Impact of DNMT1 and DNMT3a forebrain knockout on depressive- and anxiety like behavior in mice

Abstract

DNA methylation has been shown to impact certain forms of synaptic and behavioral plasticity that have been implicated in the development in psychiatric disorders. DNA methylation is catalyzed by DNA methyltransferase (DNMT) enzymes that continue to be expressed in postmitotic neurons in the forebrain. Using a conditional forebrain knockout of DNMT1 or DNMT3a we assessed the role of these DNMTs in anxiety and depressive-like behavior in mice using an array of behavioral testing paradigms. Forebrain deletion of DNMT1 had anxiolytic and antidepressant-like properties as assessed by elevated plus maze, novelty suppressed feeding, forced swim, and social interaction tests. DNMT3a knockout mice, by contrast, did not exhibit significant behavioral alterations in these tests. Given the putative role of altered DNA methylation patterns in the development of schizophrenia, we also assessed DNMT1 and DNMT3a knockout mice in a prepulse inhibition task and found an enhanced prepulse inhibition of startle in DNMT1 knockouts relative to wild type mice, with no change evident in DNMT3a knockout mice. Our data suggest that DNMT1 and DNMT3a are distinctly involved in affective behavior and that DNMT1 may ultimately represent a potential target for treatment of certain affective behavioral disorders.

Keywords: Anxiety; Behavior; DNA methylation; Depression; Prepulse inhibition.

Figure 1

Forebrain-specific knockout of DNMT1 and DNMT3a. DNMT1 and DNMT3a mRNA were reduced (~70–85%) as compared to CTL mice, with no change in the cerebellum, consistent with a forebrain-specific knockout.

Figure 2

Depressive- and anxiety-like behavior in DNMT 1 and DNMT3a KO mice. (A–D) In the open field test neither DNMT1 (n = 10) nor DNMT3a knockout (KO) mice (n = 10) exhibited significant differences in time spent in the zones of the open field (A, B) or in total distance moved within the open field (C, D) relative to control littermates (CTL; n = 10, 10); “per” = peripheral region, “non-per” = non-peripheral region. (E, F) In the elevated plus maze DNMT1 KOs (n = 13) spent significantly greater time in the open arms of the maze relative to CTL [n = 12; t(23) = 2.098, p = 0.048] but significantly less in the closed arms [t(23) = 2.346, p = 0.028]; DNMT3a KO mice (n = 10) were indistinguishable from CTL mice (n = 10). (G–J) In the forced swim test, DNMT1 KO mice (n =12) exhibited a reduction in total immobility time compared with CTLs [n = 12; t(22) = 2.189, p =0.04] and a significantly increased latency to first immobility posture [t(22) = 3.52, p =0.002]. Although a trend was apparent, DNMT3a KO mice (n = 10) did not differ from CTL mice (n = 10) in terms of total time spent immobile [t(18) = 1.967, p = 0.07] and no differences in latency to immobility were observed.

Figure 3

Anxiolytic and antidepressant-like effects of DNMT1 KO. (A) In the novelty suppressed feeding test DNMT1 KO mice (n = 12) showed a significantly reduced latency to feed relative to CTL mice [n = 12; t(22) = 2.144, p = 0.043] but no change in latency to feed in the home cage following food deprivation. (B) Relative to CTL mice (n = 8) in the social interaction paradigm DNMT1 KO mice (n = 8) had an increased ratio of time spent in the interaction zone with target mouse present as compared to time spent with no target mouse present [t(14) = 2.198, p = 0.045]

Figure 4

Prepulse inhibition and acoustic startle in DNMT1 and DNMT3a KO mice. (A, B) Realtive to CTL mice (n = 8) DNMT1 KO mice (n = 8) exhibited enhanced prepulse inhibition of startle at each stimulus magnitude tested [PP4: t(14) = 2.209, p = 0.044; PP8: t(14) = 2.694, p = 0.017; PP16: t(14) = 2.193, p = 0.046] whereas DNMT3a KO mice (n = 10) were not different from CTL mice (n = 10). (C, D) Assessment of startle amplitudes revealed a trend for decreased startle response at 110 dB (p = 0.197). DNMT3a KO mice (n = 10) showed no change in startle response at any stimulus magnitude tested as compared to CTL (n = 10).