The type III inositol 1,4,5-trisphosphate receptor is phosphorylated by cAMP-dependent protein kinase at three sites

Abstract

IP₃ (inositol 1,4,5-trisphosphate) receptors form tetrameric, IP₃-gated Ca²⁺ channels in endoplasmic reticulum membranes, and are substrates for several kinases, including PKA (cAMP-dependent protein kinase). Activation of PKA has been reported to either enhance or inhibit type III IP₃ receptor Ca²⁺-channel activity, but, as yet, the sites of phosphorylation remain unknown. Here, we reveal that PKA phosphorylates the type III IP₃ receptor at Ser⁹¹⁶, Ser⁹³⁴ and Ser¹⁸³², and that, intriguingly, each site is located close to a putative surface-exposed peptide loop. Furthermore, we demonstrate that Ser⁹³⁴ is considerably more susceptible to PKA-dependent phoshorylation than either Ser⁹¹⁶ or Ser¹⁸³². These findings define the sites at which the type III IP₃ receptor is phosphorylated by PKA, and provide the basis for exploring the functional consequences of this regulatory event.

INTRODUCTION

IP₃ (inositol 1,4,5-trisphosphate) receptors (IP₃Rs) are a family of proteins which form Ca²⁺ channels in endoplasmic reticulum membranes, and which govern Ca²⁺ release from this organelle by opening in response to IP₃ and Ca²⁺ binding [1]. In mammals, type I, II and III IP₃Rs (IP₃R1, IP₃R2 and IP₃R3) are highly homologous, and have the same overall structure, with an N-terminal ‘ligand-binding domain’, a C-terminal ‘channel domain’, and an intervening ‘coupling domain’ containing putative binding sites for several regulatory factors [1,2]. Despite the structural similarities, expression of IP₃R1 to IP₃R3 varies considerably between tissues [3–5]. IP₃R1, for example, is very widely expressed, with particularly high levels in neuronal cells, whereas IP₃R2 and IP₃R3 are expressed more sporadically, but with considerable abundance, in the liver and pancreas respectively [3–7].

There are also subtle, but clear, differences in the properties of the three IP₃R types. IP₃R3, for example, has a lower affinity for IP₃ [8,9] and ATP [10,11] than IP₃R1 or IP₃R2, but is more sensitive to Ca²⁺ [11]. Furthermore, functional differences between the three receptor types have been attributed to their differential phosphorylation by PKA (cAMP-dependent protein kinase) [12–14]. To date, however, only the effects of PKA on IP₃R1 have been characterized. Two serine residues (1588 and 1755) are phosphorylated [15–17], and these modifications appear to enhance Ca²⁺ channel activity in intact cells [18]. In contrast, whereas phosphorylation of IP₃R2 [12,19] and IP₃R3 [12–14,20] may modulate Ca²⁺ channel activity, reported results have often been contradictory; for example, for IP₃R3, PKA has been suggested to have either stimulatory [13] or inhibitory [14] effects. Furthermore, nothing is currently known about the sites at which these receptors are phosphorylated.

IP₃R3 has been implicated in a variety of Ca²⁺-dependent physiological processes. For example, it may play an important role in generating the Ca²⁺ signals that modulate glucose-stimulated insulin secretion in pancreatic β cells [7], and initiate exocytosis and electrolyte secretion in pancreatic acinar cells [20], and therefore its phosphorylation could have important physiological consequences. To provide a basis for fully understanding the functional consequences of IP₃R3 phosphorylation, we set out to define the sites at which it is phosphorylated. We report here that IP₃R3 is phosphorylated by PKA at serine residues 916, 934 and 1832, and that, intriguingly, these sites are close to putative surface-exposed peptide loops.

EXPERIMENTAL

Materials

[³²P]P_i (H₃PO₄, carrier-free) and [γ-³²P]ATP (6000 Ci/mmol) were from PerkinElmer, and the PKA catalytic subunit was from Promega. All other chemicals were from Sigma. TL3, a mouse monoclonal antibody directed against N-terminal amino acids 22–230 of IP₃R3, was from BD Biosciences. CT1 and CT3, affinity-purified rabbit antisera directed against the C-termini of IP₃R1 and IP₃R3 respectively, were prepared as described previously [3].

Mutagenesis

The constructs pCWI (which encodes mouse SI+/SII+IP₃R1) and pCI₃ (which encodes rat IP₃R3; kindly given by Dr Graeme Bell, University of Chicago, IL, U.S.A.), have been described previously [7,17]. Mutagenesis of Ser⁹¹⁶, Ser⁹³⁴ and Ser¹⁸³² to alanine was achieved by PCR and pairs of mutagenic primers. Briefly, pCI₃ was mutated to introduce alanine at positions 934, 916 and 1832 using the primer pairs 5′-CCATGGTGCTATCGCGAAAGCAGGCTGTATTTGGTGC-3′ and 5′-GCACCAAATACAGCCTGCTTTCGCGATAGCACCATGG-3′, 5′-GCAAGAACGTGCGCAGGGCCATCCAGG-3′ and 5′-CCTGGATGGCCCTGCGCACGTTCTTGC-3′, and 5′-CCACCAAAGGACGCGTGTCCGCCTTCTCCATGC-3′ and 5′-GCATGGAGAAGGCGGACACGCGTCCTTTGGTGG-3′ respectively. These primer pairs also introduced NruI, FspI and MluI sites, respectively, to facilitate screening, and correct introduction of the desired mutations was confirmed by sequencing.

Phosphorylation of IP₃Rs in intact cells

Wells of HEK (human embryonic kidney) cells were transfected [17], incubated with approx. 100 μCi of [³²P]P_i for 2 h in 750 μl of serum-free, phosphate-free Dulbecco's modified Eagle's medium, and then stimulated to induce IP₃R phosphorylation. Cells were then solubilized with ice-cold, phosphatase-inhibitor-supplemented lysis buffer [17], and IP₃Rs were immunoprecipitated overnight at 4 °C with Protein A–Sepharose CL-4B beads and IP₃R1-specific CT1 or IP₃R3-specific CT3 [3]. Immune complexes were then washed three times and finally resuspended in 2× gel loading buffer [3].

Phosphorylation of purified IP₃R3

Transfected cells were solubilized with ice-cold, phosphatase-inhibitor-supplemented lysis buffer, IP₃R3s were immunoprecipitated, and in vitro phosphorylation of the purified receptors was then performed in 200 μl of phosphorylation buffer [12] supplemented with approx. 20 μCi of [γ-³²P]ATP, 5 μM non-radioactive ATP and 40 units of PKA catalytic subunit for 15 min at 30 °C. The immune complexes were then washed three times with ice-cold phosphorylation buffer supplemented with 1 mM ATP, and finally resuspended in 2× gel loading buffer.

Chymotrypsin treatment of purified IP₃R3

IP₃R3 immune complexes were washed twice with ice-cold 20 mM Tris/HCl, pH 8.0, containing 120 mM KCl and 1 mM EDTA, and were then resuspended in the same buffer supplemented with 1 μg/ml α-chymotrypsin (Type I-S) and incubated at 25 °C for the indicated times, before the addition of 4× gel loading buffer.

Electrophoresis, staining and autoradiography

IP₃R1 (≈270 kDa) and IP₃R3 (≈260 kDa) were electrophoresed and stained with Coomassie Blue [3], and associated radioactivity was assessed initially by autoradiography of dried gels, and then quantified by excision and scintillation counting of the radioactive bands [17].

RESULTS AND DISCUSSION

Identification of phosphorylation sites on IP₃R3

To identify sites phosphorylated by PKA, wild-type and mutant IP₃R3s were expressed in HEK cells and endogenous PKA was activated by exposing the cells to a maximally effective dose of PGE₁ (prostaglandin E₁), an approach we have used previously to analyse PKA-mediated IP₃R1 phosphorylation [17]. We mutated 33 serine and 18 threonine residues at sites resembling the PKA consensus sequences R(R/K)X(S/T), RX₂(S/T) or RX(S/T) [21]. As shown in Figure 1(A), we found that mutation of Ser⁹¹⁶, Ser⁹³⁴ and Ser¹⁸³² suppressed PGE₁-induced phosphorylation of IP₃R3. In comparison with wild-type receptor (IP₃R3^SSS; lane 4), phosphorylation of IP₃R3 mutated at Ser⁹¹⁶ (IP₃R3^ASS; lane 6) or Ser¹⁸³² (IP₃R3^SSA; lane 10) was reduced by approx. 25%, whereas phosphorylation of IP₃R3 mutated at Ser⁹³⁴ (IP₃R3^SAS; lane 8) was reduced by approx. 50%. Furthermore, phosphorylation of IP₃R3 mutated at both Ser⁹¹⁶ and Ser⁹³⁴ (IP₃R3^AAS; lane 12) reduced phosphorylation by approx. 75%, and mutation of all three serine residues (IP₃R3^AAA; lane 14) completely abolished PGE₁-induced phosphorylation. Therefore, under conditions of maximal PKA activation, Ser⁹¹⁶, Ser⁹³⁴ and Ser¹⁸³² represent the entire complement of sites phosphorylated by PKA. The fidelity of this mutagenic approach is demonstrated by the fact that IP₃R3 phosphorylation was unaffected by mutations at several other consensus phosphorylation sites (e.g. at Thr²²⁰¹ and Ser²⁴⁸³), and at sites close to Ser⁹¹⁶, Ser⁹³⁴ and Ser¹⁸³² (e.g. at Ser¹⁸³¹, which lies adjacent to Ser¹⁸³², and at Ser⁹²⁵ and Ser⁹³⁰, which lie between Ser⁹¹⁶ and Ser⁹³⁴; results not shown).

Figure 1. Site specificity of PKA-mediated IP₃R3 phosphorylation.

(A) ³²P-labelled cells transfected with empty vector (lanes 1 and 2), or IP₃R3 cDNAs encoding wild-type receptor (IP₃R3^SSS; lanes 3 and 4), or receptors with mutations to alanine at Ser⁹¹⁶ (IP₃R3^ASS; lanes 5 and 6), Ser⁹³⁴ (IP₃R3^SAS; lanes 7 and 8), Ser¹⁸³² (IP₃R3^SSA; lanes 9 and 10), Ser⁹¹⁶ and Ser⁹³⁴ (IP₃R3^AAS; lanes 11 and 12), or Ser⁹¹⁶, Ser⁹³⁴ and Ser¹⁸³² (IP₃R3^AAA; lanes 13 and 14) were exposed to 1 μM PGE₁ for 5 min as indicated, and IP₃R3s were immunoprecipitated, electrophoresed and assessed for staining capacity (upper panel) and associated radioactivity (lower panel). Exogenous IP₃R3 phosphorylation (histogram; means±S.E.M., n≥3) was calculated by subtracting radioactivity associated with endogenous IP₃R3s (lanes 1 and 2) from that in lanes 3–14, and is expressed as a percentage of the radioactivity associated with IP₃R3^SSS in the presence of PGE₁ (lane 4). (B) ³²P-labelled cells transfected with IP₃R3 (lanes 1–3) or IP₃R1 (lanes 4 and 5) cDNA were exposed to 1 μM PGE₁ (P) or 10 μM forskolin (F) for 5 min as indicated, and IP₃R phosphorylation was visualized as in (A). (C) IP₃R3s were immunoprecipitated from non-stimulated transfected cells and were phosphorylated with PKA catalytic subunit in vitro. The approx. 200–300 kDa region of gels are shown in (A), (B) and (C), and represent typical data from at least two independent experiments.

We next sought to validate our results by demonstrating that the data in Figure 1(A) represent efficient phosphorylation of the majority of exogenous IP₃R3s, rather than modification of an insignificant fraction of receptors. Figure 1(B) demonstrates that 1 μM PGE₁ elicited maximal IP₃R3 phosphorylation, since its effects (lane 3) were equal to those of 10 μM forskolin (lane 2), which maximally activates adenylate cyclase and PKA in HEK cells [22]. Importantly, IP₃R3 (lane 2) was phosphorylated with a stoichiometry similar to that seen for IP₃R1 (lane 5), which contains two well-characterized PKA phosphorylation sites and which, under these experimental conditions, incorporates at least 1 mol of phosphate/mol of receptor [17]. Thus IP₃R3s are phosphorylated with high efficiency under these conditions.

To explore IP₃R3 phosphorylation further, immunopurified receptors were phosphorylated with the catalytic subunit of PKA in vitro (Figure 1C). Surprisingly, whereas mutation of Ser⁹³⁴ reduced PKA-mediated phosphorylation by approx. 80% (IP₃R3^SAS; lane 4), mutation of either Ser⁹¹⁶ (IP₃R3^ASS; lane 3) or Ser¹⁸³² (IP₃R3^SSA; lane 5) had little or no effect on PKA-mediated phosphorylation, indicating that Ser⁹³⁴ is the only site efficiently phosphorylated in vitro. Thus IP₃R3, like IP₃R1 [17], is phosphorylated differently in vitro than in intact cells. This discrepancy may reflect the absence of regulatory factors in vitro, or alteration of IP₃R3 structure following immunopurification, and reinforces the need for caution when assessing protein phosphorylation on the basis of in vitro experiments.

Conservation of phosphorylation sites in IP₃R3

Analysis of the sequences surrounding Ser⁹¹⁶, Ser⁹³⁴ and Ser¹⁸³² (Figure 2) reveals that these serine residues, and the adjacent amino acids likely to specify PKA-mediated phosphorylation, are almost completely conserved between rodent and human IP₃R3s, indicating that the results obtained herein for rat IP₃R3 can be extrapolated to human receptors. Furthermore, the 916/934 region is very well conserved between IP₃Rs 1 to 3, whereas the 1832 region is not conserved at all. Intriguingly, despite the overall conservation of the 916/934 region, the two PKA consensus sequences in IP₃R3 are not retained in IP₃R2 or IP₃R1. This is consistent with existing knowledge of IP₃R1 phosphorylation by PKA, which occurs only at Ser¹⁵⁸⁸ and Ser¹⁷⁵⁵ [15–17], and suggests that IP₃R2 will not be phosphorylated in the 916/934 region. It is also interesting to note that, in IP₃R1, two sites particularly sensitive to proteolysis by trypsin are within the 916/934 region [24] (Figure 2, arrows). Furthermore, the 1832 region lies immediately N-terminal to an additional trypsin cleavage site in IP₃R1 (within approx. 25 amino acids) [24]. These protease-sensitive sites in IP₃R1 have been proposed to be surface-exposed loops, as opposed to tightly folded regions that are resistant to proteolysis [10,24]. The corresponding regions in IP₃R3, therefore, may also be exposed loops.

Figure 2. Amino acid sequences surrounding Ser⁹¹⁶, Ser⁹³⁴ and Ser¹⁸³².

Rat [7], mouse [23] and human [5,6] IP₃R1, IP₃R2 and IP₃R3 amino acid sequences were aligned and examined for regions of identity (shaded areas). The IP₃R1 and IP₃R2 sequences corresponding to the 1832 region of IP₃R3 are not shown, as these possessed no significant identity. Ser⁹¹⁶, Ser⁹³⁴ and Ser¹⁸³² of IP₃R3 are emboldened and underlined, and the sites at which trypsin cleaves IP₃R1 [24] are indicated by arrows.

Chymotrypsin treatment of IP₃R3

To examine whether the 916/934 region of IP₃R3 is a protease-sensitive, surface-exposed loop, we immunoprecipitated phosphorylated IP₃R3s from HEK cells and subjected them to partial proteolysis with chymotrypsin. It has been shown previously that chymotrypsin cleaves baculovirus-generated IP₃R3 into four pieces (fragments I–IV) that migrate at 105, 70, 35 and 95 kDa respectively [10]. Our data confirm this finding, since we obtained fragments at 105 kDa, 70 kDa and 30/32 kDa when wild-type or mutant IP₃R3s were cleaved (Figure 3A). Surprisingly, a 95 kDa fragment (IV) was not detected by Coomassie Blue staining (Figure 3A), apparently because it was rapidly degraded; low amounts of this fragment, however, were detectable at early time points by immunoblotting with the C-terminus-directed antibody CT3 [3] (results not shown). Similarly, the 105 kDa fragment (I) was confirmed to be the receptor's N-terminus by immunoblotting with TL3 (results not shown). Note that, although the absolute amounts of fragments I, II and III were slightly higher for the mutant IP₃R3s (lanes 5–8) than for wild-type IP₃R3 (lane 4) (due to variation in receptor expression level), the relative abundance of fragments I, II and III appeared to be constant for each receptor, indicating that the mutations do not have a significant effect on proteolysis. Analysis of the extent to which the fragments were phosphorylated (Figure 3B) showed that, for wild-type IP₃R3 (IP₃R3^SSS; lanes 1–4), fragments I, II and III were all phosphorylated (lanes 2–4), that mutation of Ser⁹¹⁶ (IP₃R3^ASS; lane 5) abolished phosphorylation of fragment I, that mutation of Ser⁹³⁴ (IP₃R3^SAS; lane 6) abolished phosphorylation of fragment II, that mutation of Ser¹⁸³² (IP₃R3^SSA; lane 7) abolished phosphorylation of fragment III, and that mutation of the three serine residues (IP₃R3^AAA; lane 8) abolished phosphorylation of fragments I, II and III. Furthermore, and importantly, mutation of any one site did not diminish phosphorylation of the other two sites, indicating that phosphorylation at Ser⁹¹⁶, Ser⁹³⁴ and Ser¹⁸³² are independent events. These data lead to the model shown in Figure 3(C), in which the chymotrypsin cleavage site that separates fragments I and II lies between Ser⁹¹⁶ and Ser⁹³⁴. The exact cleavage site could not be determined, however, as chymotrypsin cuts at the C-terminal side of residues with large hydrophobic side chains [25], and several of these are present between Ser⁹¹⁶ and Ser⁹³⁴. Overall, these data indicate that the 916/934 region is a surface-exposed loop and, since this region is so well conserved among IP₃Rs 1 to 3, suggest that it may be critical to receptor function. It is likely to be especially important in IP₃R3 regulation, as it contains two PKA phosphorylation sites.

Figure 3. Chymotrypsin treatment of phosphorylated IP₃R3.

³²P-labelled cells transfected with IP₃R3 cDNA were exposed to 1 μM PGE₁ for 5 min, and IP₃R3s were immunoprecipitated, digested with 1 μg/ml α-chymotrypsin for the times indicated, and electrophoresed. (A) Coomassie Blue stain showing IP₃R3 fragments in kDa generated upon digestion. (B) Autoradiograph of the stained gel showing radioactivity associated with each IP₃R3 fragment. (C) Schematic demonstrating the proximity of phosphorylation sites (closed circles) to the proposed sites of chymotrypsin-mediated cleavage (arrows) and the fragments (I–IV; size givens in kDa) generated. The transmembrane (TM) region of IP₃R3, spanning amino acids 2204–2516, is shaded.

PKA preferentially phosphorylates Ser⁹³⁴

Since Ser⁹³⁴ contributed more to overall IP₃R3 phosphorylation than Ser⁹¹⁶ or Ser¹⁸³² under conditions of maximal PKA activation (Figures 1A and 3B), we sought to determine whether the three sites differed in their sensitivity to PKA. To do this, we measured the amount of phosphorylation in IP₃R3s containing only Ser⁹¹⁶ (IP₃R3^SAA), Ser⁹³⁴ (IP₃R3^ASA) or Ser¹⁸³² (IP₃R3^AAS) in response to a range of PGE₁ concentrations. Figures 4(A) and 4(B) show that Ser⁹³⁴ was phosphorylated approximately twice as much as that of Ser⁹¹⁶ or Ser¹⁸³² at all doses of PGE₁ and at all times tested, and that all three sites were phosphorylated maximally within 2 min. Furthermore, Figure 4(A) demonstrates that Ser⁹³⁴ was slightly more sensitive to PGE₁ than Ser⁹¹⁶ or Ser¹⁸³², since EC₅₀ values for PGE₁ were approx. 60, 130 and 130 nM for Ser⁹³⁴, Ser⁹¹⁶ and Ser¹⁸³² respectively. Taken together, these data indicate that PKA preferentially phosphorylates Ser⁹³⁴. The reasons for this remain unknown, but could be due to several factors. First, the structure of IP₃R3 could account for the differences in phosphorylation. Ser⁹³⁴ may simply be more accessible to PKA, and thus be more readily phosphorylated than Ser⁹¹⁶ or Ser¹⁸³². Although this may explain why Ser⁹³⁴ is phosphorylated more readily than Ser¹⁸³², it does not explain why Ser⁹³⁴ would be more sensitive to PKA than Ser⁹¹⁶, since they are in such close proximity. A better explanation may be found in the sequences surrounding the phosphorylation sites. Ser⁹³⁴, in the sequence RKQS, falls within the optimal PKA consensus sequence R(R/K)X(S/T), whereas Ser⁹¹⁶, in the sequence RRS, and Ser¹⁸³², in the sequence RVSS, fall within less optimal consensus sequences [21].

Figure 4. Ser⁹³⁴ is preferentially phosphorylated by PKA.

³²P-labelled cells transfected with cDNA encoding IP₃R1 (○), IP₃R3^SSS (●), IP₃R3^SAA (▲), IP₃R3^ASA (■), or IP₃R3^AAS (▼) were exposed to PGE₁ as indicated. (A) Dose-dependence of phosphorylation after 5 min incubations. (B) Time course of phosphorylation. Stimulated IP₃R phosphorylation was calculated by subtracting radioactivity associated with IP₃Rs in the absence of PGE₁ and is expressed as a percentage of the radioactivity associated with wild-type IP₃Rs in the presence of 10 μM (A) or 1 μM (B) PGE₁ (means±S.E.M., n≥3; error bars may be occluded by symbols).

Interestingly, whereas Figure 1(B) shows that wild-type IP₃R1 and IP₃R3 are phosphorylated equally under conditions of maximal PKA activation, Figure 4(A) indicates that IP₃R1 is ≈5-fold more sensitive to PKA than is IP₃R3, since EC₅₀ values for PGE₁ were approx. 15 and 75 nM for IP₃R1 and IP₃R3 respectively. The reason behind this difference remains unknown. It could be due to a type-specific regulatory protein, such as a PKA-anchoring protein, an example of which has been shown to associate with IP₃R1, but not IP₃R3 [26]. Alternatively, it could be due to the fact that only one site (Ser⁹³⁴) in IP₃R3 falls within the optimal PKA consensus sequence R(R/K)X(S/T) [21], whereas both PKA-dependent phosphorylation sites in IP₃R1 (Ser¹⁵⁸⁸ and Ser¹⁷⁵⁵) meet this requirement. Regardless of the reasons, this difference should be taken into account when examining the functional effects of IP₃R phosphorylation in systems that express both receptor types. This is especially important, considering that IP₃R1 and IP₃R3 are often co-expressed [3,4], that heterotetrameric IP₃R channels readily form in intact cells [1,3,4], and that PKA may have different functional effects on IP₃R1 and IP₃R3.

Phosphorylation of IP₃R3 by PKG (cGMP-dependent protein kinase)

Finally, we sought to define the effects of PKG, since PKA and PKG share similar consensus sequences [21] and both can phosphorylate IP₃R1 [16,17] in intact cells. Activation of PKG with 8-bromo-cGMP (Figure 5A) resulted in a slight increase in IP₃R3 phosphorylation, approx. 20% of that seen when PKA was activated with 8-bromo-cAMP. Similarly to PKA, although there was some PKG-dependent phosphorylation on fragments I (Ser⁹¹⁶) and III (Ser¹⁸³²), the majority of phosphorylation occurred on fragment II (Ser⁹³⁴; Figure 5B). The effects of 8-bromo-cGMP were not due to ‘cross-activation’ of PKA, since that does not occur under these conditions [17]. We conclude, therefore, that PKG phosphorylates IP₃R3 predominately at Ser⁹³⁴, but with significantly less efficiency than PKA.

Figure 5. Effects of PKG activation.

³²P-labelled cells transfected with cDNA encoding wild-type IP₃R3 were exposed to 1 mM 8-bromo-cAMP (A) or 8-bromo-cGMP (G) for 30 min, as indicated. (A) IP₃R3 phosphorylation expressed as a percentage of the radioactivity associated with IP₃R3 in the presence of 8-bromo-cAMP (means±S.E.M., n=5). (B) Autoradiograph of phosphorylated IP₃R3 fragments generated as in Figure 3.

Conclusions

In summary, we have determined that PKA phosphorylates IP₃R3 at Ser⁹¹⁶, Ser⁹³⁴ and Ser¹⁸³² in an independent manner, and that each site is located at or close to a putative surface-exposed loop. Further, we have shown that Ser⁹³⁴ is preferentially phosphorylated, that IP₃R1 is ≈5-fold more sensitive to PKA-dependent phosphorylation than IP₃R3, and that PKG phosphorylates IP₃R3, but with significantly less efficiency than PKA. This is the first study to address the sites of cyclic-nucleotide-dependent phosphorylation of IP₃R3, and provides a basis for studying the functional consequences of this regulatory event.