after 6 months (Eriksson et al. 2016). These results suggest that some inflammatory responses may contribute to CNS protection, while other components of the inflammatory response may be associated with brain tissue damage.

Suppression of the immune responses is essential for reducing incidence of late detrimental intrathecal inflammation. Therefore, the kinetics of the different parts of the immune responses in the brain must be defined and separated. When this has been accomplished, the most appropriate time point for immunosuppressive treatment can be determined. In mice, one study showed that early administration of corticosteroids was not at all beneficial compared with administration 3 days later. This finding indicates that selecting the opportune time point for immunosuppressive treatment during the disease course is essential when aiming to regulate different immune responses to HSE (Sergerie et al. 2007).

The described effects of AF1 and of its complex relation to complement factors suggest that AF compleasomes are part of the innate immune system. The N-terminal residues 5–188 of the sequence coding for AF1 contain a motif related to the von Willebrand factor A (vWA) domain. It has been described elsewhere that proteins incorporating von Willebrand factor (vWF) motifs participate in several biological events via interaction with other proteins (Edwards and Perkins 1996). Complement factor 3 plays a central role in the complement system via conformational changes to activated fragments, resulting in C3c, the physiological down-regulation product of C3 (Gros et al. 2008). Thus, vWF acts as a cofactor for factor I-mediated cleavage of C3b to inactive iC3b, thereby shutting down complement activation (Feng et al. 2015). Noone et al. suggest that vWF has a protective effect on endothelial cells and complement-mediated injury by functioning as a complement regulator (Noone et al. 2016).

The present results demonstrate that AF compleasomes are formed in CSF during HSV-1 infection. The levels of compleasome in CSF were correlated to the number of days after onset of symptoms; at acute phase of infection, the HSE patients demonstrated higher levels of CSF compleasomes compared with the later stage of infection. The rapid formation of compleasomes in response to the HSV-1 infection suggests an important role in the host defence immune system of this specific complex.