NMR study of the human NCK2 SH3 domains structure determination, binding diversity, folding and amyloidogenesis 6

Chapter A novel nucleolar transcriptional activator ApLLP for long-term memory formation is intrinsically unstructured but functionally active Part of this chapter has been published in: Liu J, Song J A novel nucleolar transcriptional activator ApLLP for long-term memory formation is intrinsically unstructured but functionally active Biochem Biophys Res Commun 2008 Feb 8;366(2):585-91 Epub 2007 Dec 17 PMID: 18078811 146 6.1 Introduction The role of activity-dependent synaptic plasticity in learning and memory is a central issue in neuroscience Changing the strength of connections between neurons is widely assumed to be the mechanism by which memory traces are encoded and stored in the central nervous system (S J Martin, 2000) Numerous results suggest that neural activity could induce an expression of genes, including transcription factors (M.W Jones, 2001; J Hall, 2001; B Bozon, 2003; H Lin, 2003; P Gass, 2004; R.M Weragoda, 2004) Long-term memory requires a new protein and mRNA synthesis, and is associated with the growth of new synaptic connections (Bailey, C.H.,1983; Montarolo, P.G., 1986; Glanzman, D.L., 1990) Transcriptional and translational regulatory factors as well as their signalling pathways that underlie long-term synaptic plasticity will be crucial to consolidate new generated synaptic connections Inducible or modification factors for long-term memory formation have received intense studies in recent years The learning-induced activation of transcription factors, such as CREB, c-fos, EGR-1 and C/EBP, has been well reported (P.J Colombo, 2004) Once the inducing signal has been transmitted to the nucleus through different signalling pathways, regulated transcription factors must be activated In Aplysia, early evidence indicates that the CRE (cAMP-Response Element) was a critical enhancer for this gene induction (Dash, P K., 1990) A later cloning of the Aplysia CREB gene provided a confirmation that this inducible transcription factor is an early element of the cascade of gene activation required for the establishment of long-lasting, synaptic facilitation (Bartsch, D., 1998) In mammals, the situation is complicated by the existence of several alternatively spliced and heterodimerised CREB-like transcription factors, but 147 CREB and CRE-driven transcriptions appear to have a similarly central role (Bourtchouladze, R., 1994; Pittenger, C., 2002) Recently, Hyoung and Bong-Kiunin have identified a new transcription factor, ApLLP (Aplysia LAPS18-like protein), which induces an ApC/EBP expression that is required for long-term synaptic facilitation in Aplysis neurons.( Alberini, C.M., 1994; Hyoung Kim, 2003; Hyoung Kim, 2006) The ApLLP encodes 120 amino acids and has a 57% identity with LAPS18 Most of the ApLLP is localised in the nucleoli, suggesting that NLSs exist in the sequence The PSORT WWW server also confirmed that there are two putative NLSs in both ends of the ApLLP (Hyoung Kim, 1994) Interestingly, when there is overexpressing ApC/EBP in Aplysia sensory neurons, the application of a single pulse of 5-HT which normally induces only short-term facilitation, now induces long-term facilitation In addition, microinjecting dsRNAi gives rise to a blocked, long-term facilitation in a sequence-specific manner These data show that it serves as a molecular switch from short-term to long-term synaptic plasticity (Jin-A Lee, 2001) Bong-Kiunin’s group also used the Aplysia culture system to examine the effect of depolarisation on the ApLLP, and the ability of the ApLLP to increase synaptic strength They found that the ApLLP is an activitydependent transcriptional activator, which switches an STF to an LTF, by inducing ApC/EBP expression (Hyoung Kim, 2006) Although the purified ApLLP was shown to directly bind the CRE of the ApC/EBP promoter using electromobility shift assay (EMSA), its underlying molecular mechanism still remains elusive For example, to date no other protein bindingpartner has been identified for ApLLP Accordingly, structural characterization of 148 ApLLP may offer enlightening insights into the molecular details associated with ApLLP functions Due to the fact that no previous structural characterization has been conducted on both ApLLP and its homologs, in this study we cloned and expressed the entire and the two dissected ApLLP forms and subsequently characterized their structural properties and binding interactions with the CRE DNA element by the use of CD and NMR spectroscopy 6.2 Materials and Methods 6.2.1 Cloning, expression, and purification of ApLLP protein To achieve a high-level protein expression, DNA fragments encoding ApLLP (Hyoung Kim, 2003) with Escherichia coli-preferred codons were obtained by a PCRbased de novo gene synthesis approach, as has been previously described (Wei, Z., 2005), using DNA oligos PRIMER1:5’CAGATGCGCAACGTGAAACGCGAACATTTTGCGAAAAAAGATCTGGATCG CCTGAAACGCCTGGCGAGCAAAGCGCAGGAACTGG-3’; PRIEMR2:5’CGCGCTGGTGCTCGGTTTGTTTTTAATTTCTTCCGCGCTTTTCATGGTCACC ACGTTATCCAGATCCAGTTCCTGCGCTTTGCTC-3’; PRIMER3:5’ACAAACCGAGCACCAGCGCGAGCGATGCGGATAAAGGCATGGAAGTGGA TAACACCAAAAAAGTGTTTAAAAAAAAAACCCAGCA-3’; 149 PRIMER4:5’CACCGCGCGCTGGTTCATCCACTGCGGATAATGGCCATCTTCGTTCTGCTG GGTTTTTTTTTTAAA-3’; PRIMER5:5’-GGCGGCGGATCCCAGATGCGCAACGTG-3’; PRIMER6:5’-GGCGGCCTCGAGTCACACCGCGCGCTGGTTCATC-3’ The protein fragment cloned excludes the N- and C-terminal NLSs described by others (Hyoung Kim, 2003) The full length ApLLP (120AA) was cloned into a histag expression vector pET32a (Novagen), using three more primers: PRIMER7:5’GGCGGCGGATCCATGGCGAAAAGCATTCGCAGCAAACATCGCCGCCAGA TGCGCAACGTG-3’; PRIMER8:5’GGTTTTCAGTTTCGCCACTTTCACTTTCTGTTTTTTCACCGCGCGCTGGTT3’; PRIMER9: 5’GGCGGCCTCGAGTCACCATTTAATTTTTTTGCCAATTTTTTTTTTGGTTTTC AGTTTCGC-3’ The 87AA fragment and 55AA fragment were cloned in the GST expression vector pGEX-4T-1(Amersham Biosciences) The GST-ApLLP (87AA), GST-ApLLP (55AA) and HIS-ApLLP (120AA) proteins were overexpressed by using the Escherichia coli BL21 strain in an LB medium containing 100 μM/ml of ampicillin Cells were grown at 37℃ and the induction of the expression of the protein was performed by an addition of 0.4 mM IPTG Cells were grown at 20℃ for 16 hours, collected by centrifugation, re-suspended in a 50 mM phosphate buffer, pH 7.4, 0.15 M NaCl The recombinant ApLLP containing proteins (87AA fragment and 55AA fragment) were purified by GST affinity chromatography under native conditions, followed by an in- 150 gel thrombin cleavage to remove the GST tag Full-length HIS-ApLLP (120AA) was expressed into the inclusion body and was pre-purified by Ni2+-affinity chromatography under a denatured condition The released ApLLP (120AA) and ApLLP fragments (87AA and 55AA) were further purified by HPLC on a reversephase C8 column and C18 (Vydac) respectively The collected sample was lyophilised for at least 48 hours After lyophilisation, the powder was kept at -20℃ before use For the NMR isotope labelling, recombinant proteins were prepared by growing the cells in an M9 medium with additions of (15NH4)2SO4 for 15N labelling The identities of ApLLP were verified by MALDI-TOF mass spectrometry 6.2.2 Oligonucleotide synthesis The asymmetric CRE cis-elements of the ApC/EBP promoter (CRE2F: 5’TGACGTCT-3’; CRE2R: 5’-AGACGTCA-3’) were synthesised The powder of the oligonucleotide was dissolved in pure water The equal molar of the forward and reverse-complementary oligonucleotides were mixed together at room temperature The mixture was then heated to 95℃ for five minutes, and cooled down to 4℃ The double strand CRE element solution pH was further adjusted to 6.8, using a pH meter equipped with a long stem glass body electrode Aliquot solutions in a ml tubes were kept at -20℃ before using 6.2.3 Circular Dichroism Far-UV CD spectra were performed with a series of ApLLP (87AA)/CRE molar ratios: 2, 1, 1/2, 1/3 and 1/6 All the spectra were subtracted by the reference spectra only, including oligonucleotide with the same concentrations of CRE Near-UV 151 spectra of the ApLLP dissolved in 8M urea and ApLLP dissolved in a 20mM phosphate buffer were also acquired in parallel 6.2.4 NMR sample preparation and NMR experiments Similarly NMR samples of ApLLP-87 and ApLLP-55 were prepared in 20 mM phosphate buffer (with the final pH value of 6.8) while those of the full-length ApLLP was dissolved in the salt-free water (with the final pH value of 3.8) About 10% D2O was added to provide the deuterium lock signal for the NMR spectrometers All NMR experiments with protein concentrations of ~200 μM were acquired on an 800 MHz or 500 MHz Bruker Avance spectrometer equipped with pulse field gradient units at 298 K The HSQC NMR titrations were carried out to assess the binding of the CRE DNA fragment to the full-length ApLLP and two dissected ApLLP proteins The NMR spectra were processed and analysed using an NMRPipe (F Delaglio, 1995) and NMRView (B Johnson, 1994) 6.2.5 Disorder Predictions The disorder prediction of ApLLP proteins were done by online tools, the VSL2 predictor (http://www.ist.temple.edu/disprot/predictorVSL2.php), which exploits the length-dependent (heterogenous) amino acid compositions and sequence properties of the intrinsically disordered regions to improve prediction performance 6.3 Results 6.3.1 Bioinformatics analysis of ApLLp Initially, PSORT analysis showed that two nuclear localization signals were located 152 over the N-terminal residues 1–11 and C-terminal residues 99–120, as previously reported [Hyoung Kim, 2003] Further analysis by SMART and ELM didn’t demonstrate any known domain, repeat and motif with high confidence Nevertheless, a basic leucine zipper domain was implied with low confidence in the region from residue 12 to 66 As shown in Figure 1A, the ApLLP protein has the characteristic of low hydrophobicity with only 25.83% hydrophobic residues Accordingly, the globularity and disordered region of the ApLLP protein were further assessed by DISPROT Surprisingly, as shown in Fig 1B, in most of the region of ApLLP, the disorder probabilities are larger than 0.8, strongly suggesting that the ApLLP protein might be intrinsically unstructured Based on the above analysis, we thus cloned and expressed three ApLLP forms for detailed structural and binding investigations, namely the full-length ApLLP; the ApLLP-87 (residues 12–98) with both nuclear localization signals deleted and ApLLP-55 (residues 12–66) mainly containing the putative leucine zipper domain Interestingly, the full-length ApLLP protein with about 25.83% hydrophobicity was only found in inclusion body and was insoluble in buffer even after affinity and HPLC purifications (Figure6.2A) However, it could be dissolved in salt free water as we previously discovered on other insoluble proteins [M Li, 20061; M Li, 20062; M Li2, 2007] On the other hand, both ApLLP-87 and ApLLP-55 with hydrophobicity of 25.29% and 29.09%, respectively, were highly soluble in buffer The molecular weights of three recombinant proteins were further determined by MALDI-TOF MS (Figure6.2B) 153 Figure 6.1 Aplysia ApLLP encodes 120 amino-acids A) Two Nuclear Localisation Signals are boxed as shown The full-length ApLLP (120AA), truncated ApLLP (87AA) and ApLLP-55 are shown below the sequence The Lysine Rich Motif in the C-terminal end ApLLP is underlined A schematic cartoon is presented at the bottom, with each region corresponding to the sequences described above B) Analysis results of the globularity and disordered regions of the ApLLP by the VSL2B 154 Figure 6.2 HPLC profile and MS spectra of ApLLP A) ApLLP purification analytical profile, using reverse-phase C18 column (Vydac) B) MALDI-TOF MS profile of the ApLLP protein The main peak value is zoomed in and placed in the central area 155 6.3.2 CD and NMR experiments indicate that ApLLP is an intrinsically unstructured protein The CD study was first carried out for the three recombinant ApLLPs As shown in Figure 6.3A, the far-UV spectra of the full-length ApLLP, ApLLP-87 and ApLLP120 showed similarly typical random-coil properties To fully evaluate whether a tertiary contact exists despite absence of secondary structure, the near-UV CD spectrum was also acquired As shown in Figure 6.3B-D, profiles of the three recombinant ApLLPs in the presence of 8M urea are very similar to those in the absence of 8M urea, indicating that all of them lack a tertiary packing Three ApLLP forms were also 15 N-isotope labelled and further characterized by HSQC spectroscopy As shown in Figure 3, all three proteins had HSQC spectra with very narrow 1H spectral dispersions (