Exposure to explosive blast causes a specific form of traumatic brain injury (TBI), termed blast-induced TBI or bTBI (1). The most frequent form of bTBI is mild; due to the mild and transient nature of clinical symptoms that follow. Soldiers are frequently exposed to a second, third, or more mild blasts (2). Exposure to additional blasts within the window of increased cerebral vulnerability, i.e., when the brain is still recovering from the initial impact, can have a cumulative effect that results in more severe acute symptoms and long-term pathological implications (3). Post-concussive symptoms (PCS) following mbTBI, and the possibility that symptoms are compounded by repeated exposures, have been extensively reported in the literature (4–7). Symptom onset can be immediate but also transient, or develop over time and become chronic (4, 5). PCS can include cognitive impairments, such as problems with memory or concentration, visual disturbances, tinnitus and headaches, as well as mood disorders, such as anxiety or depression, increased irritability or rage (8, 9). Some of these symptoms are also observed in post-traumatic stress disorder (PTSD) that can be caused by exposure(s) to psychological stress without evidence of physical injury (9). The exact pathobiology of PTSD is currently poorly understood, but alterations in the catecholamine, serotonin, and galanin systems have been implicated in the neurobehavioral abnormalities (10).

Animal models have played a fundamental role in the identification of disease mechanisms, diagnostic and molecular targets, and in testing therapeutic approaches. Animal models have also played a key role in recreating the physical and biological components of bTBI under controlled, reproducible conditions (11, 12). Apart from using free-field blast exposures with explosives, there are two basic models for generating blast overpressure that can be employed within the confinements of laboratory conditions. One is based on the design by Clemedson, which uses low quantities of explosives in a modified artillery piece (or blast tube) to generate a blast wave (13, 14). We refer to this model, used at the Karolinska Institutet, as the “KI model.” The other, more widely used model utilizes compressed air (or another driver gas, such as Helium) to generate a shock wave through the sudden burst of a membrane separating the compression and expansion chambers of the shock tube (15). This model, first described by Elsayed (16) and used at the Uniformed Services University (USU) and Walter Reed Army Institute of Research (WRAIR), is herein referred to as the “USU/WRAIR” model.

We have previously reported that exposing rats to a single mild blast in the KI model results in changes in the classical noradrenergic and serotonergic neurotransmitter systems in various forebrain regions, and of their rate-limiting enzymes in the lower brainstem (17). In a separate study, we also presented changes in the transcripts of the neuropeptide galanin and its receptors following exposure to a single mild blast (18). We consider these findings important, because they provide the rationale for developing potential pharmacological treatments to mitigate PCS; however, our findings need to be independently replicated. In addition, how these transmitter systems respond to repeated mild blasts remains to be examined.

Using the USU/WRAIR model, we have previously found that single exposure to mild blast overpressure triggers neurobehavioral changes, including increased anxiety, impaired learning and memory (19–21). At the molecular level, we identified metabolic and neuroinflammatory changes, as well as axonal, neuronal, and glial damage in behaviorally-relevant brain regions and in serum. We also found that repeated exposure to mild blast overpressure has a cumulative effect on the brain as indicated by both proteomics and diffusion tensor imaging (DTI) (22). Again, these findings need to be independently verified using additional outcome measures for bTBI.

An important consideration is that the majority of TBI studies are carried out using different blast models, strains and ages of rodents, to investigate various outcome measures at a range of terminal time points. Furthermore, where there are repeated exposures, the latency between injuries ranges from minutes to days between studies. Therefore, bTBI studies have produced results that are as heterogeneous as the disease itself. This paper is a step toward increasing translatability between experimental bTBI models and increasing confidence in the field's findings by replicating, combining, and building on previous findings. In line with this aim, we employed two well-established models of bTBI—the KI blast tube vs. USU/WRAIR shock tube—in two different laboratories using identical rat strains, ages, and anesthesia to compare the effects of single and repeated mild blast exposure on the transcript levels of the rate-limiting enzymes tyrosine hydroxylase (TH) and tryptophan hydroxylase 2 (TPH2), and of galanin in the lower brainstem using in situ hybridization (ISH). We also used immunohistochemistry to look for degenerating neurons and signs of axonal injury in the hippocampus.

In this study, we directly compared the short-term effects of single and repeated mild injuries on the transcriptional response of genes involved in catecholamine, serotonin and galanin neurotransmission in the lower brainstem using two well-characterized models of bTBI. These models showed partial similarities with respect to single and repeated blast exposure relative to sham groups. But we found no cumulative effects after repeated exposure (i.e., single vs. double blast) under the experimental conditions used. Our findings highlight the importance of taking into account not only physical and technical parameters, but also environmental stressors when modeling mild bTBI. The importance of time, both in terms of inter-injury interval and post-injury termination, further emphasize the role of the tested transmitter systems in mediating responses in mild bTBI.

Modeling bTBI in the laboratory poses unique challenges due to the complexity of the physical forces resulting from explosive blast (12, 24, 25). Blast creates a highly complex injurious environment consisting of shockwaves and the blast wind, a high velocity (supersonic) air movement that can cause kinetic type of injury (26, 23). The closest to laboratory modeling of explosive blast affecting soldiers is using real explosives in a confined environment, as the blast tube used in the KI model (14, 23). This model, designed by Clemedson and Criborn (27). uses explosives resulting in high peak pressure and short positive phase duration (550 kPa and positive phase duration of 0.2 ms). The USU/WRAIR model is the more commonly used device that uses compressed air to generate the shock wave (12, 23). This model generates a lower peak pressure (average peak pressure 88.9 kPa) and a longer positive phase duration (8–9 ms) than the KI model. Whether the above differences in the physical parameters of blast generated by the two models can trigger different physiological responses is currently not known. However, the fact that the peak values of all analyzed transcripts were similar in all exposed animal groups, regardless of the blast model, suggest that both models trigger similar responses. This is true at least for our current outcome measures focusing on mRNA transcript levels for the rate limiting enzymes of noradrenergic and serotonergic neurotransmitters and galanin in the lower brainstem.

In contrast to the comparable injury-induced changes in transcript levels, sham values were significantly different in the KI and USU/WRAIR models. KI sham groups had significantly lower transcript levels than their corresponding USU/WRAIR sham groups. Statistically significant differences between exposed and sham animals of the KI model were likely due to relatively low sham values. Conversely, most of the values detected in the shams of the USU/WRAIR model were significantly higher than their respective KI shams. In fact, many sham values in the USU/WRAIR model were as high as exposed values seen in the KI model, implicating factors beyond physical blast parameters.

Among these factors are psychological stressors, which play a critical role in our selected neurotransmitter systems and can explain the detected differences in sham values. As detailed in the Methods, animals in the USU/WRAIR model were exposed to significant environmental stressors, associated with extended transportation, blast interleukin 6 levels in noises, and repeated handling. Consequently, in our earlier studies using the USU/WRAIR model we detected elevated corticosterone, interferon-γ, and interleukin 6 levels in sham animals relative to naïve animals not exposed to any stressors (28, 21). Using a battery of behavioral tests, the influence of different laboratory environments has also been demonstrated despite rigorous standardization between labs (29–31). Importantly, the transmitter systems we used as outcome measures have been known to be sensitive mediators of various stressors (32).

Intense activation of the noradrenergic system in the LC in response to stressful conditions has been clearly defined in the literature, and associated with adaptation to acute and chronic stress (33, 34). This in turn leads to activation of the noradrenergic terminals of the forebrain regions, including the cortex, with increases in noradrenaline (NA) turnover (35). A similar increase in TH mRNA levels in the LC was found following single or repeated (7x) daily immobilization stress (36). The neuropeptide galanin, which in rat co-exists with NA and 5-HT, in the LC and DRN (37), respectively, is also influenced by stress exposure (38). Chronic social stress studies revealed increased preprogalanin mRNA levels in the LC of subordinate rats that correlated with the number of wounds (39). Another study revealed an interesting effect on preprogalanin in two different strains of rats in the LC and central nucleus of the amygdala following acute and chronic restraint stress (40). With regard to the role of increased galanin in stress, it has been suggested that this peptide promotes stress resilience (41). An additional, important consideration when interpreting our results is that we only analyzed the transcriptional response at a single, acute time point. Our previous study has shown that the transcriptional response to mild blast using the KI model changes over time (17, 18).

The overwhelming majority of bTBIs are mild (42) and repeated (43). Clinical observations have shown that exposures to repeated mild blasts (as well as repeated concussions) can have cumulative effects resulting in more severe symptomatology after consecutive insults. Using the USU/WRAIR model, we have found evidence of a cumulative effect on brain microstructure using diffusion tensor imaging (DTI) (22). Diffusion tensor parameters had significant blast x no. of events (i.e., single vs. multiple) interactions in several subcortical brain regions including the stria terminalis, a key mediator of various autonomic and behavioral responses. Using the same model we also found significant behavioral and physiological changes after repeated mild blast exposure, including increased anxiety- and depression-related behaviors, elevated heart rate (up to 24 h post-injury), and increases in injury-related blood-based biomarkers (e.g., glial fibrillary acidic protein, neuron-specific enolase, and neurofilament) (44, 45). These and other studies (46) have demonstrated the critical role of inter-injury interval in causing cumulative damage.

The exact period of increased cerebral vulnerability after injury is currently not unknown, but it can be as long as several weeks in humans. Rats have a very different physiology compared to humans, and experiments have shown that such a window can be as short as 30–60 min in rats (28, 22, 47, 48). However, these and other experiments have indicated that metabolic, inflammatory, axonal, and vascular functions (among other processes) have differing periods of increased cerebral vulnerability after injury. Importantly, and of particular relevance for this study, isoflurane post-conditioning has been explored as a possible therapeutic, as it can reduce brain infarction size and attenuate neurological deficits (49). Given that our blast-exposed animals were repeatedly anesthetized with isoflurane, this could account for the absence of a cumulative effect and/or the mitigation of subsequent injury mechanisms in the double blast groups.

There is an ongoing and important discussion concerning a link between long-term neurodegenerative pathologies, such as chronic traumatic encephalopathy (CTE). and repeated mild TBI. This concern includes both blast exposure and occupational aspects of the use of heavy weapons. While the majority of CTE cases have been reported in contact sports like American football or Ice hockey. there are a few cases among military veterans with a history of blast exposure (50). There are also a few interesting reports in tau related pathology in mice models of concussion or blast exposure (51, 52). The current study was not designed to reveal specific information on possible neurodegenerative changes at later time points, so we cannot make statements on a possible relation to CTE. However, some studies suggest a link between the function of the monoaminergic systems and the progression of dementia (53).

This is the first study to our knowledge that has directly compared two well characterized and frequently used experimental rodent models of blast, examining changes in select neurotransmitter and peptide systems following single and repeated exposures. The limitations of our study include the relatively low number of subjects used in the different experimental groups, the single post-injury time point of sampling for analysis, and the single, short inter-injury interval separating repeated exposures. Also, we have only investigated blast effects at the mRNA levels of TH, TPH2 and galanin. How these transcriptional changes translate into changes at the protein level remains to be studied.

There have been significant efforts to elucidate the neurological underpinnings of mbTBI and its long-term sequelae. The monoamine systems have since long been associated with mood disorders (54), and recent studies indicate that this may also be true for galanin (55). However, translation between rodent models and clinical cases is difficult. Our findings highlight the critical importance of mimicking not only the physical parameters of blast, but also the environmental factors. They can, as our data show, majorly affect post-injury outcomes. Given its blast profile and optimal animal handling conditions, the KI model offers a useful experimental system for determining the period of increased cerebral vulnerability, and dissecting the effects of physical blast forces on this and other transmitter systems, from bTBI and concomitant psychological insults.

The views, opinions, and/or findings contained herein are those of the authors and should not be construed as an official position, policy, or decision of the Department of the Army or the Department of Defense.

LK carried out all tissue preparation at KI and all in situ hybridization and immunohistochemistry experiments. LK was also responsible for all the data analysis and was involved in planning and writing. AK carried out blast experiments and tissue preparation at USU. AK was also involved in involved with planning, and writing. JL associations and use of blast facilities at WRAIR. UA carried out blast experiments at KI. TH senior professor involved with planning, writing, and interpreting data. DA senior author at USU and KI, involved with planning, writing, and interpreting data. MR senior author at KI, involved with planning, writing, and interpreting data.