Elevated serum levels of FGF-2, NGF and IGF-1 in patientswith manic episode of bipolar disorder

Xiaohua Liu a,n, Tianhong Zhang a, Shen He a, Bo Hong a, Zheng Chen a, Daihui Peng a,Yan Wu a, Hui Wen a, Zhiguang Lin b, Yiru Fang a, Kaida Jiang a

a Department of Psychiatry, Shanghai Mental Health Center, Shanghai Jiao Tong University School of Medicine, 600 South Wanping Road,Shanghai 200030, PR Chinab Biochemistry Laboratory, Shanghai Mental Health Center, Shanghai Jiao Tong University School of Medicine, Shanghai, PR China

a r t i c l e i n f o

Article history:Received 18 September 2013Received in revised form15 February 2014Accepted 29 March 2014

Keywords:Manic episode of bipolar disorderAngioneurinsNeurotrophicSerumStress

a b s t r a c t

Multiple neurotrophic and/or growth factors, recently nominated as “angioneurins”, play the key roles inmood modulation and neuroplasticity, and their dysfunction might be involved in the pathophysiology andtreatment of mood disorders. We examined serum levels of vascular endothelial growth factor (VEGF),fibroblast growth factor (FGF)-2, nerve growth factor (NGF) and insulin-like growth factor (IGF)-1 in 70 drug-naïve or drug-free patients with manic episode of bipolar disorder and 50 healthy controls, using theenzyme-linked immunosorbent assay (ELISA) method. The results showed that mean serum levels of VEGF,FGF-2, NGF and IGF-1 were 168.137225.61 pg/ml, 279.097378.62 pg/ml, 61.387171.67 pg/ml and162.01772.00 ng/ml in patients, and 140.807143.71 pg/ml, 275.467235.29 pg/ml, 36.34715.14 pg/mland 138.90780.11 ng/ml in healthy controls, respectively. Serum levels of FGF-2, NGF and IGF-1 in patientswere significantly higher than those in healthy controls, though there was no statistical difference in serumVEGF level between two groups. Moreover, serum NGF level in patients was significantly correlated withduration of disorder and times of previous manic episodes. We conclude that the increase of serum FGF-2,NGF and IGF-1 levels in manic state of bipolar disorder may be associated with their compensatory roles ofneuroprotection and angiogenesis, and these angioneurins may be involved in the pathophysiology ofbipolar disorder.

& 2014 Elsevier Ireland Ltd. All rights reserved.

1. Introduction

Recent studies have demonstrated that there are impairments ofneuron–glia–vascular network in the brain, so called neuroplasticity,and that these impairments are associated with mood disorders(Pittenger and Duman, 2008). Emerging evidence has shown thatmultiple neurotrophic and/or growth factor systems, such as brain-derived neurotrophic factor (BDNF), vascular endothelial growth factor(VEGF), fibroblast growth factor (FGF)-2, nerve growth factor (NGF),insulin-like growth factor (IGF)-1, etc., play the key roles in neuroplas-ticity. And the dysfunction of neurotrophic/growth factor systemsmight be critically involved in the pathophysiology and treatmentof mood disorders (Evans et al., 2004; Hashimoto et al., 2004; Warner-Schmidt and Duman, 2007). It is noteworthy that, acting as molecularand functional basis of the neurovascular link, these growth factorswere recognized to exert the effect on both neural and vascularcell functions, hence their recent nomination as “angioneurins”(Quaegebeur et al., 2011; Segura et al., 2009; Zacchigna et al., 2008).

In a broad sense, angioneurins might play the role in regulating theangiogenesis, blood–brain barrier (BBB) integrity, neuroprotection,neuroregeneration and neuroplasticity. Here, several growth factorsincluding VEGF, FGF-2, NGF and IGF-1 attracted our particular atten-tion due to such dual neurovascular properties.

VEGF, as a prototypic angioneurin, is able to induce vascularendothelial proliferation, migration, survival and vaso-permeabilityin different cell types (Ferrara et al., 2003). Beyond its role inangiogenesis, VEGF has neurogenic and neurotrophic effects (Jinet al., 2002; Sun and Guo, 2005). VEGF has also been demonstratedto promote the survival of adult neurons, increase the proliferationof neurons, enhance neurite extension, stimulate synaptic transmis-sion and remyelination, and so on (Cao et al., 2004; Newton et al.,2003; Segi-Nishida et al., 2008; Warner-Schmidt and Duman, 2007;Zacchigna et al., 2008). The findings that genetic variation of VEGFinfluences human hippocampal morphology implied that the pleio-tropic effects of VEGF have been extended to humans (Blumberget al., 2008). FGF-2 is the first and also known as basic member ofFGF family (Abraham et al., 1986). Growing evidence has revealedthat the FGFs play the important role in the central nervous system(CNS), expanding from their classical involvement in development,neuronal repair and protection to a potential influence on learning

Contents lists available at ScienceDirect

journal homepage: www.elsevier.com/locate/psychres

Psychiatry Research

http://dx.doi.org/10.1016/j.psychres.2014.03.0420165-1781/& 2014 Elsevier Ireland Ltd. All rights reserved.

n Corresponding author. Tel.: þ86 21 64387250x73528; fax: þ86 21 64387986.E-mail address: drliuxiaohua@gmail.com (X. Liu).

Please cite this article as: Liu, X., et al., Elevated serum levels of FGF-2, NGF and IGF-1 in patients with manic episode of bipolar disorder.Psychiatry Research (2014), http://dx.doi.org/10.1016/j.psychres.2014.03.042i

Psychiatry Research ∎ (∎∎∎∎) ∎∎∎–∎∎∎

and neuronal plasticity (Reuss and von Bohlen und Halbach,2003b). As well, FGFs have been considered to have vasculo-protective effects on quiescent vessels. Among them, FGF-2 canmaintain the integrity of BBB by regulating the expression ofendothelial-junction molecules (Bendfeldt et al., 2007; Reusset al., 2003a). Recent postmortem studies, which reported dysre-gulated expressions of FGF transcripts including FGF-2 in majordepression, provided further evidence that neurotrophic moleculesmight be relevant to the etiology and treatment of mood disorders(Evans et al., 2004; Riva et al., 2005). NGF, represents the firstisolated and best characterized member of neurotrophins family, isessential for the development, maintenance and functional integrityof neurons in the peripheral and CNS (Aloe et al., 1997). NGF viabinding and activation of its receptors, TrkA and p75, triggersnumerous intracellular signaling cascades including Ras/Raf/Mek/Erk1/2, PI3K/PDK1/Akt, PLCγ/PKC, etc., which facilitate neuropro-tective functions and neural repair (Sofroniew et al., 2001). Notably,these signal pathways mediating neurotrophic and neuroplasticfunctions get involved in pathophysiology of mood disorders, andare also targets by mood stabilizers and antipsychotics (Chen andManji, 2006; Einat et al., 2003). In addition, NGF plays a modulatoryrole in the hypothalamic–pituitary–adrenal (HPA) axis, and alsoexerts its effect on the maintenance of a balanced interplay betweenthe nervous, endocrine and immune systems (Aloe et al., 2002).Likewise, NGF can enhance the angiogenesis, either by promotingvascular cell growth directly, or by increasing the VEGF expressionindirectly (Zacchigna et al., 2008). A recent review highlighted thatNGF is implicated in neurobehavioral responses associated withstress-related events and psychiatric disorders (including depres-sion) (Bersani et al., 2000). As for IGF-1, it is the best characterizedone of several dozen IGFs. IGF-1 is a multifaceted growth factor, andits receptor is expressed by multiple types of cells includingendothelial cells, smooth muscle cells, neurons and glia. Unlikeother angioneurins, IGF-1 could promote the angiogenesis byregulating the proliferation of vascular smooth muscle cells(VSMCs) (Cheng and Du, 2007). IGF-1 may stimulate the pre- andpost-synaptic components and increase the neurogenesis andsynaptogenesis besides its effects on somatic growth and develop-ment (O'Kusky et al., 2000). So it is also a crucial mediator ofneurogenesis, myelination and remyelination, neuroprotection andneuromodulation. The IGF-1 signaling cascade was considered toactivate the PI3K and ras/MAPK pathways, which demonstrated theimplication in neuroplasticity, learning, memory, and mood mod-ulation (Einat et al., 2003; McIntyre et al., 2008; Sonntag et al.,2005).

Taken altogether, mood disorders can result at least partially fromgeneralized neurotrophic impairments which may eventually lead tothe structural and functional abnormalities of neural networks in thebrain (Duman et al., 1997; Pittenger and Duman, 2008). It might beplausible that these growth factors, namely angioneurins, participatein the pathogenesis of mood disorders through their multiple effectson vasculature and/or neuroplastic pathways. Yet, there remainsinsufficient in clinical data about the levels of these four factors inpatients with mood disorders, especially manic episode. As well as,the results so far have been controversial. Herein, we tested ourspeculation and investigated possible alterations of serum concen-trations of VEGF, FGF-2, NGF and IGF-1 in patients with manicepisode of bipolar disorder by comparing to those in healthy controls.

2. Methods

2.1. Subjects

We recruited 70 patients who met the criteria of the Diagnostic and StatisticalManual of Mental Disorders, 4th edition (DSM-IV, American Psychiatry Association,1994) for bipolar disorder, manic episode, and 50 healthy controls in Shanghai

Mental Health Center, Shanghai Jiao Tong University School of Medicine. The studypopulation, between the ages 16 and 65 years, consisted entirely of ethnic Chinesepeople. All patients were either drug-naïve or drug-free for at least four weeksprior to entering the present study.

Self-designed questionnaire was applied to collect the sociodemographic andclinical information. This questionnaire was also used for screening and excludingthe subjects with previous or current physical diseases including cardiovasculardiseases (i.e. hypertension, coronary artery disease), cerebrovascular diseases (i.e.stroke), endocrine and metabolic diseases (i.e. diabetes mellitus, thyroid diseases),as well as the patients with psychiatric comorbidities including schizophrenia orother psychotic disorders, alcohol and substance abuse, anxiety disorder, etc. In themeanwhile, the medical records of all subjects were referred. The healthy controlswere required to have no personal or familial psychiatric histories. And the subjectscurrently pregnant or lactating were also excluded. The severity of manicsymptoms was measured using Bech-Rafaelsen Mania Rating Scale (BRMS) andYoung Mania Rating Scale (YMRS). All clinical assessments were performed by twotrained psychiatrists independently.

This study protocol was approved by the Ethics Committee of Shanghai MentalHealth Center, Shanghai Jiao Tong University School of Medicine. Written informedconsent was obtained from all included subjects after a complete and extensivedescription of the study profile.

2.2. Blood sample collection and measurement of four factors

Blood samples were drawn from the antecubital veins of all participants inanticoagulant-free tubes between 7 and 9 a.m. after an overnight fast. They werekept at room temperature for 2 h, followed by centrifugation (1710g for 20 min at4 1C) for serum separation. Serum samples were stored at �80 1C till the time ofassay. Repeated freeze–thaw cycles were avoided.

Serum levels of VEGF, FGF-2, NGF and IGF-1 for each sample were measured induplicate by the enzyme-linked immunosorbent assay (ELISA) method usinghuman VEGF DuoSet (Catalog#: DY293B, R&D system, USA), human bFGF ELISAKit (Catalog#: EL10048, Anogen, Canada), ChemiKineTM NGF Sandwich ELISA Kit(Catalog#: CYT304, Chemicon (Millipore), USA) and human IGF-1 DuoSet (Cata-log#: DY291, R&D system, USA), according to the manufacturer's protocols,respectively. The concentrations of VEGF, FGF-2 and NGF were expressed as pg ofprotein/ml of serum, and that of IGF-1 was expressed as ng of protein/ml of serum.The detection limits for VEGF, FGF-2, NGF and IGF-1 were 31.2 pg/ml, 80 pg/ml,10 pg/ml and 31.2 pg/ml, respectively. The intra- and inter- assay coefficients ofvariation were below 9%.

2.3. Statistical analysis

Serum levels of VEGF, FGF-2, NGF and IGF-1 were not normally distributedbased on the Kolmogorov–Smirnov test (Z¼0.251, P¼0.000; Z¼0.379, P¼0.000;Z¼0.430, P¼0.000; Z¼0.117, P¼0.000, respectively), and we failed to normalizethe distribution of data after log-transformation. Therefore, continuous variableswere described as means7standard deviation (S.D.) and analyzed using two-tailedindependent samples Mann–Whitney U-test. Categorical variables were describedas frequencies (proportions) and analyzed using chi-square tests. The relationshipsbetween serum levels of four factors and sociodemographic and clinical data wereanalyzed using two-tailed Spearman's correlation test. All statistical analyses wereperformed using SPSS 16.0 software for Windows, with the cutoff P value setat 0.05.

3. Results

The sociodemographic and clinical characteristics of all subjectsare shown in Table 1. No significant differences were found in theage, gender and educational level but marital status betweenmanic patients and healthy controls. In our patients, 91.4%was recurrent and the mean age at first-onset was 26 years. Theclear times (shown as means7S.D.) of previous episodes were3.9878.99 with manic episodes (n¼63) and 1.0970.76 withdepressive episodes (n¼66), according to the description fromthe patients or informants. Most of the patients were prescribed toone or a combination of antidepressants, antipsychotics, lithium,anticonvulsants or electroconvulsive therapy during their previousepisodes. However, the detailed information about the name anddosage of medications could not be recalled accurately. At present,all cases (including six drug-free patients and 64 drug-naïvepatients) suffered from moderate to severe manic episode foralmost eight weeks.

X. Liu et al. / Psychiatry Research ∎ (∎∎∎∎) ∎∎∎–∎∎∎2

Please cite this article as: Liu, X., et al., Elevated serum levels of FGF-2, NGF and IGF-1 in patients with manic episode of bipolar disorder.Psychiatry Research (2014), http://dx.doi.org/10.1016/j.psychres.2014.03.042i

The serum levels (shown as means7S.D.) of VEGF, FGF-2, NGFand IGF-1 were 168.137225.61 pg/ml, 279.097378.62 pg/ml,61.387171.67 pg/ml and 162.01772.00 ng/ml in mania patients,and 140.807143.71 pg/ml, 275.467235.29 pg/ml, 36.34715.14 pg/ml and 138.90780.11 ng/ml in healthy controls, respectively.The serum levels of FGF-2, NGF and IGF-1 in patients weresignificantly higher than those in healthy controls (Z¼�2.896,P¼0.004; Z¼�2.050, P¼0.040; Z¼�2.188, P¼0.029, respectively),though there was no significant difference in serum VEGF levelbetween two groups (Z¼�0.468, P¼0.639) (see Fig. 1).

Correlation analysis showed that serum NGF level in maniapatients was negatively correlated with duration of disorder(rs¼�0.241, P¼0.044) and times of previous manic episodes(rs¼�0.275, P¼0.029), whereas no other correlations weredemonstrated between serum levels of these four factors, andclinical characteristics (Table 2) (P all 40.05).

4. Discussion

“Angioneurins” newly nominated has been recognized to affectboth neural and vascular cell process by closely developing theneurovascular crosstalk (Quaegebeur et al., 2011; Segura et al.,2009; Zacchigna et al., 2008). VEGF is one striking example which

is a key regulator of angiogenesis. Some molecules such as BDNFand NGF have been demonstrated to promote angiogenesisalthough originally they were discovered as neurotrophic factors.In addition, other molecules including FGF-2 and IGF-1 that canexert the effect on neurons and vascular cells either directly orindirectly have also been attributed to angioneurins (Zacchignaet al., 2008). Over the past years, the relationships between serum/plasma BDNF levels and mood disorders, mostly depression, havelong been given extensive attention. There seems to be insufficientand inconsistent exploration about levels of VEGF, FGF-2, NGF andIGF-1 in patients with mood disorders. Thus, we determinedserum levels of these four growth factors in patients with currentmanic episode of bipolar disorder and healthy people. To the bestof our knowledge, this is the first study to detect FGF-2 levels inmanic patients. And to date, there are few studies on VEGF, NGFand IGF-1 levels and bipolar disorder.

Our findings revealed that there was no significant difference inserum VEGF level between manic patients and healthy controls,although Lee and Kim (2012) found that both depressive andmanic patients had significantly elevated plasma VEGF levels (Leeand Kim, 2012). Here, the differences on clinical characteristics ofthe patients might contribute to explain this discrepancy. In thepatients of Lee and Kim's study, the mean duration of illness was24.2 months, while the present episode lasted for almost 2 months

Table 1Sociodemographic and clinical characteristics of manic patients and healthy controls.

Mania (n¼70) HC (n¼50) Statistics

Age (years) 37.91714.53 36.78711.20 Z¼�0.139, P¼0.890a

Sex χ2(1)¼0.025, P¼0.875b

Male 29 (41.4%) 20 (40.0%)Female 41 (58.6%) 30 (60.0%)

Education χ2(2)¼4.303, P¼0.116b

Elementary school and below 18 (25.7%) 11 (22.0%)High school 21 (30.0%) 8 (16.0%)University and above 31 (44.3%) 31 (62.0%)

Marital status χ2¼8.104, P¼0.027cn

Married 34 (48.6%) 32 (64.0%)Single 27 (38.6%) 18 (36.0%)Divorced 8 (11.4%) 0 (0.0%)Widow 1 (1.4%) 0 (0.0%)

Age of first-episode (years) 26.13710.34 – –

Duration of disorder (weeks) 634.547665.95 – –

Duration of the present episode (weeks) 7.80710.71 – –

PatientsDrug-free 6 (8.6%) – –

Drug-naïve 64 (91.4%) – –

With first episode 6 (8.6%) – –

With recurrent episode 64 (91.4%) – –

Times of previous episodesManic 3.9878.99, n¼63d –

Depressive 1.0970.76, n¼66d –

Psychotic symptomsWith 6 (8.6%) – –

Without 64 (91.4%) – –

Familial psychiatric historyWith 22 (31.4%) – –

Without 48 (68.6%) – –

BRMS scores 23.0176.99 – –

YMRS scores 29.9979.84 – –

BRMS: Bech-Rafaelsen Mania Rating Scale; YMRS: Young Mania Rating Scale; HC: healthy controls.Data are presented as mean7standard deviation (S.D.) or frequency (proportion).

a Mann–Whitney U-test.b Pearson chi-square test.c Fisher's Exact test.d Clear times and type of previous episodes that patients or informants could recall and sample sizes.n Po0.05.

X. Liu et al. / Psychiatry Research ∎ (∎∎∎∎) ∎∎∎–∎∎∎ 3

Please cite this article as: Liu, X., et al., Elevated serum levels of FGF-2, NGF and IGF-1 in patients with manic episode of bipolar disorder.Psychiatry Research (2014), http://dx.doi.org/10.1016/j.psychres.2014.03.042i

in our study even though total duration spanned more than 12years. And different from Lee and Kim's study, only the drug-freeand drug-naïve patients were included in the present study, thusthis generally ruled out the effects of medications that have beenshown to enhance the neuroprotection and angiogenesis (Chiuet al., 2013). From a physiological perspective, VEGF is a prominentfactor that promotes angiogenic process in which new capillariesgenerate on existing blood vessels via directed proliferation andmigration of endothelial progenitor cells, further provide neuro-trophic support for concurrent neurogenesis and synaptogenesis(Beck and Plate, 2009; Ferrara et al., 2003). And VEGF signalinginfluences multiple functions of neurovascular units where thevasculature can affect neuronal activity, vice versa, dynamicallyadjust to its change (Licht et al., 2011; Nowacka and Obuchowicz,2012). Mounting evidence has indicated that several systems suchas monoamine neurotransmitters, cytokines, HPA axis, even the

neuroimmunologic networks that are composed of the abovesystems participate in the modulation of VEGF signaling (Fornaroet al., 2013). There might be a homeostatic mechanism involvingVEGF signaling, that is, VEGF signaling is down-regulated duringacute stress due to hypercortisolism and up-regulated because ofsustained stress and/or compensatory mechanisms (Fornaro et al.,2013). Therefore, we speculate that this may be the critical reasonthat the inconsistency between Lee and Kim's and our studyappeared since our patients were probably in the transition zonefrom acute to chronic stress.

FGF-2 level in manic patients was firstly shown to be higherthan that in healthy controls according to the present study.Recent postmortem studies reported the decreased expressionsof FGF-2 transcript in frontal cortical regions and decreaseddensity of FGF-2 mRNAþ cells in hippocampus from subjectswith major depression, but not bipolar disorder or schizophrenia

Fig. 1. Comparisons between serum levels of VEGF, FGF-2, NGF and IGF-1 in mania patients and healthy controls. Mania: n¼70; HC (healthy controls): n¼50. Serum levels ofVEGF, FGF-2 and NGF were expressed as pg/ml, and serum level of IGF-1 was expressed as ng/ml. (A) Comparison between serum VEGF levels in mania patients and healthycontrols (mania: 168.137225.61 pg/ml; HC: 140.807143.71 pg/ml). (B) Comparison between serum FGF-2 levels in mania patients and healthy controls (mania:279.097378.62 pg/ml; HC: 275.467235.29 pg/ml). (C) Comparison between serum NGF levels in mania patients and healthy controls (mania: 61.387171.67 pg/ml; HC:36.34715.14 pg/ml). (D) Comparison between serum IGF-1 levels in mania patients and healthy controls (mania: 162.01772.00 ng/ml; HC: 138.90780.11 ng/ml). The barsrepresent the mean and the error bars represent the standard deviation (S.D.). nSignificant differences between mania patients and healthy controls, Mann–Whitney U-test atPo0.05.

Table 2Correlations between serum levels of VEGF, FGF-2, NGF and IGF-1 and clinical characteristics in manic patients (n¼70).

VEGF (pg/ml) FGF-2 (pg/ml) NGF (pg/ml) IGF-1 (ng/ml)

Age (years) rs¼0.078, P¼0.519 rs¼�0.130, P¼0.283 rs¼�0.163, P¼0.179 rs¼0.216, P¼0.073Age of first-episode (years) rs¼0.075, P¼0.535 rs¼�0.118, P¼0.331 rs¼0.159, P¼0.190 rs¼0.099, P¼0.417Duration of disorder (weeks) rs¼0.035, P¼0.771 rs¼�0.031, P¼0.802 rs¼�0.241, P¼0.044n rs¼0.147, P¼0.226Duration of present episode (weeks) rs¼0.021, P¼0.862 rs¼0.005, P¼0.964 rs¼0.113, P¼0.353 rs¼0.008, P¼0.945Times of manic episodesa rs¼�0.023, P¼0.86 rs¼�0.052, P¼0.686 rs¼�0.275, P¼0.029n rs¼�0.046, P¼0.719Times of depressive episodesb rs¼0.000, P¼0.996 rs¼�0.068, P¼0.587 rs¼�0.048, P¼0.703 rs¼0.191, P¼0.125BRMS scores rs¼0.058, P¼0.636 rs¼0.000, P¼0.994 rs¼0.074, P¼0.542 rs¼0.019, P¼0.876YMRS scores rs¼0.010, P¼0.938 rs¼�0.021, P¼0.865 rs¼0.074, P¼0.541 rs¼�0.044, P¼0.720

rs: Spearman's correlation coefficient.BRMS: Bech-Rafaelsen Mania Rating Scale; YMRS: Young Mania Rating Scale.

a Clear times of previous episodes that patients or informants could recall, n¼63.b Clear times of previous episodes that patients or informants could recall, n¼66.n Po0.05.

X. Liu et al. / Psychiatry Research ∎ (∎∎∎∎) ∎∎∎–∎∎∎4

Please cite this article as: Liu, X., et al., Elevated serum levels of FGF-2, NGF and IGF-1 in patients with manic episode of bipolar disorder.Psychiatry Research (2014), http://dx.doi.org/10.1016/j.psychres.2014.03.042i

(Evans et al., 2004; Gaughran et al., 2006). However, several linesof evidence have confirmed that a decrease in the number ofcortical and hippocampal glial cells which express some membersof FGF family occurred in mood disorders (Rajkowska, 2000). Forthis reason, the decline of the FGFs levels in major depressionmight be a consequence of chronic progression. On the other side,stressful situations which represent the major risks of mooddisorders also can regulate the expression levels of FGF-2 in theadult brain (Riva et al., 2005). An acute stress, similar as activationof HPA axis in current manic episode, induces a significant increaseof FGF2 mRNA levels in various brain regions, while this responsemight imply an attempt to cope with an adverse event (Molteniet al., 2001). Hypothetically, other compensatory mechanisms ofvascular protection or maintenance of BBB integrity lead to anincrease of serum FGF-2 level in manic patients.

In the present study, serum NGF level in patients significantlyincreased when compared to those in healthy controls, and thisresult appears to be contradictory to the findings from Barbosaet al.'s study. Barbosa et al. (2011) found that plasma NGFconcentration of bipolar disorder patients in manic state waslower than that in controls. Moreover, patients in manic statehad lower NGF levels than euthymic patients or controls, and NGFlevel was negatively correlated with the severity of mania (Barbosaet al., 2011). Yet our results demonstrated that serum NGF level inmanic patients was elevated notably. And serum NGF level wasnegatively correlated with duration of disorder and times ofprevious manic episodes, but was not associated with severity ofmanic symptoms. Recent data has suggested that the changes inneuroplasticity could be associated with cumulative mood epi-sodes, especially manic episodes. A number of neurotrophicfactors including BDNF might play a key role in the transductionof stress and recurrent episodes into pathophysiology of bipolardisorder (Kapczinski et al., 2008). Probably NGF, similar as BDNF,also plays a modulatory role in the HPA axis and exerts its effect onthe maintenance of a dynamic balance between the neuro-endo-crine–immune systems (Aloe et al., 2002). Therefore, plasma orserum NGF levels in patients with bipolar mania might reflect, to acertain extent, the cumulative effect of mood episodes, particularlymanic episodes (Kapczinski et al., 2008). As a result, the discre-pancy may have occurred because the clinical characteristics weredifferent between Barbosa et al.'s study and ours. All includedpatients in their study were medicated by mood stabilizers and/ortypical antipsychotics though they thought medications had noeffect on NGF concentrations, whereas our patients were eitherdrug-naïve or drug-free for at least four weeks. In addition, totalduration of Barbosa et al.'s sample was longer than that of ours,combined with our finding in correlation analysis, which made uspostulate that decreased NGF levels in their study might notreflect the condition of the acute or early manic phase in bipolardisorder.

As far as IGF-1 was concerned, the present study revealed thatthe manic patients had significantly higher level than healthycontrols. This result, consistent with the findings from Kim et al.(2013), corroborates the view that IGF-1 plays a crucial role in thepathophysiology of bipolar disorder (Kim et al., 2013), although itwas inconsistent with another recent study that there was nosignificant alteration of IGF-1 levels at different stages of bipolardisorder (Palomino et al., 2013). Besides, Squassina et al. (2013)compared gene expression levels in lymphoblastoid cell linesfrom bipolar disorder with lithium responders, lithium non-responders and healthy controls, and found that expression ofIGF-1 was up-regulated significantly in the patients with lithiumresponders. Their findings indicated that IGF could be importantfor lithium to exert its mood stabilizing effect (Squassina et al.,2013). Another postmortem study reported a significantly down-regulated expression of IGF binding protein (IGFBP)-2 in prefrontal

cortex from patients with bipolar disorder compared to controls,particularly in patients remotely treated or not treated withlithium at time of death. IGFBP-2 is the principle carrier proteinof IGF-1 in CNS, and the decreased IGFBP-2 expression may be amanifestation of adaptive response to low bioavailability of IGF.Moreover, Bezchlibnyk et al. (2007) did not observe significantdifference in IGFBP-2 expression between patients with majordepressive disorder and controls. It is possible that alteredexpression of IGFBP-2 represents a trait-specific marker for bipolardisorder, and IGFBP-2 level can be a function of mood state, withalterations being specific to mania (Bezchlibnyk et al., 2007).Furthermore, the central effects of IGF-1 in feedback loop ofperipheral IGF-1 production probably are weakened by a decreasein sensitivity of IGF receptor (Trejo et al., 2007). Then it seems tobe reasonable that increased IGF-1 levels may be representative orconsequence of low IGF bioavailability. At least in some extent,Squassina et al.'s, Bezchlibnyk et al.'s and our study broadlysuggest that IGF system may be involved in the pathogenesisand therapeutic mechanisms of bipolar disorder.

There were several limitations in the present study. Firstly, oursample size was relatively small and negatively impacted thedetection of possible differences between manic patients andhealthy controls. Secondly, this study just observed the increasedserum levels of VEGF, FGF-2, NGF and IGF-1 in manic episode ofbipolar disorder at one time point, thus whether these angioneur-ins are state biomarkers or trait biomarkers could not be deter-mined. Thirdly, recent studies have indicated that some variables,such as BMI, menstrual phase, exercise or cigarette smoking, etc.,may influence the serum levels of these factors (Bullo et al., 2007;Lee and Kim, 2012; Symvoulakis et al., 2010; Yang et al., 2013).However, we could not control these possible confounding vari-ables. So the considerable inter-individual variability of serumlevels in this study might be partially influenced. Moreover, thesources of these neurotrophic factors were not clear. For instance,IGF-1 in serum largely results from hepatic production in responseto growth hormone (GH) binding to its membrane-bound receptor,also is produced at nonhepatic sites (Franz et al., 1999). Never-theless, serum IGF-1 can regulate numerous brain mechanismsincluding cell energy modulation, growth, survival and cognition(Carro and Torres-Aleman, 2006). In the same way, serum FGF-2levels may be directly related to those in the brain (Deguchi et al.,2000). In view of these, serum levels of these factors could stillreflect the changes in the brain, at least to some extent. Finally, thepatients with significant physical and psychiatric diseases wereexcluded to the utmost in this study for exploring possiblealterations of these four factors in bipolar disorder itself. However,comorbidity with medical conditions is very common in bipolardisorder (Magalhaes et al., 2012). So this may limit the extensionof our results to the general clinic population.

In summary, this is the preliminary study on relationshipsbetween serum levels of VEGF, FGF-2, NGF and IGF-1 and manicepisode of bipolar disorder. Our findings tentatively corroboratethe speculation that NGF, FGF-2 and IGF-1 might be relevant to thepathophysiology of bipolar disorder. We cautiously conclude thatthe elevation of serum angioneurins levels in manic state ofbipolar disorder may be associated with their compensatoryneuroprotective and/or angiogenic role, and to a certain extent,reflect the alterations of these factors in CNS. Undoubtedly, theresults are warranted to be confirmed in more well-designedstudies controlling for possible confounding factors.

Conflict of interest

All authors declare that they have no conflict of interest.

X. Liu et al. / Psychiatry Research ∎ (∎∎∎∎) ∎∎∎–∎∎∎ 5

Please cite this article as: Liu, X., et al., Elevated serum levels of FGF-2, NGF and IGF-1 in patients with manic episode of bipolar disorder.Psychiatry Research (2014), http://dx.doi.org/10.1016/j.psychres.2014.03.042i

Acknowledgments

This work was supported by projects from the SMC-morningstar excellent young teachers (C plan) in Shanghai JiaoTong University, Shanghai Health Bureau(2009098), the NationalNatural Science Foundation of China(81000588), “Shanghai healthsystem young talents training plan” in Shanghai Health Bureau(XYQ2011016), and the National Key Clinical Disciplines at Shang-hai Mental Health Center(OMA-MH, 2011-873). We thank thepatients and healthy volunteers for participation and the medicalstaff for collecting specimens. We are also grateful for the supportfrom division of mood disorders in Shanghai Mental Health Center,Shanghai Jiao Tong University School of Medicine.

References

Abraham, J.A., Mergia, A., Whang, J.L., Tumolo, A., Friedman, J., Hjerrild, K.A.,Gospodarowicz, D., Fiddes, J.C., 1986. Nucleotide sequence of a bovine cloneencoding the angiogenic protein, basic fibroblast growth factor. Science 233,545–548.

Aloe, L., Alleva, E., Fiore, M., 2002. Stress and nerve growth factor: findings inanimal models and humans. Pharmacology, Biochemistry, and Behavior 73,159–166.

Aloe, L., Bracci-Laudiero, L., Bonini, S., Manni, L., 1997. The expanding role of nervegrowth factor: from neurotrophic activity to immunologic diseases. Allergy 52,883–894.

Barbosa, I.G., Huguet, R.B., Neves, F.S., Reis, H.J., Bauer, M.E., Janka, Z., Palotas, A.,Teixeira, A.L., 2011. Impaired nerve growth factor homeostasis in patients withbipolar disorder. World Journal of Biological Psychiatry: The Official Journal ofthe World Federation of Societies of Biological Psychiatry 12, 228–232.

Beck, H., Plate, K.H., 2009. Angiogenesis after cerebral ischemia. Acta Neuropatho-logica 117, 481–496.

Bendfeldt, K., Radojevic, V., Kapfhammer, J., Nitsch, C., 2007. Basic fibroblast growthfactor modulates density of blood vessels and preserves tight junctions inorganotypic cortical cultures of mice: a new in vitro model of the blood-brainbarrier. Journal of Neuroscience: The Official Journal of the Society forNeuroscience 27, 3260–3267.

Bersani, G., Iannitelli, A., Fiore, M., Angelucci, F., Aloe, L., 2000. Data and hypotheseson the role of nerve growth factor and other neurotrophins in psychiatricdisorders. Medical Hypotheses 55, 199–207.

Bezchlibnyk, Y.B., Xu, L., Wang, J.F., Young, L.T., 2007. Decreased expression ofinsulin-like growth factor binding protein 2 in the prefrontal cortex of subjectswith bipolar disorder and its regulation by lithium treatment. Brain Research1147, 213–217.

Blumberg, H.P., Wang, F., Chepenik, L.G., Kalmar, J.H., Edmiston, E., Duman, R.S.,Gelernter, J., 2008. Influence of vascular endothelial growth factor variation onhuman hippocampus morphology. Biological Psychiatry 64, 901–903.

Bullo, M., Peeraully, M.R., Trayhurn, P., Folch, J., Salas-Salvado, J., 2007. Circulatingnerve growth factor levels in relation to obesity and the metabolic syndrome inwomen. European Journal of Endocrinology/European Federation of EndocrineSocieties 157, 303–310.

Cao, L., Jiao, X., Zuzga, D.S., Liu, Y., Fong, D.M., Young, D., During, M.J., 2004. VEGFlinks hippocampal activity with neurogenesis, learning and memory. NatureGenetics 36, 827–835.

Carro, E., Torres-Aleman, I., 2006. Serum insulin-like growth factor I in brainfunction. Keio Journal of Medicine 55, 59–63.

Chen, G., Manji, H.K., 2006. The extracellular signal-regulated kinase pathway: anemerging promising target for mood stabilizers. Current Opinion in Psychiatry19, 313–323.

Cheng, J.Z., Du, J., 2007. Mechanical stretch simulates proliferation of venoussmooth muscle cells through activation of the insulin-like growth factor-1receptor. Arteriosclerosis, Thrombosis, and Vascular Biology 27, 1744–1751.

Chiu, C.T., Wang, Z., Hunsberger, J.G., Chuang, D.M., 2013. Therapeutic potential ofmood stabilizers lithium and valproic acid: beyond bipolar disorder. Pharma-cological Reviews 65, 105–142.

Deguchi, Y., Naito, T., Yuge, T., Furukawa, A., Yamada, S., Pardridge, W.M., Kimura, R.,2000. Blood–brain barrier transport of 125I-labeled basic fibroblast growthfactor. Pharmaceutical Research 17, 63–69.

Duman, R.S., Heninger, G.R., Nestler, E.J., 1997. A molecular and cellular theory ofdepression. Archives of General Psychiatry 54, 597–606.

Einat, H., Yuan, P., Gould, T.D., Li, J., Du, J., Zhang, L., Manji, H.K., Chen, G., 2003. Therole of the extracellular signal-regulated kinase signaling pathway in moodmodulation. Journal of Neuroscience: The Official Journal of the Society forNeuroscience 23, 7311–7316.

Evans, S.J., Choudary, P.V., Neal, C.R., Li, J.Z., Vawter, M.P., Tomita, H., Lopez, J.F.,Thompson, R.C., Meng, F., Stead, J.D., Walsh, D.M., Myers, R.M., Bunney, W.E.,Watson, S.J., Jones, E.G., Akil, H., 2004. Dysregulation of the fibroblast growthfactor system in major depression. Proceedings of the National Academy ofSciences of the United States of America 101, 15506–15511.

Ferrara, N., Gerber, H.P., LeCouter, J., 2003. The biology of VEGF and its receptors.Nature Medicine 9, 669–676.

Fornaro, M., Rocchi, G., Escelsior, A., Contini, P., Ghio, M., Colicchio, S., De Berardis,D., Amore, M., Fornaro, P., Martino, M., 2013. VEGF plasma level variations induloxetine-treated patients with major depression. Journal of Affective Dis-orders 151, 590–595.

Franz, B., Buysse, D.J., Cherry, C.R., Gray, N.S., Grochocinski, V.J., Frank, E., Kupfer, D.J., 1999. Insulin-like growth factor 1 and growth hormone binding protein indepression: a preliminary communication. Journal of Psychiatric Research 33,121–127.

Gaughran, F., Payne, J., Sedgwick, P.M., Cotter, D., Berry, M., 2006. Hippocampal FGF-2 and FGFR1 mRNA expression in major depression, schizophrenia and bipolardisorder. Brain Research Bulletin 70, 221–227.

Hashimoto, K., Shimizu, E., Iyo, M., 2004. Critical role of brain-derived neurotrophicfactor in mood disorders. Brain research reviews 45, 104–114.

Jin, K., Zhu, Y., Sun, Y., Mao, X.O., Xie, L., Greenberg, D.A., 2002. Vascular endothelialgrowth factor (VEGF) stimulates neurogenesis in vitro and in vivo. Proceedingsof the National Academy of Sciences of the United States of America 99,11946–11950.

Kapczinski, F., Frey, B.N., Kauer-Sant’Anna, M., Grassi-Oliveira, R., 2008. Brain-derived neurotrophic factor and neuroplasticity in bipolar disorder. ExpertReview of Neurotherapeutics 8, 1101–1113.

Kim, Y.K., Na, K.S., Hwang, J.A., Yoon, H.K., Lee, H.J., Hahn, S.W., Lee, B.H., Jung, H.Y.,2013. High insulin-like growth factor-1 in patients with bipolar I disorder: atrait marker? Journal of Affective Disorders 151, 738–743.

Lee, B.H., Kim, Y.K., 2012. Increased plasma VEGF levels in major depressive ormanic episodes in patients with mood disorders. Journal of Affective Disorders136, 181–184.

Licht, T., Goshen, I., Avital, A., Kreisel, T., Zubedat, S., Eavri, R., Segal, M., Yirmiya, R.,Keshet, E., 2011. Reversible modulations of neuronal plasticity by VEGF.Proceedings of the National Academy of Sciences of the United States ofAmerica 108, 5081–5086.

Magalhaes, P.V., Kapczinski, F., Nierenberg, A.A., Deckersbach, T., Weisinger, D.,Dodd, S., Berk, M., 2012. Illness burden and medical comorbidity in theSystematic Treatment Enhancement Program for Bipolar Disorder. Acta Psy-chiatrica Scandinavica 125, 303–308.

McIntyre, R.S., Vagic, D., Swartz, S.A., Soczynska, J.K., Woldeyohannes, H.O.,Voruganti, L.P., Konarski, J.Z., 2008. Insulin, insulin-like growth factors andincretins: neural homeostatic regulators and treatment opportunities. CNSDrugs 22, 443–453.

Molteni, R., Fumagalli, F., Magnaghi, V., Roceri, M., Gennarelli, M., Racagni, G.,Melcangi, R.C., Riva, M.A., 2001. Modulation of fibroblast growth factor-2 bystress and corticosteroids: from developmental events to adult brain plasticity.Brain Research Reviews 37, 249–258.

Newton, S.S., Collier, E.F., Hunsberger, J., Adams, D., Terwilliger, R., Selvanayagam, E.,Duman, R.S., 2003. Gene profile of electroconvulsive seizures: induction ofneurotrophic and angiogenic factors. Journal of Neuroscience: The OfficialJournal of the Society for Neuroscience 23, 10841–10851.

Nowacka, M.M., Obuchowicz, E., 2012. Vascular endothelial growth factor (VEGF)and its role in the central nervous system: a new element in the neurotrophichypothesis of antidepressant drug action. Neuropeptides 46, 1–10.

O’Kusky, J.R., Ye, P., D’Ercole, A.J., 2000. Insulin-like growth factor-I promotesneurogenesis and synaptogenesis in the hippocampal dentate gyrus duringpostnatal development. Journal of Neuroscience: The Official Journal of theSociety for Neuroscience 20, 8435–8442.

Palomino, A., Gonzalez-Pinto, A., Martinez-Cengotitabengoa, M., Ruiz de Azua, S.,Alberich, S., Mosquera, F., Matute, C., 2013. Relationship between negativesymptoms and plasma levels of insulin-like growth factor 1 in first-episodeschizophrenia and bipolar disorder patients. Progress in Neuro-Psychopharmacology and Biological Psychiatry 44C, 29–33.

Pittenger, C., Duman, R.S., 2008. Stress, depression, and neuroplasticity: a conver-gence of mechanisms. Neuropsychopharmacology: Official Publication of theAmerican College of Neuropsychopharmacology 33, 88–109.

Quaegebeur, A., Lange, C., Carmeliet, P., 2011. The neurovascular link in health anddisease: molecular mechanisms and therapeutic implications. Neuron 71, 406–424.

Rajkowska, G., 2000. Postmortem studies in mood disorders indicate alterednumbers of neurons and glial cells. Biological Psychiatry 48, 766–777.

Reuss, B., Dono, R., Unsicker, K., 2003a. Functions of fibroblast growth factor (FGF)-2and FGF-5 in astroglial differentiation and blood-brain barrier permeability:evidence from mouse mutants. Journal of Neuroscience: The Official Journal ofthe Society for Neuroscience 23, 6404–6412.

Reuss, B., von Bohlen und Halbach, O., 2003b. Fibroblast growth factors and theirreceptors in the central nervous system. Cell and Tissue Research 313, 139–157.

Riva, M.A., Molteni, R., Bedogni, F., Racagni, G., Fumagalli, F., 2005. Emerging role ofthe FGF system in psychiatric disorders. Trends in Pharmacological Sciences 26,228–231.

Segi-Nishida, E., Warner-Schmidt, J.L., Duman, R.S., 2008. Electroconvulsive seizureand VEGF increase the proliferation of neural stem-like cells in rat hippocam-pus. Proceedings of the National Academy of Sciences of the United States ofAmerica 105, 11352–11357.

Segura, I., De Smet, F., Hohensinner, P.J., Ruiz de Almodovar, C., Carmeliet, P., 2009.The neurovascular link in health and disease: an update. Trends in MolecularMedicine 15, 439–451.

Sofroniew, M.V., Howe, C.L., Mobley, W.C., 2001. Nerve growth factor signaling,neuroprotection, and neural repair. Annual Review of Neuroscience 24,1217–1281.

X. Liu et al. / Psychiatry Research ∎ (∎∎∎∎) ∎∎∎–∎∎∎6

Please cite this article as: Liu, X., et al., Elevated serum levels of FGF-2, NGF and IGF-1 in patients with manic episode of bipolar disorder.Psychiatry Research (2014), http://dx.doi.org/10.1016/j.psychres.2014.03.042i

Sonntag, W.E., Ramsey, M., Carter, C.S., 2005. Growth hormone and insulin-likegrowth factor-1 (IGF-1) and their influence on cognitive aging. Ageing ResearchReviews 4, 195–212.

Squassina, A., Costa, M., Congiu, D., Manchia, M., Angius, A., Deiana, V., Ardau, R.,Chillotti, C., Severino, G., Calza, S., Del Zompo, M., 2013. Insulin-like growthfactor 1 (IGF-1) expression is up-regulated in lymphoblastoid cell lines oflithium responsive bipolar disorder patients. Pharmacological Research: TheOfficial Journal of the Italian Pharmacological Society 73, 1–7.

Sun, F.Y., Guo, X., 2005. Molecular and cellular mechanisms of neuroprotection byvascular endothelial growth factor. Journal of Neuroscience Research 79,180–184.

Symvoulakis, E.K., Vardavas, C.I., Fountouli, P., Stavroulaki, A., Antoniou, K.M.,Duijker, G., Tzatzarakis, M.N., Sfiridaki, K., Bolonaki, E., Alegakis, T., Tsatsakis,A.M., 2010. Time interval from cigarette smoke exposure to blood donation and

markers of inflammation: should a smoking cut-off be designated? Xenobio-tica; the Fate of Foreign Compounds in Biological Systems 40, 613–620.

Trejo, J.L., Piriz, J., Llorens-Martin, M.V., Fernandez, A.M., Bolos, M., LeRoith, D.,Nunez, A., Torres-Aleman, I., 2007. Central actions of liver-derived insulin-likegrowth factor I underlying its pro-cognitive effects. Molecular Psychiatry 12,1118–1128.

Warner-Schmidt, J.L., Duman, R.S., 2007. VEGF is an essential mediator of theneurogenic and behavioral actions of antidepressants. Proceedings of theNational Academy of Sciences of the United States of America 104, 4647–4652.

Yang, M.J., Tseng, J.Y., Chen, C.Y., Yeh, C.C., 2013. Changes in maternal serum insulin-like growth factor-I during pregnancy and its relationship to maternal anthro-pometry. Journal of the Chinese Medical Association 76, 635–639.

Zacchigna, S., Lambrechts, D., Carmeliet, P., 2008. Neurovascular signalling defectsin neurodegeneration. Nature Reviews. Neuroscience 9, 169–181.

X. Liu et al. / Psychiatry Research ∎ (∎∎∎∎) ∎∎∎–∎∎∎ 7

Please cite this article as: Liu, X., et al., Elevated serum levels of FGF-2, NGF and IGF-1 in patients with manic episode of bipolar disorder.Psychiatry Research (2014), http://dx.doi.org/10.1016/j.psychres.2014.03.042i