Soluble Glycoprotein 130Fc Reduces Controlled Cortical Impact-Induced Cognitive Deficits in Rats

Introduction

Traumatic brain injury (TBI) survivors often face cognitive and motor impairments. ¹ These deficits vary in severity and chronicity, impacting overall quality of life. ¹ A key contributor to these impairments is chronic uncontrolled inflammation,^2,3 a secondary injury response,^4,5 and for which interleukin (IL)-6 pathology has been implicated clinically in the acute^6–10 and chronic ² phases post-injury. Despite progress in the TBI field in understanding acute, subacute, or chronic phases of patient management and diagnosis,^11,12 there are no effective disease-modifying therapies for long-term TBI outcomes.

Modulating early and prolonged inflammatory responses after TBI may improve long-term outcomes. ⁴ One key target is IL-6, an immunomodulatory cytokine elevated after TBI and linked to poor outcomes.^{2,6–10,13–16} Clinically, IL-6 is elevated in the days after TBI and is negatively associated with recovery outcomes.^9,17 Higher cerebrospinal fluid IL-6 over the first 5 days post-injury (dpi) is associated with increased serum IL-6 levels and increased odds of an unfavorable Glasgow Outcome Scale score at 6 months post-injury. ⁸ A higher serum inflammatory burden after severe TBI, as measured by an increased IL-6/IL-10 ratio (pro-inflammatory/anti-inflammatory), is associated with an increased odds of unfavorable outcome at 6 months post-injury. ² Preclinical literature also demonstrates acute increases in systemic and central nervous system (CNS) IL-6 after TBI across various injury models and severities.^18–22 Thus, the potential role IL-6 plays with both systemic and CNS dysregulation and pathology after TBI makes it a possible therapeutic target to promote neurorecovery.

Among healthy individuals, IL-6 levels are low; however, levels can rise dramatically—up to six orders of magnitude—in disease, highlighting its role in disease pathology. ²³ IL-6 is implicated in immune system pathologies including rheumatoid arthritis, cytokine release syndrome, Crohn’s disease, irritable bowel syndrome, cancer, sepsis, and COVID-19.^24–27 IL-6 is also linked to neurodegenerative diseases such as Alzheimer’s disease (AD), multiple sclerosis, and Parkinson’s disease. In AD, higher IL-6 correlates with poorer cognition and hypothalamic/hippocampal volumes. ²⁸ In a mouse AD model, IL-6 neutralization improved memory and normalized IL-6 plasma levels. ²⁸

However, IL-6 affects both homeostasis and disease, making targeting IL-6 a nuanced process. Critical to this point, the involvement of IL-6 in both anti-inflammatory and pro-inflammatory processes ²⁹ is dependent on the signaling pathway. IL-6 can act via classical or trans-signaling. In classical signaling, IL-6 binds to membrane-bound IL-6 receptor (mIL-6R) and membrane-bound glycoprotein 130 (gp130). In trans-signaling, soluble IL-6R (sIL-6R), which is generated by constitutive and (e.g., injury) inducible proteolytic cleavage by ADAM10 and ADAM17, respectively,^30,31 forms a complex with IL-6 to activate gp130. ³² gp130 is ubiquitously expressed, but IL-6R is limited to specific cell types like hepatocytes, T cells, and leukocytes. ²³ Thus, trans-signaling can occur in virtually all cell types, while classical signaling is limited to IL-6R-expressing cells. Classical signaling is associated with anti-inflammatory and protective processes including regeneration, bacterial infection defense, and cell proliferation. ²³ Trans-signaling induces pro-inflammatory responses and is associated with multiple CNS pathologies.^23,25,33

IL-6 signaling can be blocked through antibodies to the receptor complex or inhibitors of signaling molecules, but these may have off-target effects. For example, inhibition of Janus kinase/signal transducer and activator of transcription (JAK/STAT)—downstream signaling molecules—can impact many signaling cascades beyond IL-6. JAK inhibitors such as baricitinib, tofacitinib, and filgotinib are currently in clinical trials for autoimmune, allergic, and inflammatory diseases. ³⁴ Tocilizumab and sarilumab, IL-6R antibodies to IL-6R,^35–37 block both IL-6 signaling pathways, which may limit efficacy or cause unintended risks such as blocking protective classical signaling and causing immunosuppression that leads to detrimental clinical outcomes such as increased nosocomial infection risk. A promising alternative is soluble (s)gp130Fc, a fusion of soluble glycoprotein (sgp)130 and the Fc portion of immunoglobulin (IgG). It binds the IL-6/sIL-6R complex and impedes gp130 signaling, thus selectively blocking IL-6 trans-signaling without affecting classical signaling.^32,38,39 Inhibition of IL-6 trans-signaling benefits several disease states including acute pancreatitis, ⁴⁰ bone fracture healing, ⁴¹ myocardial infarction, ⁴² and TBI. ⁴³ We previously showed in a mouse TBI model that sgp130Fc reduced TBI-induced cognitive deficits, anxiety, and brain inflammation. ⁴³ To begin to extend this work, we hypothesized that targeting IL-6 trans-signaling may reduce TBI-associated behavioral deficits in a severe controlled cortical impact (CCI) model in rats.

Methods

Motor function

Motor function was assessed on the beam balance and beam walk 1–7 dpi (Fig. 1).^48–50 Rats were pretrained and baseline tested before surgery. Beam balance performance was measured as the time (max 60 sec/trial) each rat remained on an elevated (90 cm) beam (1.5 cm wide; 34 cm long). Beam walk was assessed by time (max 60 sec/trial) and distance to cross a beam (2.5 cm wide; 100 cm long). A bright light and noise were used as a stimulus to promote beam traversal and entrance into a goal box at the opposite end. Each task included three trials/day.

OPEN IN VIEWER

FIG. 1.

Motor function 1–7 dpi. (A) Time spent balancing on the beam. (B) Time to traverse the beam walk. Arrows depict sgp130Fc treatment on 1, 4, and 7 dpi (after testing). Points represent mean ± SEM. *p < 0.05 Sham + VEH versus CCI + VEH; ^p < 0.05 Sham + sgp130Fc versus CCI + sgp130Fc. SEM, standard error of the mean; VEH, vehicle; CCI, controlled cortical impact.

Spatial learning and memory

Spatial learning was assessed using the Morris water maze (MWM) on 14–18 dpi (Fig. 2).^46,50–53 The maze consisted of a pool (180 cm diameter; 60 cm high) and a Plexiglass platform (10 cm diameter; 26 cm high). The pool was filled with water (26 ± 1°C) to a depth of 28 cm such that the hidden platform was submerged 2 cm below the water surface. Extra-maze cues were present. Each rat was tested on four trials/day with 4-min intertrial intervals for 5 consecutive days (14–18 dpi; 120 sec/trial max). For each trial, the rat started from a different pseudo-randomized cardinal position of the pool.

OPEN IN VIEWER

FIG. 2.

MWM spatial learning and memory during acquisition trials (14–18 dpi). (A) Latency to locate the submerged platform. (B) Path length traveled to locate the submerged platform. (C) Time spent in the peripheral (thigmotaxis) zone. Points represent mean ± SEM. *p < 0.05 Sham + VEH versus CCI + VEH; ^p < 0.05 Sham + sgp130Fc versus CCI + sgp130Fc; #p < 0.05 CCI + VEH versus CCI + sgp130Fc. MWM, Morris water maze; SEM, standard error of the mean; VEH, vehicle; CCI, controlled cortical impact.

Visible platform (VP) trials were completed 19 dpi (Fig. 3).^46,51–53 The platform was raised 2 cm above the surface of the water for visibility. Each rat was tested on four trials with 4-min intertrial intervals (120 sec/trial max). The latency to find the platform, peripheral (thigmotaxic) zone time, path length, and swim speed were recorded and analyzed using ANY-maze software.

OPEN IN VIEWER

FIG. 3.

MWM visible platform trials (19 dpi). (A) Latency to locate the visible platform. (B) Path length traveled to locate the visible platform. (C) Time spent in the peripheral (thigmotaxis) zone. (D) Average swim speed over all testing trials (14–19 dpi). Bars represent mean ± SEM. *p < 0.05. MWM, Morris water maze; SEM, standard error of the mean.

Search strategies across all MWM trials (n = 874) were qualitatively analyzed using ANY-maze tracking plots. Two blinded experimenters classified the predominant swim strategy per trial.^46,52,54,55 Discrepancies were resolved by a third experimenter. Classifications were based on swim path complexity and platform-finding efficacy and grouped as spatial, nonspatial, or thigmotaxis (Fig. 4). Spatial strategies included direct and self-orienting paths. Nonspatial included scanning, circling, or random searches. Thigmotaxis was defined as wall-hugging behavior (Fig. 4). Strategy use per rat was calculated as a percentage of total trials.

OPEN IN VIEWER

FIG. 4.

MWM swim strategy during acquisition (14–18 dpi) and VP (19 dpi) trials. Each trial was scored as either (1) spatial, (2) nonspatial, or (3) thigmotaxis, and the percent of trials is represented for each experimental group. MWM, Morris water maze; VP, visible platform.

Results

Impact of sgp130Fc on lesion volume in CCI rats

There was no significant effect of treatment on lesion volume or hemispheric tissue loss comparing CCI + sgp130Fc rats with CCI + VEH at 21 dpi (p = 0.1161, Fig. 5A, and p = 0.6558, Fig. 5B, respectively).

OPEN IN VIEWER

FIG. 5.

Histopathology after CCI. Top row shows representative images of the lesion site in CCI + VEH (left) and CCI + sgp130Fc (right) rats. (A) Lesion volume and (B) hemispheric tissue loss at 21 dpi. Bars represent mean ± SEM. CCI, controlled cortical impact; VEH, vehicle; SEM, standard error of the mean.

Discussion

Uncontrolled inflammation after TBI leads to poor chronic outcomes. Since IL-6 is a reliable stress- and injury-induced cytokine, targeting its signaling mechanism—specifically IL-6 trans-signaling—offers a promising therapeutic opportunity with less side effects (e.g., immunosuppression) than pan-IL-6 inhibition. As expected, CCI induced acute motor deficits (beam walk and beam balance) and impaired cognition (MWM). sgp130Fc improved learning and memory but did not impact motor outcomes or lesion/hemispheric volumes. These findings expand on our previous work, ⁴³ demonstrating efficacy of sgp130Fc in rats as well as mice. ³⁸ Overall, our data suggest that intermittent systemic administration of sgp130Fc after TBI may be a viable treatment to promote lasting neurobehavioral recovery.sgp130Fc has shown success in reducing inflammation and affecting outcomes across a variety of conditions. For example, sgp130Fc (olamkicept) is in phase II clinical trials, showing efficacy in treating active ulcerative colitis and inflammatory bowel disease.^32,38 In the ulcerative colitis trial, 12 weeks of olamkicept improved clinical remission and mucosal healing compared with placebo. ⁵⁸ In a mouse sepsis model with elevated IL-6, sgp130Fc enhanced survival, lowered inflammatory cytokines, and reduced cognitive decline. ⁵⁹ sgp130Fc also has shown benefits in infectious diseases like COVID-19, reducing symptoms, lung damage, and mortality in mice. ⁶⁰ Moreover, we found that intermittent sgp130Fc administration improved neurorecovery after CCI in mice, reducing learning deficits, anxiety-like behaviors, and brain inflammation. ⁴³ Notably, a recent report suggests neuroprotection in a mouse MCAO model but with sex-specific differences in effective dosing. ⁶¹ Interestingly, the investigators found that females had higher plasma IL-6 and sIL-6R levels at 24 h post-MCAO; thus, a higher dose of sgp130Fc—1 mg/kg in females compared with 0.5 mg/kg in males—was needed to improve functional outcomes at 7 and 28 dpi. ⁶¹ These findings will inform future work exploring sex differences in response to injury and sgp130Fc treatment.

In our rat study, we observed a transient motor deficit after CCI that resolved within a week. sgp130Fc did not reduce motor deficits after CCI. Since IL-6 rises quickly post-injury, earlier sgp130Fc administration should be tested for potential impacts on motor function. Prior work shows that pan-IL-6 inhibition improved motor coordination in the rotarod test 7–12 dpi, ¹⁹ suggesting cognitive-motor tasks like the rotarod may be informative in future studies. However, motor recovery by 7 dpi on beam balance and walking tasks suggests motor deficits did not confound cognitive outcomes in our study, which is further supported by no significant MWM swim speed differences.

At 18 dpi, CCI + sgp130Fc rats showed reduced latency and path length to find the submerged platform versus CCI + VEH rats, suggesting improved cognitive function. This finding reflects a more efficient, spatially directed swim strategy, with less nonspatial and/or thigmotaxis selections. Overall, we show ∼50% reduction in thigmotaxis strategy selection in CCI + sgp130Fc rats versus CCI + VEH rats, with no differences in peripheral zone time. These data suggest that decreased thigmotaxis selection is likely linked to a more spatially directed, efficient swim strategy rather than anxiety-like phenotypes. In a mouse CCI model, intermittent doses of 0.25 mg or 1 mg (i.p.) starting 1 dpi reduced latency and peripheral zone time. ⁴³ These species differences in sgp130Fc efficacy may be, in part, due to IgG fusion tail immunogenicity as the bioconjugates used differed between rats and mice (mouse IgG_2a vs. IgG1, respectively). Additional consideration is needed to understand effector functions ⁶² and how this might impact sgp130Fc effectiveness in TBI models. Future studies should also explore sgp130Fc’s effects on additional behavioral outcomes, to address working memory, both short- and long-term retention, cognitive flexibility, and psychological health-related outcomes (e.g., anxiety, anhedonia).sgp130Fc may support cognitive recovery after injury by reducing brain inflammation. We previously showed sgp130Fc lowered chemokines at 21 dpi. ⁴³ Pan-IL-6 inhibition also reduces inflammation at 1 dpi. ¹⁹ Region- and cell-type-specific effects may contribute. For instance, IL-6-driven STAT3 activation in the hippocampal CA1 impairs memory. ⁶³ As CA1 was severely damaged in our CCI model, other regions may mediate sgp130Fc’s benefits. Indeed, tocilizumab reduces neuronal death in the hippocampus and cortex post-hemorrhage. ⁶⁴ Since microglia express gp130 and are involved in chronic inflammation after TBI,^5,65–67 they may be key targets with reducing glial-associated IL-6 trans-signaling, wherein cells with constitutive signaling capabilities may be vulnerable to trans-signaling under pathological conditions and in an sgp130:sIL-6R-dependent manner. ⁶⁸ Although lesion volume changes were modest, sgp130Fc’s mechanism and neuroanatomical effects in TBI warrant further study in our rat model.

This study has limitations and raises important questions. First, although IL-6 rises sharply after TBI, mIL-6R and gp130 expression are only marginally affected.^32,69 Due to assay limitations, we could not track biomarkers in our CCI model, though our prior work supports an acute IL-6 rise and the rationale for targeting trans-signaling. ⁴³ Second, the extent of hippocampal damage in our model limited additional pathology measures. While statistically significant, our findings were modest potentially due to the dose selected and variability in IgG fusion molecule between rat and mouse bioconjugates. Though lower than doses in other models (e.g., sepsis, COVID-19),^59,60,70 we previously observed cognitive benefits at similar levels in mice. ⁴³ Dose–response studies are needed to assess pharmacodynamics, pharmacokinetics, and target engagement. Timing may also matter: the blood–brain barrier (BBB) is acutely comprised post-CCI^71,72 but recovers by 5 dpi, ²⁶ limiting large molecule access. However, persistent BBB disruption in clinical TBI⁷³ suggests this deserves further investigation. Lastly, this study included only males; future work should examine females, given sex differences in neuroinflammatory responses.^5,73,74

In conclusion, IL-6 trans-signaling is a manipulable target that is associated with multiple inflammatory conditions. We demonstrate that sgp130Fc treatment reduced CCI-induced cognitive deficits in rats, highlighting its therapeutic potential to promote TBI-associated neurorecovery.

Transparency, Rigor, and Reproducibility

All animals were handled and treated in accordance with the NIH Guide for the Care and Use of Laboratory Animals, and experiments were approved by the University of Pittsburgh IACUC. Sample size calculations were based on previous work.^46,52 Forty rats underwent procedures; two were excluded for technical reasons and one died. Rats were randomly assigned to groups prior to intervention. Investigators were blinded to experimental group. Data normality was tested using the Shapiro–Wilk test. Study data are available in a FAIR data repository (odc-tbi.org).

Authors’ Contributions

Conceptualization: A.K.W., A.L.R.; Investigation: E.G.D., A.M., L.N., N.C., J.A.S., V.S.P., I.G.G.; Analysis: A.L.R., J.C.C., A.K.W.; Data interpretation: A.L.R., P.M.K., A.K.W.; Writing—original draft: A.L.R., J.C.C., A.K.W.; Review and editing—original draft: A.L.R., V.S.P., J.A.S., P.M.K., A.K.W.; Funding: A.K.W.