SEARCH

SEARCH BY CITATION

Keywords:

  • lens development;
  • GABAA receptor subunits;
  • GABAB receptor subunits;
  • membrane GABA transporters (GAT1-4);
  • vesicular GABA transporter (VGAT);
  • non-neuronal GABA;
  • GAD;
  • intracellular calcium

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS AND DISCUSSION
  5. EXPERIMENTAL PROCEDURES
  6. Acknowledgements
  7. REFERENCES

Gamma-aminobutyric acid (GABA), the major inhibitory neurotransmitter in the vertebrate nervous system, serves as a signaling molecule modulating diverse processes during embryonic development. Earlier we have demonstrated that different forms of glutamic acid decarboxylase (GAD) are differentially regulated during mouse lens development. Here we show that the developing lens expresses also components of GABA signaling downstream of GAD. Multiple GABAA and GABAB receptor subunits as well as the GABA transporters show expression profiles highly correlated with the expression of different GADs. GABA receptors (GABAR) and the vesicular GABA transporter localize at the apical/basal membranes of the lens epithelia and differentiating fibers and may be involved in conventional GABAR-mediated signaling, while the membrane GABA transporters may also function as Na+/Cl/GABA carriers. The functionality of GABAR was verified by calcium imaging in whole lenses. Our data suggest that GABA synthesized locally by GAD, acts through GABA receptors by modulating the intracellular calcium levels. Developmental Dynamics 237:3830–3841, 2008. © 2008 Wiley-Liss, Inc.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS AND DISCUSSION
  5. EXPERIMENTAL PROCEDURES
  6. Acknowledgements
  7. REFERENCES

The vertebrate lens is composed of two populations of cells—a layer of cuboidal epithelial cells that covers the anterior surface, and elongated, terminally differentiated fiber cells that constitute the bulk of the tissue. In a sequence of events precisely regulated by multiple transcription and growth factors, single epithelial cells of the germinative zones proliferate and move to the equator of the lens. In the transitional zone (lens bow region) found below the equator, the epithelial cells elongate and differentiate into fiber cells that move in to fill up the lumen of the lens (Lang,1999,2004; McAvoy et al.,1999; Ogino and Yasuda,2000; Lovicu and McAvoy,2005). Concomitant loss of all membranous organelles and increase of lens crystallins' concentration provides for the lens transparency required for proper vision (reviewed in Piatigorsky,1981; Bassnett,2002; Menko,2002; Bloemendal et al.,2004). This regional compartmentalization makes the lens an attractive model system to study signaling mechanisms during cellular differentiation.

It is widely accepted that γ-aminobutyric acid (GABA) is a trophic factor during central nervous system (CNS) development modulating the proliferation, migration and differentiation in a paracrine way, by activating nonsynaptic GABAA and GABAB receptors triggering membrane excitation and rise in intracellular Ca2+ ([Ca2+]i) (Barker et al.,1998; Varju et al.,2001; Owens and Kriegstein,2002a,b; Represa and Ben-Ari,2005). Glutamic acid decarboxylase (GAD) and GABA as well as GABA transporters and receptors are expressed from the earliest stages of nervous system development and in numerous non-neuronal cells like the chromaffin cells of the adrenal gland, endocrine cells of the gastrointestinal tract, tubular epithelium of the kidney, hepatocytes, pancreas, in chondrocytes of the growth plate, embryonic nasal epithelium and the lens (Erdo and Wolff,1990; Li et al.,1995; Watanabe et al.,2002; Hagiwara et al.,2003; Gladkevich et al.,2006; Kwakowsky et al.,2007). The molecular mechanisms underlying GABA release and downstream effects in immature cells of the nervous system and non-neuronal cell types are not fully understood although results from gene knock-outs of GABA signaling components, leading to (often lethal) developmental defects clearly indicate that GABA has an indispensable morphogenetic function in both CNS and periphery (Culiat et al.,1995; Asada et al.,1997; Homanics et al.,1997; Kash et al.,1997; Ji et al.,1999; Burt,2003; Wojcik et al.,2006).

The presence of GAD67, EGAD, and GABA was originally described in the developing rat lens by (Li et al.,1995). We have recently shown that GABA, GAD, the enzyme catalyzing GABA synthesis, and Dlx2, Dlx5, known as upstream regulators of GAD are expressed in the developing mouse lens from early embryonic stages (Kwakowsky et al.,2007). Here we report the expression patterns of different GABAA and GABAB receptor subunits, vesicular (VGAT) and membrane GABA transporters (GATs) in the developing mouse lens. We find that different molecular components of the GABA signaling pathway are spatiotemporaly regulated in a way suggesting that specific combinations are coordinately expressed during different stages of lens development. Furthermore, we demonstrate that externally added GABA could evoke transient increase of intracellular Ca2+ concentration in intact lenses.

This report is the first ever description of functional GABAergic signaling system in the lens and the first detailed characterization of the dynamic regulation of multiple components of this pathway downstream of GAD providing evidence for their developmental stage-specific co-regulation.

RESULTS AND DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS AND DISCUSSION
  5. EXPERIMENTAL PROCEDURES
  6. Acknowledgements
  7. REFERENCES

Multiple GABA Signaling Components Are Expressed and Temporaly Regulated During Lens Development: Evidence for Co-regulation

The GABA signaling machinery encompasses molecular components involved in GABA synthesis (GAD), release (vesicular GABA transporter-VGAT and membrane transporter-GAT), binding (GABAA and GABAB receptors), uptake (GAT), and degradation (GABA transaminase-GABA-T). In this study, we present data on the temporal expression profile of mRNAs encoding 11 different GABAAR, subunits (α1–4; β1–3; γ1–3, and δ), the two GABABR subunits, VGAT and GAT1-4, respectively, during developmental stages E14.5–P30 obtained by semiquantitative reverse transcriptase-polymerase chain reaction (RT-PCR; Fig. 1A). These expression profiles were analyzed and grouped according to developmental stage (embryonic, fetal, neonatal, and postnatal; Fig. 2A). Furthermore, applying Spearman's correlation (Kotlyar et al.,2002), we obtained the correlation coefficient for several gene pairs with highly similar profiles (Fig. 2B1–B6).

thumbnail image

Figure 1. Expression of γ-aminobutyric acid A (GABAA) and GABAB receptor subunits and GABA transporters in the developing mouse lens as revealed by semiquantitative reverse transcriptase-polymerase chain reaction (RT-PCR). Left panel: Representative gel images of RT-PCR products of total mRNA amplified from embryonic (E14.5–17.5) and postnatal (P0, 7, 14, 30) lenses and adult mouse brain (MB). Sequences of the specific exon-spanning primer pairs and product size for each cDNA are listed in Table 1. The β-actin (*) was co-amplified with the target genes. Right panel: Gene expression map of GABA receptor subunits and GABA transporters, based on calculations of intensities of RT-PCR products for each gene relative to the intensity of the co-amplified β-actin and represented on a grey scale as % of the maximum level. Analyses were performed in ImageJ v. 1.40, an open-access public domain Java image-processing program for Mac OS X.

Download figure to PowerPoint

thumbnail image

Figure 2. Correlated developmental stage-specific expression profiles of γ-aminobutyric acid (GABA) signaling components in the developing mouse lens. A: Reverse transcriptase-polymerase chain reaction (RT-PCR) data of GABAergic components shown in Figure 1 and Figure 1A from Kwakowsky et al. (2007) are presented according to developmental stages (E14.5–P30) as follows: embryonic (E14.5); fetal-E15.5–E17.5; P0- neonatal; postnatal-P7–P30. Relative expression levels are indicated as: +++, high; ++, medium; +, low; −, no expression. BB6: Correlated expression curves of selected components: EGAD vs. GAD65 (B1), vesicular GABA transporter (VGAT) vs. GAD65 (B2), GABAAα2 vs. GAD65 (B3), GABAAγ2 vs. GABAAα2 (B4), GAT-2 vs. GAD67 (B5) and GABAAα1 vs. GAD67 (B6). Pair-wise correlation analysis of relative gene expression levels at different developmental stages was performed using Prism software. Spearman's rank correlation coefficient (rs) for each gene pair is indicated on the scatter plots. The best fit (solid line) is shown with 95% confidence intervals (dotted lines). Numbers on x- and y-axes represent percentage of relative gene expression levels presented in Fig. 1 (right panel).

Download figure to PowerPoint

Table 1. Primers and Amplification Parameters Used in the Semiquantitative Reverse Transcriptase-Polymerase Chain Reaction Analysesa
mRNAPositionPrimer SequenceProduct Size (bp)Annealing temp.(°C)Extention time (sec)Reference
  • a

    (1) Varju et al.,2002; (2) Ma et al.,1993; (3) Liu et al., 1998; (4) Drescher et al.,1993; (5) Sim et al.,2000; (6) Behar et al.,2001; (7) Chen et al.,2007; (8) Voutsinos et al.,1998; (9) Westmoreland et al.,2001; (*) Gene-specific primers were designed for this study according to the published sequences as appeared in Ensembl (www.ensembl.org). Primer pairs for all studied genes were selected from different exons. The positions are relative to the ATG codons of the corresponding cDNAs. F: forward; R: reverse.

Actin609F 5′-AGCTGAGAGGGAAATCGTGC-3′49955-6260-90(Varju et al.,2002)
1107R 5′-GATGGAGGGGCCGGACTCAT-3′
GABAAα151F:5′-TCTGAGCACACTGTCGGGAAG-3′5846260Modified from (Ma et al.,1993)
634R:5′-ACCCATCTTCTGCTACAACCACTG-3′
GABAAα2798F:5′-TCTCTCCCAAGTGTCATTCTGGCTG-3′5835860*
1380R:5′:-GCCCAAAAGTAACCAAGTCTA-3′
GABAAα3217F:5′-CGGCTTTTGGATGGCTATG-3′7376090*
953R:5′-ATGGTGAGAACAGTGGTGACA-3′
GABAAα41045F:5′-CAGAAAGCCAAAAAGAAGATA-3′5505890*
1594R:5′-TAAATGCTCCAAATGTGACTG-3′
GABAAα5164F:5′-GACTCTTGGATGGCTATGAC-3′5605890*
723R:5′-TGTGCTGGTGCTGATGTTCTC-3′
GABAAα652F:5′-CAAGCTCAACTTGAAGATGAAGG-3′4205860(Liu and Burt,1998)
419R:5′-TCCATCCATAGGGAAGTTAACC-3′
GABAAβ1581F:5′-ATGGAGGAGAGGGAGCAGTAA-3′8065590(Drescher et al.,1993)
1386R:5′-AGAAAAGGTGATGGGGAAGAA-3′
GABAAβ2856F:5′-ACCACAATCAACACCCACCT-3′2995860*
1154R:5′-CCCATTACTGCTTCGGATGT-3′
GABAAβ3934F:5′-GGCTGCTTTGTCTTTGTATTC-3′3375560Modified from (4)
1270R:5′-TGTGCGGGATGCTTCTGTCTC-3′
GABAAγ11077F:5′-CAACAATAAAGGAAAAACCACCAGA-3′2935860Modified from (Ma et al.,1993)
1369R:5′-CCAGATTGAACAAGGCAAAAGC-3′
GABAAγ2553F:5′-ATTGATGCTGAGTGCCAGTTGC-3′4126060(Sim et al.,2000)
964R:5′-TGGCTATGGTGCTTAAAGTTGTC-3′
GABAAγ3646F:5′-TGGCGGCTCTATCAGTTTG-3′2256060*
870R:5′-GATGCCTAATGTTGTTCTTGC-3′
GABAAδ1079F:5′-TCGTCCTTTTCTCCCTCTCAG-3′4545860*
1532R:5′-AGCCCATCCTGTTCCATCTA-3′
GABABR1859F:5′-ACGCATCCATCCGCCACAC-3′3475860(Behar et al.,2001)
1183R:5′-AACCAGTTGTCAGCATACCACCC-3′
GABABR22171F:5′-CCCTGGTCATCATCTTCTGTAG-3′2305860*
2400R:5′-CTTCATTCGTAGGCGGTGG-3′
GABACρ1−25F:5′-GAATCTATGTTGGCTGTCCAGA-3′7756090(Chen et al.,2007)
751R:5′-TGGTGTGGAATTCTTGAAT GAG-3′
GAT1147F:5′-ACGCTTCGACTTCCTCATGTCCTGC-3′6996290(Voutsinos et al.,1998)
845R:5′-GAATCAGACAGCTTTCGGAAGTTG-3′
GAT2294F:5′-CTGTGGCATCCCAGTGTTC-3′3136060*
606R:5′-GACAGGCGAGGTAAAGTTCTC-3′
GAT3517F:5′-AATGTGACCTCCGAGAATGC-3′5336290*
1049R:5′-ACCTCAGATATGGGCACACC-3′
GAT41130F:5′-ACCCCAAGGCTGTCACTATG-3′7346290*
1864R:5′-CTGTGATGGCAGAGATGGTG-3′
VGAT−31F:5′TTCTGTCCTTTTCTCCCGCCCCGCCGCC-3′57268180(Westmoreland et al.,2001)
541R:5′GCACCACCTCCCCGTCTTCGTTCTCCT-3′

The functional GABAAR is a heteropentamer assembled from two α, two β and one γ (or δ, ϵ, π, θ) subunits (Rudolph et al.,2001; Watanabe et al.,2002). Of six α subunits (α1–6), we failed to detect α5 and α6 at any stage. α2 and α3 subunits were abundantly expressed during embryonic stages (E15.5–P0) in contrast to α1, which is mostly postnatal (Figs. 1, 2A). α4 had a steady elevated expression beyond E16.5 (Fig. 1). Of the three known β subunits, β1 displayed low expression throughout all studied stages with a moderate elevation at postnatal day (P) P0 (Fig. 1). β3, by far the most abundant subunit at embryonic stages was dramatically reduced after P0, while β2 was low at E14.5 and transiently up-regulated at E16.5–P7. All three γ and the δ subunits were expressed from E14.5 (Fig. 1). The γ1 was predominantly prenatal and almost undetectable beyond P0, while γ2, γ3 and δ were present at all studied stages (Fig. 1,2A). γ3 is by far the most abundant of all γ subunits, which contrasts its highly restricted expression in the brain (Laurie et al.,1992; Wisden et al.,1992; Pirker et al.,2000). However, this is not followed by any of the GABA-binding β subunits (Fig. 1; 2A) suggesting that the relative levels of active GABAAR in the postnatal lens may be determined by the availability of β2/3 subunits. We failed to amplify the retina-specific GABACR subunits ρ1 and ρ2 at any stage, which demonstrates that all RT-PCR products shown in Figure 1 are result of strictly lens-specific mRNA amplification.

The functional GABABR is an obligatory heterodimer composed of GABABR1 harboring the GABA binding site and GABABR2, responsible for dimer formation (Bettler et al.,2004; Couve et al.,2004). In brain and peripheral tissues GABABR1 subunit is much more abundant than GABABR2 (Thuault et al.,2004). Similar correlation was observed at E15.5–E17.5 lens (Fig. 1; 2A). The consistent high expression of GABABR1 and extremely low level of GABABR2 in the postnatal lens may suggest the existence also of an alternative assembly of GABABR1 with other protein partners (Calver et al.,2000; Balasubramanian et al.,2004).

The four known membrane GABA transporters (GAT1–4) expressed in both glia and neurons can mediate GABA re-uptake from the extracellular space, but can also release GABA in exchange for 2Na+ and 1Cl (Jursky and Nelson,1996; Richerson and Wu,2003; Wu et al.,2007). The vesicular transporter VGAT mediates GABA (and glycine) transport into synaptic or synaptic-like vesicles used for GABA release from GABAergic cells of the nervous system and periphery (Chaudhry et al.,1998; Gammelsaeter et al.,2004; Wojcik et al.,2006). All GABA transporters showed abundant expression in the developing lens, with distinct patterns (Fig. 1, 2A). GAT1–3 were low at early stages, then GAT2/3 increased after E16.5, while GAT1 was transiently elevated at E17.5–P0. GAT4 was highly and nearly uniformly expressed throughout all stages. VGAT showed a striking profile being much more abundant before birth, with a transient peak at E16.5, but hardly detectable after birth (Fig. 1, 2A).

The striking similarities between the expression patterns of different GABAergic components prompted their grouping according to developmental stage (Fig. 2A). During embryonic stages, mostly GAD65 and the EGAD are expressed (Kwakowsky et al.,2007) and this was paralleled by predominant expression of VGAT and the membrane GAT4 (Fig. 2A). The prevalent configuration of the GABAAR at this stage is α2,3,432,3. Most of the GABA signaling components are greatly up-regulated in the fetal lens: GAD65 and EGAD continued to predominate (Kwakowsky et al.,2007), paralleled at the beginning by VGAT and GAT4, and later by GAT3 and GAT1 (Fig. 1,2A). The predicted GABAAR pentamer during this period is α2,3,42/31,2,3 with all these subunits highly expressed.

P0 was singled out not only because around this stage GABAAR β1 and GAT1 were selectively and transiently up-regulated, but it also coincided with the suppression of embryonic/fetal components (Fig. 2A).

The postnatal stage is clearly characterized by fewer and highly expressed components. Remarkably, these stages are predominated solely by GAD67, while GAD65 is last detected at extremely low levels at P14 (Kwakowsky et al.,2007). The GAD65–GAD67 switch was paralleled by an almost complete down-regulation of VGAT, GABAAR α2, α3, β1, β3, γ1, γ2, δ, GABABR2 with concomitant up-regulation of GAT1–3 and the GABAAR α1 subunit, the most abundant synaptic α subunit in adult brain (Kralic et al.,2002). This observation gained further support from the Spearman correlation curves (Fig. 2B1–B6), which clearly showed high expression correlation between the predominantly prenatal GAD65 and EGAD (rs = 0.94), GAD65-GABAA2 (rs = 0.80), GABAA2-GABAA2 (rs = 0.86) and GAD65-VGAT (rs = 0.67; Fig. 2B1–B4). GAD67 expression was highly correlated with that of GAT2 (rs = 0.99) and GABAAR α1 (rs = 0.96) (Fig. 2B5, B6), all expressed postnatally.

Our data indicate a clear switch from embryonic to postnatal components correlating with the switch from primary to secondary fiber differentiation. The GABAAR subunit switch from the predominant embryonic α2,331,2,3/δ to the predominant postnatal α1,423 should result in changes of the receptor kinetics. For instance, an ontogenetic or compensatory replacement of α2 and/or α3 with α1 results in a change from a slow decay to fast decay receptor kinetics (Okada et al.,2000; Ramadan et al.,2003). Similar compensatory up-regulation of α23 (and down-regulation of β2/3 and γ2) subunits was observed in αmath image mutants (Kralic et al.,2002). This suggests that α1,2,332 subunits are coordinately regulated during development. The total absence of GABAAR α5 and the exceptionally high level of expression of γ3 (Fig. 1,2), encoded by clustered genes of the same locus on mouse Chr 7, indicate that they are oppositely regulated at the transcriptional level throughout lens development. The presence and selective up-regulation of δ subunit, which confers high affinity to GABA, slow rate of desensitization and sensitivity to neuroactive steroids (Mihalek et al.,1999; Peng et al.,2004) at late gestational stages may indicate a specific function in the maturation of primary fibers that may be modulated by endogenous steroids.

In the lens, the GABAAR subunit switch strongly correlates with the switch from mainly apo-GAD65 and EGAD to the constitutively active GAD67 and gradual substitution of VGAT with membrane GABA transporters (GATs). In the adult brain, GAD65 is greatly enriched in the presynaptic terminals and is thought to contribute mainly to the synaptic GABA pool (Battaglioli et al.,2003). Our data suggest that, in the embryonic lens, GABA synthesized by GAD65 is released predominantly by VGAT and binds to GABAAR α2,3 subunit-containing receptors, while GABA synthesized by GAD67 in the postnatal lens is probably released mostly by the membrane GAT2 and binds to α1-containing GABAAR. GABA produced by the enzymatically active embryonic GAD44, which is co-expressed with GAD65 (correlation coefficient 0.94, Fig. 2B1) and is the embryonic counterpart of GAD67 (Szabo et al.,1994), may be released by inversion of GAT, as previously suggested (Szabo et al.,1994; Varju et al.,2001).

The high level of GAT1-4 suggests that the membrane transporters are involved in active GABA uptake/release coupled with Na+/Cl transport independent of GABAR-mediated signaling and thus may participate in maintenance of the lens ion homeostasis and transparency. Especially noteworthy is the robustly expressed GAT4, which is predominantly radial glia-specific in brain (Jursky and Nelson,1996) and is pharmacologically unique among all transporters being able to operate as a Cl channel in the absence of Na+ (Karakossian et al.,2005). Future experiments will be needed to clarify the possible roles of the membrane GABA transporters in the lens ion transport.

GABA Receptors and Transporters Exhibit Spatially Restricted Localization in the Developing Lens

The cellular distribution of VGAT, GATs, and the two main GABAR subtypes were studied by immunohistochemistry using specific antibodies (Figs. 3, 4). At embryonic day (E) 11.5 VGAT, GAT1, GAT3, GABAAβ3 and GABABR showed strong labeling in the lens vesicle (Fig. 3A,D,G,J,M), comprised at this stage of a single epithelial layer (Lovicu and McAvoy,2005). Except for GAT3, which displayed a patchy appearance over the lateral and apical membranes, the staining of all other proteins was identical and greatly enriched at the apical (luminal) side of the lens epithelial cell monolayer (Fig. 3A,D,G,J,M). At E16.5, the lens is already polarized with the lens epithelium and anterior suture at the anterior pole, and the posterior suture formed by fusion of the posterior fiber tips at the posterior pole (Kaufman,1992; Lang,1999; Ogino and Yasuda,2000; Zampighi et al.,2000). The lens nucleus (Fig. 3E) is composed exclusively of primary lens fiber cells at various stages of final differentiation and is covered superficially by elongating secondary fibers. At this stage, staining with all antibodies was weak in the lens nucleus and much stronger in the equatorial secondary fibers and overlaying lens epithelium (Fig. 3B,E,H,K,N). GABAAR, GABABR, and VGAT were greatly enriched in the apical/basal membranes of both epithelium and fiber cells showing especially prominent staining in the posterior sutures (Fig. 4A,B,G–J). GAT1 and GAT3 showed similar predominantly apical/basal localization in the lens epithelium, but their patterns in the fibers differed significantly. GAT1 was detected in both apical/basal and lateral membranes (Fig. 4C,D), while GAT3 showed a unique pattern, being localized in apical and lateral membranes, but not in the basal tips of fibers forming the posterior suture (Fig. 4E,F).

thumbnail image

Figure 3. Cellular localization of gamma-aminobutyric acid (GABA) transporters and receptors in the developing mouse lens. Coronal cryosections of embryonic day (E) 11.5, E16.5, and postnatal day (P) 0 mouse lenses stained with antibodies specific for vesicular GABA transporter (VGAT), membrane GABA transporter-1 (GAT1), GAT3, GABAAβ3, and GABABR2. A,D,G,J,M: At E11.5, labeling is localized in the lens epithelial cells comprising the lens vesicle (LV) and is greatly enriched at their luminal (apical) and to a lesser extent at the basal ends (arrows in D). GAT3 is also localized at the lateral membranes conferring a patchy appearance of the LV (G). B,E,H,K,N: Both epithelial cells and the tips of secondary fiber cells of the equatorial region exhibit strong labeling for GABA receptors (GABAR) and GABA transporters at E16.5 (arrows). Primary fibers forming the lens nucleus (LN) were stained for both GAT1 and GAT3 (E,H, empty arrowheads in E). Compared with E16.5, at P0 GAT1 showed increased expression in the secondary fibers, but was undetectable in the lens nucleus, composed of primary lens fibers (F vs. E). C,F,I,L,O: GAT1 also exhibited more expanded expression in the cortical region containing secondary fibers compared with VGAT, GAT3, GABAAR, and GABABR2 (arrowheads). VGAT and GAT3 expression was decreased and was absent from the lens nucleus compared with E16.5 (B,H vs. C,I). EQ, equatorial region of the lens; L, lens; LE, lens epithelial cells; LN, lens nucleus. LV, lens vesicle; R, retina. Scale bars = 50 μm in A–O.

Download figure to PowerPoint

thumbnail image

Figure 4. A–J: Subcellular localization of the gamma-aminobutyric acid (GABA) receptors and transporters in the embryonic day (E) 16.5 lens: high magnification images of coronal cryosections of E16.5 embryos representing part of the equatorial lens region containing elongating secondary fiber cells (A,C,E,G,I) and posterior lens pole with the forming suture (B,D,F,H,J). Immunoreactivity for GABAAβ3 and GABABR2 receptor subunits and vesicular GABA transporter (VGAT) is predominantly localized to the apical/basal membranes of lens epithelial cells (white arrowheads in A,G,I), apical ends of equatorial fibers (arrows in A,G,I) and their basal tips forming the posterior lens suture (arrows in B,H,J). Membrane GABA transporter-1 (GAT1) and GAT3 are also seen in the lateral secondary fiber membranes (yellow arrowheads in C,E). Note that GAT1 staining shows less polarized pattern compared with VGAT in the epithelium and apical fiber tips (C vs. A), but is enriched at the posterior suture—at the ends of elongating fibers (arrows in D). GAT3 shows a polarized apical/basal expression in the epithelium (arrowheads in E) and enrichment at the apical tips of equatorial fibers (arrow in E), but is not detectable at the posterior suture (F). C and I are confocal images. Scale bars = 50 μm in A,C,E,G,I, 20 μm in B,D,F,H,J.

Download figure to PowerPoint

The E16.5 expression pattern was virtually preserved through P0, except for GAT1, which spanned a larger expression domain compared with E16.5 (Fig. 3F), in accordance with the elevated expression of the transcript at this stage (shown in Fig. 1). Both GAT1 and GAT3 were abundantly expressed at the lateral fiber membranes (Fig. 3F,I). In comparison, VGAT was highly enriched at the apical/basal membranes but was never detected on the lateral surfaces (compare Fig. 3C to F,I) and this pattern of staining was also followed by the two GABA receptors (Fig. 3L,O).

The spatiotemporal expression of GAD, GABA, VGAT, and GABAR in the developing lens show striking similarities with the expression patterns of some typical neuronal synaptic vesicular transport-related proteins like synapsin, synaptotagmin, synaptophysin (Frederikse et al.,2004; Min et al.,2004), and Snap-25 (Wride et al.,2003), which may suggest the existence of a GABA signaling similar to the synaptic GABA transmission in the nervous system. Like synapsin, VGAT, GABAAR, and GABABR are present in clusters at the tips of elongating fibers and apical membranes of the lens epithelium reminiscent of focal regions where vesicles interact with the fiber cell surface.

GABA Is Produced in Epithelial and Fiber Cells and Induces Rise of [Ca2+]i in Equatorial Cells of Intact Newborn Lenses

In the P0 mouse lens, GABA was predominantly localized in the lens epithelium and fiber cells with the strongest labeling found at the equatorial region and the forming sutures (Fig. 5, Ia, c, g1–g3), a pattern entirely consistent with the reported expression of the two GAD forms (Kwakowsky et al.,2007). An inward gradient of GABA staining in the lens cortex and nucleus follows the progressive fiber maturation and denucleation: the signal declined toward the border of secondary/primary fibers as secondary fibers loose their nuclei (Fig. 5, Ic, g1) but became strong again in external, still nucleated primary fibers (Fig. 5, Ic–f; g1–g3). The nuclei followed a concomitant change in shape and chromatin compaction (Fig. 5, Ic–f). Our results show that GAD (Kwakowsky et al.,2007) and GABA synthesis is turned off shortly before denucleation of primary and secondary fibers.

thumbnail image

Figure 5. Gamma-aminobutyric acid (GABA) expression and function in modulation of [Ca2+]i in the neonatal lens. I: Staining with GABA antibody of the neonatal mouse lens (a, c, e- Cy3; g1–g3-DAB). Nuclei were visualized by DAPI (4′,6-diamidine-2-phenylidole-dihydrochloride) staining (a,c,d,f). GABA immunoreactivity is visible predominantly in equatorial secondary fibers possessing intact nuclei (arrowhead in c,d) and in primary fibers with nuclei in the process of disintegration (arrow in c,d). The area between the dashed lines in e and f includes fiber cells undergoing denucleation and decrease of GABA staining. The border between primary and secondary fibers (arrowheads in g1) and the lens sutures (arrows in g1–g2) are well demarcated by strong GABA staining as is the equatorial lens region containing the newly generated secondary fibers (hollow arrowhead in g3). A P0 lens loaded with Fluo-4/AM (b). II: Images from a representative recording of the fluorescence emission in three cells (shown in D in three different colors) of the equatorial region of a Fluo-4/AM loaded newborn mouse lens in buffer alone (A), at the addition of 1 mM GABA (B) and after GABA washout (C). The cells responding to GABA are circled in A, B, and C in colors matching the respective recordings shown in D. arb.U, arbitary units; EQ, equator; F, fluorescence; LE, lens epithelium; t, time. Scale bars = 100 μm in a,b, 50 μm in c–f,g1, 20 μm in g2, g3.

Download figure to PowerPoint

To test the functionality of the GABA signaling, we performed Ca2+ imaging in intact lenses using laser-scanning microscopy and a specially designed perfusion system (Fig. 5, IIA–D). GABA evoked an increase in fluorescence intensity mostly in cells of the equatorial region (Fig. 5, IIA–EQ) that are undergoing transition from epithelial to fiber cells. Usually, several equatorial cells reacted nearly simultaneously with a moderate increase in fluorescence. Application of KCl after GABA washout induced further increase in fluorescence demonstrating the viability of the responding cells.

The responding cells in our studies may correspond to the population of chicken annular pad cells responding to muscarinic acetycholine receptor stimulation with a [Ca2+]i rise (Oppitz et al.,2003). We suggest that GABA and Ach may operate in concert as described previously for the chick retinal ventricular zone (Pearson et al.,2002) where the two neurotransmitters act synergistically to induce [Ca2+]i rise. Neurotransmitter coexpression and corelease is a common feature also of undifferentiated neuronal progenitors that may provide for compensatory mechanisms during early development (Demarque et al.,2002; Owens and Kriegstein,2002a).

Possible Functions of the GABA Signaling in the Developing Lens

Structurally, the molecular components of the GABA signaling are co-expressed in the germinative and migratory epithelial cells and immature lens fibers (Fig. 6), suggesting that GABAR-mediated paracrine signaling may potentially modulate the cell proliferation, migration and fiber elongation through modulation of the [Ca2+]I. An independent function in the ion exchange, regulation of the ion homeostasis and cell volume (Schousboe et al.,2004), can be postulated for the membrane GATs on the basis of their abundant expression, specific temporal regulation, and selective targeting. Finally, the lens affords an ideal model system to study the underlying molecular mechanisms and downstream effects of GABA signaling that can be addressed in viable genetically modified mouse models.

thumbnail image

Figure 6. Expression patterns of different gamma-aminobutyric acid (GABA) signaling components in the neonatal mouse lens: a schematic presentation. GAD65 is localized almost exclusively in the primary lens fibers comprising the lens nucleus, while GAD67 is mostly found in the elongating secondary fibers and lens epithelium cells (Kwakowsky et al.,2007). GABAA and GABAB receptor subunits, vesicular GABA transporter (VGAT), and membrane GABA transporter-1 (GAT1) show preferential expression in the apical/basal membranes (including the lens sutures) of both lens epithelial cells and elongating secondary fibers at the equatorial region. GAT3 is expressed in the apical/basal membranes of epithelial cells and the apical and lateral membranes of fiber cells, but not expressed in the basal tips of elongating secondary fibers. GAT2 may be predominantly localized to lateral membranes (see the Discussion section). Dlx5, an upstream regulator of GAD genes is expressed in the nuclei of the epithelial cells and fibers expressing GAD and GABA.

Download figure to PowerPoint

EXPERIMENTAL PROCEDURES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS AND DISCUSSION
  5. EXPERIMENTAL PROCEDURES
  6. Acknowledgements
  7. REFERENCES

Animals

All studies were carried out on wild-type C57Bl6 or FVB/Ant (Errijgers et al.,2007) embryonic and postnatal mice housed in the Medical Gene Technological Unit of the Institute of Experimental Medicine in Budapest, Hungary. All experiments with animals were conducted in compliance with NIH (NIH Publication #85-23, 1985) and EC (86/609/EEC/2) guidelines and approved by in-house and national committees.

RNA Preparation and PCR Amplification

Lens tissue dissected free of cilliary epithelium or retina in ice-cold DEPC (diethyl pyrocarbonate, Sigma-Aldrich)-treated phosphate buffered saline (PBS) was stored in RNA later (Sigma-Aldrich). Total RNA was isolated using TRI-Reagent (Sigma-Aldrich) according to the manufacturer's protocol. Three micrograms of total brain or lens RNA was reverse transcribed using Minus First Strand cDNA Synthesis Kit (Fermentas) with random hexamer primers followed by PCR amplification as follows: initial denaturation for 5 min at 95°C, followed by 30–35 cycles of 1 min at 94°C, 1–3 min at 55–68°C, 1–2 min at 72°C, and final extension for 10 min at 72°C. As control, the same quantity of total RNA was amplified under identical conditions. The housekeeping gene β-actin was co-amplified for 16–20 cycles either in the same tube or separately at identical conditions for quantification.

Specific primers from different exons were used for each cDNA (Table 1) based on previously published sequences or designed based on Ensemble sequence (www.ensembl.org). Amplification products obtained with novel primers were cloned and sequenced to verify the specificity of the PCR reaction (data not shown). Quantification of data was performed exactly as in (Kwakowsky et al.,2007). For each experimental group, relative gene expression levels were expressed as a % of maximum level (100%). To find possible co-regulation among the genes involved in the GABA-signaling pathway, we performed nonlinear pair-wise correlation tests of relative gene expressions spanning all studied developmental stages (Kotlyar et al.,2002). Data were analyzed with Spearman correlation test using GrapPad Prism 5.0 software (GraphPad Software Inc. CA). In all cases, P < 0.05 was considered significant.

Immunohistochemistry

Embryos were derived from timed-pregnant females mated overnight to wild-type males. The day of the vaginal plug was considered embryonic day 0.5 (E0.5). For staining with anti-GABA antibody, tissue was fixed in 4% (w/v) paraformaldehyde (PFA), 0.1% glutaraldehyde in PBS, pH 7.4. For all other antibodies fixation was with 4% PFA in PBS overnight at 4°C. The tissue was equilibrated in 20% sucrose-PBS at 4°C, embedded in Cryo-gel (Instrumedics) and sectioned at 25 μm on a Microm HM550 (Microm).

Cryostat sections were dried, post-fixed in ice-cold 4 % PFA for 10 min followed by 2 × 5-min washes with TBS (Tris-buffered saline), pH 7.5. Sections were subsequently blocked in 1% bovine serum albumin (BSA)-TBS containing 0.02% saponin (Sigma-Aldrich) for 1 hr at room temperature. The sections were incubated overnight at 4°C with the following primary antibodies: rabbit anti-GABA (Sigma-Aldrich, 1:5,000), rabbit anti-GABAAβ3 (Novus Biologicals, 1:1,000), rabbit anti-GABABR2 (BD Biosciences, 1:500), guinea pig anti-VGAT (Merck, 1:5,000), and anti-mouse GAT1 (a kind gift from Nathan Nelson, Tel Aviv University, 1:1,000), rabbit anti-GAT3 (Abcam, 1:1,000) diluted in 1% BSA-TBS containing 0.005% saponin (TBS-S). After 3 × 10-min washes with TBS-S, the sections were incubated for 2 hr with anti-rabbit (Vector Laboratories, 1:500) or anti-guinea pig (Fisher Scientific, 1:500) biotinylated secondary antibodies in 1% BSA-TBS-S, followed by 2 × 10-min washes in TBS-S and 1 × 10 min in TBS. Finally, streptavidin-conjugated Cy3 (Jackson ImmunoResearch) diluted at 1:5,000 in 0.1% BSA-TBS was applied for 90 min at room temperature, followed by washes in TBS and finally in PBS. Sections were cover-slipped and examined under Zeiss Axioscop-2 microscope equipped with AxioCam HRc digital camera using AxioVision 4.6 software (Carl Zeiss).

[Ca2+]i Imaging Studies

Freshly isolated lenses from neonatal mice were washed in Hepes-buffered saline solution (HBSS) containing (in mM): 137 NaCl, 5 KCl, 20 HEPES, 10 glucose, 1.4 CaCl2, 3 NaHCO3, 0.6 Na2HPO4, 0.4 KH2PO4 at pH 7.4. Intact lenses were loaded for 30 min at 37°C with 5 μM Fluo-4/AM in the presence of 0.1% pluronic acid F-127 (both from Molecular Probes). Subsequently, the lenses were secured by means of a mesh to the bottom of a plastic perfusion chamber mounted on a laser-scanning microscope (LSM, Olympus FV-500). Recordings were at 488/515 nm excitation/emission. Image acquisition was set at one image/1.8 sec. Cells were superfused with HBSS buffer at a rate of 2 ml/min using a peristaltic pump (Gilson) initially for 2 min with HBSS to wash out excess dye. Additional 2 min perfusion with HBSS was performed to obtain a steady baseline, then 1 mM GABA in HBSS was applied for 2 min, followed by 2 min HBSS and finally HBSS-20 mM KCl. Analysis of fluorescence intensity was performed using the imaging software of the Olympus FV500 (FluoView).

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS AND DISCUSSION
  5. EXPERIMENTAL PROCEDURES
  6. Acknowledgements
  7. REFERENCES

The authors thank Katalin Döme and the staff of the Medical Gene Technology Division of the Institute of Experimental Medicine for excellent technical assistance.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS AND DISCUSSION
  5. EXPERIMENTAL PROCEDURES
  6. Acknowledgements
  7. REFERENCES