- Email: [email protected]
Identification of gene expression profiles in rat ears with cDNA microarrays
Available online at www.sciencedirect.com R
Hearing Research 175 (2003) 2^13 www.elsevier.com/locate/heares
Identi¢cation of gene expression pro¢les in rat ears with cDNA microarrays Jizhen Lin a
a;
, Masashi Ozeki a , Eric Javel a , Zhenfen Zhao b , Wei Pan c , Eileen Schlentz a , Samuel Levine a
Department of Otolaryngology, University of Minnesota Medical School, Division of Biostatistics, Minneapolis, MN 55455, USA b Department of Pediatrics, University of Minnesota Medical School, Division of Biostatistics, Minneapolis, MN 55455, USA c School of Public Health, University of Minnesota, Minneapolis, MN 55455, USA Received 18 March 2002; accepted 18 September 2002
Abstract The physiological processes of hearing implicate thousands of molecules acting in harmony; however, their identities are only partially understood. We used cDNA microarrays containing 1,176 genes to identify s 150 genes expressed in rat middle and inner ear tissue. Expressed genes covered several gene families and biological pathways, many of which have previously not been described. Transcription factor genes that were expressed included inhibitors of DNA binding protein (Id). These were localized to the spiral ganglion, organ of Corti and stria vascularis, and they are possibly involved in neurogenesis and angiogenesis. Transcriptional factors that were highly expressed included Gax (homeobox) and I-UB, which inhibit cellular proliferation. Their presence suggests that inhibitory programs for cell proliferation are enforced in the ear. Ion channel genes that were expressed included voltage-dependent L-type calcium channels (LTCC) and proton-gated cation channels (PGCC). Genes involved in neurotransmitter production and release included glutamic acid decarboxylase (GAD1). Genes involved in postsynaptic inhibition included neuropeptide Y5 receptors (NPY5) and GAD1. Due to the existence of receptors and/or enzymes involved in their biochemical synthesis, neurotransmitters associated with these might include serotonin, glutamide, acetylcholine, Q-aminobutyric acid (GABA), neurotensin, and dopamine. ; 2002 Elsevier Science B.V. All rights reserved. Key words: Cochlea; Gene pro¢le; Microarray; Rat
1. Introduction The inner ear is a complex sensory organ possessing
* Corresponding author. 2001 Sixth Street S.E., 216 Lions Research Bldg., Minneapolis, MN 55455, USA. Tel.: +1 (612) 6245059; Fax: +1 (612) 626-9871. E-mail address: [email protected] (J. Lin). Abbreviations: Id, inhibitor of DNA binding protein or inhibitor of di¡erentiation; LTCC, voltage-dependent L-type calcium channel; PGCC, proton-gated cation channel; GAD1, glutamic acid decarboxylase; NPY5, neuropeptide Y5 receptor; GABA, Q-aminobutyric acid; ABR, auditory brainstem response; DLU, digitalized light unit; RT-PCR, reverse transcription-polymerase chain reaction; MAKP1, mitogen-activated kinase protein 1; Gax, growth arrest-speci¢c protein; NF-UB, nuclear factor-kappa B; I-UB, inhibitor of nuclear factor-kappa B; bHLH, basic helix-loop-helix
receptor or hair cells, various types of supporting cells, and yet other cells that provide nutrition and maintain the ionic environment. Hair cells are specialized mechanoreceptors that convert mechanical vibration into electrical potentials. Supporting cells provide sca¡olds for hair cells and, in conjunction with cells of the stria vascularis, maintain the inner ear’s unique ionic environment. In contrast, the middle ear is a structure that passively ampli¢es and ¢lters sound pressure. It possesses a specialized mucosa that contains polarized epithelial and secretory cells. The functional roles played by cells of the middle and inner ear implicate a large ensemble of molecules. Only a few of the genes or proteins involved in auditory structure and function have been identi¢ed to date. Knowledge of the complete repertoire of genes expressed in the ear is important be-
0378-5955 / 02 / $ ^ see front matter ; 2002 Elsevier Science B.V. All rights reserved. PII: S 0 3 7 8 - 5 9 5 5 ( 0 2 ) 0 0 7 0 4 - 9
HEARES 4022 18-12-02
J. Lin et al. / Hearing Research 175 (2003) 2^13
3
Table 1 Genes of interest in rat ear by category, in decreasing order of expression intensity Class
GB no.
Gene/protein name
Transcription factors (8/24)
Z17223
Gax
4889
X63594
I-UB
4270
D10862 D13374
Id1 Nm23-M2
3687 3038
+
D10864 M91597
Id3 NDK-B
2046 1473
+
L26267
NF-UB
843
X91810
Stat3
754
Growth factors
M13969
758
Ion water pumps and channel proteins (8/145)
M59786
IGF-II, insulin-like growth factor II LTCC
1079
M28647 M91808 D84450 U09402
Naþ /Kþ ATPase K1 sodium channel L1 Naþ /Kþ ATPase L3 P2U purinoceptor 1, ATP receptor
1024 937 864 801
J02701 AF013598 L06096
Naþ /Kþ -transporting ATPase L1 PGCC, sensory neuron speci¢c IP3R
M69055
rIGFBP-6
1620
M86870 M17528 J04486
CABP2 GNAI2 IGFBP-2; Brl-2
1590 1127 987
L35921 U34959
G(i)/G(s)/G(o) Q-9 G(i)/G(s)/G(t) L2
M83676
rab12, ras-related GTPase
U77918 D10874
TBP1 ATP6C
4949 3353
M83681
rab16, ras-related GTPase
2357
U43175 D10021
ATPase, subunit F, vacuolar ATP5H
1020 1018
M19044
(ATP5B)
D13124 M61177 M17086 L24907 D30040 J02592
P2 gene, subunit c MAPK1 cAMP-dependent PK type I-K Caþ /calmodulin-dependent PK type I RAC-PK-K K GSTM2
754 11846
U03763
calcium-dependent PLA2
11701
M33648
hydroxymethylglutaryl-CoA synthase
Ion and other binding proteins (N/A)
GTPase ATPase (N/A)
Protein kinases (N/A)
Other enzyme proteins (N/A)
Level (DLU)
763 693 610
PCR
Functions and characteristics
+
inhibition of cellular growth and proliferation inhibitor of NF-UB, inhibition of cellular growth neurogenesis and angiogenesis oncogene c-myc-related transcription factor neurogenesis and angiogenesis c-myc-related transcription factor, cellular growth cellular growth in response to mitogenic signals signal transducers, critical for cell growth cellular growth and di¡erentiation
+
+ +
954 875
10946
973
HEARES 4022 18-12-02
852 1620 1048 876
6515
+
+
ion transport, Ca2þ channel, voltage-dependent Naþ /Kþ pump Naþ pump Naþ /Kþ pump ion transport, secretion and absorption of cations Naþ /Kþ pump Hþ pump, detection of pH value a channel releasing intracellular Ca2þ insulin-like growth factor binding protein calcium binding protein guanine nucleotide binding protein insulin-like growth factor binding protein G-proteins/GTP binding proteins, involved signal transduction together with rab12 and 16 (see below), to modulate cGMP-gated cation channels involved in signal transduction via G-proteins ATPase, proteasomal, liver vacuolar ATP synthase, neurotransmitter tra⁄c involved in signal transduction via G-proteins energy metabolism mitochondrial ATP synthase D subunit mitochondrial ATP synthase L subunit precursor ATP synthase, energy metabolism mitogen signaling molecule cAMP signaling system Ca2þ signaling system Rac-K serine/threonine kinase glutathione S-transferase Yb subunit enzyme for arachidonic acid metabolism mitochondrial energy production
4
J. Lin et al. / Hearing Research 175 (2003) 2^13
Table 1 (Continued). Class
Receptors (14/197)
Stress-related proteins (6/35)
GB no.
Gene/protein name
M34445 X02904 X82679 M21410
GAD1 GST7-7 phospholipid hydroperoxide glutathione peroxidase 5-hydroxytryptamine 2C receptor
1911
L10073
5-hydroxytryptamine receptor 5B
1900
X97121
neurotensin receptor 2 (NTR2)
1023
Z11932
vasopressin V2 receptor
987
U20907
5-HT receptor 4; 5-HT4L
824
M35077 M35162 U66274
dopamine receptor GABA receptor N precursor NPY5
812 803 792
+
AF016297
neuropilin (NP)
762
+
D14869
prostaglandin E2 receptor EP3
753
M31838
neurokinin A receptor, Tacr2
633
+
AF005099
neuronal pentraxin receptor, NPR
578
+
M63901 D12820
secretogranin V, SGV GUST27
s 536 528
+ +
S45392 J03752 Y00404 M86389
heat shock protein 84 (HSP84) GST12; MGST1 Cu-Zn superoxide dismutase heat shock 27-kDa protein (HSP27) heat shock 70-kDa protein (HSP70)
14330 2867 2243 2028
Z27118
Skeleton and motility (3/7)
others
Level (DLU) 3521 2178 1786
Functions and characteristics
+
GAD1 glutathione S-transferase P subunit neurotransmitter metabolism
+
1034
X62660 X62908 X67788 U58858
glutathione transferase, subunit 8 co¢lin ezrin; cytovillin; villin 2 (VIL2) plakoglobin
910 4090 987 976
U34959
transducin L-2 subunit (TDB2)
5718
U62326
MIF
4156
M84716
Fte-1; RPS3A
2536
X74806
von ebner’s gland protein 2, VEGP
2090
L31884
TIMP2
1434
M25889 M24105
myelin basic protein S synaptobrevin-2 (SYB2)
973 947
M96377
non-processed neurexin II-L major
929
HEARES 4022 18-12-02
PCR
+
+
serotonin receptor, water and ion metabolism serotonin receptor, water and ion metabolism G-protein coupled receptor, signaling molecule blood vessel (£ow) control and regulation serotonin receptor, metabolism of water and ions neurotransmitter conduction neurotransmitter conduction neuropeptide responding to PKC and cAMP axonal growth, fasciculation, branching, and etc. cellular growth and ion/water metabolism linked to Ca2þ and cAMP responses a member of the neural uptake system induction of dopamine release a protein for purinergic signaling system chaperones and/or enzymes involved in protection of the cells under stress conditions such as heat, trauma, high intensity noise exposure, infection, and ototoxicity drug administration; some of these proteins are also important components for neurotransmitter metabolism co¢lin is a microvillar cytoplasmic peripheral protein; cytovillin is a cytoskeletal linker protein linked to lamins, keratins, and neurophilaments; plakoglobin and plectin are involved in cell^cell adhesion, and adherent junctions G-proteins, involved in signal transduction macrophage migration inhibitory factor putative v-fos transformation e¡ector protein serious gland product in response to stress, etc. tissue inhibitor of metalloproteinase structural protein of myelin sheath vesicle associated proteins involved in vesicle fusion and transmission (see syntaxins below) neuronal cell surface proteins, axonal insulation
J. Lin et al. / Hearing Research 175 (2003) 2^13
5
Table 1 (Continued). Class
GB no.
Gene/protein name
Level (DLU)
PCR
Functions and characteristics
X76489
CD9, platelet surface protein
894
+
protocadherin 4 (PC4)
802
+
M20373
neuropeptide Y precursor (NPY)
783
+
K03242
myelin P0 protein precursor, MPZ
762
surface marker for neural progenitors decisive adhesion molecule in synaptogenesis neuropeptide responding to cAMP and PKC extracellular matrix, structural protein of myelin
AB004276
Genes listed in the table had a signal v 750 DLU, that is, at least 50% higher than background (500 DLU). Certain genes expressed less highly were included because their existence in the ear had been con¢rmed by RT-PCR. This is based on a criterion adapted by most gene expression pro¢ling studies. Genes in bold face represent ones previously not described in ear tissue and their presence and expression con¢rmed by RTPCR (except for Stat3). N/A, not applicable; GB no., GenBank accession number. Full names of the genes in the table and detailed information regarding their functions may be obtained at the website (www.ncbi.nlm.nih.gov/) using their accession numbers.
cause it improves understanding of the molecular mechanisms involved in hearing and permits the design of innovative strategies for reversing or preventing cochlear hearing loss. Defects in any of the genes involved in auditory structure and function may upset the balance of several intricate physiological processes known to exist in hearing, among which are maintenance of a 150-mV potential di¡erence across hair cell membranes and generation of rapid motile responses at rates that can approach 100 kHz. Accumulated evidence to date indicates that more than 40 genomic DNA loci are linked to non-syndromic hearing loss in humans, and these are distributed throughout the genome. Only a handful of these genes have been cloned and characterized (Avraham, 1998). The pro¢le of genes expressed in the ear is not well understood. It is generally accepted that about 30,000^ 40,000 genes exist in genomic DNA, and approximately 5^10% of these are likely to be expressed in the ear (http://www.ncbi.nlm.nih.gov/disease). Some genes are expressed in the developing ear but are inactivated postnatally, and others are expressed only in mature ears. We were curious about the genes expressed by mature ear tissue because they are more likely to encode ‘working’ molecules that mediate biophysical processes of hearing and maintain unique structure of the ear. Genes that are highly active or abundantly expressed by ear tissue may be detected by chip-based technologies. cDNA microarrays or ‘gene chips’ (Schena et al., 1995) are nylon membranes or slides dotted with potentially thousands of gene-speci¢c cDNA probes or oligos. Hybridization of these gene-speci¢c cDNA fragments with cDNA or cRNA probes prepared from a target tissue or organ permits quantitative analysis of all oligos contained on the chip. It is, therefore, a very useful and powerful tool to detect major gene families or biochemical pathways in a given target tissue. cDNA microarrays have a wide range of applications (Nelson,
1996), among which are investigating normal biological and disease processes (Efron et al., 2000), pro¢ling differential gene expression (Lomax et al., 2000), and discovering potential therapeutic drug targets. Otologic investigators have recently applied this technique for addressing the issue of gene expression during development in mice (Chen and Cpreu, 2001), gene expression pro¢les in subfractions of the auditory system, namely, the rat cochlea, cochlear nucleus and inferior colliculus (Cho et al., 2002), and the pathology of middle ear infectious disease in rats (Lin et al., 2002; Pan et al., 2002). In this report, we used cDNA microarrays to identify major gene families expressed by rat ear. We found that several genes were expressed which have previously not been described in the ear.
2. Materials and methods Healthy Sprague^Dawley rats weighing 230^250 g were used in this study. The rats had no previous exposure to noise or speci¢c pathogens. Auditory brainstem response (ABR) recordings in response to calibrated acoustic signals were obtained in seven randomly selected rats to verify that auditory sensitivity to rarefaction clicks and tone bursts over a 1^16 kHz range laid within normal limits. Following this, animals were anesthetized and decapitated, both bullae were dissected out, and all exterior muscle and connective tissues were completely removed. Twelve bullae were immediately submerged in liquid nitrogen for microarray experiments. Another 12 bullae were used for dissection of the cochlear and middle ear tissues using a stereomicroscope. The vestibular system was not targeted. An additional six bullae were ¢xed in 10% formalin, embedded in para⁄n, and processed for in situ hybridization as previously described (Lin et al., 1999a,b,c).
HEARES 4022 18-12-02
6
J. Lin et al. / Hearing Research 175 (2003) 2^13
2.1. cDNA microarray protocols Procedures for acquiring and analyzing data using cDNA microarray (Atlas1 rat 1.2 array, Clontech, Palo Alto, CA, USA) were the same as described previously (Lin et al., 2002). This cDNA microarray was selected for its extensive coverage of genes involved in various biochemical reactions and pathways. Brie£y, mRNA was isolated from combined middle and inner ear tissue using stardard oligo dT methods. cDNA probes were synthesized from the mRNA by speci¢c CDS primer mix (provided by the cDNA array kit) and radiolabeled with 32 P. Total radioactivity of cDNA probes was 3.8U106 cpm. The Atlas cDNA array membrane was hybridized with the above cDNA probes, washed stringently, and exposed to a phosphorimager. Expression of genes and their intensities were quanti¢ed using OptiQuant software (version 03.00, vPackard) and presented as digitalized light unit (DLU). A blank area of the same chip was used for determining background radioactivity, and the intensity of each location on the grid that exhibited a level higher than background was measured. DLU values for each expressed gene were plotted as a three-dimensional (3-D) graph using the SigmaPlot software 5.0 according to the original coordinates of the chip (visit www. clontech.com for details). 2.2. Scatter plot analysis The entire experiment was duplicated to assess repeatability. To do this, the gene expression pattern from one of the mRNA pools (representing 12 bullae from six rats) was plotted against the expression pattern from the other mRNA pool. Digitized gene expression values were exported to an Excel spreadsheet, and the DLU value of each grid location was normalized by dividing by background intensity. The result was then exported for the scatter plot analysis. The scatter plot consisted of the gene index expressed by the ¢rst cDNA microarray experiment, plotted as a function of the gene index expressed by the second experiment. This provided a two-dimensional representation of the gene expression pattern to evaluate possible variations in gene expression between individual cDNA microarrays. 2.3. Con¢rmation of the microarray results by reverse transcription-polymerase chain reaction (RT-PCR) and/or in situ hybridization To avoid possible false positive results from cDNA microarrays, RT-PCR and/or in situ hybridization were selectively performed for candidate genes (see Tables 1 and 2) using methods described elsewhere (Lin et al., 2001, 1999). To study whether candidate genes were
only expressed in the ear, various tissues (eye, brain, nasopharynx, trachea, lung, liver, heart, kidney, spleen, intestine, stomach, and muscle) were used as controls. Total RNA was isolated from these tissues. To study the cellular location of candidate mRNA transcripts, the cochlear tissue of both one-day-old and adult rats was used for in situ hybridization. For RT-PCR, speci¢c gene sequence information was obtained from the website of the National Center of Biologic Information (http://www.ncbi.nlm.nih.gov), speci¢c primer pairs for genes of interest were designed (see Table 2), and their speci¢city was con¢rmed by BLAST search using the above website. RT-PCR was performed in the same manner as previously described (Lin et al., 1999a,b,c). Annealing temperature for individual primer pairs was determined according to the temperature (Tm ) of sense and antisense primers. PCR products were routinely sequenced for approximately up to 500 bp using an autosequencer. For in situ hybridization, the PCR products of candidate genes were cloned into the vector (pT-Adv) using a PCR cloning kit (Clontech) after sequence analysis. The cDNA probe in the vector was re-con¢rmed by DNA sequencing and then transcribed into a 35 S-radiolabeled antisense riboprobe using T7 promoter sequence with a sense riboprobe as control. This was then hybridized with ear tissue sections using procedures described earlier (Lin et al., 2001, 1999). Sections were autoradiographed and counterstained with hematoxylin.
3. Results Representative ABRs obtained from one rat in response to 100-Ws rarefaction clicks are shown in the left column of Fig. 1. The right-hand panel indicates auditory sensitivity across a 1^16 kHz frequency range for seven of the rats used in this study. Solid and dashed lines indicate mean ABR thresholds T 1 standard deviation (S.D.) in a di¡erent set of normal-hearing rats. These data indicate that the animals used here had essentially normal hearing, suggesting a fully functional complement of receptor and spiral ganglion cells. As shown in Fig. 2A, hybridization of the cDNA probes from pooled mRNA of 12 ears exhibited a unique, repeatable expression pattern. Hybridizing signals exceeded background levels of approximately 500 DLU for s 200 grid locations. The locations covered many gene families such as transcription factor, growth factor, ion channel, receptor, etc. Genes whose expression either attained a level of 50% greater than background or whose expression levels fell between 500 and 750 DLU but were con¢rmed by RT-PCR, are listed in Table 1. Seventy-six genes met these criteria. An addi-
HEARES 4022 18-12-02
J. Lin et al. / Hearing Research 175 (2003) 2^13
7
Table 2 Primers pairs used to examine gene expression Gene
Primer
Oligonucleotide sequence
Annealing Tm
Size (bp)
Gax
sense antisense sense antisense sense antisense sense antisense sense antisense sense antisense sense antisense sense antisense sense antisense sense antisense sense antisense sense antisense sense antisense sense antisense sense antisense sense antisense sense antisense sense antisense sense antisense sense antisense sense antisense sense antisense sense antisense sense antisense
5P-AGCGACAGTTCAGATTCCC-3P 5P-TCCCTTTTTCACATTCACCAG-3P 5P-GCCATCTCGCGCTGCGCTGGG-3P 5P-GGTGGAGAGGGCGAGACTCCG-3P 5P-CCCCTGCGGCGGCTCCCTCC-3P 5P-CCAGGCCGGAGACACCTGGGG-3P 5P-GCCATTGCGCGAGGCCGGGG-3P 5P-CCTTCGCTCTCGGGCTCCAG-3P 5P-CGGCCGCACACCTGGCCAAC-3P 5P-GGCGCCGTCGGGCTGCGGCGG-3P 5P-CCCGAGCACATCCCTACTCCAGGGG-3P 5P-CCCTCCCAGAGGGTTGGCCCCGTCA-3P 5P-GGCCTTGCCCGGATTGCTGACCC-3P 5P-GGTGCCCCTGGCTGGAAGCGGGC-3P 5P-CTGGTGCCCGCTTCCGGCGC-3P 5P-GGGCACAGCCGCCATGCCTT-3P 5P-TGCTCCCCTTTCACCCTGAC-3P 5P-GGGGCGGGTAAGACACAGGC-3P 5P-GGGCCAGCCCCCCGCACCACCTGGC-3P 5P-GGAGGGCTGGGTCTGGGTCTGGGGG-3P 5P-CGTGCTCAGCTACTACAGGTACCA-3P 5P-GGGAGGCGTTGAGATTGGA-3P 5P-ATATTAAAACCTGGCGGCCA-3P 5P-GGCCGCTCTGTCTTCATCAG-3P 5P-GGTACCATCAGCTTCTGCCC-3P 5P-GATCTGGCTGGCATCCTTGT-3P 5P-CCACCTTCTCTGATGGGAGG-3P 5P-GTTTTTGCCATGCCGGTAGA-3P 5P-GAAAGAGAGGAGCCGATCCA-3P 5P-CAGTTCCGTCTCGAATTGCC-3P 5P-GCCTGGGAATCGTGTCAAAG-3P 5P-CCCCCAGGTCTAAGTTCTGAATC-3P 5P-CCACCAGGCCATGAATCTTG-3P 5P-ACTGGAATTCTCGGCTGAACTC-3P 5P-GTGCTTGGGAACTTGCTCATC-3P 5P-CGATCACAGGCCATTAAAGTGA-3P 5P-AGCAGCTCCGGAACCAGAT-3P 5P-CAGAGGGATGACGTGGACCTT-3P 5P-TCTGTGACTCCCATGAAAATCAA-3P 5P-CTGCCCACGAGTTTTGCAAT-3P 5P-TCACAGTGCATGCTGGGATT-3P 5P-GGGCAGATGTCCGAGATAAACT-3P 5P-ATGCTGCAAGCCATGATCTTG-3P 5P-TCTTTCTTGCCCGACTTGTTG-3P 5P-GCTCTATCCCTGCTCGTGTGT-3P 5P-CACCACATGGAAGGGTCTTCA-3P 5P-CTCGTTCCTTTTTCCTCCTT-3P 5P-CTGTGTGCGCAAAGACTAGC-3P
51‡C
301
66.7‡C
732
69‡C
668
66‡C
480
71.5‡C
473
70‡C
561
70.4‡C
572
68.3‡C
590
62‡C
720
74.7‡C
543
60‡C
296
59‡C
151
58.5‡C
151
59‡C
251
58.5‡C
251
59‡C
251
59‡C
251
58.5‡C
251
59‡C
251
59‡C
251
58.5‡C
251
59.5‡C
251
58.5‡C
251
55.5‡C
296
Id1 Id2 Id3 Id4 LTCC MAKP1 GAD1 NPY5 PGCC IP3Ra RAC-PKa NTR2a Nrp2a Tacr2a NPRa SGVa GUST 27a TDB2a VEGPa CD9a PC4a NPYa S29 a
PCR result is marked in Table 1 (both cochlear tissue and hair cell progenitors are positive) and the rest of PCR results is presented in Fig. 3.
tional 47 genes exhibited expression levels of 500^749 DLU, but their presence was not con¢rmed by RTPCR. These are not listed in Table 1. Some signals were especially strong ( s 10 000 DLU), some were very weak ( 6 500 DLU), and yet others were expressed at intermediate levels. Fig. 2A shows the intensity in DLU of gene expression for 12 ears in six normal rats. The scatter plot in Fig. 2B shows that the expression pattern (in terms of the distribution of dots) was
symmetric between two replications using cDNA microarrays. Thus, gene expression was highly repeatable across cDNA microarray experiments in normal rats. Of 24 transcription factors studied, eight were expressed in the middle and inner ears of mature rats (see Table 1). Gax (growth arrest-speci¢c protein) and I-UB (inhibitor of nuclear factor-kappa B) were highly expressed. These are involved in cell cycle arrest and cellular di¡erentiation. Expression of some of these
HEARES 4022 18-12-02
8
J. Lin et al. / Hearing Research 175 (2003) 2^13
Fig. 1. Auditory sensitivity of rats used in this study. Left: Representative ABRs elicited by 100-Ws rarefaction clicks at intensities of 10^90 dB SPL. Right: Symbols connected by solid lines indicate ABR thresholds at octave intervals from 1 to 16 kHz for seven of the rats in this study. Heavier solid and dashed lines represent mean thresholds T 1 S.D. for a di¡erent series of rats with normal hearing.
genes in cochlear and middle ear tissue was veri¢ed by RT-PCR. Speci¢c genes chosen for RT-PCR analysis are indicated in Table 1. Some of the PCR products are presented in Fig. 3A. These genes included inhibitor of DNA binding protein (Id)1^4 (Nagata and Todokoro, 1994), voltage-dependent L-type calcium channel (LTCC) alpha 1C (Koch et al., 1990), mitogen-activated kinase protein 1 (MAKP1) (Maisonpierre et al., 1991), glutamic acid decarboxylase (GAD1) (Wyborski et al., 1990), neuropeptide Y5 receptor (NPY5) (Hu et al., 1996), and proton-gated cation channel (PGCC, sensory neuron speci¢c) (Waldmann et al., 1997a,b). These genes were obviously expressed in inner ear tissue but were either unexpressed or weakly expressed in the middle ear. For example, Id1 gene expression was relatively high in the inner ear, compared to that in the middle ear and in all other tissues except for lung and kidney (Fig. 3B, lanes 7 and 10, Id1). Since the expression pattern of the Id1 and Id3 genes in the inner ear is poorly understood, we further performed in situ hybridization to reveal their cellular locations. The results, shown in Fig. 4, indicate that Id1 and Id3 genes are expressed in the spiral ganglion, the organ of Corti and the stria vascularis of postnatal rats but not discernible in the counterparts of adult rats by in situ hybridization. It suggests that Id1 and Id3 genes are involved in the development of cochlea and their expression in adult cochlear tissue is downregulated. Other genes expressed in the ear included members of
several gene families (see Table 1). For example, eight of 145 genes encoding ion channels were expressed, as were genes encoding ion pumps (Na+/K+ ATPase), calcium binding protein such as calmodulin, and GTPases such as rab12 and 16. Fourteen of 197 genes encoding receptors were expressed, as were genes encoding enzymes such as glutathione S-transferase, growth factors such as insulin-like growth factor II, and mitogenic kinases such as MAPK1. Three of seven genes encoding the hair cell cytoskeleton were expressed (e.g. co¢lin), six of 35 stress-related genes were expressed (e.g. HSP90 and HSP27), and housekeeping proteins such as ribosomal protein S29 were also expressed. Many other genes known to be expressed by ear tissue are not included in the particular gene chip used in this study. Among these are Brn-3, Pax2, and Math-1. Due to the limited sensitivity of cDNA microarrays, some of the genes included in this gene chip and which were expected to be expressed in ear tissue, ¢broblast growth factor for example, were in fact undetectable in this study.
4. Discussion We have demonstrated in this study that cDNA microarrays are useful in discovering gene expression patterns in the ear. We identi¢ed several genes that previ-
HEARES 4022 18-12-02
J. Lin et al. / Hearing Research 175 (2003) 2^13
Fig. 2. Gene expression pattern pro¢les of the normal ears from six SPF rats. Presentation of the gene expression pro¢les in rat bullae involves matching each dot on the array, quantifying the intensity of each dot, and exporting the data to the SigmaPlot 5.0 spreadsheet to create a three-dimensional graph corresponding to the original coordinate of each gene on the rat 1.2 array. X and Y axes in this graph represent the numeric order and alphabetical order on the original array, respectively. The Z axis indicates gene expression magnitude in a color-increasing manner (from low to high, this is black, pink, green, yellow, purple, red, blue, gray, dark red, and black again). This distinct gene expression pattern (A) is speci¢c to SPF rat ears. It is therefore termed the gene expression pattern signature. The high degree of reproducibility across animals is indicated by data in B, which show the index of genes expressed by the ¢rst set, plotted as a function of gene index expressed by a second set. Gene intensity has been arbitrarily cut o¡ at 15 000 DLU to facilitate identi¢cation of genes that are less highly expressed.
ously have not been known to be expressed by ear tissue. Identi¢cation of major gene families in wholeear tissue is a ¢rst but important step toward understanding basic biophysical and molecular mechanism involved in hearing. The gene expression pattern in the ear tissue is
9
unique (see Fig. 2A, gene expression signature) and is probably not matched by any other tissue. Characteristic features of transcription factors expressed in mature ear tissue are strongly with cell cycle arrest and cellular di¡erentiation (e.g. Gax and I-UB) and weakly with cellular growth and proliferation (e.g. nuclear factor-kappa B (NF-UB), Id genes, and oncogene-related transcription factors). It appears from this that mature ear tissue is biologically blocked for cellular proliferation. NF-UB is actively involved in cellular growth and proliferation, and the action of many mitogens and growth factors is linked to NF-UB or through the NF-UB pathway, but its expression level is much lower than its inhibitor I-UB. An observation favoring the concept that genes involved in cell cycle arrest and cellular di¡erentiation are expressed in mature ear tissue is that homeobox transcriptional factor Gax is highly expressed. The purpose of this gene is to keep well-di¡erentiated or highly polarized cells quiescent and stable, which is very important for hearing. On the other hand, Gax expression obviously does not facilitate hair cell regeneration. We believe that Gax and I-UB need to be downregulated if hair cell regeneration is to occur. Other homeobox transcription factors that were expressed in the ear included Id1 and Id3. Homeobox genes are known to be involved in the development of ears, and previously documented homeobox genes include Nkx 5.1, SoHo-1, Dlx-2 and 3, Msx-C and D, Hoxa-1 and 2, and Otx-1 and 2 (Rivolta, 1997). To our knowledge, Gax has not been previously described as being expressed by ear tissue. Noteworthy genes that were expressed include Id genes, LTCC, PGCC (sensory neuron speci¢c), neuropeptide Y5 receptor (NPY5), and several others (bold face in Table 1). The Id genes are known to function as negative regulators of the basic helix-loop-helix transcription factors (bHLH, a basic region plus two amphipathic K helices separated by an intervening loop). They play a critical role in the developmental processes for various cells (Weintraub et al., 1991). Id genes inhibit the functions of these transcription factors in a dominant-negative manner by blocking their DNA binding activity through the HLH domain (Benezra et al., 1990). Unlike other bHLH members, Id proteins contain only the HLH domain but miss the basic region adjacent to the HLH domain that is essential for speci¢c DNA binding. Therefore, Id proteins bind to bHLH transcription factors and inhibit their binding to the DNA ‘E box’ consensus sequence (Benezra et al., 1990; Littlewood and Evan, 1994; Murre et al., 1989). bHLH transcription factors activate the expression of genes involved in cell di¡erentiation, whereas Id proteins promote cell growth and proliferation (Norton et al., 1998). This suggests that interaction of Id proteins and bHLH transcription factors may play an im-
HEARES 4022 18-12-02
10
J. Lin et al. / Hearing Research 175 (2003) 2^13
portant role in controlling tissue development and differentiation. Recent studies have demonstrated that Id genes are required for neurogenesis and angiogenesis of tumor xenografts (Lyden et al., 1999), suggesting that they play a role in the development and maintenance of the blood vessel and nerve tissues. This role perhaps accounts for the expression of Id1 and Id3 in the organ of Corti, cochlear ganglion, and stria vascularis (Fig. 3). The LTCC gene is another entity prominently expressed in ear tissue. Its expression in cochlear tissue of mice and chickens has been reported previously (Green et al., 1996; Kollmar et al., 1997a,b). LTCC encodes voltage-gated L-type calcium channel protein and is actively involved in calcium movement across cell membranes. It is known that calcium movement is associated with the release of neurotransmitter at pre-
Id1 Id2 Id3 Id4
LTCC MAKP1 GAD1
NPY5 PGCC
A
synaptic terminals. The LTCC gene is clearly important in hearing, since its knockout in mice leads to degeneration of outer and inner hair cells and consequent deafness (Platzer et al., 2000). GAD1 is the enzyme responsible for the production of Q-aminobutyric acid (GABA) (Bain et al., 1993), an important inhibitory neurotransmitter (Arnold et al., 1998; Mu et al., 1994; Vale and Sanes, 2000). It is generally accepted that GABA is produced in e¡erent nerve endings, where its role is to modulate activity elicited in the a¡erent pathway. PGCC gene expression has not previously been described for inner ear tissue. It encodes a sensory ion channel protein that senses acid (Waldmann et al., 1997a,b) and therefore may be related to homeostasis or pain sensation. In vitro studies have shown that low extracellular pH can evoke inward currents in both the
1
5
8
IE
ME S29 9
10
Control
7
Intestine Stomach Muscle
6
Kidney Spleen
4
Liver Heart
3
Cochlea Middle ear Eye Brain Nasopharynx Trachea Lung
B
2
732
Id1 Id3
480
S29
296 1
2
3
4
5
6
7
8
9
10 11 12 13 14 15
Fig. 3. RT-PCR con¢rmation of partial genes identi¢ed by cDNA microarrays. mRNAs that were expressed in cochlear tissue included Id1 (lane 1, IE), Id3 (lane 3, IE), LTCC (lane 5), MAPK1 (lane 6), GAD1 (lane 7), NPY5 (lane 8), and PGCC (lane 9). Lane 10 represents a negative PCR control and omission of tissue from the middle ear. The bottom panel represents housekeeping gene ribosomal protein S29 (S29). Id2 and Id4 genes that were not detectable by cDNA microarrays were also not detected by RT-PCR. Lower panel represents the expression patterns of Id1 and Id3 genes in various tissues of rats with high levels in the cochlea (lane 1), brain (lane 4), and spleen (lane 11) tissues. Some more genes identi¢ed by microarrays and con¢rmed by RT-PCR are not presented in this ¢gure but indicated in Table 1.
HEARES 4022 18-12-02
J. Lin et al. / Hearing Research 175 (2003) 2^13
11
Fig. 4. Expression of Id genes in the cochlear tissue of rats (postnatal day 1) by in situ hybridization (panels a^f) but not detectable in the adult cochlear tissue (panels h and j). Sense riboprobes of Id1 (panel a) and Id3 (panel d) showed no hybridizing signals. Antisense riboprobes of Id1 (panels b and e) and Id3 (panels c and f). Arrows point to stria vascularis, organ of Corti and cochlear ganglion, indicating hybridizing signals for Id1 mRNA transcripts (panel e). Panels d, e, and f are ampli¢ed versions of panels a, b, and c. Note that Id1 gene has the same expression pattern as the Id3 gene, and both are expressed in stria vascularis, organ of Corti, and spiral ganglion (panels b and c, or e and f). Panels g and i, stained with hematoxylin and eosin (H and E), are serial sections of panels h and j hybridized with antisense Id1 and Id3 riboprobes, respectively. ISH, in situ hybridization; CD, cochlear duct; CM, cochlear modulus; SV, stria vascularis; SG, spiral ganglion; OHC, outer hair cells; bar = 10 W at same scale for panels a, b and c, for panels d, e, and f, for panels g and h, and for panels i and j.
HEARES 4022 18-12-02
12
J. Lin et al. / Hearing Research 175 (2003) 2^13
central nervous system and peripheral sensory neurons (Kovalchuk et al., 1990; Krishtal and Pidoplichko, 1980). The NPY5 gene is primarily expressed in the central nervous system and has been implicated in food intake (Gerald et al., 1996) and reduction of synaptic excitation (Ho et al., 2000). Although expression of this gene has not been previously described for ear tissue, the functional role it plays in audition is unclear. Since NPY is tied to calcium and cAMP signaling pathways (Higuchi et al., 1988), it is possible that NPY and NPY5 are e¡ector molecules of the calcium and cAMP signaling system in ear tissue. It has been noted in this study that the Ca2þ /calmodulin- and/or cAMPdependent pathways as well as G-protein coupled signal transduction system, as expected, are notable in the ear. As shown in Table 1, many genes in addition to the ones described above were also detected in ear tissue. Insulin-like growth factor II gene was expressed, which is essential for cellular growth and survival. MAPK1 was highly expressed, but we found it to be expressed primarily in the middle ear, where epithelium renewal is active, and not in cochlear tissue, where cellular renewal is inhibited. The genes listed in Table 1 represent only a portion of the gene repertoire expressed in the ear that we deem to be of interest. These genes are also expressed in developing cochlear tissue from embryonic days 12^18 although their expression levels vary and some of them (see Table 1, marked in bold face) are expressed in progenitor hair cell lines (data not shown), suggestive of their expression in hair cells. Many other genes were highly expressed in ear tissue but are not listed in Table 1 because of their housekeeping roles in cell metabolism. Genes in this category include those encoding cytochrome c oxidases, ribosomal proteins, actins, glyceraldehyde 3-phosphate dehydrogenase, and polyubiquitin. Some of the non-abundantly expressed genes or families have previously been described as being expressed by hair cells. Among these are Math-1 (bHLH transcription factor). Yet other genes known to be expressed by hair cells are not included in the cDNA microarrays we used. Among these are Nkx 5.1, SoHo-1, Dlx-2 and 3, Msx-C and D, Hoxa-1 and 2, and Otx-1 and 2. Further studies are warranted to amplify and study gene expression in precisely dissected subpopulations such as hair cells and supporting cells using techniques such as laser capture microdissection. An advantage of the relatively small gene chip used in this study is the low false positive rate it generates. We used RT-PCR to con¢rm the presence of some of the genes not previously described as being expressed in ear tissue. RT-PCR results are presented in Fig. 3 or indicated in Table 1. Most of the genes listed in Table 1
are expressed in developing ears of rats in our most recent studies, but their expression levels vary dramatically (data not shown). A disadvantage of this gene chip is the relatively small number of cDNAs on it. The limited number permits analysis of only a portion of the full gene repertoire expressed in rat ears. A full understanding of the complete repertoire of genes expressed in the ear and their expression dynamics in response to acoustic stimulation would require a chip possessing a substantially larger number of genes. Application of cDNA microarrays at subgenome or genome levels would permit analysis of the expression patterns of most, if not all, of the genes representing the major biochemical pathways and biophysical processes in the ear. Future studies employing expanded gene chips will be extremely helpful for (1) understanding the intricate and precise biochemical/biophysical processes that underlie hearing at a molecular level and (2) permitting rational design of gene therapy strategies for preventing, ameliorating or reversing hearing loss.
Acknowledgements This work was supported by grants from the National Organization for Hearing Research, the Lions 5M Hearing Foundation, and the Viking Children’s Fund, and in part by grants R01-DC03433 (microarray administrative supplement) and P30 DC04660 from NIDCD. References Arnold, T., Oestreicher, E., Ehrenberger, K., Felix, D., 1998. GABA(A) receptor modulates the activity of inner hair cell a¡erents in guinea pig cochlea. Hear. Res. 125, 147^153. Avraham, K.B., 1998. Hear come more genes! Nat. Med. 4, 1238^ 1239. Bain, G., Ramkumar, T.P., Cheng, J.M., Gottlieb, D.I., 1993. Expression of the genes coding for glutamic acid decarboxylase in pluripotent cell lines. Brain Res. Mol. Brain Res. 17, 23^30. Benezra, R., Davis, R.L., Lockshon, D., Turner, D.L., Weintraub, H., 1990. The protein Id: a negative regulator of helix-loop-helix DNA binding proteins. Cell 61, 49^59. Chen, Z.Y., Cpreu, D.P., 2001. An inner ear gene expression database. JARO 3, 140^148. Cho, Y., Gong, E.L., Stover, T., Lomax, M.I., Altschuler, R.A., 2002. Gene expression pro¢les of the rat cochlea, cochlear nucleus, and inferior colliculus. JARO 3, 54^67. Efron, B., Tibshirani, R., Goss, V., Chu, G., 2000. Microarrays and their use in a comparative experiment. In: Technical report, Statistics Department, Stanford University. Gerald, C., Walker, M.W., Criscione, L., Gustafson, E.L., Batzl-Hartmann, C., Smith, K.E., Vaysse, P., Durkin, M.M., Laz, T.M., Linemeyer, D.L., Scha¡hauser, A.O., Whitebread, S., Hofbauer, K.G., Taber, R.I., Branchek, T.A., Weinshank, R.L., 1996. A receptor subtype involved in neuropeptide-Y-induced food intake. Nature 382, 168^171.
HEARES 4022 18-12-02
J. Lin et al. / Hearing Research 175 (2003) 2^13 Green, G.E., Khan, K.M., Beisel, D.W., Drescher, M.J., Hat¢eld, J.S., Drescher, D.G., 1996. Calcium channel subunits in the mouse cochlea. J. Neurochem. 67, 37^45. Higuchi, H., Yang, H.Y., Sabol, S.L., 1988. Rat neuropeptide Y precursor gene expression. mRNA structure, tissue distribution, and regulation by glucocorticoids, cyclic AMP, and phorbol ester. J. Biol. Chem. 263, 6288^6295. Ho, M.W., Beck-Sickinger, A.G., Colmers, W.F., 2000. Neuropeptide Y(5) receptors reduce synaptic excitation in proximal subiculum, but not epileptiform activity in rat hippocampal slices. J. Neurophysiol. 83, 723^734. Hu, Y., Bloomquist, B.T., Corn¢eld, L.J., DeCarr, L.B., Flores-Riveros, J.R., Friedman, L., Jiang, P., Lewis-Higgins, L., Sadlowski, Y., Schaefer, J., Velazquez, N., McCaleb, M.L., 1996. Identi¢cation of a novel hypothalamic neuropeptide Y receptor associated with feeding behavior. J. Biol. Chem. 271, 26315^26319. Koch, W.J., Ellinor, P.T., Schwartz, A., 1990. cDNA cloning of a dihydropyridine-sensitive calcium channel from rat aorta. Evidence for the existence of alternatively spliced forms. J. Biol. Chem. 265, 17786^17791. Kollmar, R., Fak, J., Montgomery, L.G., Hudspeth, A.J., 1997a. Hair cell-speci¢c splicing of mRNA for the alpha1D subunit of voltagegated Ca2+ channels in the chicken’s cochlea. Proc. Natl. Acad. Sci. USA 94, 14889^14893. Kollmar, R., Montgomery, L.G., Fak, J., Henry, L.J., Hudspeth, A.J., 1997b. Predominance of the alpha1D subunit in L-type voltage-gated Ca2+ channels of hair cells in the chicken’s cochlea. Proc. Natl. Acad. Sci. USA 94, 14883^14888. Kovalchuk, Y.N., Krishtal, O.A., Nowycky, M.C., 1990. The protonactivated inward current of rat sensory neurons includes a calcium component. Neurosci. Lett. 115, 237^242. Krishtal, O.A., Pidoplichko, V.I., 1980. A receptor for protons in the nerve cell membrane. Neuroscience 5, 2325^2327. Lin, J., Ho, S.B., Schekels, L.L., Paparella, M.M., Kim, Y., 1999a. Mucin expression in the rat middle ear: an improved method for RNA harvest. Ann. Otol. Rhinol. Laryngol. 108, 762^768. Lin, J., Kawano, H., Paparella, M.M., Ho, S.B., 1999b. Improved RNA analysis for immediate autopsy of temporal bone soft tissues. Acta Otorhinolaryngol. 119, 787^795. Lin, J., Tsprun, V., Kawano, H., Paparella, M.M., Zhang, Z.-Q., Anway, R., Ho, S.B., 2001. Characterization of mucins in human middle ear and eustachian tube. Am. J. Physiol. Lung Cell Mol. Physiol. 280, L1157^1167. Lin, J., Tsuboi, Y., Pan, W., Giebink, S.G., Adams, G.L., Kim, Y., 2002. Analysis by cDNA microarrays of altered gene expression in middle ears of rats following pneumococcal otitis media. Int. J. Ped. Otolarynogol. 65, 203^211. Lin, J., Vambutas, A., Haruta, A., Paparella, M.M., Giebink, G.S., Kim, Y., 1999c. Pneumococcus activation of the 5-lipoxygenase pathway and production of glycoproteins in the middle ear of rats. J. Infect. Dis. 179, 1145^1151. Littlewood, T.D., Evan, G.I., 1994. Transcription factors 2: helixloop-helix. Protein Pro¢le 1, 635^709. Lomax, M.I., Huang, L., Cho, Y., Gong, T.L., Altschuler, R.A.,
13
2000. Di¡erential display and gene arrays to examine auditory plasticity. Hear. Res. 147, 293^302. Lyden, D., Young, A.Z., Zagzag, D., Yan, W., Gerald, W., O’Reilly, R., Bader, B.L., Hynes, R.O., Zhuang, Y., Manova, K., Benezra, R., 1999. Id1 and Id3 are required for neurogenesis, angiogenesis and vascularization of tumour xenografts. Nature 401, 670^677. Maisonpierre, P.C., Le Beau, M.M., Espinosa, R., Ip, N.Y., Belluscio, L., de la Monte, S.M., Squinto, S., Furth, M.E., Yancopoulos, G.D., 1991. Human and rat brain-derived neurotrophic factor and neurotrophin-3: gene structures, distributions, and chromosomal localizations. Genomics 10, 558^568. Mu, M.Y., Pao, X., Jie, T., 1994. SEM observation and GABA immunoreactivity of human embryonic organ of Corti tissue culture. Acta Otolaryngol. 114, 516^519. Murre, C., McCaw, P.S., Baltimore, D., 1989. A new DNA binding and dimerization motif in immunoglobulin enhancer binding, daughterless, MyoD, and myc proteins. Cell 56, 777^783. Nagata, Y., Todokoro, K., 1994. Activation of helix-loop-helix proteins Id1, Id2 and Id3 during neural di¡erentiation. Biochem. Biophys. Res. Commun. 199, 1355^1362. Nelson, N., 1996. Microarrays pave the way to 21st century medicine. J. Natl. Cancer Inst. 88, 1803^1805. Norton, J.D., Deed, R.W., Craggs, G., Sablitzky, F., 1998. Id helixloop-helix proteins in cell growth and di¡erentiation. Trends Cell Biol. 8, 58^65. Pan, W., Lin, J., Le, C.T., 2002. Model-based cluster analysis of cDNA microarray data. Genome Biol.3, research 0022.1^0022.10. Platzer, J., Engel, J., Schrott-Fischer, A., Stephan, K., Bova, S., Chen, H., Zheng, H., Striessnig, J., 2000. Congenital deafness and sinoatrial node dysfunction in mice lacking class D L-type Ca2+ channels. Cell 102, 89^97. Rivolta, M.N., 1997. Transcription factors in the ear: molecular switches for development and di¡erentiation. Audiol. Neurootol. 2, 36^49. Schena, M., Shalon, D., Davis, R.W., Brown, P.O., 1995. Quantitative monitoring of gene expression patterns with a complementary DNA microarray. Science 270, 467^470. Vale, C., Sanes, D.H., 2000. A¡erent regulation of inhibitory synaptic transmission in the developing auditory midbrain. J. Neurosci. 20, 1912^1921. Waldmann, R., Bassilana, F., de Weille, J., Champigny, G., Heurteaux, C., Lazdunski, M., 1997a. Molecular cloning of a non-inactivating proton-gated Na+ channel speci¢c for sensory neurons. J. Biol. Chem. 272, 20975^20978. Waldmann, R., Champigny, G., Bassilana, F., Heurteaux, C., Lazdunski, M., 1997b. A proton-gated cation channel involved in acid-sensing. Nature 386, 173^177. Weintraub, H., Davis, R., Tapscott, S., Thayer, M., Krause, M., Benezra, R., Blackwell, T.K., Turner, D., Rupp, R., Hollenberg, S. et al., 1991. The myoD gene family: nodal point during speci¢cation of the muscle cell lineage. Science 251, 761^766. Wyborski, R.J., Bond, R.W., Gottlieb, D.I., 1990. Characterization of a cDNA coding for rat glutamic acid decarboxylase. Mol. Brain Res. 8, 193^198.
HEARES 4022 18-12-02
Hearing Research 175 (2003) 2^13 www.elsevier.com/locate/heares
Identi¢cation of gene expression pro¢les in rat ears with cDNA microarrays Jizhen Lin a
a;
, Masashi Ozeki a , Eric Javel a , Zhenfen Zhao b , Wei Pan c , Eileen Schlentz a , Samuel Levine a
Department of Otolaryngology, University of Minnesota Medical School, Division of Biostatistics, Minneapolis, MN 55455, USA b Department of Pediatrics, University of Minnesota Medical School, Division of Biostatistics, Minneapolis, MN 55455, USA c School of Public Health, University of Minnesota, Minneapolis, MN 55455, USA Received 18 March 2002; accepted 18 September 2002
Abstract The physiological processes of hearing implicate thousands of molecules acting in harmony; however, their identities are only partially understood. We used cDNA microarrays containing 1,176 genes to identify s 150 genes expressed in rat middle and inner ear tissue. Expressed genes covered several gene families and biological pathways, many of which have previously not been described. Transcription factor genes that were expressed included inhibitors of DNA binding protein (Id). These were localized to the spiral ganglion, organ of Corti and stria vascularis, and they are possibly involved in neurogenesis and angiogenesis. Transcriptional factors that were highly expressed included Gax (homeobox) and I-UB, which inhibit cellular proliferation. Their presence suggests that inhibitory programs for cell proliferation are enforced in the ear. Ion channel genes that were expressed included voltage-dependent L-type calcium channels (LTCC) and proton-gated cation channels (PGCC). Genes involved in neurotransmitter production and release included glutamic acid decarboxylase (GAD1). Genes involved in postsynaptic inhibition included neuropeptide Y5 receptors (NPY5) and GAD1. Due to the existence of receptors and/or enzymes involved in their biochemical synthesis, neurotransmitters associated with these might include serotonin, glutamide, acetylcholine, Q-aminobutyric acid (GABA), neurotensin, and dopamine. ; 2002 Elsevier Science B.V. All rights reserved. Key words: Cochlea; Gene pro¢le; Microarray; Rat
1. Introduction The inner ear is a complex sensory organ possessing
* Corresponding author. 2001 Sixth Street S.E., 216 Lions Research Bldg., Minneapolis, MN 55455, USA. Tel.: +1 (612) 6245059; Fax: +1 (612) 626-9871. E-mail address: [email protected] (J. Lin). Abbreviations: Id, inhibitor of DNA binding protein or inhibitor of di¡erentiation; LTCC, voltage-dependent L-type calcium channel; PGCC, proton-gated cation channel; GAD1, glutamic acid decarboxylase; NPY5, neuropeptide Y5 receptor; GABA, Q-aminobutyric acid; ABR, auditory brainstem response; DLU, digitalized light unit; RT-PCR, reverse transcription-polymerase chain reaction; MAKP1, mitogen-activated kinase protein 1; Gax, growth arrest-speci¢c protein; NF-UB, nuclear factor-kappa B; I-UB, inhibitor of nuclear factor-kappa B; bHLH, basic helix-loop-helix
receptor or hair cells, various types of supporting cells, and yet other cells that provide nutrition and maintain the ionic environment. Hair cells are specialized mechanoreceptors that convert mechanical vibration into electrical potentials. Supporting cells provide sca¡olds for hair cells and, in conjunction with cells of the stria vascularis, maintain the inner ear’s unique ionic environment. In contrast, the middle ear is a structure that passively ampli¢es and ¢lters sound pressure. It possesses a specialized mucosa that contains polarized epithelial and secretory cells. The functional roles played by cells of the middle and inner ear implicate a large ensemble of molecules. Only a few of the genes or proteins involved in auditory structure and function have been identi¢ed to date. Knowledge of the complete repertoire of genes expressed in the ear is important be-
0378-5955 / 02 / $ ^ see front matter ; 2002 Elsevier Science B.V. All rights reserved. PII: S 0 3 7 8 - 5 9 5 5 ( 0 2 ) 0 0 7 0 4 - 9
HEARES 4022 18-12-02
J. Lin et al. / Hearing Research 175 (2003) 2^13
3
Table 1 Genes of interest in rat ear by category, in decreasing order of expression intensity Class
GB no.
Gene/protein name
Transcription factors (8/24)
Z17223
Gax
4889
X63594
I-UB
4270
D10862 D13374
Id1 Nm23-M2
3687 3038
+
D10864 M91597
Id3 NDK-B
2046 1473
+
L26267
NF-UB
843
X91810
Stat3
754
Growth factors
M13969
758
Ion water pumps and channel proteins (8/145)
M59786
IGF-II, insulin-like growth factor II LTCC
1079
M28647 M91808 D84450 U09402
Naþ /Kþ ATPase K1 sodium channel L1 Naþ /Kþ ATPase L3 P2U purinoceptor 1, ATP receptor
1024 937 864 801
J02701 AF013598 L06096
Naþ /Kþ -transporting ATPase L1 PGCC, sensory neuron speci¢c IP3R
M69055
rIGFBP-6
1620
M86870 M17528 J04486
CABP2 GNAI2 IGFBP-2; Brl-2
1590 1127 987
L35921 U34959
G(i)/G(s)/G(o) Q-9 G(i)/G(s)/G(t) L2
M83676
rab12, ras-related GTPase
U77918 D10874
TBP1 ATP6C
4949 3353
M83681
rab16, ras-related GTPase
2357
U43175 D10021
ATPase, subunit F, vacuolar ATP5H
1020 1018
M19044
(ATP5B)
D13124 M61177 M17086 L24907 D30040 J02592
P2 gene, subunit c MAPK1 cAMP-dependent PK type I-K Caþ /calmodulin-dependent PK type I RAC-PK-K K GSTM2
754 11846
U03763
calcium-dependent PLA2
11701
M33648
hydroxymethylglutaryl-CoA synthase
Ion and other binding proteins (N/A)
GTPase ATPase (N/A)
Protein kinases (N/A)
Other enzyme proteins (N/A)
Level (DLU)
763 693 610
PCR
Functions and characteristics
+
inhibition of cellular growth and proliferation inhibitor of NF-UB, inhibition of cellular growth neurogenesis and angiogenesis oncogene c-myc-related transcription factor neurogenesis and angiogenesis c-myc-related transcription factor, cellular growth cellular growth in response to mitogenic signals signal transducers, critical for cell growth cellular growth and di¡erentiation
+
+ +
954 875
10946
973
HEARES 4022 18-12-02
852 1620 1048 876
6515
+
+
ion transport, Ca2þ channel, voltage-dependent Naþ /Kþ pump Naþ pump Naþ /Kþ pump ion transport, secretion and absorption of cations Naþ /Kþ pump Hþ pump, detection of pH value a channel releasing intracellular Ca2þ insulin-like growth factor binding protein calcium binding protein guanine nucleotide binding protein insulin-like growth factor binding protein G-proteins/GTP binding proteins, involved signal transduction together with rab12 and 16 (see below), to modulate cGMP-gated cation channels involved in signal transduction via G-proteins ATPase, proteasomal, liver vacuolar ATP synthase, neurotransmitter tra⁄c involved in signal transduction via G-proteins energy metabolism mitochondrial ATP synthase D subunit mitochondrial ATP synthase L subunit precursor ATP synthase, energy metabolism mitogen signaling molecule cAMP signaling system Ca2þ signaling system Rac-K serine/threonine kinase glutathione S-transferase Yb subunit enzyme for arachidonic acid metabolism mitochondrial energy production
4
J. Lin et al. / Hearing Research 175 (2003) 2^13
Table 1 (Continued). Class
Receptors (14/197)
Stress-related proteins (6/35)
GB no.
Gene/protein name
M34445 X02904 X82679 M21410
GAD1 GST7-7 phospholipid hydroperoxide glutathione peroxidase 5-hydroxytryptamine 2C receptor
1911
L10073
5-hydroxytryptamine receptor 5B
1900
X97121
neurotensin receptor 2 (NTR2)
1023
Z11932
vasopressin V2 receptor
987
U20907
5-HT receptor 4; 5-HT4L
824
M35077 M35162 U66274
dopamine receptor GABA receptor N precursor NPY5
812 803 792
+
AF016297
neuropilin (NP)
762
+
D14869
prostaglandin E2 receptor EP3
753
M31838
neurokinin A receptor, Tacr2
633
+
AF005099
neuronal pentraxin receptor, NPR
578
+
M63901 D12820
secretogranin V, SGV GUST27
s 536 528
+ +
S45392 J03752 Y00404 M86389
heat shock protein 84 (HSP84) GST12; MGST1 Cu-Zn superoxide dismutase heat shock 27-kDa protein (HSP27) heat shock 70-kDa protein (HSP70)
14330 2867 2243 2028
Z27118
Skeleton and motility (3/7)
others
Level (DLU) 3521 2178 1786
Functions and characteristics
+
GAD1 glutathione S-transferase P subunit neurotransmitter metabolism
+
1034
X62660 X62908 X67788 U58858
glutathione transferase, subunit 8 co¢lin ezrin; cytovillin; villin 2 (VIL2) plakoglobin
910 4090 987 976
U34959
transducin L-2 subunit (TDB2)
5718
U62326
MIF
4156
M84716
Fte-1; RPS3A
2536
X74806
von ebner’s gland protein 2, VEGP
2090
L31884
TIMP2
1434
M25889 M24105
myelin basic protein S synaptobrevin-2 (SYB2)
973 947
M96377
non-processed neurexin II-L major
929
HEARES 4022 18-12-02
PCR
+
+
serotonin receptor, water and ion metabolism serotonin receptor, water and ion metabolism G-protein coupled receptor, signaling molecule blood vessel (£ow) control and regulation serotonin receptor, metabolism of water and ions neurotransmitter conduction neurotransmitter conduction neuropeptide responding to PKC and cAMP axonal growth, fasciculation, branching, and etc. cellular growth and ion/water metabolism linked to Ca2þ and cAMP responses a member of the neural uptake system induction of dopamine release a protein for purinergic signaling system chaperones and/or enzymes involved in protection of the cells under stress conditions such as heat, trauma, high intensity noise exposure, infection, and ototoxicity drug administration; some of these proteins are also important components for neurotransmitter metabolism co¢lin is a microvillar cytoplasmic peripheral protein; cytovillin is a cytoskeletal linker protein linked to lamins, keratins, and neurophilaments; plakoglobin and plectin are involved in cell^cell adhesion, and adherent junctions G-proteins, involved in signal transduction macrophage migration inhibitory factor putative v-fos transformation e¡ector protein serious gland product in response to stress, etc. tissue inhibitor of metalloproteinase structural protein of myelin sheath vesicle associated proteins involved in vesicle fusion and transmission (see syntaxins below) neuronal cell surface proteins, axonal insulation
J. Lin et al. / Hearing Research 175 (2003) 2^13
5
Table 1 (Continued). Class
GB no.
Gene/protein name
Level (DLU)
PCR
Functions and characteristics
X76489
CD9, platelet surface protein
894
+
protocadherin 4 (PC4)
802
+
M20373
neuropeptide Y precursor (NPY)
783
+
K03242
myelin P0 protein precursor, MPZ
762
surface marker for neural progenitors decisive adhesion molecule in synaptogenesis neuropeptide responding to cAMP and PKC extracellular matrix, structural protein of myelin
AB004276
Genes listed in the table had a signal v 750 DLU, that is, at least 50% higher than background (500 DLU). Certain genes expressed less highly were included because their existence in the ear had been con¢rmed by RT-PCR. This is based on a criterion adapted by most gene expression pro¢ling studies. Genes in bold face represent ones previously not described in ear tissue and their presence and expression con¢rmed by RTPCR (except for Stat3). N/A, not applicable; GB no., GenBank accession number. Full names of the genes in the table and detailed information regarding their functions may be obtained at the website (www.ncbi.nlm.nih.gov/) using their accession numbers.
cause it improves understanding of the molecular mechanisms involved in hearing and permits the design of innovative strategies for reversing or preventing cochlear hearing loss. Defects in any of the genes involved in auditory structure and function may upset the balance of several intricate physiological processes known to exist in hearing, among which are maintenance of a 150-mV potential di¡erence across hair cell membranes and generation of rapid motile responses at rates that can approach 100 kHz. Accumulated evidence to date indicates that more than 40 genomic DNA loci are linked to non-syndromic hearing loss in humans, and these are distributed throughout the genome. Only a handful of these genes have been cloned and characterized (Avraham, 1998). The pro¢le of genes expressed in the ear is not well understood. It is generally accepted that about 30,000^ 40,000 genes exist in genomic DNA, and approximately 5^10% of these are likely to be expressed in the ear (http://www.ncbi.nlm.nih.gov/disease). Some genes are expressed in the developing ear but are inactivated postnatally, and others are expressed only in mature ears. We were curious about the genes expressed by mature ear tissue because they are more likely to encode ‘working’ molecules that mediate biophysical processes of hearing and maintain unique structure of the ear. Genes that are highly active or abundantly expressed by ear tissue may be detected by chip-based technologies. cDNA microarrays or ‘gene chips’ (Schena et al., 1995) are nylon membranes or slides dotted with potentially thousands of gene-speci¢c cDNA probes or oligos. Hybridization of these gene-speci¢c cDNA fragments with cDNA or cRNA probes prepared from a target tissue or organ permits quantitative analysis of all oligos contained on the chip. It is, therefore, a very useful and powerful tool to detect major gene families or biochemical pathways in a given target tissue. cDNA microarrays have a wide range of applications (Nelson,
1996), among which are investigating normal biological and disease processes (Efron et al., 2000), pro¢ling differential gene expression (Lomax et al., 2000), and discovering potential therapeutic drug targets. Otologic investigators have recently applied this technique for addressing the issue of gene expression during development in mice (Chen and Cpreu, 2001), gene expression pro¢les in subfractions of the auditory system, namely, the rat cochlea, cochlear nucleus and inferior colliculus (Cho et al., 2002), and the pathology of middle ear infectious disease in rats (Lin et al., 2002; Pan et al., 2002). In this report, we used cDNA microarrays to identify major gene families expressed by rat ear. We found that several genes were expressed which have previously not been described in the ear.
2. Materials and methods Healthy Sprague^Dawley rats weighing 230^250 g were used in this study. The rats had no previous exposure to noise or speci¢c pathogens. Auditory brainstem response (ABR) recordings in response to calibrated acoustic signals were obtained in seven randomly selected rats to verify that auditory sensitivity to rarefaction clicks and tone bursts over a 1^16 kHz range laid within normal limits. Following this, animals were anesthetized and decapitated, both bullae were dissected out, and all exterior muscle and connective tissues were completely removed. Twelve bullae were immediately submerged in liquid nitrogen for microarray experiments. Another 12 bullae were used for dissection of the cochlear and middle ear tissues using a stereomicroscope. The vestibular system was not targeted. An additional six bullae were ¢xed in 10% formalin, embedded in para⁄n, and processed for in situ hybridization as previously described (Lin et al., 1999a,b,c).
HEARES 4022 18-12-02
6
J. Lin et al. / Hearing Research 175 (2003) 2^13
2.1. cDNA microarray protocols Procedures for acquiring and analyzing data using cDNA microarray (Atlas1 rat 1.2 array, Clontech, Palo Alto, CA, USA) were the same as described previously (Lin et al., 2002). This cDNA microarray was selected for its extensive coverage of genes involved in various biochemical reactions and pathways. Brie£y, mRNA was isolated from combined middle and inner ear tissue using stardard oligo dT methods. cDNA probes were synthesized from the mRNA by speci¢c CDS primer mix (provided by the cDNA array kit) and radiolabeled with 32 P. Total radioactivity of cDNA probes was 3.8U106 cpm. The Atlas cDNA array membrane was hybridized with the above cDNA probes, washed stringently, and exposed to a phosphorimager. Expression of genes and their intensities were quanti¢ed using OptiQuant software (version 03.00, vPackard) and presented as digitalized light unit (DLU). A blank area of the same chip was used for determining background radioactivity, and the intensity of each location on the grid that exhibited a level higher than background was measured. DLU values for each expressed gene were plotted as a three-dimensional (3-D) graph using the SigmaPlot software 5.0 according to the original coordinates of the chip (visit www. clontech.com for details). 2.2. Scatter plot analysis The entire experiment was duplicated to assess repeatability. To do this, the gene expression pattern from one of the mRNA pools (representing 12 bullae from six rats) was plotted against the expression pattern from the other mRNA pool. Digitized gene expression values were exported to an Excel spreadsheet, and the DLU value of each grid location was normalized by dividing by background intensity. The result was then exported for the scatter plot analysis. The scatter plot consisted of the gene index expressed by the ¢rst cDNA microarray experiment, plotted as a function of the gene index expressed by the second experiment. This provided a two-dimensional representation of the gene expression pattern to evaluate possible variations in gene expression between individual cDNA microarrays. 2.3. Con¢rmation of the microarray results by reverse transcription-polymerase chain reaction (RT-PCR) and/or in situ hybridization To avoid possible false positive results from cDNA microarrays, RT-PCR and/or in situ hybridization were selectively performed for candidate genes (see Tables 1 and 2) using methods described elsewhere (Lin et al., 2001, 1999). To study whether candidate genes were
only expressed in the ear, various tissues (eye, brain, nasopharynx, trachea, lung, liver, heart, kidney, spleen, intestine, stomach, and muscle) were used as controls. Total RNA was isolated from these tissues. To study the cellular location of candidate mRNA transcripts, the cochlear tissue of both one-day-old and adult rats was used for in situ hybridization. For RT-PCR, speci¢c gene sequence information was obtained from the website of the National Center of Biologic Information (http://www.ncbi.nlm.nih.gov), speci¢c primer pairs for genes of interest were designed (see Table 2), and their speci¢city was con¢rmed by BLAST search using the above website. RT-PCR was performed in the same manner as previously described (Lin et al., 1999a,b,c). Annealing temperature for individual primer pairs was determined according to the temperature (Tm ) of sense and antisense primers. PCR products were routinely sequenced for approximately up to 500 bp using an autosequencer. For in situ hybridization, the PCR products of candidate genes were cloned into the vector (pT-Adv) using a PCR cloning kit (Clontech) after sequence analysis. The cDNA probe in the vector was re-con¢rmed by DNA sequencing and then transcribed into a 35 S-radiolabeled antisense riboprobe using T7 promoter sequence with a sense riboprobe as control. This was then hybridized with ear tissue sections using procedures described earlier (Lin et al., 2001, 1999). Sections were autoradiographed and counterstained with hematoxylin.
3. Results Representative ABRs obtained from one rat in response to 100-Ws rarefaction clicks are shown in the left column of Fig. 1. The right-hand panel indicates auditory sensitivity across a 1^16 kHz frequency range for seven of the rats used in this study. Solid and dashed lines indicate mean ABR thresholds T 1 standard deviation (S.D.) in a di¡erent set of normal-hearing rats. These data indicate that the animals used here had essentially normal hearing, suggesting a fully functional complement of receptor and spiral ganglion cells. As shown in Fig. 2A, hybridization of the cDNA probes from pooled mRNA of 12 ears exhibited a unique, repeatable expression pattern. Hybridizing signals exceeded background levels of approximately 500 DLU for s 200 grid locations. The locations covered many gene families such as transcription factor, growth factor, ion channel, receptor, etc. Genes whose expression either attained a level of 50% greater than background or whose expression levels fell between 500 and 750 DLU but were con¢rmed by RT-PCR, are listed in Table 1. Seventy-six genes met these criteria. An addi-
HEARES 4022 18-12-02
J. Lin et al. / Hearing Research 175 (2003) 2^13
7
Table 2 Primers pairs used to examine gene expression Gene
Primer
Oligonucleotide sequence
Annealing Tm
Size (bp)
Gax
sense antisense sense antisense sense antisense sense antisense sense antisense sense antisense sense antisense sense antisense sense antisense sense antisense sense antisense sense antisense sense antisense sense antisense sense antisense sense antisense sense antisense sense antisense sense antisense sense antisense sense antisense sense antisense sense antisense sense antisense
5P-AGCGACAGTTCAGATTCCC-3P 5P-TCCCTTTTTCACATTCACCAG-3P 5P-GCCATCTCGCGCTGCGCTGGG-3P 5P-GGTGGAGAGGGCGAGACTCCG-3P 5P-CCCCTGCGGCGGCTCCCTCC-3P 5P-CCAGGCCGGAGACACCTGGGG-3P 5P-GCCATTGCGCGAGGCCGGGG-3P 5P-CCTTCGCTCTCGGGCTCCAG-3P 5P-CGGCCGCACACCTGGCCAAC-3P 5P-GGCGCCGTCGGGCTGCGGCGG-3P 5P-CCCGAGCACATCCCTACTCCAGGGG-3P 5P-CCCTCCCAGAGGGTTGGCCCCGTCA-3P 5P-GGCCTTGCCCGGATTGCTGACCC-3P 5P-GGTGCCCCTGGCTGGAAGCGGGC-3P 5P-CTGGTGCCCGCTTCCGGCGC-3P 5P-GGGCACAGCCGCCATGCCTT-3P 5P-TGCTCCCCTTTCACCCTGAC-3P 5P-GGGGCGGGTAAGACACAGGC-3P 5P-GGGCCAGCCCCCCGCACCACCTGGC-3P 5P-GGAGGGCTGGGTCTGGGTCTGGGGG-3P 5P-CGTGCTCAGCTACTACAGGTACCA-3P 5P-GGGAGGCGTTGAGATTGGA-3P 5P-ATATTAAAACCTGGCGGCCA-3P 5P-GGCCGCTCTGTCTTCATCAG-3P 5P-GGTACCATCAGCTTCTGCCC-3P 5P-GATCTGGCTGGCATCCTTGT-3P 5P-CCACCTTCTCTGATGGGAGG-3P 5P-GTTTTTGCCATGCCGGTAGA-3P 5P-GAAAGAGAGGAGCCGATCCA-3P 5P-CAGTTCCGTCTCGAATTGCC-3P 5P-GCCTGGGAATCGTGTCAAAG-3P 5P-CCCCCAGGTCTAAGTTCTGAATC-3P 5P-CCACCAGGCCATGAATCTTG-3P 5P-ACTGGAATTCTCGGCTGAACTC-3P 5P-GTGCTTGGGAACTTGCTCATC-3P 5P-CGATCACAGGCCATTAAAGTGA-3P 5P-AGCAGCTCCGGAACCAGAT-3P 5P-CAGAGGGATGACGTGGACCTT-3P 5P-TCTGTGACTCCCATGAAAATCAA-3P 5P-CTGCCCACGAGTTTTGCAAT-3P 5P-TCACAGTGCATGCTGGGATT-3P 5P-GGGCAGATGTCCGAGATAAACT-3P 5P-ATGCTGCAAGCCATGATCTTG-3P 5P-TCTTTCTTGCCCGACTTGTTG-3P 5P-GCTCTATCCCTGCTCGTGTGT-3P 5P-CACCACATGGAAGGGTCTTCA-3P 5P-CTCGTTCCTTTTTCCTCCTT-3P 5P-CTGTGTGCGCAAAGACTAGC-3P
51‡C
301
66.7‡C
732
69‡C
668
66‡C
480
71.5‡C
473
70‡C
561
70.4‡C
572
68.3‡C
590
62‡C
720
74.7‡C
543
60‡C
296
59‡C
151
58.5‡C
151
59‡C
251
58.5‡C
251
59‡C
251
59‡C
251
58.5‡C
251
59‡C
251
59‡C
251
58.5‡C
251
59.5‡C
251
58.5‡C
251
55.5‡C
296
Id1 Id2 Id3 Id4 LTCC MAKP1 GAD1 NPY5 PGCC IP3Ra RAC-PKa NTR2a Nrp2a Tacr2a NPRa SGVa GUST 27a TDB2a VEGPa CD9a PC4a NPYa S29 a
PCR result is marked in Table 1 (both cochlear tissue and hair cell progenitors are positive) and the rest of PCR results is presented in Fig. 3.
tional 47 genes exhibited expression levels of 500^749 DLU, but their presence was not con¢rmed by RTPCR. These are not listed in Table 1. Some signals were especially strong ( s 10 000 DLU), some were very weak ( 6 500 DLU), and yet others were expressed at intermediate levels. Fig. 2A shows the intensity in DLU of gene expression for 12 ears in six normal rats. The scatter plot in Fig. 2B shows that the expression pattern (in terms of the distribution of dots) was
symmetric between two replications using cDNA microarrays. Thus, gene expression was highly repeatable across cDNA microarray experiments in normal rats. Of 24 transcription factors studied, eight were expressed in the middle and inner ears of mature rats (see Table 1). Gax (growth arrest-speci¢c protein) and I-UB (inhibitor of nuclear factor-kappa B) were highly expressed. These are involved in cell cycle arrest and cellular di¡erentiation. Expression of some of these
HEARES 4022 18-12-02
8
J. Lin et al. / Hearing Research 175 (2003) 2^13
Fig. 1. Auditory sensitivity of rats used in this study. Left: Representative ABRs elicited by 100-Ws rarefaction clicks at intensities of 10^90 dB SPL. Right: Symbols connected by solid lines indicate ABR thresholds at octave intervals from 1 to 16 kHz for seven of the rats in this study. Heavier solid and dashed lines represent mean thresholds T 1 S.D. for a di¡erent series of rats with normal hearing.
genes in cochlear and middle ear tissue was veri¢ed by RT-PCR. Speci¢c genes chosen for RT-PCR analysis are indicated in Table 1. Some of the PCR products are presented in Fig. 3A. These genes included inhibitor of DNA binding protein (Id)1^4 (Nagata and Todokoro, 1994), voltage-dependent L-type calcium channel (LTCC) alpha 1C (Koch et al., 1990), mitogen-activated kinase protein 1 (MAKP1) (Maisonpierre et al., 1991), glutamic acid decarboxylase (GAD1) (Wyborski et al., 1990), neuropeptide Y5 receptor (NPY5) (Hu et al., 1996), and proton-gated cation channel (PGCC, sensory neuron speci¢c) (Waldmann et al., 1997a,b). These genes were obviously expressed in inner ear tissue but were either unexpressed or weakly expressed in the middle ear. For example, Id1 gene expression was relatively high in the inner ear, compared to that in the middle ear and in all other tissues except for lung and kidney (Fig. 3B, lanes 7 and 10, Id1). Since the expression pattern of the Id1 and Id3 genes in the inner ear is poorly understood, we further performed in situ hybridization to reveal their cellular locations. The results, shown in Fig. 4, indicate that Id1 and Id3 genes are expressed in the spiral ganglion, the organ of Corti and the stria vascularis of postnatal rats but not discernible in the counterparts of adult rats by in situ hybridization. It suggests that Id1 and Id3 genes are involved in the development of cochlea and their expression in adult cochlear tissue is downregulated. Other genes expressed in the ear included members of
several gene families (see Table 1). For example, eight of 145 genes encoding ion channels were expressed, as were genes encoding ion pumps (Na+/K+ ATPase), calcium binding protein such as calmodulin, and GTPases such as rab12 and 16. Fourteen of 197 genes encoding receptors were expressed, as were genes encoding enzymes such as glutathione S-transferase, growth factors such as insulin-like growth factor II, and mitogenic kinases such as MAPK1. Three of seven genes encoding the hair cell cytoskeleton were expressed (e.g. co¢lin), six of 35 stress-related genes were expressed (e.g. HSP90 and HSP27), and housekeeping proteins such as ribosomal protein S29 were also expressed. Many other genes known to be expressed by ear tissue are not included in the particular gene chip used in this study. Among these are Brn-3, Pax2, and Math-1. Due to the limited sensitivity of cDNA microarrays, some of the genes included in this gene chip and which were expected to be expressed in ear tissue, ¢broblast growth factor for example, were in fact undetectable in this study.
4. Discussion We have demonstrated in this study that cDNA microarrays are useful in discovering gene expression patterns in the ear. We identi¢ed several genes that previ-
HEARES 4022 18-12-02
J. Lin et al. / Hearing Research 175 (2003) 2^13
Fig. 2. Gene expression pattern pro¢les of the normal ears from six SPF rats. Presentation of the gene expression pro¢les in rat bullae involves matching each dot on the array, quantifying the intensity of each dot, and exporting the data to the SigmaPlot 5.0 spreadsheet to create a three-dimensional graph corresponding to the original coordinate of each gene on the rat 1.2 array. X and Y axes in this graph represent the numeric order and alphabetical order on the original array, respectively. The Z axis indicates gene expression magnitude in a color-increasing manner (from low to high, this is black, pink, green, yellow, purple, red, blue, gray, dark red, and black again). This distinct gene expression pattern (A) is speci¢c to SPF rat ears. It is therefore termed the gene expression pattern signature. The high degree of reproducibility across animals is indicated by data in B, which show the index of genes expressed by the ¢rst set, plotted as a function of gene index expressed by a second set. Gene intensity has been arbitrarily cut o¡ at 15 000 DLU to facilitate identi¢cation of genes that are less highly expressed.
ously have not been known to be expressed by ear tissue. Identi¢cation of major gene families in wholeear tissue is a ¢rst but important step toward understanding basic biophysical and molecular mechanism involved in hearing. The gene expression pattern in the ear tissue is
9
unique (see Fig. 2A, gene expression signature) and is probably not matched by any other tissue. Characteristic features of transcription factors expressed in mature ear tissue are strongly with cell cycle arrest and cellular di¡erentiation (e.g. Gax and I-UB) and weakly with cellular growth and proliferation (e.g. nuclear factor-kappa B (NF-UB), Id genes, and oncogene-related transcription factors). It appears from this that mature ear tissue is biologically blocked for cellular proliferation. NF-UB is actively involved in cellular growth and proliferation, and the action of many mitogens and growth factors is linked to NF-UB or through the NF-UB pathway, but its expression level is much lower than its inhibitor I-UB. An observation favoring the concept that genes involved in cell cycle arrest and cellular di¡erentiation are expressed in mature ear tissue is that homeobox transcriptional factor Gax is highly expressed. The purpose of this gene is to keep well-di¡erentiated or highly polarized cells quiescent and stable, which is very important for hearing. On the other hand, Gax expression obviously does not facilitate hair cell regeneration. We believe that Gax and I-UB need to be downregulated if hair cell regeneration is to occur. Other homeobox transcription factors that were expressed in the ear included Id1 and Id3. Homeobox genes are known to be involved in the development of ears, and previously documented homeobox genes include Nkx 5.1, SoHo-1, Dlx-2 and 3, Msx-C and D, Hoxa-1 and 2, and Otx-1 and 2 (Rivolta, 1997). To our knowledge, Gax has not been previously described as being expressed by ear tissue. Noteworthy genes that were expressed include Id genes, LTCC, PGCC (sensory neuron speci¢c), neuropeptide Y5 receptor (NPY5), and several others (bold face in Table 1). The Id genes are known to function as negative regulators of the basic helix-loop-helix transcription factors (bHLH, a basic region plus two amphipathic K helices separated by an intervening loop). They play a critical role in the developmental processes for various cells (Weintraub et al., 1991). Id genes inhibit the functions of these transcription factors in a dominant-negative manner by blocking their DNA binding activity through the HLH domain (Benezra et al., 1990). Unlike other bHLH members, Id proteins contain only the HLH domain but miss the basic region adjacent to the HLH domain that is essential for speci¢c DNA binding. Therefore, Id proteins bind to bHLH transcription factors and inhibit their binding to the DNA ‘E box’ consensus sequence (Benezra et al., 1990; Littlewood and Evan, 1994; Murre et al., 1989). bHLH transcription factors activate the expression of genes involved in cell di¡erentiation, whereas Id proteins promote cell growth and proliferation (Norton et al., 1998). This suggests that interaction of Id proteins and bHLH transcription factors may play an im-
HEARES 4022 18-12-02
10
J. Lin et al. / Hearing Research 175 (2003) 2^13
portant role in controlling tissue development and differentiation. Recent studies have demonstrated that Id genes are required for neurogenesis and angiogenesis of tumor xenografts (Lyden et al., 1999), suggesting that they play a role in the development and maintenance of the blood vessel and nerve tissues. This role perhaps accounts for the expression of Id1 and Id3 in the organ of Corti, cochlear ganglion, and stria vascularis (Fig. 3). The LTCC gene is another entity prominently expressed in ear tissue. Its expression in cochlear tissue of mice and chickens has been reported previously (Green et al., 1996; Kollmar et al., 1997a,b). LTCC encodes voltage-gated L-type calcium channel protein and is actively involved in calcium movement across cell membranes. It is known that calcium movement is associated with the release of neurotransmitter at pre-
Id1 Id2 Id3 Id4
LTCC MAKP1 GAD1
NPY5 PGCC
A
synaptic terminals. The LTCC gene is clearly important in hearing, since its knockout in mice leads to degeneration of outer and inner hair cells and consequent deafness (Platzer et al., 2000). GAD1 is the enzyme responsible for the production of Q-aminobutyric acid (GABA) (Bain et al., 1993), an important inhibitory neurotransmitter (Arnold et al., 1998; Mu et al., 1994; Vale and Sanes, 2000). It is generally accepted that GABA is produced in e¡erent nerve endings, where its role is to modulate activity elicited in the a¡erent pathway. PGCC gene expression has not previously been described for inner ear tissue. It encodes a sensory ion channel protein that senses acid (Waldmann et al., 1997a,b) and therefore may be related to homeostasis or pain sensation. In vitro studies have shown that low extracellular pH can evoke inward currents in both the
1
5
8
IE
ME S29 9
10
Control
7
Intestine Stomach Muscle
6
Kidney Spleen
4
Liver Heart
3
Cochlea Middle ear Eye Brain Nasopharynx Trachea Lung
B
2
732
Id1 Id3
480
S29
296 1
2
3
4
5
6
7
8
9
10 11 12 13 14 15
Fig. 3. RT-PCR con¢rmation of partial genes identi¢ed by cDNA microarrays. mRNAs that were expressed in cochlear tissue included Id1 (lane 1, IE), Id3 (lane 3, IE), LTCC (lane 5), MAPK1 (lane 6), GAD1 (lane 7), NPY5 (lane 8), and PGCC (lane 9). Lane 10 represents a negative PCR control and omission of tissue from the middle ear. The bottom panel represents housekeeping gene ribosomal protein S29 (S29). Id2 and Id4 genes that were not detectable by cDNA microarrays were also not detected by RT-PCR. Lower panel represents the expression patterns of Id1 and Id3 genes in various tissues of rats with high levels in the cochlea (lane 1), brain (lane 4), and spleen (lane 11) tissues. Some more genes identi¢ed by microarrays and con¢rmed by RT-PCR are not presented in this ¢gure but indicated in Table 1.
HEARES 4022 18-12-02
J. Lin et al. / Hearing Research 175 (2003) 2^13
11
Fig. 4. Expression of Id genes in the cochlear tissue of rats (postnatal day 1) by in situ hybridization (panels a^f) but not detectable in the adult cochlear tissue (panels h and j). Sense riboprobes of Id1 (panel a) and Id3 (panel d) showed no hybridizing signals. Antisense riboprobes of Id1 (panels b and e) and Id3 (panels c and f). Arrows point to stria vascularis, organ of Corti and cochlear ganglion, indicating hybridizing signals for Id1 mRNA transcripts (panel e). Panels d, e, and f are ampli¢ed versions of panels a, b, and c. Note that Id1 gene has the same expression pattern as the Id3 gene, and both are expressed in stria vascularis, organ of Corti, and spiral ganglion (panels b and c, or e and f). Panels g and i, stained with hematoxylin and eosin (H and E), are serial sections of panels h and j hybridized with antisense Id1 and Id3 riboprobes, respectively. ISH, in situ hybridization; CD, cochlear duct; CM, cochlear modulus; SV, stria vascularis; SG, spiral ganglion; OHC, outer hair cells; bar = 10 W at same scale for panels a, b and c, for panels d, e, and f, for panels g and h, and for panels i and j.
HEARES 4022 18-12-02
12
J. Lin et al. / Hearing Research 175 (2003) 2^13
central nervous system and peripheral sensory neurons (Kovalchuk et al., 1990; Krishtal and Pidoplichko, 1980). The NPY5 gene is primarily expressed in the central nervous system and has been implicated in food intake (Gerald et al., 1996) and reduction of synaptic excitation (Ho et al., 2000). Although expression of this gene has not been previously described for ear tissue, the functional role it plays in audition is unclear. Since NPY is tied to calcium and cAMP signaling pathways (Higuchi et al., 1988), it is possible that NPY and NPY5 are e¡ector molecules of the calcium and cAMP signaling system in ear tissue. It has been noted in this study that the Ca2þ /calmodulin- and/or cAMPdependent pathways as well as G-protein coupled signal transduction system, as expected, are notable in the ear. As shown in Table 1, many genes in addition to the ones described above were also detected in ear tissue. Insulin-like growth factor II gene was expressed, which is essential for cellular growth and survival. MAPK1 was highly expressed, but we found it to be expressed primarily in the middle ear, where epithelium renewal is active, and not in cochlear tissue, where cellular renewal is inhibited. The genes listed in Table 1 represent only a portion of the gene repertoire expressed in the ear that we deem to be of interest. These genes are also expressed in developing cochlear tissue from embryonic days 12^18 although their expression levels vary and some of them (see Table 1, marked in bold face) are expressed in progenitor hair cell lines (data not shown), suggestive of their expression in hair cells. Many other genes were highly expressed in ear tissue but are not listed in Table 1 because of their housekeeping roles in cell metabolism. Genes in this category include those encoding cytochrome c oxidases, ribosomal proteins, actins, glyceraldehyde 3-phosphate dehydrogenase, and polyubiquitin. Some of the non-abundantly expressed genes or families have previously been described as being expressed by hair cells. Among these are Math-1 (bHLH transcription factor). Yet other genes known to be expressed by hair cells are not included in the cDNA microarrays we used. Among these are Nkx 5.1, SoHo-1, Dlx-2 and 3, Msx-C and D, Hoxa-1 and 2, and Otx-1 and 2. Further studies are warranted to amplify and study gene expression in precisely dissected subpopulations such as hair cells and supporting cells using techniques such as laser capture microdissection. An advantage of the relatively small gene chip used in this study is the low false positive rate it generates. We used RT-PCR to con¢rm the presence of some of the genes not previously described as being expressed in ear tissue. RT-PCR results are presented in Fig. 3 or indicated in Table 1. Most of the genes listed in Table 1
are expressed in developing ears of rats in our most recent studies, but their expression levels vary dramatically (data not shown). A disadvantage of this gene chip is the relatively small number of cDNAs on it. The limited number permits analysis of only a portion of the full gene repertoire expressed in rat ears. A full understanding of the complete repertoire of genes expressed in the ear and their expression dynamics in response to acoustic stimulation would require a chip possessing a substantially larger number of genes. Application of cDNA microarrays at subgenome or genome levels would permit analysis of the expression patterns of most, if not all, of the genes representing the major biochemical pathways and biophysical processes in the ear. Future studies employing expanded gene chips will be extremely helpful for (1) understanding the intricate and precise biochemical/biophysical processes that underlie hearing at a molecular level and (2) permitting rational design of gene therapy strategies for preventing, ameliorating or reversing hearing loss.
Acknowledgements This work was supported by grants from the National Organization for Hearing Research, the Lions 5M Hearing Foundation, and the Viking Children’s Fund, and in part by grants R01-DC03433 (microarray administrative supplement) and P30 DC04660 from NIDCD. References Arnold, T., Oestreicher, E., Ehrenberger, K., Felix, D., 1998. GABA(A) receptor modulates the activity of inner hair cell a¡erents in guinea pig cochlea. Hear. Res. 125, 147^153. Avraham, K.B., 1998. Hear come more genes! Nat. Med. 4, 1238^ 1239. Bain, G., Ramkumar, T.P., Cheng, J.M., Gottlieb, D.I., 1993. Expression of the genes coding for glutamic acid decarboxylase in pluripotent cell lines. Brain Res. Mol. Brain Res. 17, 23^30. Benezra, R., Davis, R.L., Lockshon, D., Turner, D.L., Weintraub, H., 1990. The protein Id: a negative regulator of helix-loop-helix DNA binding proteins. Cell 61, 49^59. Chen, Z.Y., Cpreu, D.P., 2001. An inner ear gene expression database. JARO 3, 140^148. Cho, Y., Gong, E.L., Stover, T., Lomax, M.I., Altschuler, R.A., 2002. Gene expression pro¢les of the rat cochlea, cochlear nucleus, and inferior colliculus. JARO 3, 54^67. Efron, B., Tibshirani, R., Goss, V., Chu, G., 2000. Microarrays and their use in a comparative experiment. In: Technical report, Statistics Department, Stanford University. Gerald, C., Walker, M.W., Criscione, L., Gustafson, E.L., Batzl-Hartmann, C., Smith, K.E., Vaysse, P., Durkin, M.M., Laz, T.M., Linemeyer, D.L., Scha¡hauser, A.O., Whitebread, S., Hofbauer, K.G., Taber, R.I., Branchek, T.A., Weinshank, R.L., 1996. A receptor subtype involved in neuropeptide-Y-induced food intake. Nature 382, 168^171.
HEARES 4022 18-12-02
J. Lin et al. / Hearing Research 175 (2003) 2^13 Green, G.E., Khan, K.M., Beisel, D.W., Drescher, M.J., Hat¢eld, J.S., Drescher, D.G., 1996. Calcium channel subunits in the mouse cochlea. J. Neurochem. 67, 37^45. Higuchi, H., Yang, H.Y., Sabol, S.L., 1988. Rat neuropeptide Y precursor gene expression. mRNA structure, tissue distribution, and regulation by glucocorticoids, cyclic AMP, and phorbol ester. J. Biol. Chem. 263, 6288^6295. Ho, M.W., Beck-Sickinger, A.G., Colmers, W.F., 2000. Neuropeptide Y(5) receptors reduce synaptic excitation in proximal subiculum, but not epileptiform activity in rat hippocampal slices. J. Neurophysiol. 83, 723^734. Hu, Y., Bloomquist, B.T., Corn¢eld, L.J., DeCarr, L.B., Flores-Riveros, J.R., Friedman, L., Jiang, P., Lewis-Higgins, L., Sadlowski, Y., Schaefer, J., Velazquez, N., McCaleb, M.L., 1996. Identi¢cation of a novel hypothalamic neuropeptide Y receptor associated with feeding behavior. J. Biol. Chem. 271, 26315^26319. Koch, W.J., Ellinor, P.T., Schwartz, A., 1990. cDNA cloning of a dihydropyridine-sensitive calcium channel from rat aorta. Evidence for the existence of alternatively spliced forms. J. Biol. Chem. 265, 17786^17791. Kollmar, R., Fak, J., Montgomery, L.G., Hudspeth, A.J., 1997a. Hair cell-speci¢c splicing of mRNA for the alpha1D subunit of voltagegated Ca2+ channels in the chicken’s cochlea. Proc. Natl. Acad. Sci. USA 94, 14889^14893. Kollmar, R., Montgomery, L.G., Fak, J., Henry, L.J., Hudspeth, A.J., 1997b. Predominance of the alpha1D subunit in L-type voltage-gated Ca2+ channels of hair cells in the chicken’s cochlea. Proc. Natl. Acad. Sci. USA 94, 14883^14888. Kovalchuk, Y.N., Krishtal, O.A., Nowycky, M.C., 1990. The protonactivated inward current of rat sensory neurons includes a calcium component. Neurosci. Lett. 115, 237^242. Krishtal, O.A., Pidoplichko, V.I., 1980. A receptor for protons in the nerve cell membrane. Neuroscience 5, 2325^2327. Lin, J., Ho, S.B., Schekels, L.L., Paparella, M.M., Kim, Y., 1999a. Mucin expression in the rat middle ear: an improved method for RNA harvest. Ann. Otol. Rhinol. Laryngol. 108, 762^768. Lin, J., Kawano, H., Paparella, M.M., Ho, S.B., 1999b. Improved RNA analysis for immediate autopsy of temporal bone soft tissues. Acta Otorhinolaryngol. 119, 787^795. Lin, J., Tsprun, V., Kawano, H., Paparella, M.M., Zhang, Z.-Q., Anway, R., Ho, S.B., 2001. Characterization of mucins in human middle ear and eustachian tube. Am. J. Physiol. Lung Cell Mol. Physiol. 280, L1157^1167. Lin, J., Tsuboi, Y., Pan, W., Giebink, S.G., Adams, G.L., Kim, Y., 2002. Analysis by cDNA microarrays of altered gene expression in middle ears of rats following pneumococcal otitis media. Int. J. Ped. Otolarynogol. 65, 203^211. Lin, J., Vambutas, A., Haruta, A., Paparella, M.M., Giebink, G.S., Kim, Y., 1999c. Pneumococcus activation of the 5-lipoxygenase pathway and production of glycoproteins in the middle ear of rats. J. Infect. Dis. 179, 1145^1151. Littlewood, T.D., Evan, G.I., 1994. Transcription factors 2: helixloop-helix. Protein Pro¢le 1, 635^709. Lomax, M.I., Huang, L., Cho, Y., Gong, T.L., Altschuler, R.A.,
13
2000. Di¡erential display and gene arrays to examine auditory plasticity. Hear. Res. 147, 293^302. Lyden, D., Young, A.Z., Zagzag, D., Yan, W., Gerald, W., O’Reilly, R., Bader, B.L., Hynes, R.O., Zhuang, Y., Manova, K., Benezra, R., 1999. Id1 and Id3 are required for neurogenesis, angiogenesis and vascularization of tumour xenografts. Nature 401, 670^677. Maisonpierre, P.C., Le Beau, M.M., Espinosa, R., Ip, N.Y., Belluscio, L., de la Monte, S.M., Squinto, S., Furth, M.E., Yancopoulos, G.D., 1991. Human and rat brain-derived neurotrophic factor and neurotrophin-3: gene structures, distributions, and chromosomal localizations. Genomics 10, 558^568. Mu, M.Y., Pao, X., Jie, T., 1994. SEM observation and GABA immunoreactivity of human embryonic organ of Corti tissue culture. Acta Otolaryngol. 114, 516^519. Murre, C., McCaw, P.S., Baltimore, D., 1989. A new DNA binding and dimerization motif in immunoglobulin enhancer binding, daughterless, MyoD, and myc proteins. Cell 56, 777^783. Nagata, Y., Todokoro, K., 1994. Activation of helix-loop-helix proteins Id1, Id2 and Id3 during neural di¡erentiation. Biochem. Biophys. Res. Commun. 199, 1355^1362. Nelson, N., 1996. Microarrays pave the way to 21st century medicine. J. Natl. Cancer Inst. 88, 1803^1805. Norton, J.D., Deed, R.W., Craggs, G., Sablitzky, F., 1998. Id helixloop-helix proteins in cell growth and di¡erentiation. Trends Cell Biol. 8, 58^65. Pan, W., Lin, J., Le, C.T., 2002. Model-based cluster analysis of cDNA microarray data. Genome Biol.3, research 0022.1^0022.10. Platzer, J., Engel, J., Schrott-Fischer, A., Stephan, K., Bova, S., Chen, H., Zheng, H., Striessnig, J., 2000. Congenital deafness and sinoatrial node dysfunction in mice lacking class D L-type Ca2+ channels. Cell 102, 89^97. Rivolta, M.N., 1997. Transcription factors in the ear: molecular switches for development and di¡erentiation. Audiol. Neurootol. 2, 36^49. Schena, M., Shalon, D., Davis, R.W., Brown, P.O., 1995. Quantitative monitoring of gene expression patterns with a complementary DNA microarray. Science 270, 467^470. Vale, C., Sanes, D.H., 2000. A¡erent regulation of inhibitory synaptic transmission in the developing auditory midbrain. J. Neurosci. 20, 1912^1921. Waldmann, R., Bassilana, F., de Weille, J., Champigny, G., Heurteaux, C., Lazdunski, M., 1997a. Molecular cloning of a non-inactivating proton-gated Na+ channel speci¢c for sensory neurons. J. Biol. Chem. 272, 20975^20978. Waldmann, R., Champigny, G., Bassilana, F., Heurteaux, C., Lazdunski, M., 1997b. A proton-gated cation channel involved in acid-sensing. Nature 386, 173^177. Weintraub, H., Davis, R., Tapscott, S., Thayer, M., Krause, M., Benezra, R., Blackwell, T.K., Turner, D., Rupp, R., Hollenberg, S. et al., 1991. The myoD gene family: nodal point during speci¢cation of the muscle cell lineage. Science 251, 761^766. Wyborski, R.J., Bond, R.W., Gottlieb, D.I., 1990. Characterization of a cDNA coding for rat glutamic acid decarboxylase. Mol. Brain Res. 8, 193^198.
HEARES 4022 18-12-02