Simian Varicella Virus Infection and Reactivation in Rhesus Macaques Trigger Cytokine and Aβ40/42 Alterations in Serum and Cerebrospinal Fluid

Simian varicella virus (SVV) produces peripheral inflammatory responses during varicella (primary infection) and zoster (reactivation) in rhesus macaques (RM). However, it is unclear if peripheral measures are accurate proxies for central nervous system (CNS) responses. Thus, we analyzed cytokine and Aβ42/Aβ40 changes in paired serum and cerebrospinal fluid (CSF) during the course of infection. During varicella and zoster, every RM had variable changes in serum and CSF cytokine and Aβ42/Aβ40 levels compared to pre-inoculation levels. Overall, peripheral infection appears to affect CNS cytokine and Aβ42/Aβ40 levels independent of serum responses, suggesting that peripheral disease may contribute to CNS disease.


Introduction
Varicella-zoster virus (VZV) is an alphaherpesvirus that exclusively infects humans, with a global infection rate of nearly 90% (Gershon et al. 2015).Initial exposure to VZV leads to primary infection, characterized by a disseminated rash known as varicella (chickenpox), during which the virus infects and establishes latency in cranial and spinal ganglionic neurons (Mahalingam et al. 1990;Richter et al. 2009; Zerboni et al. 2014).As individuals age or experience immunosuppression, VZV can reactivate, resulting in the typical dermatomal-associated skin rash known as shingles (zoster).In rare cases, VZV can cause central nervous system (CNS) diseases such as encephalitis, meningitis, myelopathy, and vasculopathy (Gilden et al. 2010;Nagel et al. 2020).Given the rarity of these fulminant CNS manifestations, primary VZV infection and reactivation is largely considered a peripheral disease.Thus, the CNS response to primary infection and reactivation during the typical clinical VZV presentation is not well-characterized.
In this study, we used simian varicella virus (SVV) infection in rhesus macaques (RMs), a non-human primate model of human VZV infection, to describe the concurrent peripheral and CNS response during primary infection, latency, and reactivation (Mahalingam et al. 2022;Mahalingam et al. 2001; Mahalingam et al. 1990; Messaoudi et al. 2009).We measured cytokine and amyloid-β (Ab)-40 and − 42 levels in paired serum and cerebrospinal uid (CSF) samples of six RMs inoculated with SVV.We con rmed previous reports of a proin ammatory peripheral response during primary infection and reactivation but also describe a dynamic CNS in ammatory response that, in some instances, preceded the peripheral response.These data suggest a transient yet robust CNS in ammatory response during all stages of VZV pathogenesis in the absence of an obvious fulminant lytic CNS infection.

Methods
Ethics Statement.RMs used in this study were housed at the Tulane National Primate Research Center (TNPRC) in Covington, LA.All animal housing, care, and research were performed in compliance with The National Research Council, Guide for the Care and Use of Laboratory Animals and the Animal Welfare Act.The Tulane National Primate Research Center (TNPRC) is fully accredited by AAALAC International (Association for the Assessment and Accreditation of Laboratory Animal Care), Animal Welfare Assurance No. A4499-01.All studies were reviewed and approved by the Tulane University Institutional Animal Care and Use Committee (IACUC) under protocol number P0177.All clinical procedures were carried out under the direction of experienced laboratory animal veterinarians in the TNPRC Division of Veterinary Medicine.Clinical procedures were performed under anesthesia using approved anesthetics, and all efforts were made to minimize stress, improve housing conditions, and provide enrichment opportunities (e.g., objects to manipulate in cages, varied food supplements, foraging, and task-oriented feeding methods, and interaction with caregivers).
SVV inoculation and establishment of latency.Six SVV-seronegative adults, male, Indian RMs (Macaca mulatta), 7.0 years of age (KA59, KC22, LA16, LC23, LE26, LF30), were inoculated intratracheally with 1.5 × 10 5 plaque-forming units (pfu) of wild type-SVV-infected rhesus lung broblast (Frhl-2) cells (Mahalingam et al. 1992).All RMs were anesthetized and monitored for well-being by physical exams with blood collections every 3 to 7 days and CSF samples, by cisterna magna punctures, collected weekly for the rst two weeks and then biweekly thereafter until the establishment of latency.All six RMs developed a typical varicella rash within 7-10 days post-inoculation (dpi).Skin scrapings and punch biopsy (4 mm) samples from skin were collected upon the appearance of typical varicella rash lesions.Scrapings were processed for DNA (as shown below).
Immunosuppressive treatment and euthanasia.Four months after primary infection and clearance of SVV viremia, viral latency was con rmed in all RMs.At that time, four Experimental RMs (KA59, KC22, LA16, and LE26) were transported (from TNPRC in Covington, LA, USA, to the School of Veterinary Medicine, Radiation Oncology facility at Louisiana State University in Baton Rouge, LA).Once there, each RM was anesthetized and exposed to a single dose of 200-cGy total body X-irradiation and afterward treated daily with oral tacrolimus (Prograf; 80 µg/kg of body weight/day) and prednisone (2 mg/kg/day) to induce virus reactivation.All animals were monitored by physical exams for well-being and zoster rash with blood collections every 7 days post-treatment (dpX) and CSF collections every 7 or 14 dpX.Skin scrapings and punch biopsy (4 mm) samples were collected upon the appearance of typical zoster rash lesions and processed as described.The Treatment Control RM, LC23, was transported with the Experimental group but did not receive immunosuppressive treatment.Following SVV infection and latency, LF30, the SVV Latent Control RM, neither received immunosuppressive treatment nor was transported.All RMs were euthanized ve months post-treatment.PBMC Isolation, DNA extraction, and qPCR.EDTA anticoagulated blood samples were collected throughout acute infection, latency, treatment, and reactivation.Samples were centrifuged, plasma aliquoted, and peripheral blood mononuclear cells (PBMCs) isolated with Ficoll Hypaque®(Sigma-Aldrich) according to the manufacturer's protocol for standard density gradient centrifugation.DNA was then isolated from the PBMCs using a commercial DNA extraction kit per the manufacturer's instructions (Qiagen, Germantown, MD, USA), followed by quantitative PCR (qPCR) for SVV open reading frame (ORF) 61 DNA in PBMCs as described previously (Messaoudi et al., 2009).

Results
Primary infection and reactivation pro le of rhesus macaques.Six RMs were intrabronchially inoculated with SVV (Fig. 1a).All six animals developed varicella rash at 7-10 dpi (Fig. 1b), which was completely cleared by 21 dpi.Representative varicella rash in monkeys KC22 and LF30 are shown in Fig. 1b.Viremia was detected in one monkey (KA59) by 2 dpi and in the rest by 4 dpi.Viremia disappeared by 14 dpi in all monkeys except one (KC22), which persisted until 12 weeks post-inoculation.The establishment of latency was con rmed by the absence of SVV DNA in blood for two consecutive weeks (Fig. 1c).Latent infection was established in animals KA59, LA16, LC23, LE26, and LF30 by 21 dpi.Viremia persisted in animal KC22 for 12 weeks post-inoculation.(Fig. 1c).Three months post-inoculation, animals were immunosuppressed using one full-body X-irradiation and daily tacrolimus and prednisone treatment (Fig. 2a).Experimental animals (KA59, KC22, LA16, LE26) received X-irradiation and drug treatments.There were two control animals: LC23 was transported to the irradiation facility but neither received irradiation nor immunosuppressive drugs, and LF30 was not transported, nor immunosuppressed (Fig. 2a).Three experimental animals (KA59, KC22, LA16) developed zoster rash 7 weeks post immunosuppression.Representative zoster rash in monkeys KC22 and KA50 are shown in Fig. 2b.Two of the experimental animals (KC22, LE26) and the transportation control animal (LC23) had detectable SVV DNA in the blood at 49-, 42-, and 37-dpX, respectively (Fig. 2c).
Changes in cytokine concentrations in serum and CSF during primary infection, latency, and reactivation.IL-6 levels in serum and CSF.Our initial goal was to determine differences in cytokines levels in the periphery versus the CNS.Serial blood draws, and spinal taps for CSF were obtained from each animal during pre-infection, primary infection, latency, and reactivation (Fig. 3).Similar to our previous studies, all animals that had blood drawn during viremia (LF30 data not available) showed a spike in serum IL-6 during primary infection (Fig. 3a; Traina-Dorge et al. 2014).Overall, IL-6 in the CSF showed a potential spike during primary infection.However, mean pre-inoculation levels were highly variable (Fig. 3b).Experimental animals KA59 (Fig. 3c), LA16 (Fig. 3e), and control monkey LC23 (Fig. 3g) showed spikes in IL-6 in the CSF during primary infection that tightly aligned with serum IL-6 levels.Experimental animals KC22 and LE26 showed large uctuations in serum IL-6 levels during primary infection (Fig. 3d, f).Very little uctuation was seen in the CSF of these two animals.However, there was a small spike in CSF IL-6 following zoster in KC22 (Fig. 3d).One non-immunosuppressed control animal, LC23, had detectable viremia during reactivation (Fig. 2c).It showed a large spike in CSF IL-6 without substantial change in serum IL-6 levels (Fig. 3g).There was no change in the IL-6 levels in serum or CSF during primary infection in the other latently infected control RM LF30 in serum (Fig. 3h).IL-2 levels in serum and CSF.Similar to serum IL-6 levels during primary infection, three out of four experimental RMs (KA59, KC22, and LE26) had elevated spikes in the serum IL-2 levels (Fig. 4c, d, and f).However, very little change was seen in CSF IL-2 and remained largely unchanged in these experimental RMs (KA59, KC22, and LE26).Interestingly, IL-2 in LA16 did spike in the CSF before the onset of rash, with no change in serum (Fig. 4b-f).In addition, non-immunosuppressed RM (LC23) had a considerable spike in CSF IL-2 during primary infection, without any detectable levels of serum IL-2 (Fig. 4g).LC23 also showed an increase in IL-2 in CSF before the onset of viremia during varicella and post-zoster.The latently infected monkey LF30 also had a spike in serum IL-2, but (give the time) post varicella (Fig. 3h).
Serum and CSF levels of IL-10.Overall IL-10 in experimental animals' serum and CSF was elevated during primary infection with little overall change in latency and reactivation (Fig. 5a-b).Although lower levels of IL-10 were seen in CSF at early times during primary infection, they were more variable during latency and reactivation (Fig. 5b).Experimental animals (KA59, KC22, and LE26) and the non-immunosuppressed control monkey LC23 showed increased serum IL-10 during primary infection (Fig. 5c, d, f, g).Both control animals (LC23 and LF30) showed spikes of CSF IL-10 during primary infection (Fig. 5g, h).Further, LC23 also showed not only a spike in serum IL-10 levels during primary infection but also in latency and reactivation (Fig. 5g).
Serum and CSF levels of IL-1β.Serum IL-1β only marginally increased during primary infection and then again during reactivation in the experimental animals (Fig. 6a).There was an increase in the average CSF IL-1β during latency preceding zoster onset in experimental animals (Fig. 6b).Experimental animals had higher variability in serum IL-1β (Fig. 6c-f).Of note, serum IL-1β level in experimental animal KA59 spiked preceding zoster, and this was followed by a larger spike in CSF IL-1β at the time of zoster (Fig. 6c).The non-immunosuppressed monkey (LC23) and the latently infected monkey LF30 had the most predominant spike in serum IL-1β following primary infection (Fig. 6g, h), while only LC23 had another spike in serum IL-1β following zoster (Fig. 6g).
Serum and CSF levels of IL-8.Serum IL-8 was increased in all RMs during primary infection and remained high until latency (Fig. 7a).All animals, including the two control monkeys, had elevated IL-8 levels in serum during primary infection (Fig. 7a-g).One experimental animal (LA16) and the transportation control (LC32) showed increased serum IL-8 following zoster (Fig. 7d, f).The non-immunosuppressed monkey (LC23) also had elevated IL-8 just prior to reactivation (Fig. 7f).Only one experimental animal (LE26) showed any detectable IL-8 in the CSF, and these levels stayed consistent throughout primary infection, latency, and reactivation (Fig. 7e).
Changes in Aβ40 and Aβ42 peptides in the serum and CSF during primary infection, latency, and reactivation.
A role for Aβ42 and Aβ40 in the host's anti-microbial response has been previously proposed (Schwartz and Boles, 2013;Soscia et al., 2010).Therefore, we investigated the levels of Aβ42 and Aβ40 in both the periphery and CNS during primary SVV infection, latency, and reactivation (Fig. 8, 9).Serum Aβ42 and Aβ40 levels were slightly elevated during primary SVV infection (Fig. 8a, 9a).However, during reactivation, we observed a decrease in both Aβ42 and Aβ40 levels in the serum.A similar trend followed with slightly elevated Aβ42 and Aβ40 during primary infection and a slight decrease during reactivation (Fig. 8c-h and Fig. 9c-h).The non-immunosuppressed monkey (LC23) that developed zoster showed a similar trend as experimental animals, with a decrease in serum Aβ42 and Aβ40 following reactivation.The latently infected control monkey (LF30) had the opposite effect, with a shared transient increase of CSF Aβ42 and Aβ40 during primary infection and latency.

Discussion
This study aimed to describe the concurrent peripheral serum and CSF cytokine and Aβ42/Aβ40 responses during primary infection, latency, and reactivation of SVV-infected RMs compared to preinoculation levels to better understand the interplay between the periphery and CNS during VZV pathogenesis.Our data revealed a robust pro-in ammatory response in the serum and CSF during primary SVV infection and reactivation.Our results con rm and extend previous reports of increased cytokines within plasma of RMs during primary SVV infection and reactivation (Haberthur et  ).We found high levels of cytokines, including IL-6, IL-8, IL-10, and IL-2, during primary infection and, while more variable, also found elevated levels during reactivation.Interestingly, we also found elevated cytokines within the CSF during primary infection and reactivation.
However, these levels did not always correspond to elevated levels in the periphery.We also found elevated levels of both Aβ40 and Aβ following primary infection in the plasma, while there was an overall decrease following zoster.Levels of both peptides in the CSF were slightly elevated during primary infection and had a small decrease during zoster.
The results showed that serum IL-6 levels spiked during primary infection in all animals, while CSF IL-6 levels exhibited only in three out of six animals were elevated (Fig. 3).Animals KA59, LA16, and treatment control RM LC23 showed spikes in CSF IL-6 that aligned with serum IL-6 levels during primary infection.Animals KC22 and LE26 had uctuations in serum IL-6 levels but minimal change in CSF IL-6.
Interestingly, the control RM LC23 had a large spike in CSF IL-6 during reactivation without substantial changes in serum IL-6 levels.IL-6 has a variety of actions, including both pro-and anti-in ammatory innate immune responses (Scheller et al. 2011;Schneiders et al. 2015).Increases in IL-6 transcription during VZV-skin infection is well documented and may be considered a cellular stress response in the periphery (Jarosinski et al. 2018).CSF of patients with VZV vasculopathy had elevated IL-6 and IL-8, which may increase the risk of vascular disease (Jones et al. 2016).Serum IL-8 levels increased during primary infection and remained high until latency in all animals (Fig 7).One experimental animal and the transportation control LF30 showed a second spike in serum IL-8 following zoster.Nonimmunosuppressed RM, LC23, also had elevated IL-8 levels prior to reactivation.Only one experimental animal showed detectable IL-8 in the CSF throughout the infection.Furthermore, studies have shown that both IL-6 and IL-8 were elevated in VZV-infected vascular cells and are known to promote neutrophil and macrophage activation in VZV-infected arteries (Jones et al. 2017).Taken together, the increase in IL-6 and IL-8 may suggest that those speci c RMs may have an increased risk of CNS disease, including stroke or vasculopathy that would have otherwise gone undetected within the periphery.
Both IL-2 and IL-10 levels showed (how many) elevated spikes in serum during primary infection (Fig. 4,  5).CSF IL-2 levels remained largely unchanged in these animals, except for LA16, which exhibited a spike in CSF IL-2 before the onset of zoster (Fig. 4e).Non-immunosuppressed monkey LC23 also had a spike in CSF IL-2 during primary infection (Fig. 4g).Control animals LC23 and LF30 displayed spikes in CSF IL-10 during primary infection (Fig. 5).CSF levels of IL-2 in three experimental animals did not change, even with elevated peripheral IL-2.IL-2 presence is indicative of a T-cell activation/in ammatory process (Lan et  IL-1β plays a critical host-defense mechanism early in infection (Lopez-Castejon and Brough, 2011; Nour et al. 2011).However, serum IL-1β levels only slightly increased during primary infection and reactivation in all experimental animals (Fig. 6).CSF IL-1β increased during latency before zoster onset in all experimental animals.Notably, animal KA59 had a large spike in CSF right before the onset of zoster rash and proceeded the spike in serum IL-1β (Fig 6c).Control animals LC23 and LF30 exhibited the largest spike in serum IL-1β during primary infection, with LC23 showing an additional spike following zoster (Fig 6g, h) The variability in these results is surprising given IL-1β is important for initial host defense.However, clinical data suggest higher CSF levels of IL-1β in patients who develop post-herpetic neuralgia, suggesting a role for elevated CNS in ammatory markers and the development of chronic pain disorders associated with infection (Zhao et al. 2017).
The study also investigated the levels of Aβ42 and Aβ40 in the serum and CSF.Serum Aβ42 and Aβ40 levels were slightly elevated during primary infection and decreased during reactivation.A similar trend was observed in CSF Aβ42 and Aβ40 levels, except for the latently infected control monkey LF30, which showed a transient increase during both primary infection and latency.Measurements of CSF and plasma Aβ42 and Aβ40 peptide solubility are suggestive of clearance from the brain, and individuals with Alzheimer's disease routinely have less Aβ42 and Aβ40 in both the CSF and plasma (Dorey et al. 2015;Mehta et al. 2000).Indeed about 30-50% of Aβ peptides found in the blood are transported from the CNS via transport through the blood-brain barrier (Ovod et al. 2017;Roberts et al. 2014;Tanzi et al. 2004).A recent study found that plasma Aβ42 and Aβ40 ratios decreased by 15% in amyloid-positive patients compared to amyloid-negative patients (Klafki et al. 2022).It is of note that we found the largest change in Aβ42 and Aβ40 within the serum levels, with relatively less change within the CSF.These ndings may be indicative of total clearance of Aβ42 and Aβ40 from the CNS into the bloodstream.Taken together, these ndings suggest a potential attenuation of Aβ42 and Aβ40 clearance during zoster.
Overall, the results suggest differences in cytokine and Aβ42/Aβ40 levels between the periphery and CNS during SVV infection in RMs.The ndings provide insights into the immune response and potential involvement of cytokines and amyloidogenic peptides in the pathogenesis of SVV infection within the CNS that may otherwise be missed in peripheral assessments of VZV-associated diseases.

Figure 3 Changes
Figure 3

Figure 4 Changes
Figure 4

Figure 5 Changes
Figure 5

Figure 7 Changes
Figure 7