cDNA Microarray Analysis of Differential Gene Expression in Boar Testes during the Prepubertal Period

In an attempt to understand the biochemical mechanism for the synthesis of the anabolic steroid, 19-nortestosterone, produced by prepubertal boar testes and its physiological role, normalized complementary DNA (cDNA) from boar testes was generated. DNA sequencing of 2,016 randomly selected clones yielded 794,116 base pairs of high quality nucleotide sequence. Computer-assisted assembly of the nucleotide sequence of each clone resulted in 423 contigs and 403 singletons including several genes for steroidogenic enzymes and molecules related to steroid metabolism. Analysis of gene expression pattern by use of the presently-fabricated cDNA microarray identified a number of genes that were differentially expressed during the postnatal development period in boar testes. Two genes of unknown function were identified to be highly expressed in the testis of 2-weeks-old neonatal boar. In addition, the sequencing of open reading frames of these genes revealed their homology with human alpha hemoglobin and Homo sapiens hypothetical LOC643669, transcript variant 1. Moreover, the transcripts of these genes were also detected in porcine muscle and adipocytes, in addition to Leydig cells of pigs. (


INTRODUCTION
Growth and development of animals depend on a number of factors that regulate gene expression and cell behavior.Transient changes of gene expression in various tissues and organs during the early postnatal period have been amply demonstrated (Chen et al., 2004;Ponsuksili et al., 2007;Wu et al., 2008).Such differential expression of genes at this critical time in early development is required for proper differentiation and proliferation of cells in a certain tissue.Thus, it is important to determine the temporal profile of gene expression during the post-natal development period.
The boar has a number of unique features concerning the synthesis of uncommon steroid hormones that include unusually high serum levels of anabolic steroids such as androstenedione and 19-nortestosterone (Debruychere and Van Peteghem, 1991;Schwarzenberger et al., 1993).The boar is one of only a few males that have a high estrogen concentration in the serum (Schwarzenberger et al., 1993).It is noteworthy that production of these anabolic steroids and estrogens is dramatically increased between post-natal weeks 2~4, followed by a transient decrease to basal levels at 2 months of age until puberty (Schwarzenberger et al., 1993).Significant production of anabolic steroids in the boar results from the action of cytochrome P450 aromatase (Choi et al., 1997;Kao et al., 2000).Our previous study has shown that such transient increase of 19-nortestosterone concentration in the serum during early pre-natal development is closely related with expression of steroidogenic enzymes and estrogen receptors in boar testis (Choi et al., 2007).High concentrations of 19nortestosterone and 19-norandrostenedione have been detected between 1~2 weeks of age in the testis among various boar tissues examined (Choi et al., 2007).In addition, expression levels of cytochrome P450 aromatase, 17α-hydroxylase, and 3β-hydroxysteroid dehydrogenase genes are significantly increased at 2 weeks of age, compared with 0, 1, 3, 5, and 6 weeks of age (Choi et al., 2007).Moreover, the latter study has found that expression of estrogen receptor α and β in the boar testis is greatly increased at 2 weeks of age.Thus, up-regulation of gene expression of steroidogenic enzymes in the boar testis during the neonatal period could be responsible for significant production of anabolic steroids.Indeed, our recent study has demonstrated differential expression of a number of steroidogenic enzymes in boar testis during the neonatal and pre-pubertal periods (Choi et al., 2009).In this study, we further have documented that mRNA and protein levels of enzymes involved in the synthesis of testosterone and estrogen are significantly increased during neonatal development at 2-3 weeks of age compared with levels at pre-puberty (4 months of age).Along with the findings from a previous study (Choi et al., 2007), these observations imply that boar testis around 2 weeks of age during the post-natal period is metabolically active and undergo significant morphological and endocrinological changes.
Complementary DNA (cDNA) microarray analysis is a powerful biological technique commonly used to simultaneously detect differentially expressing genes.A number of previous researches have utilized cDNA microarray analyses to reveal differential expression of genes in various pig tissues (Kim et al., 2005;Guo et al., 2008).In the present study, we first constructed a normalized cDNA library from 2-weeks-old boar testis.Then we attempted to determine the differential gene expression profile in boar testis at post-natal ages of 3 days, 2 weeks, 7 weeks, and 13 weeks.These results were used in an attempt to characterize unknown genes found to be highly expressed at 2 weeks of age.

Animals and tissue collection
The male reproductive tract from boars (Landrace× Yorksire×Duroc) was obtained from local farms in Kyungbuk province, Republic of Korea, by surgical castration.The testes were separated from the rest of the male reproductive tract in ice-cold phosphate buffered saline (PBS) and were rapidly frozen in liquid nitrogen.The testes were stored at -80°C until required for further studies.Testes from a total of twenty-one 10~14-days-old piglets were used for construction of a normalized cDNA library.In addition, testes were obtained from boars at 3 days (n = 4), 2 weeks (n = 3), 7 weeks (n = 2), and 13 weeks (n = 6) of age for comparison of gene expression using DNA microarray analysis.

Construction of normalized cDNA library
Total RNA was isolated from testes using TRIzol TM (Invitrogen, Carlsbad, CA, USA) and stored in diethylpyrocarbonate (DEPC)-treated distilled water as described previously (Choi et al., 2007).The quantities and qualities of total RNA were determined using a ND-1000 spectrophotometer (Nanodrop Technologies, Wilmington, DE, USA) and agarose gel electrophoresis, respectively.Reverse transcription reactions were carried out to synthesize first-strand cDNA from total RNA in a mixture of 1 μg of total RNAs, oligo (dT) 18 primer, 10 mM of dNTP, 4 μl of 5×buffer, 1 μl of RNase out (RNase inhibitor, Takara, Shiga, Japan), and 1 μl of Superscript II reverse transcriptase (Invitrogen).Newly synthesized cDNAs were stored at -20°C until used later.To construct a normalized cDNA library, total RNA collected from the testes at 10~14 days-of-age were pooled and mRNAs were isolated using the Absolutely mRNA TM purification kit (Stratagene, La Jolla, CA, USA).Highly purified mRNAs were reversetranscribed to generate first-stranded DNAs using oligo-dT primer and reverse transcriptase, and second-stranded DNAs were generated using first-stranded DNA templates and DNA polymerase.These purified double-strand (ds)DNAs were then hybridized with the reversetranscribed cDNAs described above and treated with duplex-specific nuclease at 65°C for 2 min and at 68°C for 5 h to remove DNA-DNA or DNA-RNA duplexes.Isolated cDNAs were size-fractionated and collected by a gel filtration column.Quality and the size of the collected cDNAs were verified using a 5% non-denaturing acrylamide gel.To construct the cDNA library, 5'-ends of newly synthesized dscDNAs were ligated with EcoRI adapter DNA by DNA ligase, and the 3'-ends of the cDNA were trimmed by XhoI.These cDNAs were sizefractionated through a gel filtration column and separated by agarose gel electrophoresis.cDNAs of 600 bps or larger in size were collected and isolated by phenol-chloroform extraction and ethanol precipitation.Isolated cDNAs were inserted into a pBK-CMV vector and packed using a bacteriophage packaging extract (Gigapack III Gold Packaging Extract; Stratagene).After titration, a total of 300,000 independent clones were confirmed; cDNA clones were aliquoted in 384 well plates and stored at -80°C.

DNA sequencing and data analysis
Plasmid DNA from cDNAs clones used to construct the cDNA library were purified by the standard alkaline lysis method and were amplified by the dye-determinator cycling method (BigDye v3.1) using 5' universal primers in the vector (Stratagene).Then, sequences of plasmid DNAs were determined using an ABI3730 XL capillary sequencer (Applied Biosystems Inc, Foster City, CA, USA).The chromatogram data were initially submitted to Phred (Ewing et al., 1998) for base calling and quality assignment.The trace files were trimmed with trim-alt 0.12 and the sequence below 100 bps was removed.In addition, vector trimming was conducted with a cross-match software (http://www.phrap.org),and poly-A was removed using a trimmest of Emboss package (Rice et al., 2000).The prepared multifasta format data were clustered and assembled using the TGICL package (Pertea et al., 2003).Finally, the contigs and singletons were analyzed with NCBI local BLAST (Altschul et al., 1990).Sequences were searched against the NCBI NR databases (2005.05).
Matches with E-value of 1×e -10 for BLASTX were considered as putative hits.

cDNA microarray fabrication
A cDNA microarray was prepared from the cDNA library.Polymerase chain reaction (PCR) was conducted according to a standard protocol.The T7 (5'-TAATACGACTCACTATAGGG-3') and T3 (5'-ATTAACCCTCACTAAAGGGA-3') promoter primers were used for amplification in a 50 ul reaction mix consisting of Premix taqTM (TaKaRa Bio Inc., Shiga, Japan), 0.4 μM of each primer and 1 μl of DNA extract.Thermocycling conditions consisted of an initial denaturation step at 94°C for 3 min, then 35 cycles of 94°C for 1 min, annealing at 58°C for 1.5 min and extension at 72 °C for 1 min, with final extension at 72°C for 3 min.PCR products were visualized under ultraviolet (UV) light on 1.0% agarose gel stained with ethidium bromide and reaction mixtures were subjected to 35 cycles of amplification.The size and amount of the PCR products were verified using 1% agarose gel.The PCR products were purified by ethanol precipitation followed by re-suspension in 15 μl of spotting solution (GenoCheck, Korea) and spotting onto a CMT-GAPS II glass slide (Corning; Corning, NY, USA) with a pixsys 5500 arrayer (Cartesian Technologies, Irvine, CA, USA) using Stealth Micro spotting pins.For normalization of signals, a housekeeping gene (β-actin) and a positive control of Arabidopsis thaliana genes were included on each slide.The printed slides were processed according to CMT-GAPS II slide protocol.Briefly, the spots were rehydrated with 1× SSC for 1 min and then DNAs were linked using a UV cross-linker (Stratagene).The slides were soaked in succinic anhydride/sodium borate solution for 15 min with gentle agitation and then transferred to a 95°C water bath for 2 min.The slides were quickly transferred to 95% ethanol for 1 min and then dried by centrifugation at 3,000 rpm for 20 s.

DNA chip hybridization and data analysis
Total RNAs from boar testes at 3 days, 2 weeks, 7 weeks, and 13 weeks of age were isolated using TRIzol TM (Invitrogen).Fluorescence-labeled cDNA probes were prepared from 30 μg of total RNA by oligo (dT) 18 -primed polymerization using SuperScript II reverse transcriptase (Invitrogen) in a total volume of 30 μl including 25 mM each of dATP, dTTP, and dGTP, 10 mM of dCTP, and 3 mM of Cy3 or Cy5 labeled dCTP (NEN Life Science Product Inc., Boston, MA, USA).After reverse transcription, the labeled cDNAs were concentrated using ethanol precipitation followed by re-suspension of Cy3 and Cy5 labeled cDNAs in 10 μl of hybridization solution (GenoCheck, Seoul, Korea).Two cDNAs labeled with different dyes were mixed in an equal volume and denatured at 95°C for 2 min, followed by incubation at 45°C for 20 min.The cDNA mixture was then placed on a spotted slide and covered by a cover slip.The slides were hybridized in a hybridization chamber at 62°C for 12 h.The slides were sequentially washed in 2×SSC, 0.1% sodium dodecyl sulfate (SDS) for 2 min, 1×SSC for 3 min, and 0.2× SSC for 2 min at room temperature.The slides were centrifuged at 3,000 rpm for 20 s to dry.After the hybridization reaction, slides were scanned with GenePix 4000B scanner (Axon Instruments, Union City, CA, USA).The scanned images were analyzed with GenePix Pro 5.1 software (Axon Instruments) and GeneSpring GX 7.3.1 software (Agilent Technologies, Santa Clara, CA, USA).To allow for an algorithm-mediated elimination of all bad spots, no data points were eliminated by visual inspection from the initial GenePix image.To filter out unreliable data, spots with signal-to-noise ratio below 100 were not included in the data.Data were normalized by global, lowness, print-tip, and scaled normalization for data reliability.Data were clustered in groups of genes that behaved similarly across time course experiments using GeneSpring GX 7.3.1 (Agilent Technologies).We used an algorithm based on the Pearson correlation to separate the genes of similar patterns.The distance cut-off was statistically considered as significant when there were two-fold differences between time course experiments.The correlation cut-off was 0.95, and the accuracy of microarray analyses in the present study was confirmed by real-time PCR analysis.

Cell culture
Leg muscle satellite cells and Leydig cells of the testis were isolated from boars (110 kg body weight) at a local slaughtering house.Skeletal muscle was washed in PBS.Then, each muscle was minced using sterilized scissors and digested by trypsin-EDTA (GIBCO-BRL, Valencia, CA, USA) for 2 h.Digested tissue was centrifuged at 90 g for 3 min and the upper phase was filtered using 40 μm pore size cell strainer.The filtrate was centrifuged at 2,500 rpm at room temperature for 20 min.The collected cell pellet was washed three times using Dulbecco's Modified Eagle's Medium (DMEM; HyClone Laboratories, Logan, UT, USA) containing 1% penicillin/ streptomycin (GIBCO-BRL).The muscle satellite cells were cultured in DMEM supplemented with 10% fetal bovine serum (FBS; GIBCO-BRL), 1% penicillin/ streptomycin, and 0.1% Fungi Zone (GIBCO-BRL) in a 5% CO 2 incubator at 37 o C. Porcine testes were digested by trypsin-EDTA and the digested cells were cultured in DMEM/Ham's F-12 (1/1, v/v) (GIBCO-BRL) supplemented with 10% FBS.After 3 h, the medium was removed and replaced with fresh medium.Purified Leydig cells were cultured for 4 days.

Real-time PCR analysis and statistical analysis
The first strand cDNA was synthesized using 1 μg of total RNA template isolated from boar testes, as described earlier.Real-time PCR analysis was performed on an ABI7500 apparatus (Applied Biosystems) in Power SYBR ® green PCR Master Mix TM (Applied Biosystems) according to the manufacturer's instructions.Primer sets for real-time PCR were designed by Primer 3 software (http://frodo.wi.mit.edu)from gene sequences (Table 1).The cDNA product was amplified by PCR with Taq DNA polymerase (Bioneer, Daejeon, Korea).A melting curve was analyzed to check the absence of mispriming.The sizes of the PCR products were confirmed by agarose gel electrophoresis.The fold difference of each gene expression was determined by 2 -ΔΔCt formula (Lay et al., 2002).All values are presented as mean±SEM.Data were analyzed by ANOVA according to the general linear model procedure, followed by Duncan's new multiple range test to determine significant differences at p<0.05.All statistical procedures were performed with the SAS ® ver.9.12 software package (SAS, Cary, NC, USA).

RESULTS AND DISCUSSION
A normalized cDNA library of 10~14 days old boar testis was generated in the present study.Titration of the normalized cDNA library resulted in approximately 3×10 5 independent clones.Among these, 2,016 clones were randomly selected and the 5' end sequence of each was analyzed using the universal primer in the vector.After eliminating the lower quality DNA sequence (Phred score >13) and vector sequence, the total size of the nucleotide sequences was 803,532 bps, in which 198 clones were smaller than 100 bps in size and 1,818 clones were 100-899 bps in size (Table 2).The sequencing success rate was 90.2% and the average read length was 437 bps.In addition, clustering followed by assembly of the nucleotide sequences of the 1,818 clones yielded 423 contigs and 403 singletons.This data indicated that there were 826 unique DNA sequences.Cluster 1-68 contained 5-22 sequences, and other clusters had only two sequences (Figure 1).
Of these, 203 contigs and 326 singletons showed significant matches with known-function genes, whereas the other 220 contigs and 77 singletons were unknownfunction genes.The clones of 423 contigs and 403 singletons were initially utilized to fabricate a cDNA  microarray.Sixty-two out of 826 genes were excluded from a final construction of the microarray after checking with PCR and we constructed a DNA microarray for the genes above 100 bp.The DNA microarray comprised a total of 719 genes, which contained PCR-amplified cDNAs of 709 genes, two negative control genes, and eight positive control genes.Using this fabricated cDNA microarray, analysis was carried out to determine differentially expressed genes in boar testes among different age groups (3 days, 2 weeks, 7 weeks, and 13 weeks of age).The design of the cDNA hybridization for DNA chip analysis is shown in Figure 2A, and representative images of the hybridization are shown in Figure 2B.To detect the differential gene expression between different developmental periods, hierarchical clustering of the genes expressed in testes at different stages were shown considering each stage as controlled at a time.Day 3, and week 2, 7 and 13 samples were considered as control at a time to observe up-, and down-regulated genes in the other samples (Figure 2C).Results from the microarray analysis showed that 19 genes were up-regulated and 10 genes were down-regulated in the boar testis at 2 weeks of age, compared to those at 3 days of age (Table 3).Similarly, nine genes were down-regulated and 12 genes were up-regulated at 7 weeks of age.In addition, there were nine up-regulated genes and eight down-regulated genes at 13 weeks of age.Among the 19 genes highly up-regulated at 2 weeks of age, 12 genes had known functions and seven genes had unknown functions.These known functions included bovine osteonectin, pig guanosine diphosphate dissociation inhibitor 2, pig muscle creatine kinase, and pig skeletal alpha actin gene.In addition, further analysis revealed that expression levels of seven unknown-function genes at 2 weeks of age were 2-6-fold higher than those in the other age groups (Table 3 and Figure 3).
To verify results of the cDNA microarray analysis, realtime PCR analyses were performed with the pooled RNA samples used for microarray analysis, as well as with the individual RNA samples used to make the pooled RNA samples.With pooled RNA samples, three no-hit genes (6B06, 8E11, 5B11) showed high expression levels at 2 weeks of age, compared with those at other ages (Figure 4A).However, with individual RNA samples, six of seven unknown function genes did not show significantly increased expression at 2 weeks of age (Figure 4B).Expression of a no-hit gene, 5B11, was significantly increased at 2 weeks of age.The full-length nucleotide sequencings and the blast analyses of the 8E11 and 5B11 clones showed that 8E11 and 5B11 genes had high homology with human hemoglobin alpha subunit, and Homo sapiens hypothetical LOC643669 transcript variant 1, respectively (Figure 5).Furthermore, we examined tissue expression of the presence of 5B11 and 5G07 genes in pig Leydig cells, muscle cells, and adipocyte cells (Figure 6).Real-time PCR analyses showed that both 5B11 and 5G07 genes were present in pig Leydig cells and adipocyte cells (Figure 6A and B).However, expression of only the 5B11 gene was evident in pig muscle cells (Figure 6A).
A significant production of 19-nortestosterone in the boar during early post-natal development is detected at 2 weeks of age (Schwarzenberger et al., 1993).However, the precise physiological function of 19-nortestosterone in the boar is unclear.In the present study, cDNA microarray analysis identified genes differentially expressed in boar testes during early development.A normalized cDNA library was constructed using the mRNA isolated from the    (Choi et al., 2007).An advantage of a normalized cDNA library over other cDNA library methods is the reduced redundancy of the clones in the library (Sasaki et al., 1994).The library clones presently sequenced included a number of genes coding for enzymes involved in the synthesis or metabolism of steroid hormones such as cytochrome P450 17α-hydroxylase, cytochrome P450 side chain cleavage enzyme, cytochrome b, 17β-estradiol dehydrogenase, 3β-hydroxysteroid dehydrogenase, and catechol O-methyltransferase.Interestingly, the result from clustering of the clones based upon their sequence homology revealed that the largest contig was assembled by 22 clones and contained coding sequence for carbonyl reductase 1 (CBR1).CBR1 is a member of the short-chain dehydrogenase/ reductase superfamily, which is responsible for catalyzed reduction of ketones on androgens and progesterone in porcines (Tanaka et al., 1992;Hoffmann and Master, 2007).Our recent study showed that the expression of CBR1 mRNA and protein in pig testis increases according to age during the neonatal period and is exclusively localized in Leydig cells (Choi et al., 2009).Thus, CBR1 may be involved in steroid hormone synthesis in porcine Leydig cells.In this regard, catechol-Omethyltransferase is an enzyme involved in the breakdown of dopamine and catechol-estrogens (Boulton and Eisenhofer, 1998).
In the present study, cDNA microarray analyses showed that a number of genes were expressed at high levels in boar testis at 2 weeks of age (Table 3).Among these genes, the seven no-hit genes were found to be up-regulated at 2 weeks of age.However, only three genes (6B06, 8E11, 5B11) were verified to be true when the pooled RNA samples were used for real time PCR analysis.In contrast, the 6B06 and 5B11 genes could be confirmed when individual RNA samples were used.The reason for the disagreement in gene expression pattern between the microarray analysis and real time PCR data is due to the individual variations of gene expression in the testes caused by the difference of development stage (10~14-days-of-age).Pigs display a dramatic change of testis development during the neonatal period (França et al., 2000;Moran et al., 2002   Besides, in this experiment normalized cDNA microarray fabrication was performed by using the PCR product from the constructed cDNA library.For the construction of the cDNA library, total RNA was extracted from the whole tissue of testis.Therefore, there is high probability of presence of RNA from untargeted cells and tissues.However, for the confirmation of cDNA microarray data total RNA was extracted from Leydig cell culture.This may be the reason for being able to confirm only 1 gene expression by real time RTPCR among the 7 up-regulated genes by cDNA microarray analysis.The 5B11 and 8E11 genes, initially considered as no-hit genes due to the lack of match with the gene sequence in the database, turned out to be hemoglobin and hypothetical protein, respectively.Because these genes are expressed not only in Leydig cells, but also in myogenic satellite cells and adipocyte cells, it would be of interest to learn the physiological role of these genes in boar testes during early development.

Ages of pigs
In conclusion, we have developed a useful and efficient research tool and materials which can be used to investigate differential gene expression in boar testes during early development.Our interest in the distinct molecular character of the pig testis at 2 weeks was due to detection of high concentrations of 19-nortestosterone and 19norandrostenedione at this age.With this objective, this research showed up-regulation of a higher number of genes (19 genes) in 2 weeks old pig testes compared to the other ages studied.Among 19 genes, 7 no hit genes were detected by the cDNA library and two of them were confirmed by real time RT PCR.We speculate their involvement in the high concentration of 19-nortestosterone and 19norandrostenedione in the 2 week-old boar at this point.Further studies of the known and unknown genes identified in the study will lead to a better understanding of the role of 19-nortestosterone in this species.

ACKNOWLEDGMENTS
The authors would like to thank all members of the molecular biology laboratory for their assistance.

Figure 1 .
Figure 1.The distribution of cluster size.The 1,417 ESTs were grouped into 418 clusters and 404 singletons.

Figure 2 .
Figure 2. cDNA microarray scheme.(A) Total RNAs extracted from different aged boar piglet testes (day 3, 2 weeks, 7 weeks, and 13 weeks) were synthesized to radiolabelled cDNA and hybridized to a DNA microarray.Considering a sample as control, gene expression was compared to the other three samples.(B) Hybridization was performed in six different ways among the four different cDNA samples.A representation of two color fluorescent scans showing six different combinations of hybridization is depicted; green and red spots indicate up-regulation of Cy3 and Cy5 labeled cDNA, respectively.In addition, yellow spots represent equally expressed genes.(C) Similarly, a gene expression map showing up-(red) and down-regulated (green) genes compared with control sample (day 3, 2 weeks, 7 weeks, and 13 weeks was considered as control in first, second, third, and fourth figures, respectively).

Figure 4 .
Figure 4. Real time RT PCR of no-hit genes.Real time RT-PCR was performed to confirm the cDNA chip microarray data.Individual RNA samples were synthesized to cDNA and real time PCR was performed.(A) One no-hit gene (5B11) showed significant fold difference (Duncan's new multiple range test p< 0.05) in 2-week-old piglet testes compared to the others.(B) Furthermore, real time RT-PCR performed with the pooled total RNAs showed up-regulation of three no-hit genes (6B06, 8E11, and 5B11) in 2-week-old piglet testes.Data shown are after HPRT normalization; a, b, c represent significant differences (p<0.05).
Figure 5. i) Open reading frame region of two no-hit genes.Two genes of unknown functions that were highly up-regulated in 2-week-old piglet testes were sequenced.Analysis of open reading frame regions of 8E11 (A) and 5B11 (B) revealed high homology with human hemoglobin alpha and Homo sapiens hypothetical LOC643669, transcript variant 1, respectively.

Figure 5 .
Figure 5. ii) Open reading frame region of two no-hit genes.Two genes of unknown functions that were highly up-regulated in 2week-old piglet testes were sequenced.Analysis of open reading frame regions of 8E11 (A) and 5B11 (B) revealed high homology with human hemoglobin alpha and Homo sapiens hypothetical LOC643669, transcript variant 1, respectively.

Table 2 .
The nucleotide sequence results of normalized cDNA library clones

Table 3 .
List of differentially expressed genes in boar testes at 2 weeks, 7 weeks, and 13 weeks compared to day 3 along with their annotation ).