Strategies to defeat ketamine-induced neonatal brain injury

Abstract

Studies using animal models have shown that general anesthetics such as ketamine trigger widespread and robust apoptosis in the infant rodent brain. Recent clinical evidence suggests that the use of general anesthetics on young children (at ages equivalent to those used in rodent studies) can promote learning deficits as they mature. Thus, there is a growing need to develop strategies to prevent this injury. In this study, we describe a number of independent approaches to address therapeutic intervention. Postnatal day 7 (P7) rats were injected with vehicle (sterile PBS) or the NMDAR antagonist ketamine (20 mg/kg). After 8 h, we prepared brains for immunohistochemical detection of the pro-apoptotic enzyme activated caspase-3 (AC3). Focusing on the somatosensory cortex, AC3-positive cells were then counted in a non-biased stereological manner. We found AC3 levels were markedly increased in ketamine-treated animals. In one study, microarray analysis of the somatosensory cortex from ketamine-treated P7 pups revealed that expression of activity dependent neuroprotective protein (ADNP) was enhanced. Thus, we injected P7 animals with the ADNP peptide fragment NAPVSIPQ (NAP) 15 min before ketamine administration and found we could dose-dependently reverse the injury. In separate studies, pretreatment of P6 animals with 20 mg/kg vitamin D(3) or a nontoxic dose of ketamine (5 mg/kg) also prevented ketamine-induced apoptosis at P7. In contrast, pretreatment of P7 animals with aspirin (30 mg/kg) 15 min before ketamine administration actually increased AC3 counts in some regions. These data show that a number of unique approaches can be taken to address anesthesia-induced neurotoxicity in the infant brain, thus providing MDs with a variety of alternative strategies that enhance therapeutic flexibility.

Fig. 1. Changes in gene expression following NMDAR blockade

P7 animals were injected with vehicle, MK801 (1 mg/kg) or ketamine (20 mg/kg) and at 8h total RNA was isolated for microarray analysis (N = 6 per group). Global gene expression patterns were extracted by EPIG method and revealed 6 patterns of gene expression, with most of the changes falling into Pattern 1-enhanced or Pattern 2-repressed. The numbers at left corner in each pattern are the probe sets categorized to the corresponding pattern. The vertical axes with zero at the middle are the changes in gene expression (log2 intensity) relative to the vehicle. The heatmaps on right side correspond to 496 enhanced genes of pattern 1 (upper part) and 552 repressed genes of pattern 2 (lower part). Importantly, MK801 and ketamine enhanced and repressed genes in a similar manner, suggesting that these NMDAR antagonists promote essentially identical molecular changes across a large number of functionally diverse genes. From these data we identified activity dependent neuroprotective protein (ADNP) as a candidate gene for further study (see Fig. 2).

Fig. 2. NAP prevents ketamine-induced injury

Our microarray study (see Fig. 1) suggested that we target the ADNP fragment peptide NAP for its neuroprotective potential. P7 rat pups were injected with vehicle (Veh), 20 mg/kg ketamine (K20; 4 doses over 3 hours) or ketamine plus NAP (5, 10, or 20 mg/kg; N5K20, N10K20, N20K20, respectively; NAP injected 15 min prior to ketamine). After 8h, animals were anesthetized, perfused with 4% PFA, brains removed and 60 μm coronal sections were processed for AC3-ir. AC3-positive cells in layers IV and V of the somatosensory cortex (SSC_IV-V) were counted using non-biased stereology. Data expressed as means ± SE (N = 6 for all groups). Significant differences were observed between K20 and vehicle (p<0.001***) and between N10K20 vs K20 and N20K20 and K20 (p<0.01**; ANOVA; Bonferroni post-test). No significant difference in the means was found when N5K20 was compared to K20.

Fig. 3. Vitamin D₃ protects against ketamine-induced injury

P7 rat pups were injected with vehicle (sterile PBS), vitamin D₃ (0.5 μg/kg; Vit D), 20 mg/kg ketamine (4 doses over 3 hours, K20) or ketamine plus vitamin D₃ (K20-Vit D). Vitamin D₃ was injected 15 min prior to ketamine exposure at P7 as well 24 hours earlier to allow time for receptor internalization, translocation to the nucleus, transcription, translation and protein synthesis. After 8h, animals were anesthetized, perfused with 4% PFA, brains removed and 60 μm coronal sections were processed for AC3-ir. AC3-positive cells in the SSC_IV-V were counted using non-biased stereology. Data expressed as means ± SE (N = 5 for all groups); *** p<0.001 when comparing K20 to Veh or Vit D-K20 groups; * p<0.05, when comparing Vit D-K20 to vehicle or Vit D groups (ANOVA, Bonferroni post-test comparison). A significant difference (p<0.001) was also found between K20 and Vit D groups (not indicated).

Fig. 4. Preconditioning protects against ketamine-induced injury

Neonatal rats were injected with vehicle (sterile PBS) or ketamine (5 mg/kg, at 0, 1, 2, 3h) on P6 and/or P7 as follows: P6 vehicle - P7 vehicle (vehicle control; veh); P6 5 mg/kg ketamine - P7 5 mg/kg ketamine (K5K5); P6 5 mg/kg ketamine - P7 20 mg/kg ketamine (K5K20); P7 20 mg/kg ketamine (K20). At 8h, the numbers of AC3 cells in the SSC_IV-V were estimated by non-biased stereology. Data show means ± SE (N = 3 for each group). A. AC3 counts in the SSC_IV-V. *** p<0.001, K20 vs veh; *p<0.05, K20 vs K5K5. Not indicated are non-significant differences for K5K5 vs veh, and significant differences for K5K20 vs veh (p<0.05). B. AC3 counts in the Retrosplenial cortex. ***p<0.001 K20 vs veh, **p<0.01 K20 vs K5K20. Not indicated are significant differences between K5K5 vs veh, K5K20 vs veh, and K5K20 vs K5K5 (p<0.05). Statistical analyses performed using one-way ANOVA with Bonferroni post-test comparison of means.

Fig. 5. Aspirin enhances ketamine-induced injury

P7 rats were injected with vehicle (sterile PBS; Veh), aspirin (30 mg/kg; A30), 20 mg/kg ketamine (at 0, 1, 2, and 3h; K20) or aspirin plus ketamine (A30-K20). At 8h brains were fixed and sectioned for AC3 estimations. Data show means ± SE (N = 5 for each group). A. AC3 counts in the SSC_IV-V .*** p<0.001, A30-K20 vs veh; ns not significant K20 vs A30-K20. Not indicated are significant differences between K20 vs veh (p<0.01) and K20 vs A30 (p<0.05) as well as non-significant differences between A30 vs veh. B. AC3 counts in the cingulate cortex. ***p<0.001 A30-K20 vs veh and A30-K20 vs K20. Not indicated are significant differences between K20 vs veh (p<0.01) and K20 vs A30 (p<0.05). Statistical analyses performed using one-way ANOVA with a Bonferroni post-test comparison of means.