J Neurotrauma. 2017 Mar 1;34(5):1086-1096.

Persistent Behavioral Deficits in Rats after Parasagittal Fluid Percussion Injury

Stacy L. Sell, Kathia Johnson, Douglas S. DeWitt, and Donald S. Prough

Department of Anesthesiology, University of Texas Medical Branch , Galveston, Texas.

DOI: 10.1089/neu.2016.4616



Although traumatic brain injury (TBI) is now considered a chronic disease, few studies have investigated the long-term behavioral deficits elicited by a well-established rodent model of injury. Here we evaluate behavioral measures commonly used in TBI research to determine which tests are useful for studying long-term effects of brain injury in rats. Male Sprague- Dawley rats were handled and pre-trained to neurological, balance, and motor coordination tests prior to receiving parasagittal fluid-percussion injury (FPI), sham injury, or maintenance as naive cohorts. Rats underwent neuroscore, beam-balance, and beam-walk tests for 3 days after injury. Subsequently, in separate groups at 3, 6, or 12 months, they were retested on the same tasks followed by a working memory version of the Morris water maze. On post-injury days (PIDs) 1-3, significant effects of injury on neuroscore, beam-balance, and beam-walk were observed. Differences in the beam-walk task were not detectable at any of the later time-points. However, deficits persisted in beam-balance up to three months and in neuroscore up to 6 months. Working memory deficits persisted up to 12 months, at which time a reference memory deficit was also evident. These data suggest balance and motor coordination recovered more quickly than neurological deficits. Furthermore, while deficits in working memory remained stable over the 12 month period, the late onset of the reference memory deficit points to the progressive nature of the injury, or an age/TBI interaction. In conclusion, standard behavioral tests are useful measures of persistent behavioral deficits after parasagittal FPI and provide evidence that TBI is a chronic condition that can change over time and worsen with age.


Keywords: behavior; chronic brain injury; reference memory; TBI; working memory



The vast majority of traumatic brain injury (TBI) behavioral studies in rats that evaluate learning and memory do so at 11 – 15 days after injury. Early TBI studies considered those time points as “chronic” since most outcomes were measured acutely at 24 – 48 hours after injury. More recently, as the effects of TBI in human patients are progressive and chronic over years rather than days, the need to study the chronic effects in animal models has become apparent.

In the paper presented here, we tested rats at 3, 6, and 12 months after injury on a working memory version of the Morris water maze. To supplement the results of this paper, we compared the results at 11 – 15 days post injury to three months post injury.

The working memory version of the Morris water maze is conducted in a large pool of water (1.8 m diameter, 28 cm deep). A platform (10 cm diameter) is hidden below 2 cm of water. The goal is for the rat to find the platform to escape the water. The most useful outcome measure to indicate learning and/or memory is the time it takes the rat to locate the platform, known as the duration or latency, typically measured in seconds.

There are four possible locations where the platform can be located and four different starting locations. In this test, rats experience five days of testing with four sessions each day. Each session consists of two trials in which the rat is placed in the tank at the same location. The first trial (Trial 1) is considered “learning” as the rat discovers the location of the platform. After 15 s to get his bearings, the rat is picked up and placed back in the tank at the same starting location (Trial 2). Trial 2 is considered “working memory”.  A normal rat generally will locate the platform faster on the second trial due to remembering the location based on external cues. For each subsequent session, the platform location and entry location are changed and another pair of trials is conducted. Despite different pairing and order of entry and platform locations each day, the same four platform locations are used over the course of the five days of testing. Thus, the daily averages of the Trial 1 latencies (AKA Trials 1, 3, 5, 7) become shorter over the five days, indicating the rats remember these locations. The results of Trials 1, 3, 5, and 7 can be considered a measure of “reference memory” and Trials 2, 4, 6, and 8 indicate “working memory”.

Another way to measure working memory is to calculate the difference between Trials 1 and 2, also called the “Delta.” This difference represents the change in latency from Trial 1 to Trial 2, and the greater the delta, the better the working memory.

The results shown here are from two different sets of rats. In both experiments, half the rats received a moderate, parasagital fluid-percussion injury (as described in the paper) and the other half received a sham-injury, which consists of anesthesia, surgical preparation, and craniotomy without fluid-percussion. One set was tested on post-injury days 11 – 15 (or to end at 2 weeks after injury) and the other set was tested on post-injury days 86 – 90 (or to end at 3 months after injury).

Data were analyzed using a mixed model ANOVA (R statistical package), followed by a Differences of Least Squares means. We found no Trial 1 differences between SHAM and TBI at either 2 weeks or 3 months. For Trial 2, there was an overall significant effect of injury group, SHAM vs. TBI (P=.0006) and a significant difference between SHAM and TBI for Delta (T1-T2; P=.0003). Post-hoc t-test revealed a significant difference between SHAM and TBI on each day for PIDs 11 – 15 and Day 86 for the 3 month group Trial 2, as well as Day 14 for PIDs 11 – 15 and Day 86 for the 3 month group Delta (P<.05).

These results suggest healing occurs between the two week and three month time points. In light of the previously published data at 3, 6, and 12 months, it appears the working memory ability remains stable between 3 and 12 months after injury; however, the reference memory first showed deficits at 12 months after injury, implying a slow, progressive deterioration of reference memory.



Figure: The results of the working memory water maze test are shown for separate groups of SHAM and TBI rats at 11 – 15 days after surgery (left) and 3 months after surgery (right). Average daily latencies for Trial 1 (top), Trial 2 (second row), and Delta (third row) are shown as well as the overall cumulative latencies for Trials 1 and 2 (bottom). Results of the post hoc t test are indicated on the graphs. * P<0.05 for SHAM vs TBI on the same day or overall.