Lesion Size Is Exacerbated in Hypoxic Rats Whereas Hypoxia-Inducible Factor-1 Alpha and Vascular Endothelial Growth Factor Increase in Injured Normoxic Rats: A Prospective Cohort Study of Secondary Hypoxia in Focal Traumatic Brain Injury

Abstract

Background: Hypoxia following traumatic brain injury (TBI) is a severe insult shown to exacerbate the pathophysiology, resulting in worse outcome. The aim of this study was to investigate the effects of a hypoxic insult in a focal TBI model by monitoring brain edema, lesion volume, serum biomarker levels, immune cell infiltration, as well as the expression of hypoxia-inducible factor-1 alpha (HIF-1α) and vascular endothelial growth factor (VEGF).

Materials and methods: Female Sprague-Dawley rats (n = 73, including sham and naive) were used. The rats were intubated and mechanically ventilated. A controlled cortical impact device created a 3-mm deep lesion in the right parietal hemisphere. Post-injury, rats inhaled either normoxic (22% O2) or hypoxic (11% O2) mixtures for 30 min. The rats were sacrificed at 1, 3, 7, 14, and 28 days post-injury. Serum was collected for S100B measurements using ELISA. Ex vivo magnetic resonance imaging (MRI) was performed to determine lesion size and edema volume. Immunofluorescence was employed to analyze neuronal death, changes in cerebral macrophage- and neutrophil infiltration, microglia proliferation, apoptosis, complement activation (C5b9), IgG extravasation, HIF-1α, and VEGF.

Results: The hypoxic group had significantly increased blood levels of lactate and decreased pO2 (p < 0.0001). On MRI post-traumatic hypoxia resulted in larger lesion areas (p = 0.0173), and NeuN staining revealed greater neuronal loss (p = 0.0253). HIF-1α and VEGF expression was significantly increased in normoxic but not in hypoxic animals (p < 0.05). A trend was seen for serum levels of S100B to be higher in the hypoxic group at 1 day after trauma (p = 0.0868). No differences were observed between the groups in cytotoxic and vascular edema, IgG extravasation, neutrophils and macrophage aggregation, microglia proliferation, or C5b-9 expression.

Conclusion: Hypoxia following focal TBI exacerbated the lesion size and neuronal loss. Moreover, there was a tendency to higher levels of S100B in the hypoxic group early after injury, indicating a potential validity as a biomarker of injury severity. In the normoxic group, the expression of HIF-1α and VEGF was found elevated, possibly indicative of neuro-protective responses occurring in this less severely injured group. Further studies are warranted to better define the pathophysiology of post-TBI hypoxia.

Keywords: S100B; edema; hypoxia; hypoxia-inducible factor 1; neuronal death; secondary insults; traumatic brain injury; vascular endothelial growth factor A.

Figure 1

Cytotoxic and vasogenic edema following injury. Left picture depicts the area of decreased apparent diffusion coefficient (ADC) indicating cytotoxic edema (1 day hypoxic). Right picture shows the area affected by an increase in ADC highlighting vasogenic edema (14 days normoxic). The freehand tool in ImageJ was used to define the areas (yellow).

Figure 2

Detection of expression. A person blinded for oxygenation state and survival time highlighted the region of interest surrounding the lesion (yellow circle). The protein expression, above a specific threshold, was then used to quantify the expression. In this case, the expression of C5b-9 (white) is shown in a hypoxic animal, at 1 day following CCI.

Figure 3

Immunofluorescence of the analyzed proteins. The image illustrates representative immunofluorescence to the following proteins: (A) HIF-1α (HIF) (day 1), (B) VEGF (day 1), (C) C5b-9 (day 1), (D) IgG (day 1), (E) Activated Caspase 3 (day 1), (F) Granulocytes (CD43, day 1), (G) Proliferating microglia (CD34, day 7), and (H) Macrophages (ED1, day 7) (all red, respectively) at 10× amplification from normoxic animals. NeuN was used to label neurons (green). DAPI was used as counterstaining for all cells (blue). All pictures are taken from the peri-lesional area; the cavity is located to the right of each picture. Scale bar = 100 μm.

Figure 4

Quantification of cortical neuronal cell death. The section was labeled with NeuN (white) and the cortical lesion area (yellow) devoid of neurons was measured using the “Free hand” tool in ImageJ. Scale bar = 5 mm.

Figure 5

Lesion size quantification. (A) illustrates the difference in lesion size between the normoxic (n = 6) and hypoxic group (n = 5) at day 7, 14, and 28 post-CCI presented as mean and SD. In (B), the lesion size on two representative animals is presented at day 28 following TBI; normoxia (left) and hypoxia (right). Scale bar = 5 mm. (C) illustrates changes in lesion size on MRI (y-axis, mm²) over time (x-axis, survival time in days) in hypoxic (red) and normoxic (blue) rats (R² = 0.721). The gray area represents 95% confidence interval while the line is a regression line. (D) Correlations between the neuronal death using tissue area devoid of NeuN staining (y-axis, pixels) and all available lactate levels (x-axis, mmol/l) acquired from blood gas post surgery; as lactate increases, so does the lesion size (R² = 0.318).

Figure 6

Analyzes of lesion size, edema and immunofluorescence. Lesion size, edema, and immunofluorescence data analysis in hypoxic and normoxic rats: (A), Lesion size using DAPI staining, (B) Lesion size using NeuN staining, (C) Cytotoxic edema on MRI, (D) Vasogenic edema on MRI, (E) HIF-1α (HIF), (F) C5b-9, (G) Activated caspase-3, (H) CD34, (I) VEGF, (J) IgG, (K) CD43, and (L) ED-1 immunofluorescence intensity. Y-axis represent relative intensity. X-axis represents days after injury (log). Blue dots indicate normoxic while red dots represent hypoxic rats; dotted lines delineate regression lines. The gray area surrounding each line is the 95% confidence interval. Lesion area using DAPI and NeuN, HIF, C5b-9, Caspase-3, CD34, IgG, and ED1 changed significantly over time. Lesion Size, NeuN (p = 0.02), HIF (p = 0.01), and VEGF (p = 0.03) are significantly different between normoxic/hypoxic groups.

Figure 7

Serum S100B levels over time after injury. Serum concentration of S100B (y-axis) measured at different times post-CCI (x-axis). Black line indicates median naive levels. S100B shows a trend to be significantly elevated in hypoxic compared to normoxic group at 1 day after injury (Student’s t-test p = 0.0868). No other survival time showed any significance or trend between groups, yet levels remained elevated compared to naive rats. No significant temporal trend could be detected.