- Email: [email protected]
Gene expression profiling of neuropeptides in mouse cerebellum, hippocampus, and retina
Nutrition 24 (2008) 918 –923 www.elsevier.com/locate/nut
Gene expression profiling of neuropeptides in mouse cerebellum, hippocampus, and retina Kiyotaka Akiyama, Ph.D.*, Setsuko Nakanishi, Ph.D., Nozomu H. Nakamura, Ph.D., and Takayuki Naito, Ph.D. Molecular Neuroscience Unit, Okinawa Institute of Science and Technology, Okinawa, Japan Manuscript received June 11, 2008; accepted June 13, 2008.
Abstract
Objective: We examined gene expression profiling in single neuron types and small regions of the nervous system. Methods: The RNAs were extracted from mouse cerebellar Purkinje cells, granule cell layer, hippocampal CA1 and CA3 pyramidal cell layers, and three layers of the retina (outer nuclear layer, inner nuclear layer, and ganglion cell layer) were dissected by laser capture microdissection. The gene expression profiling of each sample was examined by Affymetrix GeneChip and real-time reverse transcription polymerase chain reaction. We studied the gene expression of 62 neuropeptide and hormone genes and 387 G-protein– coupled receptor (GPCR) genes. Results: Among them, cholecystokinin and neuropeptide Y genes were the most widely expressed. The gene expression of cholecystokinin was very high in the hippocampus, suggesting that cholecystokinin transcripts might have unknown roles in the hippocampus. More than 10 neuropeptide genes were expressed in the ganglion cell layer of the retina, whereas the outer nuclear layer of the retina did not express a considerable amount of neuropeptide mRNAs. In total 12 GPCR genes were found in all tissues examined, and half were orphans (6 of 12). Conclusion: The high ratio of orphan GPCR genes suggests our limited knowledge of the ligand-receptor system in the nervous system. These results provide basic information for studying the function of neuropeptides. © 2008 Published by Elsevier Inc.
Keywords:
Gene expression profile; Neuropeptide; Purkinje cell; Hippocampus; Retina
Introduction Brain functions remain unknown at the molecular level [1]. In a gene-based approach to the study of brain functions, our major research themes are the study of activitydependent gene expression in the brain and identification of neuron types by genes expressed within them. “Activitydependent gene expression” refers to the modulation of gene expression occurring in individual cells as a result of various stimuli, and it is believed to be correlated with self-organizing activities of the brain such as neuronal plasticity. The first step of this research is to obtain gene ex-
* Corresponding author. Tel.: ⫹81-98-929-1352; fax: ⫹81-98-9290592. E-mail address: [email protected] (K. Akiyama). 0899-9007/08/$ – see front matter © 2008 Published by Elsevier Inc. doi:10.1016/j.nut.2008.06.018
pression profiling of single-type neurons and small regions of the nervous system. We designed a combinational method. The combined method, including laser capture microdissection (LCM), DNA microarray, and real-time reverse transcription polymerase chain reaction (RT-PCR), provides us with quantitative and spatial gene expression data with high sensitivity. Mouse cerebellum, hippocampus, and retina were used for this experiment due to their simple cytoarchitecture. Neuropeptides and their receptors (most are G-protein– coupled receptors [GPCRs]) [2] have characteristic expressions in different neurons, and they are believed to be involved in the activitydependent gene expression system. Therefore, they are interesting targets for the study of activity-dependent gene expression in the brain and identification of neuron types by genes expressed within them. In this report, the gene expression profile of neuropeptides and their receptors in these tissues are discussed.
K. Akiyama et al. / Nutrition 24 (2008) 918 –923
Materials and methods
919
manufacturer’s protocol. Laser-microdissected tissue specimens were collected into lysis buffer (Qiagen GmbH, Hilden, Germany). Purkinje cells were microdissected individually. The cerebellar granule cell layer, retinal outer nuclear layer (ONL), inner nuclear layer (INL), ganglion cell layer (GCL), and hippocampal CA1 and CA3 regions were collected as a cellular mass from the cell layers.
Animals Eight-week-old male mice (C57BL/6J, purchased from Charles River Laboratories, Yokohama, Kanagawa, Japan) were used for the collection of cerebella and retinas. Eighteen-week-old mice of the same strain were used for the collection of hippocampi.
RNA isolation, target labeling, and hybridization
Laser capture microdissection
The RNAs were extracted with a RNeasy Micro Kit (Qiagen GmbH). Approximately 1 ng of total RNA was labeled using GeneChip Two-Cycle Target Labeling and Control Reagents (Affymetrix, Santa Clara, CA, USA). The hybridization was done against the Affymetrix Mouse Genome 430 2.0 Array using the GeneChip Hybridization, Wash and Stain Kit (Affymetrix), according to standard protocols.
Mice were anesthetized (diethyl ether) and dissected. Dissected tissues were embedded in OCT compound. The tissues were rapidly frozen in liquid nitrogen-cooled isopentane. Cryostat sections (12 m) were used for LCM. LCM was done by Application Solutions Laser Microdissection (Leica Microsystems GmbH, Wetzlar, Germany) according to the Table 1 Gene expression profile of neuropeptides* Gene name
Cholecystokinin Neuropeptide Y Preproenkephalin 1 Tachykinin 1 Vasoactive intestinal polypeptide Corticotropin-releasing hormone Tachykinin 4 Angiotensinogen (serpin peptidase inhibitor, clade A, member 8) Adenylate cyclase activating polypeptide 1 Calcitonin-related polypeptide,  Natriuretic peptide precursor type A Pro-melanin–concentrating hormone Calcitonin/calcitonin-related polypeptide, ␣ Galanin Gastrin-releasing peptide Somatostatin Neuropeptide B Thyroid-stimulating hormone, -subunit Natriuretic peptide precursor type C Hypocretin Prodynorphin Tachykinin 2 Oxytocin Endothelin 1 Adrenomedullin Endothelin 2
Cerebellum
Hippocampus
Retina
Purkinje cell
Granule cell layer
CA1 pyramidal cell layer
CA3 pyramidal cell layer
Ganglion cell layer
Inner nuclear layer
Outer nuclear layer
719 119 132 A A A 17 105
A A 155 A A A A 118
2416 166 413 231 550 118 A A
1582 65 A 242 113 A A A
308 532 373 4758 115 1729 16 A
230 157 A 319 506 52 25 70
A A A A A 17 10 A
A
A
A
199
61
A
A A A A
A A A A
A A A A
A A A A
85 53 19 493
17 29 17 21
16 24 16 A
A A A A 67 27 25 A A A A A A
A A A A A A A A A A A A A
A A 109 51 A A A A A A A A A
A A
56 49
25 44 A A A A A 44 A A A A A
A A A A A A A A A A A A 15
48
96 A A A A A A A A A A
A 25 A A A A 2049 148 107 16 A
A, absent call (genes that represent an absent call in seven tissues not shown) * Numbers represent the signal value of the Affymetrix GeneChip. A comparison of signal value between tissues is possible but not among genes. This table was prepared by the following criteria: 1) a “marginal” call was considered a “present” call. 2) If the same probe set produced a different detection call between two analyses of the same sample, that probe set was considered to an absent call. The analysis was done twice for Purkinje cells, the granule cell layer, and CA1 and CA3 pyramidal cell layers. The analysis was done once for the inner nuclear layer, the outer nuclear layer, and the ganglion cell layer. 3) If several probe sets corresponded to one gene, the highest signal value was adopted.
920
K. Akiyama et al. / Nutrition 24 (2008) 918 –923
Data array analysis
was carried out by using the SuperScript First-Strand Synthesis System (Invitrogen, San Diego, CA, USA) with the T7-oligo (dT) primer. Double-strand cDNA synthesis was performed by using Escherichia coli RNase H, E. coli DNA polymerase, and E. coli DNA ligase (Takarabio, Otsu, Shiga, Japan). T7 RNA transcription was done by using the MEGAscript High Yield Transcription Kit (Applied Biosystems, Foster City, CA, USA). At least 0.25 ng of the amplified cDNA was used as template for a single reaction of real-time PCR.
Data were collected with the GeneChip Scanner 3000 7G (Affymetrix). Raw data were analyzed by using Affymetrix GeneChip Operating Software (GCOS 1). The present/ absent call for each data point was calculated with the Affymetrix standard algorithm. The Affymetrix Mouse Genome 430 2.0 Array carries 45 000 probe sets. Each probe set contained 11 probes. Only 26 386 probe sets (corresponding to 19 249 genes), each of which contained more than 9 probes matched to the defined cDNAs in the ENSMBL or NCBI (reference sequence) database were picked up and used for our analysis.
Real-time PCR The PCR primers were designed to be hybridized to the region of a single exon within 200 bp from the poly(A) site of the cDNAs. Real-time PCR was performed by using the SYBR PremiEx Taq and a Thermal Cycler Dice Real Time System (Takarabio). We used mouse genomic DNA as a control for counting the copy number.
Preparation of template for real-time PCR Complementary DNAs were synthesized from collected RNAs and amplified by two cycles of T7 amplification. RT Table 2 Gene expression profile of GPCRs obtained by the Affymetrix GeneChip Gene name
GPCRs found in 3 tissues* ␥-Aminobutyric acid B receptor 1 GPCR-85 Cannabinoid receptor 1 (brain) Duffy blood group, chemokine receptor Leucine-rich repeat-containing GPCR-4 GPCR-162 Endothelin receptor type B GPCR-19 Neurotensin receptor 2 GPCR-175 GPCR-107 GPCR-61 GPCRs activated by neuropeptides† Endothelin receptor type B Neurotensin receptor 2 Corticotropin-releasing hormone receptor 1 Adenylate cyclase activating polypeptide 1 receptor 1 Opioid receptor-like 1 Calcitonin receptor-like Somatostatin receptor 3 Somatostatin receptor 4 Neuropeptide Y Y5 receptor Thyrotropin-releasing hormone receptor Melanocortin 3 receptor Gonadotropin-releasing hormone receptor
Cerebellum
Hippocampus
Retina
Purkinje cell
Granular cell layer
CA1 pyramidal cell layer
CA3 pyramidal cell layer
Ganglion cell layer
Inner nuclear layer
2715 1459 222 41
466 1012 1341 199
2691 1441 1011 218
1719 956 946 486
2198 1421 1219 853
589 775 245 759
83 29 21 80
782
298
310
167
467
439
180
50 311 74 260 35 56 59
55 141 80 88 36 38 26
131 76 126 196 53 85 107
144 51 145 190 30 66 88
546 77 282 19 161 140 143
206 107 157 14 74 156 70
34 21 37 43 51 94 15
311 260 20
141 88 116
76 196 67
51 189 32
77 19 292
107 14 26
21 43 A
90
84
101
65
164
23
218 A 76 1417 379 A
202 25 118 205 320 A
431 40 14 134 A 30
50 146 A A 123 179
31 17
39 18
A
A 42
A A A 6 A A
18 33 A A A A
13 12
A
A, absent call; GPCR, G-protein– coupled receptor * Twelve GPCR expressions in all seven tissues were selected. † These 13 GPCRs activated by neuropeptide are frequently found in seven tissues.
A A
Outer nuclear layer
A 9 A 13 22 A A A A
K. Akiyama et al. / Nutrition 24 (2008) 918 –923
921
Table 3 Twelve genes showing different expression profiles by Affymetrix GeneChip and real-time RT-PCR analyses* Gene name
Cerebellum
Hippocampus
Purkinje cell Affymetrix signal value Adrenomedullin Apelin Corticotropin-releasing hormone Cortistatin Galanin Hypocretin Neuromedin B Neuropeptide B Tachykinin 1 Tachykinin 2 Tachykinin 4 Galanin-like peptide precursor Gonadotropin-releasing hormone 1 Kininogen 2 Luteinizing hormone  Pro-opiomelanocortin-␣ Relaxin 1 Relaxin 3
Copy number/1 ng
Granular cell layer
CA1 pyramidal cell layer
CA3 pyramidal cell layer
Affymetrix signal value
Copy number/1 ng
Affymetrix signal value
Copy number/1 ng
Affymetrix signal value
Copy number/1 ng
A A A
0 80† 231†
A A A
444† 688† 17†
A A 118
0 5321 15†
A A A
2238 2314 773†
A A 25 A A A A 17
0 2236 315† 582† 121† 1074 0 909† 0
A A A A A A A A
20† 0 0 667† 17† 116† 110† 0 0
A A A A 51 231 A A
1789 0 0 46† 3605 6922 7718 0 0
A A A A A 242 A A
448† 0 106† 1387 1182 2695 27† 0 0
68†
55†
0 0 42† 16† 0
15† 0 57† 25† 0
0 0 0 224† 0 0
0 2† 0 327† 428† 0
A, absent call; RT-PCR, reverse transcription polymerase chain reaction * Copy number per 1 ng of mRNA was determined by real-time RT-PCR. Supposing that one cell might contain 1 pg of mRNA, 1000 transcripts per 1 ng of mRNA were nearly equivalent to one transcript per cell. Empty cells indicate probe-sets that are not available on this microarray. † Fewer than one transcript per cell.
Results In total 62 genes for neuropeptides and hormones were picked up from the mouse genome. The Affymetrix GeneChip carried probe sets corresponding to 55 of them. Only 26 genes were detected by the Affymetrix GeneChip as “present call” and are listed in Table 1.† We found that the cholecystokinin (CCK) and neuropeptide Y (NPY) genes were the most widely and abundantly expressed, whereas some of their receptors showed restricted expression. In total 387 GPCRs were initially picked up, excluding taste and olfactory receptors, but only 12 GPCRs were expressed in all examined tissues (Table 2). Among them, six GPCRs, leucine-rich repeat-containing GPCR-4, GPCR162, GPCR-19, GPCR-175, GPCR-107, and GPCR-61, † The remaining 36 genes whose expressions were not detected by the Affymetrix GeneChip are agouti-related protein, apelin, arginine vasopressin, glycoprotein hormone ␣-subunit, cortistatin, endothelin-3, folliclestimulating hormone-, gastrin, glucagon, growth hormone–releasing hormone, ghrelin, gastric inhibitory polypeptide, kininogen-1, neuromedin B, neuromedin U, neuropeptide FF-amide peptide precursor, natriuretic peptide precursor type B, neurotensin, prepronociceptin, pancreatic polypeptide, prolactin, parathyroid hormone, peptide YY, secretin, thyrotropin-releasing hormone, urocortin, urocortin 2, urocortin 3, urotensin 2, galanin-like peptide precursor, gonadotropin-releasing hormone 1, kininogen 2, luteinizing hormone-, pro-opiomelanocortin-␣, relaxin 1, and relaxin 3.
were orphans. As expected, the ␥-aminobutyric acid B1 receptor was highly expressed in all tested tissues. However, we could not find the ␥-aminobutyric acid B2 receptor, because the ␥-aminobutyric acid B2 probe set was not available on this DNA microarray (Table 2). To verify Affymetrix GeneChip analysis, real-time RTPCR assays for the selected 62 genes were performed on RNA extracted from Purkinje cells, the granule cell layer, and the CA1 and CA3 pyramidal cell layers. In a real-time RT-PCR experiment, only the genes whose expression level was more than 1000 transcripts per 1 ng of mRNA were chosen for our analysis. Real-time RT-PCR results showed a similarity to those of DNA microarray analysis except for 11 genes (Table 3). Among them, eight genes were detected only with real-time RT-PCR. It is notable that the expression level of these eight genes was low (⬍10 transcripts per cell). Thus, these results suggest that the Affymetrix GeneChip can detect genes whose expression level is more than 10 transcripts per cell. The remaining three genes, corticotropin-releasing hormone in CA1 pyramidal cell layer and hypocretin and tachykinin 4 in Purkinje cells, were observed in DNA microarray analysis. However, their copy numbers were calculated to be fewer than 1000 copies per 1 ng of mRNA by real-time RT-PCR. The proportion of neuropeptide transcripts found in Purkinje cells, the granule cell layer, and then CA1 and CA3 pyramidal cell layers by real-time RT-PCR is shown in Figure
922
K. Akiyama et al. / Nutrition 24 (2008) 918 –923
1. Abundant neuropeptide transcripts were found in CA1 and CA3 pyramidal cell layers compared with Purkinje cells and the granule cell layer. CCK transcripts were highly expressed in CA1 and CA3 pyramidal cell layers, but moderately in Purkinje cells, the INL, and the GCL of the retina (Table 1). In contrast, CCK transcripts were absent in the cerebellar granule cell layer and the ONL of the retina. More than 80% of all neuropeptide transcripts in the CA1 and CA3 were CCK (Fig. 1). Characteristic gene expression patterns of neuropeptides and their GPCRs transcripts were found in the cerebellum,
hippocampus, and retina (Tables 1 and 2). For example, vasoactive intestinal polypeptide transcripts were expressed in the hippocampus, but not in the cerebellum, whereas angiotensinogen was expressed in the cerebellum, but not in the hippocampus. This specific gene expression pattern suggests that these peptides might have particular roles in these brain regions. Somatostatin receptor-4 was expressed highly in hippocampal CA1 (Table 2), which also specifically expressed somatostatin (Table 1). We found more than 10 neuropeptide gene expressions in
Fig. 1. Proportions of individual neuropeptide genes expressed in the (a) granule cell layer, (b) Purkinje cell, (c) CA1 pyramidal cell layer, and (d) CA3 pyramidal cell layer. The number of transcripts of all neuropeptides per 1 ng of mRNA is indicated under pie graphs. The dimension of the circle represents the number of transcripts. The quantity of transcripts measured by real-time reverse transcription polymerase chain reaction is presented in the table.
K. Akiyama et al. / Nutrition 24 (2008) 918 –923
the GCL of the retina. In contrast, ONL did not express any considerable amounts of neuropeptide transcripts. Tachykinins 1 and 2 and corticotropin-releasing hormone genes were expressed abundantly in the GCL. In the retina, the NPY gene was expressed only in the INL and GCL, but not in the ONL (Table 1). The ratio of NPY gene expression in the INL to that in the GCL was found to be 1:3.4. The presence of their transcripts in the retina suggests that they may have some roles in the retina.
Discussion Gene expression profiling of the Purkinje cell, granule cell layer, hippocampus, and retina was done by using a nanogram order of total RNA extracted from several hundred of cells prepared by LCM. For DNA microarray experiments, we successfully amplified the extracted mRNAs without considerable loss of varieties of mRNAs. The gene expression data of neuropeptides and their receptors was chosen from the DNA microarray data, and the gene expression of neuropeptides in several regions of the cerebellum and hippocampus was further studied by quantitative PCR. We found the characteristic gene expression pattern of neuropeptides and their receptors. This characteristic gene expression pattern could help us to identify neuron types by genes expressed within them. The RNAs collected by LCM may contain some intermingled RNA from neighboring cells. It is possible to estimate the ratio of intermingled RNAs, if we have information about special markers of cell types. The abundant CCK transcript expression in CA1 and CA3 pyramidal cell layers of the hippocampus is also supported by previous in situ hybridization studies that showed robust expression of CCK mRNAs in CA1 and CA3 pyramidal cell layers including pyramidal cells and interneurons [3– 6]. CCK peptides are known to be found in scattered interneurons, but not in pyramidal cells [7–9]. The distribution discrepancy between CCK transcripts and peptides in the hippocampus provides an interesting subject to be investigated. It has been shown that NPY is expressed only in the Golgi epithelial cells, but not in Purkinje cells (GENAST Project at Rockefeller University; http://www.gensat.org/ index.html, Allen Brain Atlas: http://www.brain-map.org/). However, we found NPY transcripts in Purkinje cells (Table 1). To clarify this discrepancy, further works are needed. Neuropeptide Y is synthesized by amacrine cells in the INL and those displaced in the GCL [10,11]. It has been reported that the ratio of NPY-producing amacrine cells in the INL to those in the GCL was 1:0.44 [11]. However, DNA microarray analysis showed that the expression ratio of NPY was 1:3.4. This discrepancy can be explained by the location of the cell bodies of amacrine cells in the INL. It is known that 96% of amacrine cells in the INL lie in the
923
innermost row [11]. This area may be damaged during LCM, resulting in the smaller amount of RNA. The presence of other cells producing NPY transcripts other than amacrine cells may also interfere with the ratio. We found widely distributed GPCRs in the nervous system (Table 2). The high ratio of orphan GPCRs (6 of 12) indicates our limited knowledge about the ligand-receptor system in the nervous system. Extensive research of orphan GPCRs in neuroscience is thus necessary. Over the past decade, numerous studies have identified many neuropeptides and investigated their functions in the central nervous system [12,13]. The real molecular mechanism of neuropeptide functions in the nervous system is still unclear. We believe the study of activity-dependent gene expression including the neuropeptide-receptor system is one of the major research areas for molecular neuroscience in the coming years. The precise expression data of ligands and their receptors is basic information to design experiments for the study of activity-dependent gene expression. References [1] Waites CL, Craig AM, Garner CC. Mechanisms of vertebrate synaptogenesis. Annu Rev Neurosci 2005;28:251–74. [2] Gainetdinov RR, Premont RT, Bohn LM, Lefkowitz RJ, Caron MG. Desensitization of G protein– coupled receptors and neuronal functions. Annu Rev Neurosci 2004;27:107– 44. [3] Lucas LR, Pompei P, Ono J, McEwen BS. Effects of adrenal steroids on basal ganglia neuropeptide mRNA and tyrosine hydroxylase radioimmunoreactive levels in the adrenalectomized rat. J Neurochem 1998;71:833– 43. [4] Zachrisson O, Nomikos GG, Marcus MM, Svensson TH, Lindefors N. Effects of antipsychotic drugs on cholecystokinin and preprotachykinin (substance P) mRNA expression in the rat hippocampal formation. Eur Neuropsychopharmacol 2000;10:355– 63. [5] Bräuer AU, Savaskan NE, Plaschke M, Ninnemann O, Nitsch R. Cholecystokinin expression after hippocampal deafferentiation: molecular evidence revealed by differential display–reverse transcription– polymerase chain reaction. Neuroscience 2003;121:111–21. [6] Nakamura NH, McEwen BS. Changes in interneuronal phenotypes regulated by estradiol in the adult rat hippocampus: a potential role for neuropeptide Y. Neuroscience 2005;136:357– 69. [7] Freund TF, Buzsáki G. Interneurons of the hippocampus. Hippocampus 1996;6:347– 470. [8] Freund TF. Interneuron Diversity series: Rhythm and mood in perisomatic inhibition. Trends Neurosci 2003;26:489 –95. [9] Mátyás F, Freund TF, Gulyás AI. Immunocytochemically defined interneuron populations in the hippocampus of mouse strains used in transgenic technology. Hippocampus 2004;14:460 – 81. [10] Ammar DA, Hugbes BA, Thompson DA. Neuropeptide Y and the retinal pigment epithelium: receptor subtypes, signaling, and bioelectrical responses. Invest Ophtalmol Vis Sci 1998;39:1870 – 8. [11] Sinclair JR, Nirenberg S. Characterization of neuropeptide Y– expressing cells in the mouse retina using immunohistochemical and transgenic techniques. J Comp Neurol 2001;432:296 –306. [12] Baraban SC, Tallent MK. Interneuron Diversity series: interneuronal neuropeptides— endogenous regulators of neuronal excitability. Trends Neurosci 2004;27:135– 42. [13] Ludwig M, Leng G. Dendritic peptide release and peptide-dependent behaviors. Nat Rev Neurosci 2006;7:126 –36.
Gene expression profiling of neuropeptides in mouse cerebellum, hippocampus, and retina Kiyotaka Akiyama, Ph.D.*, Setsuko Nakanishi, Ph.D., Nozomu H. Nakamura, Ph.D., and Takayuki Naito, Ph.D. Molecular Neuroscience Unit, Okinawa Institute of Science and Technology, Okinawa, Japan Manuscript received June 11, 2008; accepted June 13, 2008.
Abstract
Objective: We examined gene expression profiling in single neuron types and small regions of the nervous system. Methods: The RNAs were extracted from mouse cerebellar Purkinje cells, granule cell layer, hippocampal CA1 and CA3 pyramidal cell layers, and three layers of the retina (outer nuclear layer, inner nuclear layer, and ganglion cell layer) were dissected by laser capture microdissection. The gene expression profiling of each sample was examined by Affymetrix GeneChip and real-time reverse transcription polymerase chain reaction. We studied the gene expression of 62 neuropeptide and hormone genes and 387 G-protein– coupled receptor (GPCR) genes. Results: Among them, cholecystokinin and neuropeptide Y genes were the most widely expressed. The gene expression of cholecystokinin was very high in the hippocampus, suggesting that cholecystokinin transcripts might have unknown roles in the hippocampus. More than 10 neuropeptide genes were expressed in the ganglion cell layer of the retina, whereas the outer nuclear layer of the retina did not express a considerable amount of neuropeptide mRNAs. In total 12 GPCR genes were found in all tissues examined, and half were orphans (6 of 12). Conclusion: The high ratio of orphan GPCR genes suggests our limited knowledge of the ligand-receptor system in the nervous system. These results provide basic information for studying the function of neuropeptides. © 2008 Published by Elsevier Inc.
Keywords:
Gene expression profile; Neuropeptide; Purkinje cell; Hippocampus; Retina
Introduction Brain functions remain unknown at the molecular level [1]. In a gene-based approach to the study of brain functions, our major research themes are the study of activitydependent gene expression in the brain and identification of neuron types by genes expressed within them. “Activitydependent gene expression” refers to the modulation of gene expression occurring in individual cells as a result of various stimuli, and it is believed to be correlated with self-organizing activities of the brain such as neuronal plasticity. The first step of this research is to obtain gene ex-
* Corresponding author. Tel.: ⫹81-98-929-1352; fax: ⫹81-98-9290592. E-mail address: [email protected] (K. Akiyama). 0899-9007/08/$ – see front matter © 2008 Published by Elsevier Inc. doi:10.1016/j.nut.2008.06.018
pression profiling of single-type neurons and small regions of the nervous system. We designed a combinational method. The combined method, including laser capture microdissection (LCM), DNA microarray, and real-time reverse transcription polymerase chain reaction (RT-PCR), provides us with quantitative and spatial gene expression data with high sensitivity. Mouse cerebellum, hippocampus, and retina were used for this experiment due to their simple cytoarchitecture. Neuropeptides and their receptors (most are G-protein– coupled receptors [GPCRs]) [2] have characteristic expressions in different neurons, and they are believed to be involved in the activitydependent gene expression system. Therefore, they are interesting targets for the study of activity-dependent gene expression in the brain and identification of neuron types by genes expressed within them. In this report, the gene expression profile of neuropeptides and their receptors in these tissues are discussed.
K. Akiyama et al. / Nutrition 24 (2008) 918 –923
Materials and methods
919
manufacturer’s protocol. Laser-microdissected tissue specimens were collected into lysis buffer (Qiagen GmbH, Hilden, Germany). Purkinje cells were microdissected individually. The cerebellar granule cell layer, retinal outer nuclear layer (ONL), inner nuclear layer (INL), ganglion cell layer (GCL), and hippocampal CA1 and CA3 regions were collected as a cellular mass from the cell layers.
Animals Eight-week-old male mice (C57BL/6J, purchased from Charles River Laboratories, Yokohama, Kanagawa, Japan) were used for the collection of cerebella and retinas. Eighteen-week-old mice of the same strain were used for the collection of hippocampi.
RNA isolation, target labeling, and hybridization
Laser capture microdissection
The RNAs were extracted with a RNeasy Micro Kit (Qiagen GmbH). Approximately 1 ng of total RNA was labeled using GeneChip Two-Cycle Target Labeling and Control Reagents (Affymetrix, Santa Clara, CA, USA). The hybridization was done against the Affymetrix Mouse Genome 430 2.0 Array using the GeneChip Hybridization, Wash and Stain Kit (Affymetrix), according to standard protocols.
Mice were anesthetized (diethyl ether) and dissected. Dissected tissues were embedded in OCT compound. The tissues were rapidly frozen in liquid nitrogen-cooled isopentane. Cryostat sections (12 m) were used for LCM. LCM was done by Application Solutions Laser Microdissection (Leica Microsystems GmbH, Wetzlar, Germany) according to the Table 1 Gene expression profile of neuropeptides* Gene name
Cholecystokinin Neuropeptide Y Preproenkephalin 1 Tachykinin 1 Vasoactive intestinal polypeptide Corticotropin-releasing hormone Tachykinin 4 Angiotensinogen (serpin peptidase inhibitor, clade A, member 8) Adenylate cyclase activating polypeptide 1 Calcitonin-related polypeptide,  Natriuretic peptide precursor type A Pro-melanin–concentrating hormone Calcitonin/calcitonin-related polypeptide, ␣ Galanin Gastrin-releasing peptide Somatostatin Neuropeptide B Thyroid-stimulating hormone, -subunit Natriuretic peptide precursor type C Hypocretin Prodynorphin Tachykinin 2 Oxytocin Endothelin 1 Adrenomedullin Endothelin 2
Cerebellum
Hippocampus
Retina
Purkinje cell
Granule cell layer
CA1 pyramidal cell layer
CA3 pyramidal cell layer
Ganglion cell layer
Inner nuclear layer
Outer nuclear layer
719 119 132 A A A 17 105
A A 155 A A A A 118
2416 166 413 231 550 118 A A
1582 65 A 242 113 A A A
308 532 373 4758 115 1729 16 A
230 157 A 319 506 52 25 70
A A A A A 17 10 A
A
A
A
199
61
A
A A A A
A A A A
A A A A
A A A A
85 53 19 493
17 29 17 21
16 24 16 A
A A A A 67 27 25 A A A A A A
A A A A A A A A A A A A A
A A 109 51 A A A A A A A A A
A A
56 49
25 44 A A A A A 44 A A A A A
A A A A A A A A A A A A 15
48
96 A A A A A A A A A A
A 25 A A A A 2049 148 107 16 A
A, absent call (genes that represent an absent call in seven tissues not shown) * Numbers represent the signal value of the Affymetrix GeneChip. A comparison of signal value between tissues is possible but not among genes. This table was prepared by the following criteria: 1) a “marginal” call was considered a “present” call. 2) If the same probe set produced a different detection call between two analyses of the same sample, that probe set was considered to an absent call. The analysis was done twice for Purkinje cells, the granule cell layer, and CA1 and CA3 pyramidal cell layers. The analysis was done once for the inner nuclear layer, the outer nuclear layer, and the ganglion cell layer. 3) If several probe sets corresponded to one gene, the highest signal value was adopted.
920
K. Akiyama et al. / Nutrition 24 (2008) 918 –923
Data array analysis
was carried out by using the SuperScript First-Strand Synthesis System (Invitrogen, San Diego, CA, USA) with the T7-oligo (dT) primer. Double-strand cDNA synthesis was performed by using Escherichia coli RNase H, E. coli DNA polymerase, and E. coli DNA ligase (Takarabio, Otsu, Shiga, Japan). T7 RNA transcription was done by using the MEGAscript High Yield Transcription Kit (Applied Biosystems, Foster City, CA, USA). At least 0.25 ng of the amplified cDNA was used as template for a single reaction of real-time PCR.
Data were collected with the GeneChip Scanner 3000 7G (Affymetrix). Raw data were analyzed by using Affymetrix GeneChip Operating Software (GCOS 1). The present/ absent call for each data point was calculated with the Affymetrix standard algorithm. The Affymetrix Mouse Genome 430 2.0 Array carries 45 000 probe sets. Each probe set contained 11 probes. Only 26 386 probe sets (corresponding to 19 249 genes), each of which contained more than 9 probes matched to the defined cDNAs in the ENSMBL or NCBI (reference sequence) database were picked up and used for our analysis.
Real-time PCR The PCR primers were designed to be hybridized to the region of a single exon within 200 bp from the poly(A) site of the cDNAs. Real-time PCR was performed by using the SYBR PremiEx Taq and a Thermal Cycler Dice Real Time System (Takarabio). We used mouse genomic DNA as a control for counting the copy number.
Preparation of template for real-time PCR Complementary DNAs were synthesized from collected RNAs and amplified by two cycles of T7 amplification. RT Table 2 Gene expression profile of GPCRs obtained by the Affymetrix GeneChip Gene name
GPCRs found in 3 tissues* ␥-Aminobutyric acid B receptor 1 GPCR-85 Cannabinoid receptor 1 (brain) Duffy blood group, chemokine receptor Leucine-rich repeat-containing GPCR-4 GPCR-162 Endothelin receptor type B GPCR-19 Neurotensin receptor 2 GPCR-175 GPCR-107 GPCR-61 GPCRs activated by neuropeptides† Endothelin receptor type B Neurotensin receptor 2 Corticotropin-releasing hormone receptor 1 Adenylate cyclase activating polypeptide 1 receptor 1 Opioid receptor-like 1 Calcitonin receptor-like Somatostatin receptor 3 Somatostatin receptor 4 Neuropeptide Y Y5 receptor Thyrotropin-releasing hormone receptor Melanocortin 3 receptor Gonadotropin-releasing hormone receptor
Cerebellum
Hippocampus
Retina
Purkinje cell
Granular cell layer
CA1 pyramidal cell layer
CA3 pyramidal cell layer
Ganglion cell layer
Inner nuclear layer
2715 1459 222 41
466 1012 1341 199
2691 1441 1011 218
1719 956 946 486
2198 1421 1219 853
589 775 245 759
83 29 21 80
782
298
310
167
467
439
180
50 311 74 260 35 56 59
55 141 80 88 36 38 26
131 76 126 196 53 85 107
144 51 145 190 30 66 88
546 77 282 19 161 140 143
206 107 157 14 74 156 70
34 21 37 43 51 94 15
311 260 20
141 88 116
76 196 67
51 189 32
77 19 292
107 14 26
21 43 A
90
84
101
65
164
23
218 A 76 1417 379 A
202 25 118 205 320 A
431 40 14 134 A 30
50 146 A A 123 179
31 17
39 18
A
A 42
A A A 6 A A
18 33 A A A A
13 12
A
A, absent call; GPCR, G-protein– coupled receptor * Twelve GPCR expressions in all seven tissues were selected. † These 13 GPCRs activated by neuropeptide are frequently found in seven tissues.
A A
Outer nuclear layer
A 9 A 13 22 A A A A
K. Akiyama et al. / Nutrition 24 (2008) 918 –923
921
Table 3 Twelve genes showing different expression profiles by Affymetrix GeneChip and real-time RT-PCR analyses* Gene name
Cerebellum
Hippocampus
Purkinje cell Affymetrix signal value Adrenomedullin Apelin Corticotropin-releasing hormone Cortistatin Galanin Hypocretin Neuromedin B Neuropeptide B Tachykinin 1 Tachykinin 2 Tachykinin 4 Galanin-like peptide precursor Gonadotropin-releasing hormone 1 Kininogen 2 Luteinizing hormone  Pro-opiomelanocortin-␣ Relaxin 1 Relaxin 3
Copy number/1 ng
Granular cell layer
CA1 pyramidal cell layer
CA3 pyramidal cell layer
Affymetrix signal value
Copy number/1 ng
Affymetrix signal value
Copy number/1 ng
Affymetrix signal value
Copy number/1 ng
A A A
0 80† 231†
A A A
444† 688† 17†
A A 118
0 5321 15†
A A A
2238 2314 773†
A A 25 A A A A 17
0 2236 315† 582† 121† 1074 0 909† 0
A A A A A A A A
20† 0 0 667† 17† 116† 110† 0 0
A A A A 51 231 A A
1789 0 0 46† 3605 6922 7718 0 0
A A A A A 242 A A
448† 0 106† 1387 1182 2695 27† 0 0
68†
55†
0 0 42† 16† 0
15† 0 57† 25† 0
0 0 0 224† 0 0
0 2† 0 327† 428† 0
A, absent call; RT-PCR, reverse transcription polymerase chain reaction * Copy number per 1 ng of mRNA was determined by real-time RT-PCR. Supposing that one cell might contain 1 pg of mRNA, 1000 transcripts per 1 ng of mRNA were nearly equivalent to one transcript per cell. Empty cells indicate probe-sets that are not available on this microarray. † Fewer than one transcript per cell.
Results In total 62 genes for neuropeptides and hormones were picked up from the mouse genome. The Affymetrix GeneChip carried probe sets corresponding to 55 of them. Only 26 genes were detected by the Affymetrix GeneChip as “present call” and are listed in Table 1.† We found that the cholecystokinin (CCK) and neuropeptide Y (NPY) genes were the most widely and abundantly expressed, whereas some of their receptors showed restricted expression. In total 387 GPCRs were initially picked up, excluding taste and olfactory receptors, but only 12 GPCRs were expressed in all examined tissues (Table 2). Among them, six GPCRs, leucine-rich repeat-containing GPCR-4, GPCR162, GPCR-19, GPCR-175, GPCR-107, and GPCR-61, † The remaining 36 genes whose expressions were not detected by the Affymetrix GeneChip are agouti-related protein, apelin, arginine vasopressin, glycoprotein hormone ␣-subunit, cortistatin, endothelin-3, folliclestimulating hormone-, gastrin, glucagon, growth hormone–releasing hormone, ghrelin, gastric inhibitory polypeptide, kininogen-1, neuromedin B, neuromedin U, neuropeptide FF-amide peptide precursor, natriuretic peptide precursor type B, neurotensin, prepronociceptin, pancreatic polypeptide, prolactin, parathyroid hormone, peptide YY, secretin, thyrotropin-releasing hormone, urocortin, urocortin 2, urocortin 3, urotensin 2, galanin-like peptide precursor, gonadotropin-releasing hormone 1, kininogen 2, luteinizing hormone-, pro-opiomelanocortin-␣, relaxin 1, and relaxin 3.
were orphans. As expected, the ␥-aminobutyric acid B1 receptor was highly expressed in all tested tissues. However, we could not find the ␥-aminobutyric acid B2 receptor, because the ␥-aminobutyric acid B2 probe set was not available on this DNA microarray (Table 2). To verify Affymetrix GeneChip analysis, real-time RTPCR assays for the selected 62 genes were performed on RNA extracted from Purkinje cells, the granule cell layer, and the CA1 and CA3 pyramidal cell layers. In a real-time RT-PCR experiment, only the genes whose expression level was more than 1000 transcripts per 1 ng of mRNA were chosen for our analysis. Real-time RT-PCR results showed a similarity to those of DNA microarray analysis except for 11 genes (Table 3). Among them, eight genes were detected only with real-time RT-PCR. It is notable that the expression level of these eight genes was low (⬍10 transcripts per cell). Thus, these results suggest that the Affymetrix GeneChip can detect genes whose expression level is more than 10 transcripts per cell. The remaining three genes, corticotropin-releasing hormone in CA1 pyramidal cell layer and hypocretin and tachykinin 4 in Purkinje cells, were observed in DNA microarray analysis. However, their copy numbers were calculated to be fewer than 1000 copies per 1 ng of mRNA by real-time RT-PCR. The proportion of neuropeptide transcripts found in Purkinje cells, the granule cell layer, and then CA1 and CA3 pyramidal cell layers by real-time RT-PCR is shown in Figure
922
K. Akiyama et al. / Nutrition 24 (2008) 918 –923
1. Abundant neuropeptide transcripts were found in CA1 and CA3 pyramidal cell layers compared with Purkinje cells and the granule cell layer. CCK transcripts were highly expressed in CA1 and CA3 pyramidal cell layers, but moderately in Purkinje cells, the INL, and the GCL of the retina (Table 1). In contrast, CCK transcripts were absent in the cerebellar granule cell layer and the ONL of the retina. More than 80% of all neuropeptide transcripts in the CA1 and CA3 were CCK (Fig. 1). Characteristic gene expression patterns of neuropeptides and their GPCRs transcripts were found in the cerebellum,
hippocampus, and retina (Tables 1 and 2). For example, vasoactive intestinal polypeptide transcripts were expressed in the hippocampus, but not in the cerebellum, whereas angiotensinogen was expressed in the cerebellum, but not in the hippocampus. This specific gene expression pattern suggests that these peptides might have particular roles in these brain regions. Somatostatin receptor-4 was expressed highly in hippocampal CA1 (Table 2), which also specifically expressed somatostatin (Table 1). We found more than 10 neuropeptide gene expressions in
Fig. 1. Proportions of individual neuropeptide genes expressed in the (a) granule cell layer, (b) Purkinje cell, (c) CA1 pyramidal cell layer, and (d) CA3 pyramidal cell layer. The number of transcripts of all neuropeptides per 1 ng of mRNA is indicated under pie graphs. The dimension of the circle represents the number of transcripts. The quantity of transcripts measured by real-time reverse transcription polymerase chain reaction is presented in the table.
K. Akiyama et al. / Nutrition 24 (2008) 918 –923
the GCL of the retina. In contrast, ONL did not express any considerable amounts of neuropeptide transcripts. Tachykinins 1 and 2 and corticotropin-releasing hormone genes were expressed abundantly in the GCL. In the retina, the NPY gene was expressed only in the INL and GCL, but not in the ONL (Table 1). The ratio of NPY gene expression in the INL to that in the GCL was found to be 1:3.4. The presence of their transcripts in the retina suggests that they may have some roles in the retina.
Discussion Gene expression profiling of the Purkinje cell, granule cell layer, hippocampus, and retina was done by using a nanogram order of total RNA extracted from several hundred of cells prepared by LCM. For DNA microarray experiments, we successfully amplified the extracted mRNAs without considerable loss of varieties of mRNAs. The gene expression data of neuropeptides and their receptors was chosen from the DNA microarray data, and the gene expression of neuropeptides in several regions of the cerebellum and hippocampus was further studied by quantitative PCR. We found the characteristic gene expression pattern of neuropeptides and their receptors. This characteristic gene expression pattern could help us to identify neuron types by genes expressed within them. The RNAs collected by LCM may contain some intermingled RNA from neighboring cells. It is possible to estimate the ratio of intermingled RNAs, if we have information about special markers of cell types. The abundant CCK transcript expression in CA1 and CA3 pyramidal cell layers of the hippocampus is also supported by previous in situ hybridization studies that showed robust expression of CCK mRNAs in CA1 and CA3 pyramidal cell layers including pyramidal cells and interneurons [3– 6]. CCK peptides are known to be found in scattered interneurons, but not in pyramidal cells [7–9]. The distribution discrepancy between CCK transcripts and peptides in the hippocampus provides an interesting subject to be investigated. It has been shown that NPY is expressed only in the Golgi epithelial cells, but not in Purkinje cells (GENAST Project at Rockefeller University; http://www.gensat.org/ index.html, Allen Brain Atlas: http://www.brain-map.org/). However, we found NPY transcripts in Purkinje cells (Table 1). To clarify this discrepancy, further works are needed. Neuropeptide Y is synthesized by amacrine cells in the INL and those displaced in the GCL [10,11]. It has been reported that the ratio of NPY-producing amacrine cells in the INL to those in the GCL was 1:0.44 [11]. However, DNA microarray analysis showed that the expression ratio of NPY was 1:3.4. This discrepancy can be explained by the location of the cell bodies of amacrine cells in the INL. It is known that 96% of amacrine cells in the INL lie in the
923
innermost row [11]. This area may be damaged during LCM, resulting in the smaller amount of RNA. The presence of other cells producing NPY transcripts other than amacrine cells may also interfere with the ratio. We found widely distributed GPCRs in the nervous system (Table 2). The high ratio of orphan GPCRs (6 of 12) indicates our limited knowledge about the ligand-receptor system in the nervous system. Extensive research of orphan GPCRs in neuroscience is thus necessary. Over the past decade, numerous studies have identified many neuropeptides and investigated their functions in the central nervous system [12,13]. The real molecular mechanism of neuropeptide functions in the nervous system is still unclear. We believe the study of activity-dependent gene expression including the neuropeptide-receptor system is one of the major research areas for molecular neuroscience in the coming years. The precise expression data of ligands and their receptors is basic information to design experiments for the study of activity-dependent gene expression. References [1] Waites CL, Craig AM, Garner CC. Mechanisms of vertebrate synaptogenesis. Annu Rev Neurosci 2005;28:251–74. [2] Gainetdinov RR, Premont RT, Bohn LM, Lefkowitz RJ, Caron MG. Desensitization of G protein– coupled receptors and neuronal functions. Annu Rev Neurosci 2004;27:107– 44. [3] Lucas LR, Pompei P, Ono J, McEwen BS. Effects of adrenal steroids on basal ganglia neuropeptide mRNA and tyrosine hydroxylase radioimmunoreactive levels in the adrenalectomized rat. J Neurochem 1998;71:833– 43. [4] Zachrisson O, Nomikos GG, Marcus MM, Svensson TH, Lindefors N. Effects of antipsychotic drugs on cholecystokinin and preprotachykinin (substance P) mRNA expression in the rat hippocampal formation. Eur Neuropsychopharmacol 2000;10:355– 63. [5] Bräuer AU, Savaskan NE, Plaschke M, Ninnemann O, Nitsch R. Cholecystokinin expression after hippocampal deafferentiation: molecular evidence revealed by differential display–reverse transcription– polymerase chain reaction. Neuroscience 2003;121:111–21. [6] Nakamura NH, McEwen BS. Changes in interneuronal phenotypes regulated by estradiol in the adult rat hippocampus: a potential role for neuropeptide Y. Neuroscience 2005;136:357– 69. [7] Freund TF, Buzsáki G. Interneurons of the hippocampus. Hippocampus 1996;6:347– 470. [8] Freund TF. Interneuron Diversity series: Rhythm and mood in perisomatic inhibition. Trends Neurosci 2003;26:489 –95. [9] Mátyás F, Freund TF, Gulyás AI. Immunocytochemically defined interneuron populations in the hippocampus of mouse strains used in transgenic technology. Hippocampus 2004;14:460 – 81. [10] Ammar DA, Hugbes BA, Thompson DA. Neuropeptide Y and the retinal pigment epithelium: receptor subtypes, signaling, and bioelectrical responses. Invest Ophtalmol Vis Sci 1998;39:1870 – 8. [11] Sinclair JR, Nirenberg S. Characterization of neuropeptide Y– expressing cells in the mouse retina using immunohistochemical and transgenic techniques. J Comp Neurol 2001;432:296 –306. [12] Baraban SC, Tallent MK. Interneuron Diversity series: interneuronal neuropeptides— endogenous regulators of neuronal excitability. Trends Neurosci 2004;27:135– 42. [13] Ludwig M, Leng G. Dendritic peptide release and peptide-dependent behaviors. Nat Rev Neurosci 2006;7:126 –36.