|
Figure 1. Schematic summary of GABAB1 isoforms.A. Introns and exons of known and novel GABAB1 isoforms were compared. The numbers listed at the top are exon numbers of GABAB1a. Blocks represent exons, and lines are introns of the GABAB1 gene. Black lines and blocks are ORFs, and grey lines and blocks are UTRs. Double line is cDNA microarray probe, clone 300899, which aligns to the intron 4 of GABAB1a. Human GABAB1a, b, c, e, and rat GABAB1j are previously identified isoforms. After cloning, the following novel isoforms were found in human: GABAB1k, GABAB1l, GABAB1m, and GABAB1n. Human GABAB1j was predicted. Mouse GABAB1j and GABAB1k were also cloned. GABAB1k contains the 3′ part of intron 4 through exon 23. GABAB1l has additional splicing out of exon 15, and GABAB1m has similar splicing patterns as GABAB1l except for additional splicing out at 5′ part of exon 6. GABAB1n has an SNP on exon 8. B. ORFs of GABAB1 isoforms are shown with important functional domains. Black bars are previously known isoforms, and grey bars represent novel isoforms. Dotted lines stand for unique sequence of each isoform. GABAB1k contains GABA binding sites, seven transmembrane domains, a G-protein coupling region, and a C-terminal intracellular region. GABAB1l and GABAB1m do not have transmembrane domains, a G-protein coupling region, or C-terminal intracellular regions (same as GABAB1e). GABAB1m has a longer N-terminal than GABAB1l. GABAB1n contains only a partial GABA binding site.
|
|
Figure 2. Sequences for human, mouse, and rat GABAB1j isoforms and relative expression levels.A. Sequence alignment of human, mouse, and rat GABAB1j showed a different C-terminal pattern in human GABAB1j. B. Human and mouse GABAB1j mRNA expression levels were measured with quantitative real-time PCR. GB1 primer and probe sets were used to detect most known major GABAB1 isoforms except GABAB1j. Human and mouse GABAB1j primer and probe sets were specific for only GABAB1j. GB1 expressions were shown as fold differences from human and mouse GABAB1j. Although a previous rat GABAB1j study indicated that GB1 and GABAB1j expression levels were similar, human and mouse GABAB1j expressions were lower than GB1. Mouse GABAB1j expression was 5.5 fold lower than GB1 expression, and human GABAB1j showed 135.89 fold lower expression. If the fold differences were compared, human GABAB1j expression level was much less (25 times) than mouse GABAB1j expression. Therefore, GABAB1j expressions vary across species.
|
|
Figure 3. Strategy for cloning GABAB1k.The top figure represents exon-intron composition of the proposed GABAB1k. Numbered rectangular boxes represent known exons, and lines between boxes are introns. Two arrows on the bottom of the boxes represent ORF. Black bar is intron 4, and the two probe sequence data, NCBI accession numbers W07715 and N80593, are indicated as #1 and #2 above small rectangular boxes, respectively. They are 5′ and 3′ sequences of the microarray probe, clone 300899. A arrow above the black bar is a possible transcription initiation site. Horizontal lines indicate sequences detected after PCR. Based on microarray probe, clone 300899, the following clones were detected in human cultured cells, mouse midbrain, and rat hippocampus: #2, #3, #4, #11, and #18. From the previous microarray probe sequence analysis, the microarray probe aligned to intron 4 and was either 3′ or 5′ UTR of novel isoforms. Because there was no known human GABAB1 isoform that had intron 4 as an exon, the two possible isoform models, GABAB1j and GABAB1k, were proposed. The predicted GABAB1j and GAGAG1k contain intron 4 as 3′ or 5′ UTR, respectively. Because GABAB1j and GABAB1k are only two possible isoform models that contain clone 300899, #2, #3, #4, #11, and #18 can be partial GABAB1j or GABAB1k and show their existences. From the same RNA sources six different N-terminal partial clones, #5-1, #5-2, #6-1, #6-2, #12, and 17&19, indicated the existence of GABAB1k. They share common 5′ UTRs and partial N-terminal ORFs of GABAB1k and suggest that GABAB1k exists in human, mouse, and rat. To clone the full ORF of GABAB1k, additional primers were designed at its possible 3′ UTR region based on other known isoform sequence analyses. Full ORF containing clones were cloned from human brain and mouse midbrain. Bottom two boxes show clones which contain the full ORF (a&b, a&7, 3&7 in human and c&d in mouse).
|
|
Figure 4. The expression of GABAB1 isoforms.A. For quantitative real-time PCR, primer and probe sets were designed from human and mouse. GB1 sets were designed by the last two exons, exon 22 and 23, and were used for detecting most known GABAB1 isoforms except GABAB1j. They also represented the microarray probes that aligned to 5′ exons of GABAB1 gene. GB2 sets were for detection of GABAB1j, k, l, m, and n. They aligned to intron 4 and designed for human cDNA microarray probe, clone 300899, and its mouse homologous probes. GB3 sets targeted intron 4 through exon 6 and were specific for GABAB1k, l, m, and n. B. In human brain, GB1 showed much more dominant expression than GB2 and GB3. C. GB1 expression was also stronger than GB2 and GB3 in mouse brain. However, GB2 and GB3 expression levels in mouse brain were much higher than human brain if they were compared with GB1 expressions. D. In human, GB3 expressions were similar in prefrontal cortex, frontal cortex, and motor cortex.
|
|
Figure 5. GABAB1a/2 concentration response curve.A. cDNA and cRNA double injections were performed. For cDNAs, both GABAB1a and GABAB2 were injected into the nucleus of Xenopus oocytes. GIRK1 and GIRK2 cRNAs were also injected into the cytosol of the same oocytes. Current response of GABAB receptor and GIRK expressing oocytes were measured by two-electrode voltage clamp. 1 mM GABA was the maximum response concentration; EC50 was 1.9 uM; Hill slope was 0.59. B. For GABAB1a/2 concentration response curve measurement, GABA dissolved in 49 mM HK was applied for 1 min following a 2 min 49 mM HK application. After 2 additional min of 49 mM HK application, a 20 min wash with ND96 was followed.
|
|
Figure 6. The functional effects of three novel human isoforms, GABAB1k, l, and m.Inhibitory effects were measured in Xenopus oocytes 5 days after injecting GABAB1a, GABAB2 cDNA and GIRK1, GIRK2 cRNA along with GABAB1k, l, or m cDNA. For vector controls, vectors were injected in the same amounts as used for the isoforms. Isoform and vector injection amounts varied from 0 to 15 times of GABAB1a injection amount. Used GABA concentration was 3 mM, an approximate maximum response concentration. For each batch, individual GIRK currents from isoform injected oocytes were normalized to percentage by the average vector control current. For nonparametric tests, the individual percent values were changed to log10. After one-way ANOVA and Dunnett's tests, one-tailed unpaired Student's t-tests were performed between no isoform injected groups and the isoform injected groups. A. GABAB1k had no inhibitory effect. B. GABAB1l data showed significant differences by ANOVA and Dunnett's tests (p = 0.0058), and caused significantly decreased GIRK currents at 5× and 6× GABAB1l (**p = 0.0014 and *p = 0.0229, respectively). C. ANOVA and Dunnett's tests showed the significant differences of GABAB1m data (p = 0.0040). GABAB1m showed significant decreases at 11× and 12× GABAB1m (***p = 0.0003 and **p = 0.0064, respectively). The numbers of recorded oocytes are shown above the bars by each group. Data are expressed as the mean ± SEM.
|
