Sex differences in adult social, cognitive, and affective behavioral deficits following neonatal phlebotomy-induced anemia in mice

Abstract

Introduction: Anemia is common in prematurely born infants due to blood loss resulting from frequent phlebotomies and may contribute to their neurobehavioral deficits. Preclinical models of phlebotomy-induced anemia (PIA) have revealed metabolic and genomic changes in multiple brain structures of young mice, yet the impact of neonatal PIA on early-life and adult behavior has not been assessed.

Methods: The present study employed a range of behavioral measures in phlebotomized anemic neonatal mice to investigate short- and long-term neurodevelopmental effects. PIA from postnatal (P) days 3 to 14 caused sex-specific changes in social behavior, novelty preference, and anxiety at P17 that persisted into adulthood.

Results: Our preclinical model suggests that PIA may contribute to acute and long-term behavioral and affective deficits and warrants further substantiation of the observed behavioral phenomena in larger samples.

Conclusions: We conclude that this model is a useful tool for beginning to better understand the lasting effect that early-life PIA might have on the developing brain. The differential impact of PIA on male and female subjects warrants further exploration for the development of appropriately targeted interventions.

Keywords: anemia; animal models; behavioral deficits; phlebotomy; prematurity; sex differences.

FIGURE 1

Sociability and social novelty behavioral data in the three‐chambered social approach task at P17 and P65. Panel a: 2‐way ANOVA of sociability behavior in P17 mice (19 nonbled controls (males = 8), 21 25% Hct mice (males = 7), and 19 18% Hct mice (males = 9)). Asterisks show main effects of treatment (p < .0001) and sex (p < .01). Panel b: 2‐way ANOVA of behavior in the social novelty phase for P17 mice. Asterisks show main effects of treatment (p < .0001). There was a main effect of sex (p < .05) and a sex × treatment interaction (p < .0001). Panel c: 2‐way ANOVA of sociability behavior in P65 mice (13 nonbled controls (males = 5), 22 25% Hct mice (males = 8), and 21 18% Hct mice (males = 9)). Asterisks denote the main effect of treatment (F(2, 50) = 17.38, p < .0001); there was a significant sex × treatment interaction (F(2, 50) = 11.77, p < .0001). Panel d: 2‐way ANOVA of behavior in the social novelty phase for P65 mice. There was a main effects of sex (F(1, 50) = 16.49, p = .0002), a main effect of treatment (F(2, 50) = 29.34, p < .0001), and a significant sex × treatment interaction (F(2, 50) = 14.10, p < .0001) (denoted by ****). Simple effects not shown with ***. *p < .05; **p < .01; ***p < .001; ****p < .0001. Means ± SEM

FIGURE 2

Confidence intervals based on post hoc tests of P17 and P65 three‐chambered sociability test data. Panels a,b: sociability (a) and social novelty (b) in P17 mice. Panels c‐d: sociability (c) and social novelty (d) in P65 mice

FIGURE 3

P17 and P65 performance in the open field test. Subjects: 18 nonbled controls (males = 7), 21 25% Hct mice (males = 7), and 22 18% Hct mice (males = 10). Panel a: 2‐way ANOVA of P17 open field behavior. No differences between treatments were found. Panel b: 2‐way ANOVA of P65 open field behavior. Asterisks denote a main effect of sex (F(1, 55) = 11.28, p = .0014), treatment (F(2, 55) = 6.366, p = .033), and a significant sex × treatment interaction (F(2, 55) = 6.549, p = .0028). Tukey HSD analyses showed that 25% Hct female mice spent significantly less time in the periphery of the open field (M = 318.807, SD = 215.361, p < 000.1). Panels c–d: confidence intervals based on post hoc tests of P17 and P65 data, respectively. “Mean differences” (x‐axis) reflects the differences between means. *p < .05; **p < .01; ***p < .001; ****p < .0001. Means ± SEM

FIGURE 4

P65 elevated plus maze behavioral data from 19 nonphlebotomized controls (males = 8), 22 25% Hct mice (males = 8), and 22 18% Hct mice (males = 10). Panel a: A 2‐way ANOVA revealed a main effect of sex (F(1, 57) = 43.18, p < .0001) and treatment (F(2, 57) = 7.239, p = .0016), and a significant sex × treatment interaction (F(2, 57) = 11.50, p < .0001), denoted by ****. Panel b: confidence intervals based on post hoc tests of P65 data. “Mean differences” (x‐axis) reflects the differences between means. *p < .05; **p < .01; ***p < .001; ****p < .0001. Means ± SEM

FIGURE 5

P17 and P65 NOR novelty scores (time(novel‐time(familiar)/time(familiar) from 19 nonbled controls (males = 12), 21 25% Hct mice (males = 13), and 20 18% Hct mice (males = 9). Panel a: 2‐way ANOVA revealed main effect of sex (F(1, 54) = 12.09, p = .0010, denoted by **), a main effect of treatment (F(2, 54) = 16.98, p < .0001, denoted by ****) but no sex × treatment interaction. Of 18% Hct males had significantly lower novelty scores than male controls (M = −0.76, SD = 0.175, p = .0095, Tukey HSD). Of 18% female mice had lower novelty scores than controls (M = −0.37, SD = 0. 512, p < .0001, Tukey HSD), and the novelty scores of 18% Hct females were significantly lower than those of 25% Hct females (M = −0.37, SD = 0. 512, p = .0037, Tukey HSD). Panel b: 2‐way ANOVA revealed a main effect of treatment (F(2, 53) = 7.403, p = .0015, denoted by **) but not sex (F(1, 53) = 0.1351, p = .7146), and no significant sex × treatment interaction (F(2, 53) = 0.0933, p = .9111). 25% Hct (M = 0.2201, SD = 0.998, p = .0060, Tukey HSD) and 18% mice (M = −0.6403, SD = 0.343, p = .0025, Tukey HSD) had significantly lower novelty scores compared to controls. Phlebotomized mice did not differ from each other. Panels c–d: confidence intervals based on post hoc tests of P17 and P65 data, respectively. “Mean differences” (x‐axis) reflects the differences between means. *p < .05; **p < .01; ***p < .001; ****p < .0001