MMM-QSAR recognition of ribonucleases without alignment: Comparison with an HMM model and isolation from Schizosaccharomyces pombe, prediction, and experimental assay of a new sequence

Please download to get full document.

View again

of 15
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Similar Documents
Information Report



Views: 39 | Pages: 15

Extension: PDF | Download: 0

MMM-QSAR recognition of ribonucleases without alignment: Comparison with an HMM model and isolation from Schizosaccharomyces pombe, prediction, and experimental assay of a new sequence
  MMM-QSAR Recognition of Ribonucleases without Alignment: Comparison with anHMM Model and Isolation from  Schizosaccharomyces pombe , Prediction, andExperimental Assay of a New Sequence Guillermı´n Agu¨ero-Chapı´n, †,‡ Humberto Gonza´lez-Dı´az,* ,†,§, | Gustavo de la Riva, ⊥ Edrey Rodrı´guez, # Aminael Sa´nchez-Rodrı´guez, ‡ Gianni Podda, † and Roberto I. Vazquez-Padro´n f Dipartimento Farmaco Chimico Tecnologico, Universita´ Degli Studi di Cagliari, Cagliari, 09124, Italy, CAP,Faculty of Chemistry and Pharmacy, IBP, and CBQ, UCLV, Santa Clara 54830, Cuba,Unit for Bioinformatics & Connectivity Analysis (UBICA), Institute of Industrial Pharmacy and Departmentof Organic Chemistry, Faculty of Pharmacy, USC, Santiago de Compostela 15782, Spain,CINVESTAV-LANGEBIO, Irapuato, Guanajuato 36821, Me´xico, Caribbean Vitroplants,Santo Domingo 1464, Dominican Republic, and Vascular Biology Institute, School of Medicine,University of Miami, Miami, Florida 33136Received August 29, 2007 The study of type III RNases constitutes an important area in molecular biology. It is known that the  pac1 + gene encodes a particular RNase III that shares low amino acid similarity with other genes despite havinga double-stranded ribonuclease activity. Bioinformatics methods based on sequence alignment may fail whenthere is a low amino acidic identity percentage between a query sequence and others with similar functions(remote homologues) or a similar sequence is not recorded in the database. Quantitative structure - activityrelationships (QSAR) applied to protein sequences may allow an alignment-independent prediction of proteinfunction. These sequences of QSAR-like methods often use 1D sequence numerical parameters as the inputto seek sequence-function relationships. However, previous 2D representation of sequences may uncoveruseful higher-order information. In the work described here we calculated for the first time the spectralmoments of a Markov matrix (MMM) associated with a 2D-HP-map of a protein sequence. We used MMMsvalues to characterize numerically 81 sequences of type III RNases and 133 proteins of a control group. Wesubsequently developed one MMM-QSAR and one classic hidden Markov model (HMM) based on thesame data. The MMM-QSAR showed a discrimination power of RNAses from other proteins of 97.35%without using alignment, which is a result as good as for the known HMM techniques. We also report forthe first time the isolation of a new Pac1 protein  (  DQ647826  )  from  Schizosaccharomyces pombe  strain428-4-1. The MMM-QSAR model predicts the new RNase III with the same accuracy as other classicalalignment methods. Experimental assay of this protein confirms the predicted activity. The present resultssuggest that MMM-QSAR models may be used for protein function annotation avoiding sequence alignmentwith the same accuracy of classic HMM models. 1. INTRODUCTION RNase III is a double-strand-specific ribonuclease (dsR-Nase) that usually makes staggered cuts in both strands of adouble helical RNA, although in some cases it cleaves oncein a single-stranded bulge in the helix. 1,2 The primarybiological function of this system is the specific processingof rRNA and mRNA precursors, 3 - 5 but it has also beenimplicated in other diverse phenomena such as mRNAturnover, 6 conjugative DNA transfer, 7 antisense RNA-mediated regulation, and others. 8,9 For instance, Dicer andDrospha are type III RNases responsible for the generationof short interfering RNAs (siRNAs) from long double-stranded RNAs during RNA interference (RNAi). Also, thecellular processing of shRNAs shares common features withthe biogenesis of naturally occurring miRNA such ascleavage by nuclear type III RNase Drosha, export from thenucleus, processing by a cytoplasmic type III RNase Dicer,and incorporation into the RNA-induced silencing complex(RISC). Each step has a crucial influence on the efficiencyof RNAi. 10 - 13 It involves both RNase proteins in severalimportant biological processes as for instance the functionof Dicer on the vascular system regulating embryonicangiogenesis probably by processing miRNAs, which regu-late the expression levels of some critical angiogenicregulators in the cell. 14 Recently, RNAi has moved from apurely experimental technique to the stage of potentialclinical applications such as a possible use for the treatmentof spinocerebellar ataxia or amyotrophic lateral sclerosis. 15 Many other dsRNases have been characterized from a varietyof prokaryotic and eukaryotic sources, and RNase III from * Corresponding author phone:  + 34-981-563100; fax:  + 34-981 594912;e-mail: or Correspondingauthor address: Faculty of Pharmacy, University of Santiago de Compostela15782, Spain. † Universita´ Degli Studi di Cagliari. ‡ IBP, and CBQ, UCLV. § Institute of Industrial Pharmacy, Faculty of Pharmacy, USC. | Department of Organic Chemistry, Faculty of Pharmacy, USC. ⊥ CINVESTAV-LANGEBIO. # Caribbean Vitroplants. f University of Miami. 434  J. Chem. Inf. Model.  2008,  48,  434 - 448 10.1021/ci7003225 CCC: $40.75 © 2008 American Chemical SocietyPublished on Web 02/07/2008   Escherichia coli  is an archetype of this class of enzymes. 6,16,17 The RNase III family consists of a growing number of enzymes that includes at least 33 bacterial and 22 eukaryoticenzymes. 18 There have been numerous reports of dsRNaseactivities in eukaryotic cells, some of which exhibitedproperties consistent with a role in pre-rRNA processing. 19 - 21 One of the best candidates for eukaryotic RNase IIIhomologues is the Pac1 RNase from  Schizosaccharomyces pombe . 22 - 24 The Pac1 product is derived from  Schizosac-charomyces pombe  pac1 + gene expression, which is alsoinvolved in the regulation of sexual development, 25 possiblythrough a mechanism that involves the processing of certainsmall nucleolar RNAs (snRNAs). 26 Pac1 works in eukaryotesas dsRNase and shares a functional similarity to RNase IIIfrom  E. coli . This fact was proved either by measuring theability of Pac1 to degrade double-stranded RNA in vitro orby expressing  pac1 + in  E. coli , where it produced an activitythat converted dsRNA into acid-soluble products. 23 Despitethese observations the Pac1 gene product shows low homol-ogy with other RNAse III enzymes, particularly with thoseones belonging to bacteria. The homology between thedifferent RNase III enzymes varies in the range 20 - 84%depending on their evolutionary distance, suggesting a lowlevel of primary structure conservation. 27 It has been reportedthat antibodies prepared against Pacl RNase have failed toreact with RNase III. 23 The Pac1 gene product from Schizosaccharomyces pombe  belongs to subclass II of theRNase III family, which is characterized by the presence of an N-terminal extension and includes fungal RNase III. 27,28 This contains 363 amino acids (aa), and only its C-terminal230 residues share 25% amino acid identity with the  Escherichia coli  ribonuclease III. 23 Methods based on sequence alignment have revealed alow amino acidic identity (20 - 40%) for the  pac1 + geneproduct with other typical RNases III, either isolated frombacteria or even from species that are genetically close. 27,29 However, experimental observations show the Pac1 proteinto be a dsRNAse enzyme. This relatively low degree of conservation probably reflects the species-specificity of RNase III, which prevents genetic complementation betweenmembers of the RNase III family. 30 All of the facts discussed above hinder the prediction of the Pac1 gene product as an RNase III-like enzyme usingcomputational methods based on sequence alignment. In fact,bioinformatics methods based on sequence alignment mayfail in general for cases of low sequence homology betweenthe query and the template sequences deposited in the database. The lack of function annotation (defined biologicalfunction) for the sequences deposited in databases and usedas templates for function prediction constitutes anotherweakness of alignment approaches. 31,32 Recently, a group of researchers published in  PROTEOMICS   (2006) a review 33 on the growing importance of machine learning methods forpredicting a protein functional class independently of se-quence similarity. In this review the authors make referenceto various papers on the topic, including their own work. 34 - 45 These methods often use as the input 1D sequence numericalparameters specifically defined to seek sequence-functionrelationships. For instance, the so-called pseudoamino acidcomposition approach 46,47 based on 1D sequence couplingnumbers has been widely used to predict subcellular local-ization, enzyme family class, and structural class as well asother attributes of proteins based on their sequencesimilarity. 45,48 - 74 Alternatively, some authors generalizedmolecular indices that are classically used for small mol-ecules 75,76 to describe protein sequences, such as the gener-alization of Broto - Moreau indices by Caballero and Ferna´n-dez et al. 77 On the other hand, many authors have introduced2D or higher dimension representations of sequences priorto the calculation of numerical parameters. This constitutesan important step in order to uncover useful higher-orderinformation not encoded by 1D sequence parameters. 78 - 98 In addition, 2D graphs have been used for proteins and DNAsequences by other researchers. For example, Zupan andRandic´ used spectral-like and zigzag representations. Theseauthors suggested an algorithm for encoding long strings of building blocks (like 4 DNA bases, 20 natural amino acids,or all 64 possible base triplets) using “zigzag” or “spectrum-like” representations. 99 Hydrophobic cluster analysis (HCA)constitutes another well-known technique for the 2D repre-sentation of protein sequences. 100 Randic´ et al. ultimatelyapproached protein representations by using 2D schemesbased on nucleotide triplet codons or virtual genetic code. 101 Finally, we introduced hydrophobicity-polarity (HP) 2DCartesian or latticelike representations for proteins relatedto plant metabolism. 93 In this work, we propose to use the spectral moments of a Markov matrix (MMM) associated to a 2D-HP-graph tonumerically characterize protein sequences and seek a QSARmodel to predict type III RNAses without alignment. First,we derived hydrophobicity-polarity (HP) 2D Cartesian orlatticelike representations (also called maps or graphs) forRNase III and control group protein sequences. 93 We thencalculated the MMM values of order  k   (symbolized as  SR π  k  )to characterize the protein sequence. Spectral moments formany kinds of graphs have been used before for quantitativestructure - activity relationships (QSAR) studies on pro-teins. 102 - 112 We subsequently developed a classifier toconnect protein sequence information (represented by the SR π  k   values) with the classification of sequences as RNAseIII or not. In general, different kinds of classifiers have beenused to derive protein sequence QSAR models. 113,114 Weselected a linear discriminant analysis (LDA), which is asimple but powerful technique. 115 - 121 The use of this MMM-QSAR model enabled us to predict a novel recombinant Pac1(rPac1) protein as an RNase III-like enzyme from a newisolate of   Schizosaccharomyces pombe . Prediction was alsosupported by profile Hidden markov model (HMM) analysisand submission to BLASTp and InterPro 122 servers anddemonstrated by experimental evidence. 2. MATERIALS AND METHODS 2.1. Computational Methods.  A Markov model (MM),also called MARCH-INSIDE, was used to codify informationabout 81 RNase III protein sequences belonging to prokary-ote and eukaryote species downloaded from the GenBank database. Briefly, our methodology considers as states of theMarkov chain (MC) any atom, nucleotide, or amino acid (aa)depending on the kind of molecule to be described. 123,124 Therefore, MM deals with the calculation of the probabilities( k   p ij ) with which the charge distribution of aa moves fromany aa in the vicinity  i  at time  t  0  to another aa  j  along theprotein backbone in discrete time periods until a stationarystate is achieved. 125,126 MMM-QSAR R ECOGNITION OF  R IBONUCLEASES  J. Chem. Inf. Model., Vol. 48, No. 2, 2008   435  Each RNase III sequence was labeled by its accessionnumber; see Table 1 in the Supporting Information. Thecontrol group consists of 133 proteins, which were selectedfrom 2184 high-resolution proteins in a structurally nonre-dundant subset of the Protein Data Bank (PDB); most of the data were published by other authors to distinguishenzymes and nonenzymes without alignment 127 (see Table2 in the Supporting Information). Many researchers havedemonstrated the possibility of predicting protein functionfrom sequences, 128 and we used 2D-HP graphs to encodeinformation about RNase III amino acid sequences. 93 Wethen calculated for the first time the  HP π  k   values for thesegraphs. As can be seen from the discussion above, weselected  HP π  k   based on the utility of other nonstochasticspectral moments 103 - 112 as well as other MMMs and otherstochastic parameters. 102,129 - 131 It is important to point out that this 2D graphicalrepresentation for proteins is similar to those previouslyreported for DNA, 92,96,97 but the 20 different amino acidsare regrouped into HP classes instead of using 4 types of bases. These four groups characterize the HP physicochem-ical nature of the amino acids as polar, nonpolar, acidic, orbasic. 132 The 2D-HP graph for the deduced amino acidsequence of rPac1 protein, obtained from  Schizosaccharo-myces pombe  strain 428-4-1 (uploaded by our group withaccession number  DQ647826  ), is shown in Figure 1. It isworth noting that 363 amino acids are rearranged in a 2Dspace compacting protein representation. Each amino acidin the sequence is placed in a Cartesian 2D space startingwith the first monomer at the (0, 0) coordinates. Thecoordinates of the successive amino acids are calculated asfollows: a) increase by  + 1 the abscissa axis coordinate foran acid amino acid (rightwards-step), or b) decrease by  - 1the abscissa axis coordinate for a basic amino acid (leftwards-step), or c) increase by  + 1 the ordinate axis coordinate fora polar amino acid (upward-step), or d) decrease by  - 1 theordinate axis coordinate for a nonpolar amino acid (downward-step). 2.2. 2D-HP Graph MMMs Used as Sequence Numer-ical Descriptors.  After the representation of the sequenceswe assigned to each graph a stochastic matrix  1 Π . Note thatthe number of nodes ( n ) in the graph is equal to the numberof rows and columns in  1 Π  but may be equal to or evensmaller than the number of amino acids or DNA bases inthe sequence. The elements of   1 Π  are the probabilities  1  p ij of reaching a node  n i  with charge  Q i  moving through a walk of length  k   )  1 from another node  n  j  with charge  Q  j 133 where  R ij  equals 1 if the nodes  n i  and  n  j  are adjacent in thegraph and equal to 0 otherwise.  Q  j  is equal to the sum of the electrostatic charges of all amino acids placed at thisnode. It then becomes straightforward to carry out thecalculation of the spectral moments of   1 Π  in order tonumerically characterize the protein sequencewhere Tr is called the trace and indicates that we sum allthe values in the main diagonal of the matrices  k  Π )  ( 1 Π ) k  , Table 1.  Classification Results Derived from the Model forTraining and Validation SeriesMMM training MMM validationtotal% 97.35 RNases control RNases control 100 total%RNases 93.44  57  4  20  0 100 RNasescontrol 100 0  90  0  43  100 controlMMM all sequences HMM classictotal% 98.1 RNases control RNases control 97.50 total%RNases 95.1 77 4  80  1 98.75 RNasescontrol 100 0 133 5  128  96.24 control Table 2.  Enzymatic Assay of Double-Stranded RNase RecombinantPac1  DQ647826   Extracted from  Schizosaccharomyces pombe  Strain428-4-1conc. rPac 1 1nM 10 nM 100 nMEUV a 6.2 × 10 5 7.4 × 10 5 7.2 × 10 5 6.0 × 10 5 6.8 × 10 5 7.3 × 10 5 6.6 × 10 5 6.9 × 10 5 7.9 × 10 5 mean 6.4 × 10 5 7.0 × 10 5 7.5 × 10 5 a Enzymatic unit value for rPac 1 (U/mg). Figure 1.  2D Cartesian representation for the amino acid sequenceof the rPac1 protein  Schizosaccharomyces pombe  strain 428-4-1,GenBank accession number  DQ647826  .  Note that a node maycontain more than one amino acid, which ensures graph compact-ness.  p ij ) Q  j ∑ m ) ln R il ‚ Q l (1)MMM k  ) SR π  k  ) ∑ i )  jnk   p ij ) Tr[( 1 Π ) k  ] (2) 436  J. Chem. Inf. Model., Vol. 48, No. 2, 2008   A GU  2  ERO -C HAP ı´ N ET AL .  which are the natural powers of   1 Π . The present class of MMMs encodes in a stochastic manner the distribution of the amino acid properties (charge) through all of the nodesplaced at different distances in the 2D-HP lattice. Expansionof expression 2 for  k  ) 0 gives the order zero MMM 0  ( HP π  0 );for  k   )  1 the short-range MMM 1  ( HP π  1 ), for  k   )  2 themiddle-range MMM 2  ( HP π  2 ), and for  k   )  3 the long-rangeMMMs. This extension is illustrated for the linear graph n 1 -n 2 -n 3 , which is characteristic of the sequence (Asp-Glu-Asp-Lys); please note that the central node contains both Gluand Asp: 93 All calculations of   HP π  k   values for protein sequences of both groups were carried out with our in-house softwareMARCH-INSIDE, version 2.0, including sequence repre-sentation. 134 We proceeded to upload a row data table witheleven  HP π  k   values for each sequence ( k  ) 0, 1, 2,...10) andgrouping variable RNaseIII-score ) 1 (for RNAses) and - 1(for control group sequences) to statistical analysis soft-ware. 135 The overall methodology is represented schemati-cally in order to improve the understanding of our approach(see Figure 2). 2.3. Statistical Analysis. K-Means Cluster Analysis.  Thenegative group was selected from 2184 proteins with diversefunctions (enzymes and nonenzymes) recorded in the PDB,as mentioned before. Our negative subset was designedaccording to K-means cluster analysis  ( k-MCA). 136 Themethod consists of carrying out a partition of the startinggroup made up by a non-RNase III series of proteins intoseveral statistically representative clusters of sequences. Thus,one may select the members to conform to the negativesubset from all of these clusters. This procedure ensures thatthe main protein classes (as determined by the clustersderived from k-MCA) will be represented in the modelcontrol group, thus allowing the representation of the entire‘experimental universe’. The spectral moment series wasexplored as clustering variables in order to carry out k-MCA.The procedure described above is represented graphicallyin Figure 3, where a cluster analysis was carried out to selecta representative sample for the control group. 2.4. Linear Discriminant Analysis.  LDA forward step-wise analysis was carried out for variable selection to buildup the model. 115 - 121 All of the variables included in the modelwere standardized in order to bring them onto the same scale.Subsequently, a standardized linear discriminant equation thatallows comparison of their coefficients was obtained. 137 Thesquare of Mahalanobis’s distance (  D 2 ) and Wilk’s (  λ ) statistic(  λ  )  0 perfect discrimination, being 0  <  λ  <  1) wereexamined in order to assess the discriminatory power of themodel. Pac1 protein was submitted to BLASTp to showgraphically the similarity of the sequence compared to otherRNases III. Each sequence presented in this study was alsosubmitted to the InterPro server 122 in order to compare ourmethodology with other classical sources of predictivefunctional annotation. InterPro consists of a database of protein families, domains, and functional sites in whichidentifiable features found in known proteins can be appliedto unknown protein sequences. 3. EXPERIMENTAL SECTION 3.1. Strains and Culture Media.  The  Schizosaccharo-myces pombe  strain 428-4-1 was routinely grown in yeastextract (YEB) medium at 30  ° C during 12 h. Bacterial strain  Escherichia coli  DH5 R   was grown in luria broth (LB).Transformed bacteria were recovered in the same LB HP π  0 ) Tr[( 1 Π ) 0 ] ) Tr ( [ 1 0 00 1 00 0 1 ] ) ) 3 (2a) HP π  1 ) Tr[( 1 Π ) 1 ] ) Tr ( [ 1  p 111  p 12  0 1  p 211  p 221  p 23 0  1  p 321  p 33 ] ) ) 1  p 11 + 1  p 22 + 1  p 22  (2b) HP π  2 ) Tr[( 1 Π ) 1 ] ) Tr ( [ 1  p 111  p 12  0 1  p 211  p 221  p 23 0  1  p 321  p 33 ] ‚ [ 1  p 111  p 12  0 1  p 211  p 221  p 23 0  1  p 321  p 33 ]) ) 2  p 11 + 2  p 22 + 2  p 22 (2c) HP π  2 ) Tr[( 1 Π ) 2 ] ) Tr ( [ 1  p 111  p 12  0 1  p 211  p 221  p 23 0  1  p 321  p 33 ] ‚ [ 1  p 111  p 12  0 1  p 211  p 221  p 23 0  1  p 321  p 33 ] ‚ [ 1  p 111  p 12  0 1  p 211  p 221  p 23 0  1  p 321  p 33 ] ) ) 3  p 11 + 3  p 22 + 3  p 22 ‚  (2d) Figure 2.  Schematic representation of the steps given in this work. MMM-QSAR R ECOGNITION OF  R IBONUCLEASES  J. Chem. Inf. Model., Vol. 48, No. 2, 2008   437  medium but supplemented with carbenicillin at 100  µ g/mL.Media were also supplemented with bacteriological agarwhen required. 3.2. Total DNA Extraction.  A colony from  Schizosac-charomyces pombe  strain 428-4-1 was inoculated in 5 mLof YEB medium and grown at 30  ° C during 12 h until OD 600 ) 0.5. From this culture, 250  µ L was transferred to 50 mLof the same medium and grown overnight at the sametemperature. When the OD 600 ) 0.8, cells were collected bycentrifugation and broken using small glass pearls. A cellularpellet was resuspended in 500  µ L of sterile water at 50  ° C,and the extract was separated from cellular debris bycentrifugation. Total DNA was purified using a total DNAextraction kit (Qiagen GmbH, Germany). Total DNA solutionwas measured at 260 nm in a GENESYS 10 spectropho-tometer, reaching a concentration of 3.8  µ g/   µ L. The solutionwas also run on agarose gel (0.8%), and high integrity wasseen. 3.3. Primer Design.  Forward (PAC5 ′ ) 5 ′ -c cc  ATG G- GACGGTTTAAGAGGCATC-3 ′  and reverse (PAC3 ′ ) 5 ′ -gtgg  ggttaa cgggcaaac TTA G-3 ′  primers were designed basedon the previously reported  pac1 + coding sequence from Schizosaccharomyces pombe  mutant snm1-1. The primersequences show the restriction sites Nco1 and Kpn1 intro-duced at the 3 ′  and 5 ′  ends, i.e., the first ATG and the stopTTA codon. The coding regions are shown in capitalletters. 138 3.4. PCR Amplifications.  Amplification of the  pac1 + genefrom  Schizosaccharomyces pombe  was performed by stan-dard PCR from its total DNA. The reaction mixture contain-ing 10 ng of template, 1 mM of each dNTP, 1.5 mM MgCl 2 ,2  µ M of each PAC5 ′  and PAC3 ′  primers, 1 ×  buffer TaqPol (Gibco BRL), and 2.5 U Taq Pol (Gibco) was completedto a total volume of 50  µ L. The PCR was carried out usinga thermocycler (Perkin-Elmer 2400) programmed as fol-lows: 5 min initial template denaturation at 94  ° C, cyclesteps s 1 min template denaturation at 94  ° C, 2 min primerannealing at 45  ° C, 2 min primer extension at 72  ° C for 30cycles, plus a final extension step at 72  ° C for 5 min. 29,30,138 PCR reaction showed a band coinciding with the size of thereported  pac1 + ORF. 138 3.5. Plasmid Construction and Sequencing.  The PCRamplification product was purified using a GEL Band Purifi-cation kit (  AmershamPharmaciaBiotech ) and ligated topMOS-Blue T-vector (  AmershamPharmaciaBiotech ). Theligation was transformed into electrocompetent  E. coli  DH5 R by electroporation in 0.2 mm cuvettes using a Gene PulserMachine (BioRad) (12.5 kV, 25  µ F, 1000  ω ). The transfor-mation was plated onto LB medium supplemented with 40  µ L of 20  µ g/mL X-gal solution and 4  µ L of isopropylthio-   -D-galactoside from a 200  µ g/mL IPTG solution per plateand allowed to grow overnight at 37  ° C. White colonies s presumably carrying the recombinant pac1 gene inserted inpMOS-Blue T-vector, named pRSPac1 s were selected, andplasmid DNA was extracted for analysis of the cloned frag-ment by restriction enzymes. Sequencing of the cloned frag-ment was performed using an ABI 3700 sequencer (AppliedBiosystems), 139 and this showed a product of 1.111 Kb. 3.6. Purification of Recombinant Pac1.  A single colonyof   E. coli  DH5 R  with pRSPac1 was grown overnight at 30 ° C in 5 mL of LB medium supplemented with carbencillinat 100  µ g/mL. 250  µ L of culture was then inoculated to 250mL of the same medium supplemented with carbenicillin(100  µ g/mL) and grown under the same culture conditionsuntil OD 600  )  0.8; at this point 50  µ L of 200  µ g/mL IPTGsolution was added to the culture. Three hours after induction,cells were harvested by centrifugation and washed with 15mL of 50 mM tris-HCl (pH 8), 100 mM NaCl, and 1 mMEDTA. Cells were collected by centrifugation and stored at - 70  ° C overnight. Around 3 g of frozen cells was resus-pended in 15 mL of lysis buffer (1% NP40, 0.5% sodiumdeoxycholate, 0.1 M NaCl, 30 mM Tris-HCl (pH 8), 1 mMEDTA), and 5 mM MgCl 2  and DNase1 (10  µ g/mL) wereadded. The cell suspension was incubated on ice for 10 min.Inclusion bodies were collected by washing four times withlysis buffer and twice with 50 mM Tris-HCl 5 mM (pH 8),1 mM DTT. Finally, the sample was dissolved in 5 mL of loading buffer and boiled in a water bath for 10 min. Thetotal volume of extract was divided into five preparativePAGE electrophoresis samples containing 1 mL of proteinextract, which were run in 12% gel. The componentcorresponding to 45.5 kDa recombinant Pac1 protein wasvisualized by staining with an aqueous solution of 0.05%Coomassie brilliant blue R250. In each case the recombinantprotein was excised from polyacrylamide gel, recovered byelectroelution, combined, concentrated with Centricon-10(Amicon) to 0.5 mL, and diluted to 1.5 mL with a storagebuffer to a final composition of 500 mM NaCl, 20 mMsodium phosphate (ph 7.4), 67 mM imidazole, 1 mM DTT,1 mM EDTA, and 30% glycerol. The recPac1 preparationwas stored at  - 20  ° C. 29,30,138 Figure 3.  k-MCA procedure for control group design. 438  J. Chem. Inf. Model., Vol. 48, No. 2, 2008   A GU  2  ERO -C HAP ı´ N ET AL .
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!