Developmental vitamin D deficiency and schizophrenia: the role of animal models

Abstract

Schizophrenia is a debilitating neuropsychiatric disorder that affects 1% of the US population. Based on twin and genome-wide association studies, it is clear that both genetics and environmental factors increase the risk for developing schizophrenia. Moreover, there is evidence that conditions in utero, either alone or in concert with genetic factors, may alter neurodevelopment and lead to an increased risk for schizophrenia. There has been progress in identifying genetic loci and environmental exposures that increase risk, but there are still considerable gaps in our knowledge. Furthermore, very little is known about the specific neurodevelopmental mechanisms upon which genetics and the environment act to increase disposition to developing schizophrenia in adulthood. Vitamin D deficiency during the perinatal period has been hypothesized to increase risk for schizophrenia in humans. The developmental vitamin D (DVD) deficiency hypothesis of schizophrenia arises from the observation that disease risk is increased in individuals who are born in winter or spring, live further from the equator or live in urban vs. rural settings. These environments result in less exposure to sunlight, thereby reducing the initial steps in the production of vitamin D. Rodent models have been developed to characterize the behavioral and developmental effects of DVD deficiency. This review focuses on these animal models and discusses the current knowledge of the role of DVD deficiency in altering behavior and neurobiology relevant to schizophrenia.

The prevalence of mental illness is over 18% among adults in the USA (Substance Use and Mental Health Services Administration, 2014). Mental illness contributes significantly to premature death, lost productivity and healthcare costs to the affected individual and society as a whole. Although significant advances have been made in understanding the etiology of mental illness, there is still a great deal that is unknown about the specific factors that increase risk for these devastating diseases. Genetics has been shown to have a significant role in susceptibility for psychiatric disorders and genome-wide association and sparse sequencing studies have begun to identify risk loci that harbor polymorphisms and functional pathways for further study (Cross-Disorder Group of the Psychiatric Genomics Consortium 2013; Network & Pathway Analysis Subgroup of Psychiatric Genomics Consortium 2015; Ripke et al. 2013; Sullivan 2015). However, heritability estimates from twin studies (Kendler 2001) and more recent genome-wide single nucleotide polymorphism (SNP) analysis (Cross-Disorder Group of the Psychiatric Genomics Consortium et al. 2013) indicates that environmental factors and gene by environment interactions are also likely to contribute significantly to increased risk. As increased risk for psychiatric diseases is thought to depend, in part, on neurodevelopmental changes that occur early in life, the role of early environment including early life stressors, perinatal immune activation and dietary deficiencies have been examined as potential environmental factors. There are many excellent reviews on the role of perinatal and postnatal stress, maternal immune activation and in utero protein deficiency on disease status in adulthood (Babenko et al. 2015; Brown 2011; Knuesel et al. 2014; Tarantino et al. 2012). In this review, we will focus on the literature describing the role of vitamin D deficiency and increased risk for schizophrenia with a focus on animal models.

A slight but significant increased risk for schizophrenia has been observed in individuals born in winter or spring in human epidemiological studies (McGrath 2007; Torrey et al. 1997) and this effect increases with distance from the equator (Davies et al. 2003). Moreover, the offspring of individuals who migrate from more equatorial regions to colder climates are also at increased risk for schizophrenia (Cantor-Graae & Selten 2005), as are the offspring of city-dwellers in comparison to those who live in rural regions (March et al. 2008). There are many factors that vary across these groups including socioeconomic status, nutritional intake and exposure to pathogens, but one factor, exposure to sunlight and its effect on levels of the active form of vitamin D, 1,25-dihydroxyvitamin D₃ (hereafter referred to as vitamin D), has emerged as a focus of research. The first chemical reaction in the formation of vitamin D is driven by exposure of the skin to sunlight (Berridge 2015) and vitamin D levels are closely tied to sun exposure (Webb & Holick 1988) leading to the hypothesis that vitamin D levels are a mediating factor that increases risk for schizophrenia as well as a host of other health problems (McGrath 2001). Reports of vitamin D deficiency as a world pandemic (Holick 2008) have brought this topic into focus and accelerated the pace of research in this area.

The health benefits of exposure to sunlight were first appreciated in the last century when it was shown that exposure to the sun was curative for rickets (Hess & Unger 1921). Subsequent studies identified vitamin D as the antirachitic factor (Holick et al. 1982). Since then, our understanding of the function of vitamin D has expanded beyond its role in calcium and phosphate homeostasis, bone formation and maintenance. More recently, vitamin D has been described as a neuroactive steroid hormone involved in brain development and function (Garcion et al. 2002).

Vitamin D exerts its effects by binding to the vitamin D receptor (VDR), a member of the nuclear hormone receptor family, at vitamin D response elements (VDREs) within the promoter regions of vitamin D-responsive genes. Once activated, the VDR regulates gene expression and stimulates intracellular signaling pathways (Ryan et al. 2015). The VDR is ubiquitously expressed in a wide range of tissues including rodent and human brain (Burkert et al. 2003; Cui et al. 2007; Eyles et al. 2005) and there are thousands of VDRE in the genome (Carlberg 2014) supporting the numerous and pleiotropic effects that have been attributed to vitamin D.

The VDR is expressed in dopaminergic neurons in both human and rat substantia nigra, hippocampus and prefrontal cortex (Cui et al. 2013; Eyles et al. 2005) – regions implicated in schizophrenia. Expression in the substantia nigra in rats emerges early in development and increases until weaning (Cui et al. 2013). This finding along with the evidence for delayed dopamine (DA) cell differentiation (Kesby et al. 2013) and DA-mediated behavioral deficits in rodents exposed to developmental vitamin D (DVD) deficiency (Eyles et al. 2013) have led to the hypothesis that exposure to vitamin D deficiency could influence the development of dopaminergic neurons and have serious implications for the development of neuropsychiatric disease.

Although human epidemiological studies present compelling evidence for elevated risk for schizophrenia based on vitamin D production, the existence of numerous and often immeasurable environmental factors and the inherent ethical and logistical challenges of conducting developmental research in humans have led to the development of rodent models of DVD deficiency. These animal models are being used to elucidate the role of vitamin D during development on dopaminergic circuitry and behaviors with the ultimate goal of developing interventions that could alter the onset, severity or course of schizophrenia and other neuropsychiatric diseases.

An animal model of DVD deficiency

There are obvious hurdles that impede human studies of nutritional deficiencies during development and their effect on behavior later in life. For example, the inability to control for numerous and life-long environmental challenges that an individual encounters and that have the potential to affect the development of a psychiatric disease makes it difficult to specifically study the effects of perinatal exposure to a nutritional deficiency. Furthermore, longitudinal studies that span the perinatal period through adulthood, when most psychiatric diseases have their onset, are labor-intensive and logistically challenging. In addition to, and more relevant than these, logistical hurdles are the obvious ethical problems with manipulating nutritional intake during gestation and critical developmental periods. Moreover, the inability to access brain tissue in humans to evaluate underlying mechanistic changes presents an insurmountable challenge, although advances in neuroimaging have provided relevant insights. In order to circumvent these logistical and ethical hurdles, research has turned to the use of rodent models to study the effect of exposure to nutritional deficiencies during the perinatal period on brain development and neurobehavioral outcomes in adulthood.

Rodents are good models for these investigations for several key reasons. First, they have a relatively short gestation period and lifespan, making it more feasible to measure the effect of a perturbation during development into adulthood (around 60–80 days). Additionally, rodent models allow for control over the environment, making it possible to manipulate a single variable during a specific window of time while controlling for other variables as much as possible. Moreover, rodents have comparable neurobiological and physiological systems to humans – a feature that ultimately makes studies in rodents. Furthermore, rodent models have been used extensively to study behavior and specific assays have been developed to model certain aspects of schizophrenia (Table 1). Finally, there are advanced molecular and genomic tools available for rodents that enable assessment of underlying mechanisms (i.e. dysregulation of specific genes, changes in neurotransmitter systems and brain morphological differences). These reasons make rodent models ideal to study the effects of vitamin D deficiencies during developmental on neurobehavioral outcomes in adult offspring.

Table 1.

Commonly used animal models to measure schizophrenia endophenotypes

	Human correlate	Animal model	References
Positive	Psychomotor agitation	Locomotor activity and response to novelty	Farrow et al. (2006); Young et al. (2007)
	Disorganized behavior	Patterns of locomotor activity	Young et al. (2007)
Negative	Anhedonia	Saccharine or sucrose consumption, intracranial self-stimulation	Amitai et al. (2009); Berlin et al. (1998); Le Pen et al. (2002)
	Avolition (lack of motivation)	Progressive ratio, conditioning paradigms	Wiley and Compton (2004)
	Social withdrawal	Social interaction	Nadler et al. (2004)
		Resident intruder	Ellenbroek and Cools (2002)
		Nest building	Belforte et al. (2010)
		Home cage social behavior	Torres et al. (2005)
Cognitive	Working memory	T-maze, holeboard, radial arm maze	File and Wardill (1975); Levin et al. (1996)
	Visual learning and memory	Morris water maze, novel object recognition, Barnes maze	Barnes (1979); Chen et al. (2000); Dere et al. (2007); Ennaceur and Delacour (1988); Morris (1981); Paylor et al. (2001)
	Reasoning/problem solving	Attentional set shifting; cross-maze-set-shifting task; serial reversal acquisition paradigm, rodent gambling task	Birrell and Brown (2000); Colacicco et al. (2002); Ragozzino et al. (1999); Widholm et al. (2001); Zeeb et al. (2009)
	Attention/vigilance	5C-SRT 5C-CPT	Robbins (2002); Young et al. (2009a)
	Processing speed	5C-SRT 5C-CPT, olfactory discrimination	Slotnick (2001); Uchida and Mainen (2003)
	Social cognition	Social interaction or recognition	Ferguson et al. (2002); Nadler et al. (2004)
Endophenotypes	Increased sensitivity to psychostimulants (e.g. amphetamine and cocaine)	Measurement of locomotor activation following administration of psychostimulants	van den Buuse et al. (2005)
	Impaired sensorimotor gating (pre-attention)	PPI of startle response	Geyer et al. (2001)
	Impaired auditory evoked potentials	Auditory-evoked potential measurement	Connolly et al. (2003)
	Impaired LI	Shuttle-box, fear-conditioning	Feldon and Weiner (1991); Mongeau et al. (2007)
	Impaired mismatch negativity	Mismatch negativity in anesthetized rodents	Umbricht et al. (2005)
	Alterations in neurochemistry	Neurochemistry and microdialysis measurements	Di Matteo et al. (2008); Lodge and Grace (2008)
	Sleep disturbances	Electroencephalogram measurements, sleep-wake cycles	Cohrs (2008); Dzirasa et al. (2006);
	Mood (anxiety)	Open field, EPM, light/dark, holeboard, marble burying; stress-induced hyperthermia	Borsini et al. (1989); Crawley (1981); Gardner and Piper (1982); Njung'e and Handley (1991); Prut and Belzung (2003); Rodgers and Johnson (1995);
	Mood (depression)	Forced swim test, tail suspension test, learned helplessness	Porsolt et al. (1977); Seligman and Beagley (1975); Steru et al. (1985)
	Self-care	Grooming, body weight, coat condition, general health, barbering, etc.	Audet et al. (2006)

DVD deficiency: effects on behavior

Subsequent to the hypothesis regarding the role of DVD deficiency as a candidate risk factor for schizophrenia (McGrath 1999), McGrath and colleagues developed a rodent model of DVD deficiency to test its effects on brain development and behavior in adulthood. The DVD deficiency model exposes rodent dams to a diet completely deficient in vitamin D for 6 weeks prior to conception and through gestation until birth. After pups are born, the mother is switched back to a diet containing normal levels of vitamin D. The effects of DVD deficiency have been assessed primarily in adult offspring of outbred Sprague-Dawley rats, or in a few cases, inbred mouse strains. As DVD deficiency has primarily been associated with increased risk for schizophrenia, the behavioral assays used have focused on specific aspects of the positive (e.g. hallucinations and delusions), negative (e.g. anti-social and lack of motivation) and cognitive deficits that can be modeled in rodents. Specific endophenotypes of schizophrenia, traits that are enriched in individuals with the disease but are not necessarily symptoms, can also be assessed in rodent models. Table 2 summarizes behavioral studies assessing the effects of DVD deficiency.

Table 2.

Literature summary of the effects of exposure to DVD deficiency in rodents

Species	Strain	Sex	Deficiency	Exposure timing	Age at testing	Outcome (increase (↑) or decrease (↓) relative to control diet)	References
Rat	Sprague-Dawley	F and M	Vitamin D (0 IU/kg; Dyets, Inc, Bethlehem PA, USA)	DVD	PND0	↑ Lateral ventricles ↓ Cortical thickness ↑ Mitotic cells in DG, Hypo, BG ↓ NGF and GDNF ↓ p75NTR mRNA and protein expression	Eyles et al. (2003)
Rat	Sprague-Dawley	F and M*	Vitamin D (0 IU/kg with 2mm Ca; Dyets, Inc, Bethlehem PA, USA)	DVD vs. weaning (DVD-PND21) vs. Life (DVD-PND70)	PND70	↑ Locomotion in holeboard test and EPM in DVD group only	Burne et al. (2004a)
Rat	Sprague-Dawley	F and M*	Vitamin D (0 IU/kg; Dyets, Inc, Bethlehem PA, USA)	DVD vs. weaning (DVD-PND21) vs. Life (DVD-PND70) vs. AVD (PND 35–70)	PND35 and PND70	↓ PPI in Life group only at PND 70	Burne et al. (2004b)
Rat	Sprague-Dawley		Vitamin D (0 IU/kg; Dyets, Inc, Bethlehem PA, USA)	DVD	ED19, ED21, PND0, P7	↓ Apoptotic cells in BG, Hypo, cingulate, DG at E21 and E23 ↑ Mitotic cells in dentate at E19 and E21 and BG at P7 ↓ Pro-apoptotic gene expression ↑ Cell cycle progression (pro-mitotic) gene expression	Ko et al. (2004)
Rat	Sprague-Dawley	M	Vitamin D (0 IU/kg with 2mm Ca; ssniff Spezialdiäten GmbH, Soest, Germany)	DVD	PND56	Impaired LI (in an active avoidance task) ↓ Habituation over time in holeboard ↓ Exploratory behavior of novel olfactory stimuli (Day 5 of holeboard) ↑ Learning skill on brightness discrimination Y-chamber	Becker et al. (2005)
Rat	Sprague-Dawley	F and M†	Vitamin D (0 IU/kg; Dyets, Inc, Bethlehem PA, USA)	DVD vs. weaning (DVD-PND21)	PND70	↑ Lateral ventricles in weaning group ↓ Anterior commissure in DVD group ↓ NGF in both groups ↓ GABA-Aα4 and MAP-2 gene expression in both groups ↓ NF-L gene expression in DVD group	Feron et al. (2005)
Rat	Sprague-Dawley	M	Vitamin D (0 IU/kg; Dyets, Inc, Bethlehem PA, USA)	DVD	PND70	↑ Locomotion in holeboard test ↑ MK-801-induced locomotion (prevented with acute haloperidol) ↑ Locomotion in open field (prevented with acute haloperidol)	Kesby et al. (2006)
Rat	Sprague-Dawley	M	Vitamin D (0 IU/kg; Dyets, Inc, Bethlehem PA, USA)	DVD	PND70	↑ Locomotion in open field	Eyles et al. (2006)
Rat	Sprague-Dawley	M	Vitamin D (0 IU/kg with 2 mm Ca; ssniff Spezialdiäten GmbH, Soest, Germany)	DVD	PND56	↓ Head dipping (prevented with subchronic haloperidol and CHPG) ↓ Habituation in holeboard test (normalized with subchronic haloperidol, CHPG and risperidone)	Becker and Grecksch (2006)
Rat	Sprague-Dawley	F and M*	Vitamin D (0 IU/kg; Dyets, Inc, Bethlehem PA, USA)	DVD	PND56	↑ Locomotion in the open field (prevented with 5-min acute restraint stress before open field)	Burne et al. (2006)
Rat	Sprague-Dawley	F and M	Vitamin D (0 IU/kg; Dyets, Inc, Bethlehem PA, USA)	DVD	PND70	Dysreguiated expression of 17 proteins in the FC Dysregulated expression of 23 proteins in the Hipp Dysregulated expression of Hs7C, Tbb1, EnoG and ATPb proteins in both FC and Hipp	Almeras et al. (2007)
Rat	Sprague-Dawley	F and M	Vitamin D (0 IU/kg; Dyets, Inc, Bethlehem PA, USA)	DVD	PND0	↑ Number of neurosphere in SVZ cultures	Cui et al. (2007)
Rat	Sprague-Dawley	F and M	Vitamin D (0 IU/kg; Dyets, Inc, Bethlehem PA, USA)	DVD	PND70	Dysregulated expression of 74 genes	Eyles et al. (2007)
Rat	Sprague-Dawley	M	Vitamin D (0 IU/kg; Dyets, Inc, Bethlehem PA, USA)	Early DVD (4 weeks before mating) vs. late DVD (E0–E21) vs. full DVD	PND70	↑ MK-801-induced locomotion in late and full DVD	O'Loan et al. (2007)
Rat	Sprague-Dawley	M	Vitamin D (0 IU/kg; Dyets, Inc, Bethlehem PA, USA)	DVD	PND70	↓ Protein expression of 22 proteins in NAc ↑ Protein expression of 13 proteins in NAc	McGrath et al. (2008)
Rat	Sprague-Dawley	M	Vitamin D (0 IU/kg; Dyets, Inc, Bethlehem PA, USA)	DVD	P0	↑ DOPAC/HVA in forebrain ↓ HVA/DA in forebrain ↓ COMT expression	Kesby et al. (2009)
Rat	Sprague-Dawley	F and M†	Vitamin D (0 IU/kg; Dyets, Inc, Bethlehem PA, USA)	DVD	ED12 and ED15	↓ Nurr1 expression at ED12 and ED15 ↓ p57Kip2 expression at ED15 (Fetal mesencephalon)	Cui et al. (2010)
Rat	Sprague-Dawley	F and M	Vitamin D (0 IU/kg; Dyets, Inc, Bethlehem PA, USA)	DVD	PND35 and PND70	↑ Amphetamine-induced locomotion ↑ DAT density in CPu ↑ DAT affinity in the Acb (females only, PND 70)	Kesby et al. (2010)
Rat	Sprague-Dawley	F	Vitamin D (0 IU/kg; Dyets, Inc, Bethlehem PA, USA)	DVD	PND0 and Tested Dams	↑ Pup retrieval time ↑ Pup-directed activities (sniffing and carrying) ↓ Latency for self-grooming	Burne et al. (2011)
Rat	Sprague-Dawley	F and M	Vitamin D (0 IU/kg; Dyets, Inc, Bethlehem PA, USA)	DVD	PND150-180	↑ MK-801-induced locomotion ↑ Mk-801-induced acoustic startle ↓ MK-801 binding in the CPu	Kesby et al. (2012)
Rat	Sprague-Dawley	F and M*	Vitamin D (0 IU/kg; Dyets, Inc, Bethlehem PA, USA)	DVD	PND140	↑ Premature responses (reversed with acute clozapine) ↑ Lack responses in non-target trials (Trials 1–14) (reversed with acute clozapine) ↓ Reward latency	Turner et al. (2013)
Rat	Sprague-Dawley	F and M*	Vitamin D (0 IU/kg; Specialty Feeds, Glen Forrest, Western Australia)	DVD	PND180	↑ THC-induced PPI	Burne et al. (2014)
Rat	Sprague-Dawley	F and M	Vitamin D (0, 1.5, 3.3, 6, and 10 IU/g; Purina Mills, LLC, TestDiet, St. Louis MO, USA)	Mating till weaning (PND18)	PND35-40 and PND100-105	↑ Grooming behavior in EPM at PND35-40 in 0 group	Pan et al. (2014)
Rat	Sprague-Dawley	F and M*	Vitamin D (0 IU/kg; Specialty Feeds, Glen Forrest, Western Australia)	DVD	PND210-290	↓ Training sessions required to meet criteria in the rodent gambling task ↑ Food rewards per session (overall increased performance)	Peak et al. (2015)
Mice	129/SvJ (129) and C57BL/6J	F and M*	Vitamin D (0 IU/kg; Dyets, Inc, Bethlehem PA, USA)	DVD	PND70	↑ Head dipping in both strains ↓ Habituation to holeboard test in B6J ↑ Locomotion in open field and holeboard in 129	Harms et al. (2008)
Mice	C57BL/6J	M	Vitamin D (0 IU/kg; INRA, France)	DVD	PND210 and PND420	↓ Lateral ventricles PND210 ↓ Learning in olfactory tubing maze (Hipp-dependent memory task)	Fernandes de Abreu et al. (2010)
Mice	C57BL/6J (B6J) and 129/X1SvJ	F and M	Vitamin D (0 IU/kg; Dyets, Inc, Bethlehem PA, USA)	DVD	PND0 (MRI) and PND70 (MRI and behavior)	↓ Hipp in B6J females at PND0 ↓ Lateral ventricles in B6J males at PND70 ↑ Striatum in B6J males at PND70	Harms et al. (2012a)
Mice	C57BL/6J	F and M	Vitamin D (0 IU/kg; Specialty Feeds, Glen Forrest, Western Australia)	DVD	PND140	↑ Repetitive responses (nose poke) to target stimuli in 5C-CPT	Harms et al. (2012b)
Mice	BALB/c	F and M	Vitamin D (0 IU/kg with 25 g/kg Ca and 15.3 MJ/kg; Specialty Feeds, Glen Forrest, Western Australia)	DVD	ED14.5 and ED17.5	↓ Crown-rump length at ED17.5 ↓ Lambda-bregma length at ED17.5 ↓ Lateral ventricle volume at ED17.5 ↓ Bdnf gene expression at ED14.5 ↑ Bdnf gene expression at ED175 ↑ Tgf-β1 gene expression at ED17.5 ↓ Foxp2 gene expression at ED14.5 ↑ Foxp2 gene expression at ED17.5 ↓ Th gene expression at ED175 in F ↓ Foxp2 expression in cortex at ED17.5 in F ↓ TH expression in SN at ED175 in F	Hawes et al. (2015)
Mice	BALB/c	F and M	Vitamin D (0 IU/kg with 25 g/kg Ca and 15.3 MJ/kg; Specialty Feeds, Glen Forrest, Western Australia)	DVD	ED14.5 and ED17.5	↑ Maternal CORT levels ↑ Fetal glucocorticoid exposure ↓ Placental vascular development	Tesic et al. (2015)

Acb, accumbens; AVD, adult vitamin D deficiency; BG, basal ganglia; Ca, calcium; Cingulate, cingulate gyrus of the cortex; COMT catechol-O-methyl transferase; CORT, corticosterone; CPu, caudate putamen; DG, dentate gyrus of the hippocampus; DOPAC, 3,4-dihydroxyphenylacetic acid; ED, embryonic day; F, female; FC, frontal cortex; Foxp2, forkhead box protein P2; GDNF, glial cell line-derived neurotrophic factor; Hipp, hippocampus; HVA, homovanillic acid; Hypo, hypothalamus; M, male; NAc, nucleus accumbens; PND, postnatal day; SN, substantia nigra; SVZ, subventricular zone.

^*Female and male offspring were tested; however, data were pooled for analysis because of no significant effect of vitamin D diet × sex effects.

^†Tissues from female and males were taken and data pooled without analyzing separately.

Cognitive deficits

Cognitive deficits are among the core symptoms observed in schizophrenic patients, are strongly correlated with functional outcomes and are not well managed by current therapies (Green 1996; Green et al. 2004; Vingerhoets et al. 2013) making the need for more effective treatments imperative. The National Institute of Mental Health supported the Measurement and Treatment Research to Improve Cognition in Schizophrenia (MATRICS) initiative aimed at fostering a better understanding of the mechanisms underlying cognitive deficits in schizophrenia. The MATRICS identified seven domains: attention/vigilance, working memory, reasoning and problem solving, processing speed, visual learning and memory, verbal learning and memory and social cognition (Young et al. 2009b). Rodent models for most of these domains have been developed (Table 1) and appear to show good construct validity (Arguello & Gogos 2010; Kellendonk et al. 2009). Cognitive studies in rodents exposed to DVD deficiency have focused primarily on attention/vigilance and working memory.

Attention/vigilance

Patients with schizophrenia have deficits in attentional processing. Attention is a multidimensional construct that can be divided into selective attention, or selecting the target stimulus, sustained attention or vigilance, focusing on the stimulus for a prolonged time, and attentional control in which attention is maintained despite distractions (Young et al. 2009b). In humans, attentional processes are measured with the continuous performance test. In rodents, the 5-choice serial reaction time task (5C-SRT) (Robbins 2002) and the 5-choice continuous performance task (5C-CPT) (Young et al. 2009a) are used to assess sustained attention. To date, no effect of DVD deficiency on attentional processing has been observed in rats (Turner et al. 2013) or C57BL/6J (B6J) mice (Harms et al. 2012b) in the 5C-SRT. In the 5C-CPT test, DVD deficiency resulted in increased premature responding thought to reflect increased impulsivity, and increased responding to non-target stimuli across all testing days, indicating a lack of response inhibition (Turner et al. 2013). Both of these behaviors could be prevented with acute treatment of the antipsychotic clozapine (Turner et al. 2013). The B6J mice exposed to DVD deficiency showed increased perseverative responses to the target stimuli in the 5C-CPT indicating a deficit in response inhibition (Harms et al. 2012b).

Latent inhibition (LI) is a learning phenomenon in which pre-exposure to an inconsequential stimulus diminishes future learning of new associations with that stimulus. The LI is disrupted in patients with schizophrenia (Lubow & Gewirtz 1995). In animal models, LI can be measured as the ability to ignore irrelevant stimuli in tests such as the shuttle box or fear conditioning. Deficits in LI in rats have been reported as a result of hippocampal lesions (Grecksch et al. 1999) and treatment with antipsychotics (Feldon & Weiner 1991). The DVD deficiency results in disrupted LI in the shuttle-box paradigm in rats (Becker et al. 2005) but not in B6J mice in a fear conditioning paradigm (Harms et al. 2012b).

Prepulse inhibition (PPI) is the attenuation of a startle response when the startle stimulus is paired with a non-startling prepulse. Prepulse inhibition is a well-validated and frequently used test of sensorimotor gating, the neurological process of filtering out irrelevant stimuli. Patients with schizophrenia show deficits in PPI that respond to treatment with atypical antipsychotics (reviewed in Braff et al. 2001). Deficits in PPI have also been observed in patients with Huntington's disease and obsessive compulsive disorder (Braff et al. 2001). In rodents, manipulating neurotransmitter systems implicated in schizophrenia using DA agonists and N-methyl-d-aspartate (NMDA) antagonists results in PPI deficits (Geyer et al. 2001). A recent paper using zebrafish ablated genomic screen homeobox 1 (Gsx1) expressing neurons using optogenetics and observed disrupted PPI. Similar PPI deficits were also observed in a Gsx1 knockout mouse (Bergeron et al. 2015). Gsx1 has been identified within a vulnerability locus in a genome-wide association study of neuropsychiatric disorders. These studies highlight the construct validity of PPI.

A study on rats reported a decrease in PPI in animals that were exposed to vitamin D deficiency throughout their entire life (perinatal + adulthood) but not those exposed only during the perinatal period (Burne et al. 2004b). Subsequent studies also found no PPI deficits in response to DVD deficiency in rats (Burne et al. 2014; Kesby et al. 2006, 2012) or mice (Harms et al. 2008) exposed during the perinatal period alone. Additionally, PPI was not altered by the psychotomimetic MK-801 [NMDA receptor (NMDAR) antagonist] or the DA agonist, apomorphine, in DVD-deficient rats (Kesby et al. 2006, 2012). Prepulse inhibition is enhanced, however, after acute treatment with the psychomimetic cannabinoid agonist D⁹-tetrahydrocannabinol (THC) (Burne et al. 2014).

Schizophrenic patients also show deficits in habituation upon repeated exposure to a startling stimulus. These deficits are thought to reflect alterations in central inhibitory mechanisms (Braff et al. 1992; Geyer & Braff 1982). The DVD deficiency does not appear to have any effect on acoustic startle response in rats (Burne et al. 2004b, Kesby et al. 2006, 2012) or mice (Harms et al. 2008). Similar to unexposed animals, acute treatment with MK-801 increases acoustic startle response in DVD-deficient rats (Kesby et al. 2012) while THC and apomorphine have no effect (Burne et al. 2014; Kesby et al. 2006).

Working memory, reasoning and problem solving

Patients with schizophrenia have impairments in working memory (Lee & Park 2005). However, modeling this form of memory in animals is complicated by subtle differences in how working memory is defined in humans vs. rodents. Behavioral tasks in rodents aimed at assessing working memory vary in the specific aspect being assessed (e.g. goal maintenance, memory capacity and interference control) (reviewed by Dudchenko et al. 2013). For the purposes of this review, we have grouped together a variety of behavioral tasks used to assess learning and memory in DVD-deficient rodents.

Rats exposed to DVD deficiency exhibited no baseline deficits in spatial working memory in the radial arm maze (Becker et al. 2005), learning acquisition or retention in a shuttle-box paradigm of active avoidance (Becker et al. 2005) or delay-dependent memory in the delay match to sample task (Burne et al. 2014). Moreover, there were no effects of DVD deficiency on decision-making processes in the rodent gambling task. The DVD-exposed animals showed significantly fewer omissions than control animals, suggesting an improvement in cognitive performance on this task (Peak et al. 2015). Rats exposed to DVD deficiency also showed decreased number of errors on the relearning session of the brightness discrimination task (Becker et al. 2005) indicating improved working memory. The DVD-deficient mice exhibited no deficits in fear conditioning (Harms et al. 2012b). One study, however, did find a significant impairment on working memory in DVD-deficient B6J mice in the olfactory tubing maze, a hippocampal-dependent learning task (Fernandes de Abreu et al. 2010).

Taken together, the results of these studies assessing cognitive function suggest that exposure to DVD deficiency has very subtle effects that differ based on both species (rat vs. mouse), underlying genetics (outbred vs. inbred) and behavioral task. Furthermore, to our knowledge, none of the positive findings on cognitive deficiencies have been replicated. Therefore, future studies are required utilizing expanded behavioral measures and diverse genetic backgrounds.

Positive symptoms

Some of the positive symptoms of schizophrenia such as hallucinations and delusions are impossible to assess in rodents. However, locomotion in a novel test environment is often used as a rodent model of the positive symptoms of psychomotor agitation and disorganized behavior. Rats with neonatal hippocampal lesions, a rodent model of schizophrenia, are hyperlocomotive in a novel environment (Black et al. 1998). Rats exposed to DVD deficiency also exhibit spontaneous hyperlocomotion in various behavioral assays including the open field (Burne et al. 2006; Eyles et al. 2006; Kesby et al. 2006), elevated plus maze (EPM) (Burne et al. 2004a) and holeboard test (Burne et al. 2004a, Kesby et al. 2006). However, these findings have not been consistent across all studies (Becker & Grecksch 2006; Becker et al. 2005; Burne et al. 2014) indicating that the hyperlocomotion resulting from DVD deficiency might be sensitive to testing conditions and environment (i.e. test duration, lighting levels, etc.).

Hyperlocomotion was also reported in 129/SvJ (129), but not in B6J inbred, mice in the holeboard test and open field but not in the EPM (Harms et al. 2008) indicating that behavioral changes may be specific to both the type of test and genetic background in mice. Interestingly, spontaneous hyperlocomotion was not found in rats exposed to a vitamin D-deficient diet after weaning and into adulthood indicating that gestation and the postpartum periods may be critical for the development of this behavior (Altemus et al. 1987; Burne et al. 2004a).

Spontaneous locomotion in the open field in rats exposed to DVD deficiency is blocked by acute exposure to restraint stress immediately prior to testing (Burne et al. 2006) – a finding that does not appear to be because of differences in Hypothalamic-Pituitary-Adrenal (HPA) axis reactivity (Eyles et al. 2006). However, exposure to stress also results in DA release in the brain (Anstrom & Woodward 2005) and DVD deficiency disrupts the development and functioning of the dopaminergic system (see Dopamine section below) providing a possible mechanism for the role of DVD deficiency and stress on spontaneous locomotion.

Negative symptoms

Negative symptoms reported in patients with schizophrenia include apathy, social deficits and lack of motivation. In rodents, exploratory behavior is used as an index of apathy and can be measured as a decrease in head dipping in the holeboard test. Rats exposed to a DVD deficiency show a decrease in head dipping, indicative of less exploratory activity (Becker & Grecksch 2006; Becker et al. 2005). This decrease in exploration is restored with subchronic treatment with haloperidol or the mGluR5 metabotropic gluta-mate receptor agonist, (R,S)-2-chloro-5-hydroxyphenylglycine CHPG; (Becker & Grecksch 2006). However, not all studies in rats have reported a difference in head-dipping behavior in the holeboard test (Burne et al. 2004a; Kesby et al. 2006) and a study of 129 and B6J mice exposed to DVD deficiency reported an increase in head dipping (Harms et al. 2008). However, interpretation of head-dipping behavior across studies may be confounded by the increased locomotor behavior in DVD-deficient animals that was observed in the latter three studies. Lack of habituation over multiple testing days in the holeboard test, as measured by head-dipping behavior, has also been reported in DVD-deficient rats (Becker & Grecksch 2006; Becker et al. 2005) and B6J, but not in 129 mice (Harms et al. 2008). Habituation in the holeboard test can be restored with subchronic treatment of haloperidol, CHPG and the atypical antipsychotic drug, risperidone (Becker & Grecksch 2006).

Lack of motivation to obtain a reward can be considered a measure of anhedonia, an aspect of depressive-like behavior. Rats exposed to DVD deficiency show decreased reward latency in the 5C-CPT task and an increase in the number of food rewards per session in the rodent gambling task suggesting increased motivation for reward (Peak et al. 2015; Turner et al. 2013). However, there was no alteration in reward latency in the 5C-SRT in rats exposed to DVD deficiency (Turner et al. 2013).

The DVD deficiency had no effect on immobility in another test of depressive-like behavior, the forced swim test in rats (Burne et al. 2004a) or mice (Harms et al. 2008). These data suggest that DVD exposure does not increase depressive-like behaviors, but may in fact increase motivation to obtain a reward. These findings also suggest that the underlying circuitry may differ between measures of depressive-like behavior (i.e. anhedonia vs. behavioral despair) and that exposure to DVD deficiency can have variable effects across different behaviors.

Asocial behavior is a core symptom of schizophrenia, affecting approximately one-third of patients. Social withdrawal and isolation emerge in the premorbid stage and persist throughout the illness (Wilson & Koenig 2014). Social behavior in rodents is assessed using the social interaction test (Nadler et al. 2004). Rats with neonatal lesions to the hippocampus show decreased social interaction in adolescence and adulthood, reflecting the emergence of social deficits in the premorbid stage (Sams-Dodd et al. 1997). The effect of DVD deficiency on social behavior in the social interaction test has been studied and no effect was observed in rats (Burne et al. 2004a) or mice (Harms et al. 2008). Lack of an effect of DVD deficiency on social interaction could indicate that the underlying pathology of social withdrawal is not related to pathways influenced by vitamin D. Alternatively, other social behaviors that have not been assessed (i.e. home cage behavior or nest building) might better capture the effects of DVD deficiency.

Mechanisms

Vitamin D has pleiotropic effects on cell differentiation and brain development in addition to its role as a neuroactive steroid (reviewed by Garcion et al. 2002). Animal models have been used to explore the mechanisms by which DVD deficiency can alter the brain, resulting in behavioral deficits in endophenotypes of schizophrenia. Table 2 summarizes studies aimed at identifying changes in brain circuitry, morphology and gene expression that might contribute to behavioral effects.

Glutamate

Glutamate neurotransmission, with emphasis on the NMDAR, is hypothesized to play a role in the pathophysiology of schizophrenia. This hypothesis was formulated based on the finding that administration of NMDAR antagonists such as MK-801, ketamine and phencyclidine resulted in psychosis in otherwise healthy individuals and mimicked symptoms of schizophrenic patients (see Howes et al. 2015 for a review). In rodents, administration of NMDAR antagonist results in hyperlocomotion in the open field. The DVD deficiency has been shown to increase sensitivity to MK-801 in the open field in rats (Kesby et al. 2006, 2012; O'Loan et al. 2007), an effect that is blocked by haloperidol (Kesby et al. 2006). Increased locomotor sensitivity to MK-801 was not seen in rats whose mothers were exposed to a vitamin D-deficient diet prior to conception but not during gestation, indicating that gestation and the perinatal period constitute a critical exposure window to induce a persistent effect (O'Loan et al. 2007). Decreased MK-801 binding in the caudate putamen was also observed in DVD-exposed rats (Kesby et al. 2012), suggesting a reduction in NMDAR function. Behavioral effects in response to MK-801 were not observed in either 129 or B6J mice (Harms et al. 2012a). These data suggest that glutamatergic transmission through NMDARs is disrupted in rats exposed to DVD deficiency, a finding that compares to what is seen in schizophrenic patients.

Future studies are needed to assess gamma-aminobutyric acid (GABA) neurotransmission in relation to exposure to DVD deficiency. A reduction in GABA-synthesizing enzymes GAD65 and GAD67 in prefrontal cortex and hippocampus has been reported in patients with schizophrenia (reviewed in Lewis et al. 2012). A recent study that employed pharmacogenetic techniques (designer receptors exclusively activated by designer drugs) to specifically inhibit GABA interneurons expressing parvalbumin or GAD65 in the ventral hippocampus of mice observed hyperlocomotion and deficits in spatial working memory and PPI (Nguyen et al. 2014). Dysregulation of GABA transmission should be investigated as a possible mechanism linking exposure to DVD deficiency and hyperlocomotion. There is evidence that GABA transmission is dysregulated in DVD-deficient animals based on two studies that looked at gene expression in whole brain from adult rats exposed to DVD deficiency. These studies identified a decrease in expression of GABA-A_α4 (Feron et al. 2005) and GABA B receptor 1 (Eyles et al. 2007). However, other studies have reported no difference in GABA transmission (Almeras et al. 2007; McGrath et al. 2008), although the discrepancies between studies could be due to differences in brain regions analyzed. Clearly, future studies investigating the GABA system in DVD-deficient animals are warranted.

Dopamine

Schizophrenic patients show a clear dysregulation of the DA system that is thought to contribute to the emergence of positive symptoms. The presence of the VDR in dopaminergic neurons in the substantia nigra and its expression at appropriate time points during development (Cui et al. 2013) suggests that vitamin D can have an effect on the developing DA system in the brain.

Exposure to DVD deficiency in rodents results in dysregulation of the developing dopaminergic system in the brain. In the rat fetal mesencephalon, DVD deficiency leads to reduction of two important factors for specification of dopaminergic neurons, Nurr1 and p57Kip2 (Cui et al. 2010). In BALB/c mice, exposure to DVD deficiency increases gene expression of an important factor in neuronal survival and neuroprotection of dopaminergic cells (transforming growth factor β1, Tgf-β1) and a decrease in both gene and protein expression of the DA synthesizing enzyme, tyrosine hydroxylase (Th), in the substantia nigra at ED17.5 (Hawes et al. 2015).

While it is clear the DVD deficiency disrupts the developing dopaminergic system, this effect does not appear to be as apparent in the expression of DA signaling proteins in the adult rat brain (Almeras et al. 2007; McGrath et al. 2008). However, the functioning of these proteins could be impaired without altering protein levels. A study by Kesby et al. (2009) found that there was a decrease in conversion of the DA metabolite 3,4-dihydroxyphenylacetic acid to homovanillic acid and a decrease in expression of the needed enzyme in this reaction, catechol-O-methyltransferase in the DVD-deficient neonatal rat brain (Kesby et al. 2009). Adult female rats exposed to DVD deficiency had increased DA transport density in the caudate putamen and increased DA affinity in the nucleus accumbens indicating that the functioning of the DA proteins was enhanced (Kesby et al. 2010).

Increased psychostimulant-induced DA release in the striatum has been observed in patients with schizophrenia (Breier et al. 1997). In rodent models, psychostimulants such as amphetamine, that induce the release of DA and increase locomotor behavior, are administered to assess functioning of the dopaminergic system. The DVD deficiency in female but not in male rats results in increased locomotor response to amphetamine in adulthood (Kesby et al. 2010). This finding was not observed in juvenile rats, mimicking the adult onset that is typical of schizophrenia (Kesby et al. 2010). Interestingly, increased sensitivity to amphetamine was not seen in DVD-exposed B6J or 129 inbred mice indicating that species-specific and possibly genetic background differences exist (Harms et al. 2012a). Collectively, these studies show that vitamin D deficiency during development has a significant effect on the development and functioning of the dopaminergic system and that there is a behavioral manifestation of DA dysregulation in adulthood in rats.

Maternal care

The quality of maternal care has been shown to have a consistent and appreciable effect on behavior in adult offspring (Francis et al. 1999; Szyf 2013) leading one to question whether maternal behavior differs among mothers exposed to vitamin D deficiency. Interestingly, a study by Burne et al. (2011) showed that exposure to DVD deficiency increases maternal behaviors in the rat. The DVD-deficient mothers perform more sniffing, carrying and retrieval of pups in a pup retrieval test (Burne et al. 2011). Pups exposed to higher levels of maternal care have previously been shown to have a lower corticosterone response to a stressor and less anxiety-like behavior in adulthood than pups of dams who exhibit low amounts of maternal behaviors (see review by Francis & Meaney 1999). Furthermore, it is thought that maternal behavior is stimulated, in part, in response to pup ultrasonic vocalizations, which are mediated by the DA system (Dastur et al. 1999). Since the dopaminergic system is altered by DVD deficiency (Cui et al. 2007, 2010; Kesby et al. 2009, 2010), Burne et al. (2011) examined the effect of DVD deficiency on calling rate of isolation-induced ultrasonic vocalizations but observed no differences.

A recent study examined the effect of exposure to DVD deficiency in BALB/c mice and reported that mothers on the vitamin D-deficient diet had significantly increased corticosterone levels during pregnancy, indicating higher stress levels (Tesic et al. 2015). Tesic et al. also reported that the placenta of DVD-exposed mothers showed deficiencies in 11β-hydroxysteroid dehydrogenase type 11, which inactivates glucocorticoids and protects the fetus from overexposure, and an increase in a highly sensitive glucocorticoid factor (Tesic et al. 2015). These findings would suggest that developing pups of vitamin D-deficient BALB/c mothers are being exposed to high levels of stress hormones in utero. The findings of this study are in contrast to a previous study in rats that found no difference in corticosterone levels at baseline or after a restraint stress in pregnant dams exposed to a DVD-deficient diet or their male offspring in adulthood (Eyles et al. 2006). These conflicting results highlight the recurring finding of species differences in response to DVD deficiency and emphasize the need for future studies directed at identifying potential mechanisms.

Structural brain abnormalities

Structural brain abnormalities in patients with schizophrenia include decreased overall brain, temporal lobe and amygdala/hippocampal volume and increased lateral ventricles (reviewed by Lawrie & Abukmeil 1998). Rats exposed to DVD deficiency were also found to have enlarged lateral ventricles and a thinner cortex at birth (Eyles et al. 2003; Feron et al. 2005). However, the opposite phenotype was seen in mice. Mice exposed to DVD deficiency have smaller lateral ventricles in utero (Hawes et al. 2015) and in adulthood (Fernandes de Abreu et al. 2010; Harms et al. 2012a). The DVD deficiency also resulted in a larger striatum and smaller hippocampus in adult B6J males and a smaller hippocampus in females at birth (Fernandes de Abreu et al. 2010; Harms et al. 2012a).

Vitamin D has anti-proliferative, pro-apoptotic and differentiation properties in various cell types (Berridge 2015). Embryonic rat hippocampal neurons treated with vitamin D had a decreased percentage of cells in mitosis and increased neurite outgrowth (Brown et al. 2003) and mouse hippocampal neurons treated with vitamin D showed a decrease in cell proliferation and an increase in differentiation (Marini et al. 2010). Conversely, rats exposed to DVD deficiency exhibited an increased number of mitotic cells from gestation into the early postnatal period (Eyles et al. 2003; Ko et al. 2004), upregulation of pro-mitotic genes involved in cell cycle progression and a decrease in apoptosis and pro-apoptotic gene expression (Ko et al. 2004). Primary hippocampal cultures from rats exposed to DVD showed an increased number of neurospheres in the subventricular zone indicating increased activity of neural progenitor cells (Cui et al. 2007). Taken together, these results suggest that vitamin D plays a role in promoting healthy cellular development of the brain and a deficiency of this key regulator during brain development in utero can lead to brain abnormalities similar to those seen in schizophrenic patients. More research is needed to examine the differences in brain abnormalities in rats vs. mice and determine if this could explain some of the behavioral differences between the two species. If so, examination of these differences would provide insight into resilience vs. susceptibility to develop schizophrenia in those exposed to vitamin D deficiency during the perinatal period.

Neurotrophic factors

Neurotrophins such as nerve growth factor (NGF) and brain-derived neurotrophic factor (BDNF) have been implicated in the pathophysiology of schizophrenia. Nerve growth factor and BDNF are involved in promotion of neuronal growth and differentiation in the developing nervous system and synaptic plasticity throughout life. Peripheral BDNF and NGF levels are reduced in patients with schizophrenia both before and during treatment with antipsychotics (reviewed by Green et al. 2011; Martinotti et al. 2012; Weickert et al. 2003). However, there is some debate in the literature with regard to BDNF levels since a few studies have reported increased BDNF levels (Autry & Monteggia 2012; Martinotti et al. 2012). Differences between studies are thought to reflect treatment with antipsychotics, brain region and sample type (e.g. serum, whole blood and brain tissue). Regardless, there is general agreement in the literature that NGF and BDNF are dysregulated in schizophrenic patients and animal models of schizophrenia (Pillai 2008). However, a gap still exists linking neurotrophic dysregulation and manifestation of symptoms.

Vitamin D has been shown to result in NGF production in vitro in cultured embryonic rat hippocampal neurons (Brown et al. 2003). Conversely, as anticipated, exposure to DVD deficiency results in a decrease in NGF as well as glial cell line-derived neurotrophic factor and low affinity neurotrophin receptor p75^NTR at birth (Eyles et al. 2003; Feron et al. 2005). Decreases in NGF also persist into adulthood (Eyles et al. 2003). Exposure to DVD deficiency also results in decreased Bdnf gene expression in BALB/c mice at ED14.5 but increased expression at ED17.5 compared with control animals (Hawes et al. 2015). A decrease in neurotrophic factors in DVD-deficient rodent models mimics what is reported in patients with schizophrenia indicating that vitamin D might play a mechanistic role in neurotrophic dysregulation. It is possible that the decrease in neurotrophic factors observed in both schizophrenics and DVD-deficient rodents might contribute to cognitive behaviors such as learning and memory where synaptic plasticity plays a major role.

Dysregulation of gene and protein expression in the brain

The ability of exposures in development to induce behavioral changes in adulthood are thought to be because of altered gene expression, possibly through epigenetic mechanisms that can persist over time (Isles 2015). Identifying genes that are altered in response to a specific manipulation (i.e. vitamin D deficiency) is challenging because of the complexity of the genome. However, detecting gene expression changes in response to vitamin D deficiency in development can provide insight into the mechanisms that result in behavioral changes in adulthood and ultimately provide targets for treatment. Whole-brain tissue from adult rats exposed to DVD deficiency showed dysregulated expression of 74 genes involved in oxidative phosphorylation, redox balance, cytoskeleton maintenance, calcium homeostasis, synaptic plasticity, neurotransmission, chaperoning and posttranslational modifications, with the majority of the genes being downregulated (Eyles et al. 2007). In concordance with this finding, the same study also reported dysregulation of protein expression in the same cellular processes for 17 and 23 proteins in the frontal cortex and hippocampus, respectively, and 4 proteins in both regions (Hs7C, Tbb1, EnoG and ATPb) (Almeras et al. 2007). Again, the majority of these proteins were found to be downregulated in adult rats exposed to DVD deficiency and were localized to mitochondria, cytoskeleton and synapses (Almeras et al. 2007). A similar study examined protein expression in the nucleus accumbens and found significant downregulation of 22 proteins and upregulation of 13 proteins in rats exposed to DVD deficiency (McGrath et al. 2008). Dysregulated proteins in the nucleus accumbens included calcium binding, dynamin-like and mitochondrial proteins (McGrath et al. 2008). Of note, there was no overlap in dysregulated proteins identified in the nucleus accumbens, frontal cortex and hippocampus (Almeras et al. 2007; McGrath et al. 2008). In another study analyzing specific genes of interest via RT-PCR, exposure to a vitamin D-deficient diet in the DVD model or DVD + postnatal period resulted in decreased expression of genes involved in neuronal structure including Gabra4, Map2 and Nefl in whole brain from adult rats (Feron et al. 2005).

The biological processes identified in gene and protein expression studies from schizophrenics and rodents exposed to DVD deficiency show considerable overlap. Gene expression microarray analysis of prefrontal cortex tissue from schizophrenics identified downregulation of genes involved in presynaptic function including N-ethylmaleimide sensitive factor (Nsf) and synapsin II (Syn2) (Mirnics et al. 2000). N-ethylmaleimide sensitive factor was also downregulated in whole brain (Eyles et al. 2007) and the SYN2 protein was upregulated in the frontal cortex of rats exposed to DVD deficiency (Almeras et al. 2007). Another study compared postmortem brain tissue of schizophrenic patients and healthy controls using proteomic and microarray analysis and found that 90% of dysregulated genes were involved in mitochondrial function, specifically energy metabolism and oxidative stress (Prabakaran et al. 2004). Specific proteins (capitalized) and genes (lowercase) that are dysregulated in both brain tissue from schizophrenic patients and rats exposed to DVD deficiency include those involved in neurotransmission (MAPK2, ENOG, 143G, VIS1, GAP43, SYN2, Nsf, Dlgh1, Glul, Syt1, Cplx2 and Gabbr1), cytoskeleton maintenance (DREB, GFAP, Map2, Kic1, Gfap and Ckb), mitochondrial functioning (MDHC, COX6a1, ATPB, Atpb and Sod2), posttranslational modifications (PSA2, HS7C, UCHL1, Psa2 and Uchl1) and the calcium-binding protein, Ppp3r1 (Almeras et al. 2007; Eyles et al. 2007; McGrath et al. 2008). Apolipoprotein E (ApoE) is downregulated in hippocampal tissue from adult rats exposed to DVD deficiency and allelic differences in ApoE have also been observed in postmortem tissue from schizophrenics (Harrington et al. 1995).

The studies discussed above provide a platform on which to begin to unravel the pathways by which DVD deficiency is altering the genome. More research is needed to understand how vitamin D is altering expression of these genes (i.e. epigenetic changes) and if the changes induced by DVD deficiency interact with genetic background.

Conclusions and future directions

The studies discussed above highlight subtle and specific alterations in behavior and neural development that result from exposure to a diet deficient in vitamin D during critical periods of development. These studies have also begun to elucidate the complex underlying mechanisms for how exposure to a diet deficient in vitamin D in gestation can lead to persistent changes in behavior in adulthood. These findings include disruption of normal brain development, dysregulation of specific genes and protein systems and alterations of the dopaminergic system and key neurotrophic factors. However, it is important to note that most of these studies were conducted in Sprague-Dawley rats. There are very few studies of DVD deficiency that have been conducted in mice. Mice are not necessarily a better model system than rats, but do offer some advantages in terms of tools available for genomic analysis and mechanistic studies. Importantly, there are both congruent and disparate results between studies utilizing rats vs. mice (Table 3). It should be noted that the rat studies were conducted on a genetically heterogeneous, outbred background while the mouse studies were performed in genetically homogeneous inbred strains. The observed differences between rats and mice could indicate gene by environment interactions where the effect of exposure to vitamin D deficiency in development on a specific phenotype depends on genetic background. Therefore, one future direction is to assess the effect of DVD deficiency on a genetically diverse set of rodents such as the Collaborative Cross (Churchill et al. 2004) that has been optimized for systems genetics studies or outbred mice (Churchill et al. 2012; Schmidt 2015) to more completely assess gene by environment interactions and potential causative mechanisms.

Table 3.

Comparison of findings of DVD deficiency in rats vs. mice. ND = No Difference Observed

Behavior	Sprague-Dawley rat	Inbred mice	References
PPI	ND	ND	Burne et al. (2014); Harms et al. (2008); Kesby et al. (2006, 2012)
Attentional processing: 5C-SRT	ND	ND	Harms et al. (2012b); Turner et al. (2013)
Sustained attentional processing: response inhibition in 5C-CPT	↓	↓	Harms et al. (2012b); Turner et al. (2013)
LI	Disrupted	ND	Becker et al. (2005); Harms et al. (2012b)
Head dipping: holeboard test	↓	↑	Becker and Grecksch (2006); Becker et al. (2005); Harms et al. (2008)
Habituation: holeboard test	↓	↓ in B6J ND in 129	Becker and Grecksch (2006); Becker et al. (2005); Harms et al. (2008)
Social interaction	ND	ND	Burne et al. (2004a); Harms et al. (2008)
Anxiety-like behavior: EPM	ND	ND	Burne et al. (2004a); Harms et al. (2008)
Spontaneous hyperlocomotion	↑	↑ in 129 ND in B6J	Burne et al. (2004a, 2006); Eyles et al. (2006); Harms et al. (2008); Kesby et al. (2006)
MK-801-induced locomotion	↑	ND	Harms et al. (2012a); Kesby et al. (2006, 2012); O'Loan et al. (2007)
Amphetamine-induced locomotion	↑	ND	Harms et al. (2012a); Kesby et al. (2010)
Lateral ventricle size	↑	↓	Eyles et al. (2003); Fernandes de Abreu et al. (2010); Feron et al. (2005); Harms et al. (2012a); Hawes et al. (2015)

The behavioral effects of DVD exposure vary based on the behavioral assessment used (i.e. hyperlocomotion in various assays and diverse learning tasks). Furthermore, the behavioral effects observed in rats were not always replicable even within the same laboratory, indicating that the behavioral effects of DVD deficiency may be unstable or sensitive to subtle differences in the environment. This is particularly important based on our observation that the majority of the literature on DVD deficiency has come from the same research group at The University of Queensland in Australia. Although this group has carried out a thorough job characterizing numerous behavioral phenotypes and the underlying mechanisms responsible for persistent changes in behavior, there is a need for additional studies by other groups to determine if these subtle behavioral phenotypes can be reproduced.

Collectively, we can conclude that exposure to a diet deficient in vitamin D during development can lead to subtle behavioral changes in adulthood and that these changes are dependent on the timing of the exposure, the model system (rats vs. mice) and genetic background.

These studies highlight the potential importance of monitoring vitamin D levels in pregnant women. The World Health Organization released a report in 2012 and concluded that vitamin D supplementation was not recommended for pregnant women as part of the standard of care (World Health Organization, 2012). However, the effects of in utero exposure to low vitamin D on mental health in adulthood were not considered. More work needs to be carried out in this area if we are to truly understand the role of vitamin D deficiency in schizophrenia and gain a better understanding of the neurobiological systems that are affected and increase risk for this devastating disease.