Spatial T-maze identifies cognitive deficits in piglets 1 month after hypoxia-ischemia in a model of hippocampal pyramidal neuron loss and interneuron attrition

Abstract

Neonatal brain injury from hypoxia-ischemia (HI) causes major morbidity. Piglet HI is an established method for testing neuroprotective treatments in large, gyrencephalic brain. Though many neurobehavior tests exist for rodents, such tests and their associations with neuropathologic injury remain underdeveloped and underutilized in large, neonatal HI animal models. We examined whether spatial T-maze and inclined beam tests distinguish cognitive and motor differences between HI and sham piglets and correlate with neuropathologic injury. Neonatal piglets were randomized to whole-body HI or sham procedure, and they began T-maze and inclined beam testing 17 days later. HI piglets had more incorrect T-maze turns than did shams. Beam walking time did not differ between groups. Neuropathologic evaluations at 33 days validated the injury with putamen neuron loss after HI to below that of sham procedure. HI decreased the numbers of CA3 pyramidal neurons but not CA1 pyramidal neurons or dentate gyrus granule neurons. Though the number of hippocampal parvalbumin-positive interneurons did not differ between groups, HI reduced the number of CA1 interneuron dendrites. Piglets with more incorrect turns had greater CA3 neuron loss, and piglets that took longer in the maze had fewer CA3 interneurons. The number of putamen neurons was unrelated to T-maze or beam performance. We conclude that neonatal HI causes hippocampal CA3 neuron loss, CA1 interneuron dendritic attrition, and putamen neuron loss at 1-month recovery. The spatial T-maze identifies learning and memory deficits that are related to loss of CA3 pyramidal neurons and fewer parvalbumin-positive interneurons independent of putamen injury.

Keywords: Cognition; Hippocampus; Hypoxia; Learning; Newborn; putamen.

Figure 1.

Protocol timeline and spatial T-maze diagram. (A) Timeline of hypoxic-ischemic (HI) injury or sham procedure, T-maze and inclined beam testing, and euthanasia for histologic examination of the brain. (B) During T-maze days 1–4, the accessible milk was located near the red triangle in the east arm. Piglets alternated entering the maze through the north or south arms. (C) For reverse days 1–2, the milk was moved to the blue triangle in the west arm. Piglets again alternated entering the maze through the north or south arms.

Figure 1.

Protocol timeline and spatial T-maze diagram. (A) Timeline of hypoxic-ischemic (HI) injury or sham procedure, T-maze and inclined beam testing, and euthanasia for histologic examination of the brain. (B) During T-maze days 1–4, the accessible milk was located near the red triangle in the east arm. Piglets alternated entering the maze through the north or south arms. (C) For reverse days 1–2, the milk was moved to the blue triangle in the west arm. Piglets again alternated entering the maze through the north or south arms.

Figure 2.

Hypoxic-ischemic (HI) brain injury reduces learning in the spatial T-maze. (A) The proportion of correct turns to reach the accessible milk increased across training days in sham piglets (p=0.046). (B) Sham piglets also reached the accessible milk faster with progressive training days (p<0.001). In post-hoc comparisons, the time to reach the milk was longer on day 1 than on days 2, 3, and 4 (p<0.05 for all pairwise comparisons. (C, D) Piglets with HI brain injury did not show differences in the proportion of correct turns (p>0.05) or time to reach the milk (p>0.05) during 4 days of training.

Figure 3.

Hypoxia-ischemia (HI)-injured piglets took longer to learn the reverse maze than did the sham-operated piglets. (A) HI piglets had fewer correct turns on reverse day 1 than did the sham piglets. *p<0.05. (B–D) The proportion of correct turns on reverse day 2 and time to reach the milk on reverse days 1 and 2 did not differ between HI and sham groups. Box plots with whiskers (5–95^th percentiles) are shown. Each circle represents one pig.

Figure 4.

Time to walk across the inclined beam. (A) Over the course of six trials, time to walk up the beam decreased for both piglets that received HI brain injury (p=0.007; n=4; closed circles) and those that received sham procedure (p=0.002; n=6; open squares). (B) In post-hoc comparisons, HI piglets reached the top of the beam faster on trial 5 than on trial 1 (p=0.020). The sham piglets had faster walking times on trials 4 (p=0.030) and 5 (p=0.014) than on trial 1. The medians and interquartile ranges are shown.

Figure 5.

Surviving neurons in putamen. (A) A sham piglet had numerous neurons in putamen (arrows). (B) A piglet resuscitated from hypoxia-ischemia (HI) had fewer surviving neurons (arrows). Photos were taken at 400x. Scale bars = 20 μm. (C) HI reduced the number of neurons to below that of sham procedure (*p<0.05).

Figure 6.

Histologic evaluation of hippocampal injury. (A) Pig brain section showing the hippocampal anatomic level studied. The arrow identifies the fornix. This section was stained with hematoxylin & eosin and converted to grayscale. The photo was taken at 10x. (B) A sham piglet’s CA3 shows numerous surviving neurons (arrows). (C) Fewer surviving neurons (arrows) were observed in the CA3 of a piglet that underwent hypoxia-ischemia. The injured CA3 also had reactive changes identified by multiple small non-neuronal cells (red arrows). Though the identity of these reactive cells was not determined, their morphology is consistent with microglia. Panels B and C were photographed at 400x. Scale bars = 20 μm.

Figure 7.

Surviving neuron counts in hippocampus and the proportion of correct T-maze turns on reverse day 1. (A) Hypoxia-ischemia (HI) decreased the number of neurons in CA3 to below that of piglets that underwent sham procedure (*p<0.05). (B) The number of surviving neurons in CA3 correlated with the proportion of correct T-maze turns (r=0.759; p<0.05). (C–F) The number of neurons in CA1 and dentate gyrus did not differ between sham and HI piglets and did not correlate with the proportion of correct turns in the maze. In panels A, C, and E, each circle represents one pig. In panels B, D, and F, sham piglets are represented by open squares, and HI piglets are represented by closed circles.

Figure 8.

Parvalbumin-immunoreactive interneurons and dendrites in hippocampus. (A) Parvalbumin-immunoreactive interneurons in CA1 are identified by the arrows in a sham piglet. Large, parvalbumin-positive processes were identified as dendrites and counted. Thin processes, which were likely to be axons, were not counted. (B) Parvalbumin-immunoreactive interneurons (arrows) and dendrite counts in the CA1 of a piglet with hypoxic-ischemic brain injury. Photos were taken at 400x. Scale bars = 20 μm.

Figure 9.

Comparison of parvalbumin-immunoreactive interneurons and dendrites to hypoxic-ischemic (HI) injury or sham procedure and to T-maze performance. (A-C) Though CA3 interneuron counts did not differ between HI and sham pigs, the number of CA3 interneurons correlated with the time to reach the milk on day 4 (r= −0.870; p<0.001) and on reverse day 2 (r= ‒0.623; p=0.035). (D-I) The number of CA1 and dentate interneurons did not differ between HI and sham groups. Interneuron counts did not correlate with time to reach the accessible milk. (J) HI piglets had fewer CA1 interneuron dendrites than did shams (*p<0.05). (K, L) T-maze time and the number of CA1 interneuron dendrites showed no correlation. In the second and third columns, sham piglets are represented by open squares, and HI piglets are represented by closed circles.