An in-vivo magnetic resonance imaging study of the olfactory bulbectomized rat model of depression

(1)

Brain Research 879 (2000) 193–199

www.elsevier.com / locate / bres

Short communication

An in-vivo magnetic resonance imaging study of the olfactory bulbectomized rat model of depression

a,b ,

* ´

a c d

Aileen S. Wrynn , Clıona P. Mac Sweeney , Florence Franconi , Laurent Lemaire ,

d e b a

Daniel Pouliquen , Sandra Herlidou , Brian E. Leonard , Jean-Marc Gandon , Jacques D. de Certaines

e

aBiotrial, Preclinical Pharmacology Unit, Rennes, France

bDepartment of Pharmacology, National University of Ireland, Galway, Ireland

cService commun de RMN, Faculte des Sciences, Angers, France´

dLab. Biophysique, Faculte de Medecine, 49045 Angers, France´ ´

eLRMBM, Faculte de Medecine, Rennes, France´ ´ Accepted 27 June 2000

Abstract

The olfactory bulbectomized (OB) rat is a well-accepted animal model of depression. The present magnetic resonance imaging (MRI) investigation demonstrates alterations in signal intensities in cortical, hippocampal, caudate and amygdaloid regions in OB animals, but not in sham operated controls. Ventricular enlargement was also evident in OB animals. These alterations have implications with regard to the face and construct validity of this model.  2000 Elsevier Science B.V. All rights reserved.

Theme: Disorders of the nervous system Topic: Neuropsychiatric disorders

Keywords: Olfactory bulbectomy; Animal model of depression; Degeneration; MRI; Ventricular enlargement

The olfactory bulbectomized (OB) rat model of depres- such as the amygdala [20]. It has been suggested that AD sion has demonstrated impressive selectivity for both treatment may restore the deficits in bulbectomized ani- typical and atypical chronic but not acute antidepressant mals through an attenuation of the impact on information (AD) treatment. Such selectivity paired with alterations processing in such newly formed neuronal circuits [3].

seen in the immune and endocrine systems has led to it Clinical studies of the depressed state have shown being proposed as a realistic animal model of depression increased ventricular size [5], decreased frontal lobe, [4,18]. While bulbectomy is thought to best model agitated caudate and amygdala nuclei volumes as well as lesions in depression [9], the central questions of how bulbectomy the frontal lobes and basal ganglia in depressed patients induces depressive-like symptomology, and why antide- [12,15,19]. In light of such growing evidence of functional pressants alone are active in this model, largely remain and anatomical abnormalities in the affective disorders we unanswered. Following bulbectomy, secondary compensat- sought to investigate possible in-vivo CNS anatomical ory neuronal reorganization, sprouting or changes in abnormalities in this animal model of depression.

synaptic strength (such as denervation supersensitivity) are Rats (200–250 g) were anesthetized with 2,2,2-tri- likely to occur [1] and behavioral abnormalities are now bromoethanol (250 mg / kg i.p.) and either bilateral olfac- attributed to adaptational changes in sub-cortical regions, tory bulbectomy or a sham operation was carried out as previously described [6]. Animals were allowed 14 days recovery. For MR image analysis (N56–10 per group), anesthetized animals (4% halothane / O ; maintained with₂

*Corresponding author. Novartis Pharma AG, Mental Health Unit, WSJ

2% halothane (3 l / min)) were carefully wrapped in an

386.3.26, 4002 Basel, Switzerland.

E-mail address: aileen.wrynn@pharma.novartis.com (A.S. Wrynn). aluminum-coated polyethylene isothermal blanket which

P I I : S 0 0 0 6 - 8 9 9 3 ( 0 0 ) 0 2 6 1 9 - 6

(2)

Fig. 1. (A) Sequential MRI scans of a sham operated control animal demonstrating normal brain morphology and signal intensity.

produced no MRI artifacts. Proton MR images were matrix providing a voxel size of 0.58 mm. The total acquired on a Bruker Avance DRX 300 spectro-imaging acquisition time for the three pulse sequences was 13 min.

system with a super wide bore (15 cm), 7.05 T vertical A superparamagnetic contrast agent was used for a pilot magnet system (Bruker, Wissenbourg, D.). The probe study in a small number of animals (n52 per group) to consisted of a birdcage resonator with an internal diameter improve the delineation of possible diseased regions.

of 64 mm used for both excitation and reception and an Magnetite-dextran (MD2) nanoparticles [11] were injected actively shielded gradient system (max 15 G / cm). After at a dose of 0.8 mg Fe / kg. Once baseline images were anatomical localization by a fast GE pulse sequence (TR acquired, animals were removed, injected intravenously 4.8 ms, TE 2.6 ms, rflip angle 208, FOV 6 cm, matrix size with MD2 and then immediately placed back into the 128, 1ex) ten 1.5 mm transverse slices were simultaneously magnet for a second acquisition. Images were transferred acquired using successively both T1 weighted (TR 500 ms, to a workstation for processing (NMR Win, medical TE 10 ms, 1ex) and T2 weighted (TR 2500 ms, TE 80 ms, imaging system 3.3 1992–1994, DK FZ, Heidelberg, 2ex) spin echo pulse sequences. A 535 cm field of view Darmstadt, Germany). The intensity and area of selected and 1.5 mm slice thickness were used with a 1283128 brain regions were evaluated. Data were analysed using the

(3)

A.S. Wrynn et al. / Brain Research879 (2000) 193 –199 195

Mann–Whitney U-test and were deemed significant when in signal intensity (P,0.05). These results are shown in P#0.05. Results are expressed as median values with Table 1. MD2 particles were not detected in sham operated interquartile ranges. control animals (Fig. 2A), demonstrating an intact blood Fig. 1A and B shows signal hyperintensities in lateral brain barrier (BBB). In contrast, Fig. 2B shows MD2 (P,0.001) and 3rd (P,0.01) ventricular regions in OB particles in the paraventricular nucleus of the hypo- animals, but not in sham operated controls. As a result, it is thalamus in an OB animal.

difficult to distinguish between the lateral ventricles and There are three major findings in this study. First, the the hippocampus in OB animals. Decreased signal intensi- OB model of depression exhibits pronounced enlargement ty was found in frontal (P50.05), cingulate (P,0.01) of the lateral and 3rd ventricles. Second, decreased signal and occipital (P,0.05) cortices. Decreased area of the intensity was observed in the cortex (frontal, occipital and caudate was observed in OB animals (P,0.0001) and was cingulate), the caudate and the amygdala. Third, the accompanied by decreased signal intensity in this region presence of MD2 particles in the hypothalamus of OB (P,0.05). The amygdala exhibited a significant decrease animals may indicate BBB disruption.

Fig. 1. (continued ). (B) Sequential MRI scans of an olfactory bulbectomized rat. The bulbectomized animal exhibits pronounced enlargement of the ventricles in addition to decreased caudate volume and cortical, amygdaloid and caudate signal hypointensities (see Table 1).

(4)

Fig. 2. (A) Sequential MRI scans of a sham operated control animal, following injection of a superparamagnetic contrast agent, magnetite-dextran (MD2), showing normal morphology and absence of contrast agent deposition.

MRI signal alteration indicates a change in the water that patients with enlarged ventricles tend to have lower content of the extracellular tissue space due to either global metabolism [5].

altered tissue volume / compression or tissue damage. Ven- A correlation exists between ventricle size and cognitive tricular enlargement may occur due to decreased CSF impairment in depression [5] and it now seems that absorption secondary to inflammation associated with impaired hippocampal function as a consequence of ven- trauma following surgery [13]; alternatively, enlargement tricular enlargement and / or hippocampal oedema may can simply reflect decreased size of nearby structures as is underlie spatial memory deficits in bulbectomized animals thought to occur in clinical studies of bipolar disorder [17]. [6]. Furthermore, our finding of decreased signal intensity Whatever the mechanism, a consequence would be altered in the amygdala is interesting given that amygdaloid blood cerebral blood flow. Hence, in this model, altered metabo- flow and metabolism consistently correlate positively with lism would most likely occur in affected regions of the depression severity [2]. This region has been proposed as a cortex, amygdala and the caudate putamen, areas which are site for the behavioral alterations and AD action in the relevant to clinical investigations of the depressed state and model [4,6,20].

AD action. It is interesting that Kellner et al. suggested We believe from our findings with MD2 that the OB rat

(5)

A.S. Wrynn et al. / Brain Research879 (2000) 193 –199 197

Fig. 2. (continued ). (B) Sequential MRI scans of an olfactory bulbectomized rat following injection of a superparamagnetic contrast agent, magnetite- dextran (MD2). The scans show pronounced ventricular enlargement as previously seen in Fig. 1. Deposition of MD2 nanoparticles is evident in the paraventricular hypothalamic area, suggesting focal disruption of the blood brain barrier at this point.

may have an altered BBB. Interestingly, the disruptive and possible BBB alteration similar to clinical studies in effects of stress on the integrity of the BBB have been depression. While it may be incongruous to suggest that highlighted in both humans and animals [14] and a large the OB rat may be the closest parallel to clinical depres- percentage of depressed (37%) [10] patients have impair- sion that can be achieved, certain aspects certainly mimic ment of the blood–CSF barrier. Alterations in such barriers those seen in the clinic and it now seems that this is also provide an opportunity for peripheral immune cells and true for some structural changes in the brain. Placed cytokines to enter the nervous system and alter brain together with reports demonstrating plasticity-augmenting function. In this and other animal models of stress and effects of AD [7] and AD actions of neurotrophic factors depression, elevated plasma levels of cytokines are evi- [16] it is not unreasonable to suggest that the OB rat may dent, some of which can be attenuated by chronic AD ultimately provide a template for the future development of

treatment [8]. compounds directed toward the neurodegenerative aspects

To our knowledge this is the only animal model of of the affective disorders and possibly towards the negative depression that exhibits central structural brain changes symptoms of schizophrenia.

(6)

Table 1

In-vivo MRI analysis of specific brain nuclei in sham and olfactory bulbectomized (OB) animals. The results are expressed as median values with interquartile ranges. Data analysed by Mann–Whitney U-test

Region Group N Intensity (grey level) Area (mm )2

Median Q1–Q3 Median Q1–Q3

Cortex

Frontal Sham 8 110.0 97.1–113.4 211.5 184.0–286.0

OB 10 102.7* 96.5–106.5 197.0 174.0–244.0

Occipital Sham 7 109.2 98.4–115.7 142.0 116.0–151.0

OB 10 99.0* 94.1–104.9 144.0 123.0–193.0

Cingulate Sham 7 111.0 97.8–115.0 22.0 16.0–26.0

OB 10 100.6** 95.8–104.7 21.0 16.0–27.0

Hippocampus

Anterior Sham 7 106.5 105.5–117.2 Boundaries not clearly

OB 10 103.5 97.6–107.8 defined

Posterior Sham 6 118.7 100.2–127.8 Boundaries not clearly

OB 6 113.0 110.9–119.5 defined

Limbic structures

Caudate Sham 6 115.4 97.7–122.4 192.0 161.0–258.0

OB 10 103.8* 100.2–108.5 118.0*** 113.0–146.0

Septum Sham 6 120.0 111.4–122.1 61.5 51.0–69.0

OB 10 111.4 108.2–113.7 58.0 52.0–66.0

Thalamus Sham 7 100.4 93.0–109.0 157.0 113.0–197.0

OB 10 97.6 89.3–105.2 132.0 112.0–214.0

Hypothalamus Sham 7 72.7 64.0–91.1 88.0 72.0–107.0

OB 10 79.0 57.6–85.2 80.5 72.0–85.0

Amygdala Sham 7 131.6 119.7–136.9 251.0 245.0–344.0

OB 10 117.0** 108.7–123.0 242.5 221.0–262.0

Ventricles

Lateral Sham 6 113.9 129.5–143.7 Boundaries not clearly

OB 10 188.4*** 177.3–202.8 defined

3rd Sham 6 153.8 138.7–163.5 Boundaries not clearly

OB 10 181.5** 178.4–190.2 defined

*P#0.05, **P,0.01, ***P,0.001.

[9] A.R. Lumia, M.H. Teicher, F. Salchli, E. Ayers, B. Possidente,

References

Olfactory bulbectomy as a model for agitated hyposerotonergic depression, Brain Res. 587 (1992) 181–185.

[1] P.C. Brunjes, Lessons from lesions: the effects of olfactory bulbec- [10] A.F. Pitts, B.T. Carroll, T.L. Gehris, R.G. Kathol, Elevated CSF tomy, Chem. Senses 17 (1992) 729–763. protein in male patients with depression, Biol. Psychiatry 28 (1990) [2] W.C. Drevets, Functional neuroimaging studies of depression: the 629–637.

anatomy of melancholia, Annu. Rev. Med. 49 (1998) 341–361. [11] D. Pouliquen, C. Chouly, Magnetic microcarriers for medical [3] G. Grecksch, D. Zhou, C. Franke, U. Schroder, S. Bernhard, A.¨ applications, in: R. Arshady (Ed.), Microspheres, Microcapsules and

Becker, G. Huether, Influence of olfactory bulbectomy and sub- Liposomes, Vol. 2, Citus Press, London, 1998, pp. 339–378.

sequent imipramine treatment on 5-hydroxytryptaminergic pre- [12] G. Rajowska, J.J. Miguel-Hidalgo, J. Wei, G. Dilley, S.D. Pittman, synapses in the rat frontal cortex: behavioural correlates, Br. J. H. Meltzer, J.C. Overholser, B.L. Roth, C.A. Stockmeier, Mor- Pharmacol. 122 (1997) 1725–1731. phometric evidence for neuronal and glial prefrontal cell pathology [4] J.A. Jesberger, J.S. Richardson, Animal models of depression: in major depression, Biol. Psychiatry 45 (1999) 1085–1098.

parallels and correlates to severe depression in humans, Biol. [13] L.P. Rowland, Blood–brain barrier, cerebrospinal fluid, brain edema, Psychol. 20 (1985) 764–784. and hydrocephalus, in: E.R. Kandel, J.H. Schwartz, T.M. Jessell [5] C.H. Kellner, D.R. Rubinow, R.M. Post, Cerebral ventricular size (Eds.), Principles of Neural Science, 2nd ed., Elsevier, New York,

and cognitive impairment in depression, J. Affect. Disord. 10 (1986) 1991, pp. 837–844.

215–219. [14] H.S. Sharma, J. Cervos-Navarro, P.K. Dey, Increased blood–brain

[6] J.P. Kelly, A.S. Wrynn, B.E. Leonard, The olfactory bulbectomized barrier permeability following acute short-term forced swimming rat as a model of depression: an update, Pharmacol. Ther. 75 (1997) exercise in conscious normotensive young rats, Neurosci. Res. 10

1–17. (1991) 211–221.

[7] I. Kitayama, T. Yaga, T. Kayahara, K. Nakano, S. Murase, M. Otani, [15] Y.I. Sheline, M. Sanghavi, M.A. Mintun, M.H. Gado, Depression J. Nomura, Long-term stress degenerates, but imipramine regener- duration but not age predicts hippocampal volume loss in medically ates, noradrenergic axons in the rat cerebral cortex, Biol. Psychol. healthy women with recurrent major depression, J. Neurosci. 19

42 (1997) 687–696. (1999) 5034–5043.

[8] L.G. LeMay, A.J. Vander, M.J. Kluger, The effects of psychological [16] J.A. Siuciak, D.R. Lewis, S.J. Wiegand, R.M. Lindsay, Antidepres- stress on plasma interleukin-6 activity in rats, Physiol. Behav. 47 sant-like effects of brain-derived neurotrophic factor (BDNF),

(1990) 957–961. Pharmacol. Biochem. Behav. 56 (1997) 131–137.

(7)

A.S. Wrynn et al. / Brain Research879 (2000) 193 –199 199 [17] J.C. Soares, J.J. Mann, The anatomy of mood disorders — review of [19] P. Videbech, MRI findings in patients with affective disorder: a

structural neuroimaging studies, Biol. Psychiatry 41 (1997) 86–106. meta-analysis, Acta Psychiatr. Scand. 96 (1997) 157–168.

[18] C. Song, B.E. Leonard, Serotonin reuptake inhibitors reverse the [20] A.S. Wrynn, J.B. Sebens, T. Koch, B.E. Leonard, J. Korf, Prolonged impairments in behaviour, neurotransmitter and immune functions in c-jun expression in the basolateral amygdala following bulbectomy:

the olfactory bulbectomized rat, Hum. Psychopharmacol. 9 (1994) possible implications for antidepressant activity and time of onset,

135–146. Mol. Brain Res. 76 (2000) 7–17.