Solubility studies, rational amino acid replacements and structural analyses of streptomyces jumonjinensis isopenicillin n synthase 3

Hence, we hypothesize that by replacing certain amino acid residues at these determinant sites in sjIPNS with conserved amino acid residues naturally occurring at the corresponding sites

4.3 Elucidation of amino acid residues important for sjIPNS solubility

Approaches to enhance protein properties such as stability and solubility have been

reported for many systems (Lehmann et al., 2000; Sim and Sim, 1999) However, thus far,

only mutations that decrease solubility have been reported for IPNS isozymes (Sim and Sim, 1999; Loke and Sim, 1999b) Rarely, perhaps no study has yet been employed to replace amino acid residues for the specific aim of improving the solubility of IPNS enzymes The low solubility of sjIPNS has fortuitously made it a suitable candidate to challenge the engineering

of crucial amino acid residues to improve its solubility Therefore, the focus of this section is

to identify amino acid residues that may be potentially important for the solubility of sjIPNS and to use site-directed mutagenesis to investigate the effects of amino acid replacements on the solubility of sjIPNS It would be exciting if amino acid substitutions that can increase sjIPNS solubility were elucidated This would serve as a platform for us to understand how the effects of specific amino acid mutations translate into altered protein expression.

The results of this study have shown that although IPNS isozymes are closely related

in terms of sequence homology and function, they can be classified into two groups based on their expression profiles at 37ºC and 25ºC (Section 4.1.5) Group I consists of four bacterial IPNS, namely scIPNS, sfIPNS, sIIPNS and nIPNS, which have been shown to be highly soluble when the induction temperature is lowered from 37°C to 25°C In contrast, sjIPNS, the only member in Group II, is insoluble at both temperatures It is intriguing that despite the high nucleotide (77%-85%) and amino acid (70%-82%) sequence homology amongst the bacterial

isozymes, the solubility of sjIPNS in E coli appears to be different compared to other bacterial

isozymes This implies that the amino acid sequence of each isozyme, especially in homologous regions, may dictate their folding and expression characteristics Consequently,

non-we propose that the low solubility nature of sjIPNS, uniquely different from the rest of the

compared IPNS isozymes, is due to the non-conserved amino acid sequences between them

More specifically, it is anticipated that unique differences in the amino acid residues in sjIPNS

Chapter 4 Results

relative to the other four soluble isozymes within the non-conserved regions may be the determinants for low solubility of sjIPNS Hence, we hypothesize that by replacing certain amino acid residues at these determinant sites in sjIPNS with conserved amino acid residues naturally occurring at the corresponding sites in all the soluble bacterial isozymes may improve sjIPNS solubility To test our hypothesis, comparative sequence analysis followed by rational amino acid replacements were performed

4.3.1 Comparative amino acid sequence analysis of sjIPNS and other

soluble bacterial IPNS isozymes

In order to identify relevant amino acid residues for site-directed mutagenesis, the amino acid sequences of the four soluble bacterial IPNS enzymes from Group I were selected

as representative soluble bacterial IPNS sequences and used in our sequence comparison with sjIPNS This was carried out using CLUSTAL W Multiple Sequence Alignment Program

(Version 1.6) (Thompson et al., 1994) Fig 4.17 shows the alignment between sjIPNS

(sequence depicted in blue) and the soluble isozymes (sequences depicted in black) The conserved residues between the compared IPNS enzymes were highlighted in gray using the BOXSHADE program (Version 3.2) In addition, functionally important conserved residues shown by IPNS crystal structures and series of mutagenic experiments (reviewed in Section 2.5.4) involved in binding Fe2+ ion and ACV substrate in IPNS were highlighted by “*”

Our focus is on the non-conserved regions between sjIPNS and the soluble isozymes, which correspond to the non-highlighted regions in the alignment (Fig 4.17) Next, was to further elucidate sites within these regions that satisfied our defined criterions, which are sites whereby the amino acid residues in the soluble bacterial IPNS are wholly conserved either in identities or in a particular amino acid property (i.e hydrophobic, polar uncharged, charged acidic and charged basic; colored in pink, yellow, blue and green respectively in the alignment) but are absent in sjIPNS Based on these criterions, eight sites with conserved amino acid

identities occurring only in the soluble IPNS isozymes and not in sjIPNS were identified (marked by “♦”) These correspond to Ser19, Ala37, Asp112, Glu120, Ile156, Ala202, Arg242 and His260 of sjIPNS In addition, three more sites whereby the soluble IPNS isozymes showed a consistent preference for a particular amino acid property over that found in sjIPNS were also identified (marked by “•”) These correspond to Tyr160, Thr308 and Thr309 of sjIPNS At these three positions in all the soluble IPNS, hydrophobic amino acid residues as

oppose to the polar uncharged residues in sjIPNS are found Altogether, eleven sites scattered

throughout the primary amino acid sequence of sjIPNS were elucidated Information on these sites was tabulated in Table 4.3

4.3.2 Proposition of sites for site-directed mutagenesis to investigate its

influence on sjIPNS solubility

Eleven amino acid positions were identified whereby the soluble scIPNS, sfIPNS, slIPNS and nIPNS differ from the insoluble sjIPNS Intuitively, these elucidated sites may cause sjIPNS to be expressed in the insoluble aggregates Upon close examination, the soluble IPNS enzymes have residues with higher hydrophobicity present at seven out of the eleven identified sites (sjIPNS position 19, 120, 160, 242, 260, 308 and 309) as compared to sjIPNS (Table 4.3) Particularly, at site 160, 308 and 309, a prevalent conservation of hydrophobic property is observed in the soluble IPNS proteins Take site 309 for example, leucine, valine and isoleucine hydrophobic residues are found at this position in the various soluble IPNS (Table 4.3), indicating that hydrophobicity is preferred at this site in all the soluble IPNS However, polar threonine is found at this site in sjIPNS

Next, we pose the question of whether replacing the amino acid residues at these identified sites in sjIPNS with those found in the soluble IPNS isozymes can improve the solubility of sjIPNS To address this possibility, site-directed mutagenesis was used to construct sjIPNS single mutants for all the eleven sites The choices of amino acid changes are

Hydrophilic, acidic : D,E Hydrophilic, basic : H, K, R

Polar, uncharged : N, Q, S, T, W, Y Hydrophobic : A, C, F, G, I, L, M, P, V

♦ ♦

scIPNS 1 MPVLMPSAHVPTIDISPLFGTDAAAKKRVAEEIHGACRGSGFFYATNHGV sfIPNS 1 MPILMPSADVPTIDISPLFGDDPDAKTHVAQQINKACRGSGFFYASHHGI slIPNS 1 MPIRMPSAHVPTIDISPLFGTDPDAKAHVARQINEACRGSGFFYASHHGI nIPNS 1 MKMPSAEVPTIDVSPLFGDDAQEKVRVGQEINKACRGSGFFYAANHGV sjIPNS 1 MPILMPSAEVPTIDISPLSGDDAKAKQRVAQEINKAARGSGFFYASNHGV

∗

scIPNS 51 DVQQLQDVVNEFHGAMTDQEKHDLAIHAYNPDNPHVRNGYYKAVPGRKAV sfIPNS 51 DVQQLQDVVNEFHGTMTDEEKYDLAINAYNSANPRVRNGYYMAVEGKKAV slIPNS 51 DVRRLQDVVNEFHRTMTDQEKHDLAIHAYNENNSHVRNGYYMARPGRKTV nIPNS 49 DVQRLQDVVNEFHRTMSPQEKYDLAIHAYNKNNSHVRNGYYMAIEGKKAV sjIPNS 51 DVQLLQDVVNEFHRNMSDQEKHDLAINAYNKDNPHVRNGYYKAIKGKKAV

♦

scIPNS 101 ESFCYLNPDFGEDHPMIAAGTPMHEVNLWPDEERHPRFRPFCEGYYRQML sfIPNS 101 ESWCYLNPSFGEDHPMIRSGTPMHEVNIWPDEKRHERFRPFCEQYYRDMF slIPNS 101 ESWCYLNPSFGEDHPMIKAGTPMHEVNVWPDEERHPDFRSFGEQYYREVF nIPNS 99 ESFCYLNPSFSEDHPEIKAGTPMHEVNSWPDEEKHPSFRPFCEEYYWTMH sjIPNS 101 ESFCYLNPSFSDDHPMIKSETPMHEVNLWPDEEKHPRFRPFCEDYYRQLL

♦ •

scIPNS 151 KLSTVLMRGLALALGRPEHFFDAALAEQDSLSSVSLIRYPYLEEYPP V sfIPNS 151 QLSKTLMRGFALALGKPEDFFDANLPEDDTLSAVSLIRYPHLKAYPP V slIPNS 151 RLSKVLLRGFALALGKPEEFFENEVTEEDTLSAVSMIRYPYLDPYPEAAI nIPNS 149 RLSKVLMRGFALALGKDERFFEPELKEADTLSSVSLIRYPYLEDYPP V sjIPNS 151 RLSTVIMRGYALALGRREDFFDEALAEADTLSSVSLIRYPYLEEYPP V

♦ ∗ ∗ ∗

scIPNS 199 KTGPDGQLLSFEDHLDVSMITVLFQTQVQNLQVETVDGWRDIPTSENDFL sfIPNS 199 KTGPDGTKLSFEDHLDVSVITVLFQTEVQNLQVETVNGWQDLPTSGDDFL slIPNS 201 KTGPDGTRLSFEDHLDVSMITVLFQTEVQNLQVETVDGWQSLPTSGENFL nIPNS 197 KTGPDGEKLSFEDHFDVSMITVLYQTQVQNLQVETVDGWRDLPTSDTDFL sjIPNS 199 KTGADGTKLSFEDHLDVSMITVLYQTEVQNLQVETVDGWQDIPRSDEDFL

♦ ∗ ∗ ∗

scIPNS 249 VNCGTYMAHVTNDYFPAPNHRVKFVNAERLSLPFFLNGGHEAVIEPFVPE sfIPNS 249 VNCGTYMGYLTNDYFPAPNHRVKFINAERLSLPFFLHAGHTTLMEPFSPE slIPNS 251 INCGTYLGYLTNDYFPAPNHRVKYVNAERLSLPFFLHAGQNSVMKPFHPE nIPNS 247 VNAGTYLGHLTNDYFPSPLHRVKFVNAERLSLPFFFHAGQHTLIEPFFPD sjIPNS 249 VNCGTYMGHITHDYFPAPNHRVKFINAERLSLPFFLNAGHNSVIEPFVPE

* Mutant Ala202Pro was constructed previously in our laboratory and the mutant clone was used here for our investigation

Position

in sjIPNS

(scIPNS, sfIPNS, slIPNS & nIPNS)

sjIPNS single mutant constructed

Phe

Leu

(scIPNS)

Hydrophobic Tyr160Leu

Chapter 4 Results

guided by the identities of amino acid residues present in the soluble IPNS isozymes (Table 4.3) Only one type of substitution was carried out for all the identified positions except at positions 160 and 309 For position 160, the endogenous tyrosine residue in sjIPNS was changed to phenylalanine and leucine separately to construct two sjIPNS single mutants at this position Likewise for position 309, the naturally occurring threonine in sjIPNS was substituted with leucine and valine separately Thus, a total of thirteen sjIPNS single mutants were constructed The list of sjIPNS single mutants constructed in this study was shown in Table 4.3

Another major cogitation when deciding on the amino acid for replacement is that the changes should not affect the catalytic ability of sjIPNS enzyme adversely Since all the proposed mutations were located within the non-conserved regions, leaving the functionally important conserved residues (marked by “*”) intact, perturbation of the enzyme activities of sjIPNS single mutants was likely to be minimal

4.3.3 Construction of sjIPNS single mutants

To investigate the effects of the proposed amino acid replacements on the expression

of soluble sjIPNS in E coli, site-directed mutagenesis was carried out followed by expression

of the resultant sjIPNS mutant clones in E coli BL21(DE3) in the following Sections

4.3.3.1 Site-directed mutagenesis, selection and sequence confirmation of

sjIPNS single mutants

To probe the influence of the thirteen amino acid substitutions at the eleven proposed sites on the solubility of sjIPNS, the approach adopted is to use site-directed mutagenesis for specific alteration of these sites using QuikChangeTM site-directed mutagenesis kit (Stratagene)

that employs the in vitro PCR mutagenesis technique The detailed experimental procedures

involved are described in Section 3.2.8

For each mutation, two primers carrying the intended mutation were designed to anneal to the same sequence on the opposite strands of the plasmid Mutagenic primer sequences for each mutation were shown in Appendix II Recombinant pET-SJ plasmid was used as the template for the PCR mutagenesis During temperature cycling, the mutagenic primer pair would anneal to the same sequence on opposite strands of the sjIPNS gene (Fig

4.18a) and Pfu polymerase would extend DNA synthesis in opposite directions from the 3’

ends of both primers As such, the amplified product (~6.3kbp) would have the same size as

the plasmid template (Fig 4.18b) Subsequently, the reaction mixture was incubated with DpnI for three hours DpnI only recognizes target sequences that are methylated This selection

served to remove the methylated parental pET-SJ plasmid The final mixture was then

transformed into E coli BL21(DE3) cells

To screen for each mutation, four putative recombinant clones were randomly picked from the agar plate and overnight cultures were prepared for plasmid extraction Subsequently, the putative plasmid that carried the desired mutant gene was identified through DNA sequencing As the mutations were dispersed throughout sjIPNS gene, those mutations located near the 5’ end of the gene were sequenced using the forward primer OL48 which annealed to the T7 promoter Accordingly, mutations found at the 3’ end of sjIPNS gene were sequenced using the reverse primer OL74 that annealed to the T7 terminator This was illustrated in Fig 4.19 which showed the details of the sequencing strategy and the comparisons of the respective sequenced regions of wildtype and mutant sjIPNS Take for example, amino acid 19 of the wildtype sjIPNS sequence is coded by TCC for serine, and the mutant enzyme would have at the same position TTC codon that codes for phenylalanine This site change is clearly shown in the electropherograms presented in Fig 4.19a The entire open reading frames of all the sjIPNS single mutants were also sequenced to verify that the mutants only contained the respective intended mutations and did not harbor other random mutations

Chapter 4 Results

129

Fig 4.18 PCR mutagenesis of sjIPNS

(a) PCR mutagenic process; (b) Lane 1 shows the λHindIII DNA markers The reaction

products derived from in vitro PCR site-directed mutagenesis of the 13 individual single

mutations of sjIPNS are resolved by gel electrophoresis in lanes 2 to 14

(a)

(b)

pET-SJ 6300bp

Mutagenesis primer pair

130

Fig 4.19 Sequence analysis of sjIPNS single mutants

Sequencing strategy used to verify the specific mutational changes introduced in sjIPNS is illustrated in the diagram below Mutations analyzed via sequencing of the forward strand using OL48 that primed upstream of sjIPNS gene are shown in pink Remaining mutations were verified using OL74 (shown in green) that primed downstream of sjIPNS gene and thus the reverse strand of the gene sequence was being analyzed Sequencing results of all sjIPNS single mutants are presented in the table (a) and (b) below The specific nucleotide regions sequenced are shown and the affected codons are underlined

Mutation wild type sequence Mutant sequence

S19 A37 D112 E120 I156 Y160

Chapter 4 Results

131

Mutation wild type sequence Mutant sequence

(a) Mutations analyzed by sequencing the forward strand of sjIPNS gene

132

(b) Mutations analyzed by sequencing reverse strand of sjIPNS gene

Mutation wild type sequence Mutant sequence

Chapter 4 Results

4.3.3.2 Expression analysis of sjIPNS single mutants

To determine the effect of the introduced mutations on the soluble expression of

sjIPNS, recombinant E coli BL21(DE3) clones harboring the plasmids of the respective single

mutants were expressed at 37°C, 30°C, 28°C and 25°C, as described in Section 4.1.4 Wildtype sjIPNS was similarly expressed for comparison The total levels of wildtype and mutant sjIPNS proteins expressed at 37°C and 30°C were comparable and were all found mainly in the insoluble protein fractions (results not shown) Hence, no differences were observed in the expression profiles of wild type and mutant sjIPNS proteins at 37°C and 30°C whereby all were insoluble

Interesting variations in the production of soluble wildtype and mutant sjIPNS proteins were observed at 28°C (Fig 4.20a) For instance, while only 1-4% of wildtype sjIPNS and mutant Ser19Phe, Ala37Cys, Asp112Glu, Ile156Leu, Tyr160Leu, Ala202Pro, and His260Asn were expressed in the soluble protein fractions, observable higher levels of soluble proteins, ranging between ~9-13%, were obtained for mutant Glu120Gly, Tyr160Phe, Arg242Thr, Thr308Ala, Thr309Val and Thr309Leu at 28°C (indicated by “*” in the chart of Fig 4.20a) In addition, the amounts of soluble proteins expressed by the respective sjIPNS mutant clones at 28°C were comparable to the amounts of soluble bacterial IPNS isozymes (~10-13% of total soluble proteins) produced at the same temperature (Fig 4.8c)

When the temperature was lowered to 25°C, a considerable increase in the number of sjIPNS mutants being produced as soluble proteins was observed (Fig 4.20b) These were evident from the presence of a prominent band, corresponding to the expected 37kDa size of sjIPNS, in the lanes containing the soluble protein fractions of the respective sjIPNS mutants These highly soluble sjIPNS mutants include mutant Ser19Phe, Asp112Glu, Glu120Gly, Tyr160Phe, Arg242Thr, Thr308Ala, Thr309Val and Thr309Leu (marked by “*” in the chart of Fig 4.20b) Interestingly, mutant Glu120Gly, Tyr160Phe, Arg242Thr, Thr308Ala, Thr309Val and Thr309Leu were soluble at both 28°C and 25°C, attesting to the possibility that these

134

Fig 4.20 Expression analysis of sjIPNS single mutants

The soluble protein fractions of sjIPNS single mutants expressed at (a) 28°C, (b) 25°C and (c) 18°C were analyzed using SDS-PAGE The first lane (marked M) in all the gels shows the protein standard of various molecular sizes (kDa) The percentages of sjIPNS expressed in the various soluble protein fractions were measured by densitometric scanning and the values were plotted in the charts shown below The arrows indicate the position of sjIPNS expressed

Chapter 4 Results

mutants assumed the same folding characteristics as the soluble bacterial IPNS isozymes (Fig 4.8c) However, mutant Ser19Phe and Asp112Glu were uniquely soluble at 25°C but not at 28°C Besides an increase in the number of sjIPNS mutants becoming soluble at 25°C, a more drastic enhancement of the levels of soluble mutant enzymes produced was also observed In contrast to only ~4% of soluble wildtype sjIPNS expressed at 25°C, up to 24-32% of soluble enzymes were obtained for the various soluble sjIPNS mutants at the same temperature (refer

to chart in Fig 4.20b) Remarkably, this constituted about 6-8 fold increments in the production of soluble sjIPNS enzymes at 25°C The rest of the sjIPNS mutants, mutant Ala37Cys, Ile156Leu, Tyr160Leu, Ala202Pro and His260Asn, remained insoluble like the wildtype sjIPNS

To investigate whether extreme low temperatures can improve the solubility of the sjIPNS mutants that remained insoluble at 25°C, expression of the various mutants were carried out at 18°C, the temperature at which highly soluble wildtype sjIPNS was previously obtained SDS-PAGE analysis of the soluble protein fractions (Fig 4.20c) showed that all the thirteen sjIPNS mutants were unanimously expressed up to ~28-33% of the total soluble proteins, comparable to ~30% of soluble wildtype sjIPNS produced at 18°C Thus, sjIPNS mutant Ala37Cys, Ile156Leu, Tyr160Leu, Ala202Pro and His260Asn behave like wildtype sjIPNS in that they were only soluble at 18°C

A summary of the diverse mutagenic effect on the expression of soluble sjIPNS was presented in Table 4.4 From the expression results, it is apparent that specific single mutations can successfully override the insolubility of sjIPNS at higher induction temperatures and in some cases, even transformed its soluble expression characteristic to closely resemble that of the soluble bacterial IPNS isozymes Such mutations encompass Ser19Phe, Asp112Glu, Glu120Gly, Tyr160Phe, Arg242Thr, Thr308Ala, Thr309Val and Thr309Leu substitutions that led to a dramatic 6-8 fold improvement in the expression of soluble sjIPNS at 25°C Interestingly, these eight mutations were further found to augment the production of soluble sjIPNS but to different extents For instance, while Glu120Gly, Tyr160Phe, Arg242Thr,

Thr308Ala, Thr309Val and Thr309Leu substitutions improved sjIPNS solubility at temperature as high as 28°C; Ser19Phe and Asp112Glu substitutions enhanced the solubility of sjIPNS at temperature not higher then 25°C On the other hand, Ala37Cys, Ile156Leu, Tyr160Leu, Ala202Pro and His260Asn substitutions did not alter the expression of sjIPNS at all the temperatures studied It is interesting to note that the substitution of tyrosine at position

160 with phenylalanine transformed sjIPNS into highly soluble enzymes, however substitution

at the same site with leucine did not improve sjIPNS solubility In contrast, substitution of threonine at position 309 in sjIPNS with either valine or leucine both gave rise to highly soluble sjIPNS mutant proteins

Table 4.4 Summary on the expression results of sjIPNS single mutants in E coli

BL21(DE3)

Mutants exhibiting improved solubility at

higher temperatures dependent solubility as wildtype sjIPNS Mutants exhibiting low Soluble at 25°C,

18°C

Ser19Phe Asp112Glu Soluble at 28°C,

25°C,

18°C

Glu120Gly Tyr160Phe Arg242Thr Thr308Ala Thr309Val Thr309Leu

Soluble only at 18°C Ala37Cys

Ile156Leu Tyr160Leu Ala202Pro His260Asn

4.3.4 Construction of sjIPNS double and triple mutants

Intriguingly, results from the site-directed mutagenesis and expression analysis of the sjIPNS single mutants corroborate with our hypothesis Indeed, by replacing the amino acid residues at seven of the eleven postulated sites in sjIPNS with corresponding residues at the analogous sites in the soluble bacterial IPNS helped to overcome sjIPNS insolubility at 25 °C The data clearly advocate that Ser19, Asp112, Glu120, Tyr160, Arg242, Thr308 and Thr309 of sjIPNS are key determinants for the expression of soluble sjIPNS Evidently, a single mutation

Chapter 4 Results

is sufficient to convert the insoluble sjIPNS into a highly soluble protein Hence, it becomes pertinent to ask whether having more then one mutation can further improve the solubility of sjIPNS In other words, we were interested to investigate whether the effects of multiple mutations is additive and what consequences ensue when too many mutations were introduced Thus, to understand the effects of combined mutations on the expression of soluble sjIPNS, more site-directed mutagenesis were performed to construct sjIPNS double and triple mutants

for expression analysis in E coli BL21(DE3)

4.3.4.1 Mutagenesis, selection and sequence confirmation of sjIPNS double

and triple mutants

To create sjIPNS multiple mutants, selected mutations that increased the solubility of sjIPNS at both 28°C and 25°C (Table 4.4) were combined together in the same clone Two sjIPNS double mutants were constructed by combining the Glu120Gly and Thr309Val mutations in one, and Thr308Ala and Thr309Val mutations in the other (Table 4.5) In addition, a sjIPNS triple mutant carrying all the three mutations, Glu120Gly, Thr308Ala and Thr309Val, was conjured (Table 4.5)

The same mutagenesis technique described in Section 4.3.3.1 was employed to

introduce the designated mutations in the appropriate clones To construct sjIPNS double mutant Glu120Gly/Thr309Val, recombinant pET24a plasmid harboring sjIPNS single mutant Thr309Val gene was used as the template in the PCR mutagenesis process with mutagenic primer pair that will then introduce the second mutation, Glu120Gly In a similar manner, the same template was used with another primer pair that will specifically introduce the Thr308Ala mutation in addition to the existing Thr309Val change in the template to construct the second double mutant Thr308Ala/Thr309Val

In turn, the resultant plasmid carrying sjIPNS mutant gene with both Thr308Ala and Thr309Val mutations served as the template for PCR mutagenesis to construct the sjIPNS

140

Table 4.5 Summary of sjIPNS double and triple mutants constructed

Number of amino acid substitutions Specific

mutations

DNA template used for PCR mutagenesis

Name of resultant mutant

Double mutant

Glu120Gly Thr309Val pET24a-Thr309Val Recombinant

sjIPNS mutant gene

Glu120Gly/Thr309Val

Double mutant

Thr308Ala

Thr309Val pET24a-Thr309Val Recombinant

sjIPNS mutant gene

Thr308Ala/Thr309Val

Triple mutant

Glu120Gly Thr308Ala

Thr309Val

Recombinant pET24a- Thr308Ala/Thr309Val sjIPNS mutant gene

Glu120Gly/Thr308Ala /Thr309Val

Fig 4.21 PCR mutagenesis and sequence analysis of sjIPNS double and triple mutants

(a) Lane 1 shows the λHindIII DNA markers The reaction products derived from in vitro PCR directed mutagenesis of Glu120Gly/Thr309Val, Thr308Ala/Thr309Val and Glu120Gly/Thr308Ala/ Thr309Val sjIPNS are resolved by gel electrophoresis in lanes 2 to 4 respectively (b) Sequence confirmation of sjIPNS double and triple mutants The specific nucleotide regions sequenced are shown and the affected codons are underlined

site-1 2 3 4 kbp

23.1 9.4 6.6 4.4 2.3 2.0

~6.3kbp

(a)

4.3.4.2 Expression analysis of sjIPNS double and triple mutants

The expression of sjIPNS double and triple mutants in E coli BL21(DE3) was carried

out at 37°C to 18°C as described previously and compared with the expression of wildtype sjIPNS No notable differences between the expression of the two double mutants and the triple mutant, and wildtype sjIPNS were seen at 37°C and 30°C whereby all were produced in the insoluble protein fractions (results not shown) However at 28°C and 25°C, unlike insoluble wildtype sjIPNS, the double mutant Glu120Gly/Thr309Val and Thr308Ala/Thr309Val as well as triple mutant Glu120Gly/Thr308Ala/Thr309Val were overexpressed in the soluble protein fractions (Fig 4.22) Densitometic scanning indicated that

~10-12% of soluble double mutant Glu120Gly/Thr309Val and Thr308Ala/Thr309Val were produced at 28°C whereas another 2.5-3 fold higher amounts of soluble proteins (~29-30% of total soluble proteins) were obtained for both at 25°C (refer to chart in Fig 4.22) The same was observed for the triple mutant Glu120Gly/Thr308Ala/Thr309Val in which 12% and 30%

of soluble mutant protein was produced at 28°C and 25°C respectively Further lowering the

143

Fig 4.22 Expression analysis of sjIPNS double and triple mutants

The soluble protein fractions of sjIPNS double and triple mutants expressed at 28°C, 25°C and 18°C were analyzed using SDS-PAGE The first lane (marked M) in all the gels shows the protein standard of various molecular sizes (kDa) The percentages of sjIPNS expressed in the various soluble protein fractions were measured by densitometric scanning and the values were plotted in the charts shown below The arrows indicate the position of sjIPNS expressed

Glu120Gly/ Thr308Ala/ Glu120Gly/

Wildtype Thr309Val Thr309Val Thr308Ala/Thr309Val

Glu120Gly/ Thr308Ala/ Glu120Gly/

Wildtype Thr309Val Thr309Val Thr308Ala/Thr309Val

% of soluble wildtype sjIPNS, double and triple sjIPNS mutants

expressed at various temperatures

4.3.5 Cyclization activities of sjIPNS mutants

A major concern when using mutagenesis as a tool to improve the behavior of a protein is fortuitous alteration of other properties especially that of the catalytic activity Although the 11 sites altered in sjIPNS were ascertained to be neither the catalytic residues nor neighboring residues to functionally important sites, we still need to confirm whether the mutations affect the activity of sjIPNS Thus, the cyclization activities of sjIPNS single, double and triple mutants were determined to probe whether there were any measurable perturbations

in the mutants’ catalytic abilities

Since the mutants constructed were all highly soluble at 18°C, soluble protein fractions

of the recombinant E coli cultures expressing the respective single, double and triple mutants

at 18°C were used in the enzyme assay The enzyme assays were repeated 3 times to ensure reproducibility Parallel experiments using soluble protein fractions of wildtype sjIPNS obtained at 18°C were also done as control

Interestingly, the mutant enzymes exhibit varied activities from the wildtype sjIPNS mutants (Table 4.6) Amongst those mutants with improved solubility at higher temperatures, varying degree of increase in activities was observed for mutant Ser19Phe, Asp112Glu, Tyr160Phe, Thr308Ala, Thr309Val, Thr309Leu and Thr308Ala/Thr309Val In particular, the cyclization activities of mutant Asp112Glu and Tyr160Phe were 38.9% and 51.4% higher then

145

Table 4.6 Activities of wildtype and mutant sjIPNS enzymes as determined by bioassay

using M luteus ATCC 381 as test organism

a One unit of activity is the amount of IPNS enzyme required to form the equivalent of 1 µmol of

isopenicillin N per min at 26°C

b Relative specific activity is expressed as percentage of the specific activity of the mutant enzyme

relative to that of wildtype enzyme

protein concentration (mg/ml)

Total activity (units) a

Specific activity (units/mg total soluble proteins)

Relative specific activity (%) b

Chapter 4 Results

the wildtype enzyme respectively Reduced activities were observed for the remaining mutant enzymes of which mutant Tyr160Leu, Arg242Thr and His260Asn retained only 19.4-26.4% of the wildtype activity

4.3.6 “Reverse substitutions”: Analysis of replacing selected amino acid

residues in soluble scIPNS with residues characteristic of insoluble

sjIPNS

Thus far, the study has established that the low solubility of sjIPNS is related to the dissimilarities in amino acid identities and/or properties between sjIPNS and the soluble bacterial IPNS isozymes at specific determinant sites (i.e sites 19, 112, 120, 160, 242, 308 and

309 in sjIPNS) Converting the residues at each of these sites in sjIPNS to that found in the soluble isozymes has unequivocally improved sjIPNS solubility, causing its expression profile

to resemble that of the soluble bacterial IPNS In retrospect, we wonder whether the relationship is true vice versa That is, substituting the residues at these sites in any of the soluble bacterial isozymes with residues characteristic of insoluble sjIPNS will also cause the soluble IPNS enzyme to become insoluble, like sjIPNS In other words, the seven corresponding sites in the soluble bacterial IPNS may also be critical in maintaining the enzyme solubility To investigate this, scIPNS was used as the representative of the soluble bacterial isozymes, whereby site-directed mutagenesis to replace the amino acid residues at two selected sites in scIPNS with those occurring in sjIPNS was carried out Thus, Gly120 and Leu309 in scIPNS were mutated to Glu120 and Thr309 (endogenous residues in sjIPNS) respectively to construct two scIPNS single mutants known as Gly120Glu and Leu309Thr (Table 4.7)

4.3.6.1 Site-directed mutagenesis of Gly120 and Leu309 of scIPNS

Using recombinant pET-SC plasmid carrying the scIPNS gene (Sim et al., 1996) and

specific mutagenic primer pairs with the engineered Gly120Glu and Leu309Thr changes

(Appendix II), PCR mutagenesis was carried out as shown previously (Section 4.3.3.1) The

~6.3kbp amplified products (Fig 4.23a) were incubated with Dpn1 enzyme and thereafter transformed into E coli BL21(DE3) cells The putative scIPNS Gly120Glu mutant gene was

identified using the OL48 forward primer while the OL74 reverse primer was used in the sequencing screening of putative scIPNS Leu309Thr mutant gene (Fig 4.23b)

4.3.6.2 Expression analysis of scIPNS mutants

At 25°C, the expression of wildtype sjIPNS and scIPNS in E coli are differentiated based on their contrasting levels of soluble enzymes produced Whereas only ~4% of soluble sjIPNS were obtained at 25°C, ~25% of soluble scIPNS were produced Therefore, it is interesting to examine how the substitutions of amino acid residues at strategic sites 120 and

309 in scIPNS with residues endogenous to sjIPNS affect the expression of soluble scIPNS at

this temperature Hence, the expression of wildtype and mutant scIPNS in E coli BL21(DE3)

at 25°C were compared (Fig 4.24) Remarkably, an obvious reduction was seen in the levels

of soluble mutant Gly120Glu and Leu309Thr produced when compared with wildtype scIPNS Scanning densitometry revealed that only ~11% of mutant Gly120Glu and Leu309Thr were obtained in the soluble protein fractions (Fig 4.24) This corresponds to ~56% decrease in the synthesis of soluble scIPNS at 25°C

Although the expression of soluble mutant Gly120Glu and Leu309Thr were impaired

at 25°C, both scIPNS mutant proteins were expressed in higher levels in the soluble fractions when the induction temperature was lowered to 18°C, comparable to the wildtype levels Thus, the two specific substitutions in scIPNS have each caused it to exhibit low temperature-

Chapter 4 Results

148

Table 4.7 Summary of scIPNS mutants constructed

Amino acid position

Fig 4 23 Gel electrophoresis of the in vitro PCR site-directed mutagenesis reactions

for scIPNS single mutants and sequence confirmation of the respective mutants

(a) Lane 1 shows the λHindIII DNA marker The reaction products derived from in vitro PCR directed mutagenesis of scIPNS Gly120Glu and Leu309Thr mutants are resolved by gelelectrophoresis in lanes 2 to 3 (b) Sequencing results of the two scIPNS single mutants are presented in the table below The specific nucleotide regions sequenced are shown and the affected codons are underlined

site-1 2 3 kbp

23.1 9.4 6.6 4.4 2.3 2.0

Fig 4.24 Expression analysis of scIPNS single mutants

The soluble protein fractions of Gly120Glu and Leu309Thr scIPNS mutants expressed at (a) 25°C and (b) 18°C were analyzed using SDS-PAGE Lane M shows the protein standard of various molecular sizes (kDa) The percentages of scIPNS expressed in the various soluble protein fractions were measured by densitometric scanning and the values were plotted in the charts shown below The arrows indicate the position of scIPNS expressed

Chapter 4 Results

dependent solubility Incredibly, the two scIPNS mutants have become more like sjIPNS, being soluble only at temperature lower then 25°C Clearly the expression results indicated that by removing the amino acid conservedness shared by all soluble bacterial IPNS at strategic locations (sites 120 and 309) in scIPNS partially impaired the production of soluble scIPNS Thus, the residue makeup of elucidated sites 120 and 309 are not only causal factors for the low solubility of sjIPNS, they are also important for maintaining the highly soluble nature of scIPNS and possibly also for other soluble bacterial isozymes The same may be true for the rest of the determinant sites that were not tested here

4.4 Characterization of sjIPNS mutants via computational analysis

The varied effects of different amino acid substitutions in sjIPNS on its solubility unveiled the intricate relationship between protein sequence and structure, in that only specific changes in the primary sequence could affect protein folding and unfolding, resulting in either more soluble protein or aggregate formation Hence, it is pertinent to examine how each of these mutations possibly affects the protein structure of sjIPNS to decipher if there are any specific changes associated with observed increased solubility Therefore, in the following analysis, attempts were made to relate the primary, secondary and tertiary structures of sjIPNS

mutants to their soluble expression in E coli

4.4.1 Properties of the amino acid residues substituted in sjIPNS mutants

To investigate whether there is any specific pattern in the type of amino acids altered with respect to protein solubility, the chemical structures and selected properties of the amino acid residues involved were analyzed (Table 4.8) The sjIPNS single mutants were grouped broadly under soluble and insoluble mutants The background of the rows in Table 4.8 were

colored according to whether the amino acid replacement in each mutant constituted a conserved or non-conserved change (yellow, conserved; blue, non-conserved)

The volume changes of the respective amino acid substitutions were calculated using values cited by Zamyatin (1972) (Appendix V) Mutations that involved substitutions with residues that occupy a bigger volume space were indicated by positive values (Table 4.8) Conversely, negative values reflect mutations involving replacement with a residue that occupies a smaller volume space The hydrophobicity index changes were calculated based on the values cited by Kyte and Doolittle (1982) (Appendix V) Negative values indicate substitutions involving replacement by a less hydrophobic or more hydrophilic residue (Table 4.8) Vice versa, positive values mean that the substitutions involved replacement by a more hydrophobic or a less hydrophilic residue

It appeared that the different solubility of the mutants is not plainly related to the size

of the amino acid residues substituted For example, amongst the eight mutations that increased sjIPNS solubility, half encompass an increase in residue volume space while the other half a reduction in volume space The same was observed for the insoluble mutants As a note, the mutation of tyrosine at position 160 to phenylalanine (mutant Tyr160Phe) and leucine (mutant Tyr160Leu) both involved reduced amino acid volume space However, phenylalanine substitution with smaller volume change (-3.7) improved sjIPNS solubility but not leucine substitution which incurred a higher volume change (-26.7) On the contrary, both valine (mutant Thr309Val) and leucine (mutant Thr309Leu) substitutions at amino acid position 309 with +23.9 and +50.6 volume change respectively resulted in equally soluble mutant proteins Hence, the magnitude of residue volume change seems to be important for the outcomes of substitutions at amino acid position 160 but not for position 309

Analysis of the changes in the hydrophobic and hydrophilic properties of the amino

acid residues substituted concurs an earlier observation, that is majority of the mutations have

a positive value for hydrophobicity index change, i.e replacement by a more hydrophobic residue Particularly fascinating is that all but one of the soluble sjIPNS mutants involved substitutions by a more hydrophobic residue Hence, it would be interesting to know the spatial

b The amino acid volume changes were calculated from values cited in Zamyatin, 1972

c The hydrophobicity index changes were calculated using values cited from Kyte and Doolittle, 1982

d Rows shaded yellow depict conserved substitutions whereas rows shaded blue depict non-conserved changes

a Amino acid main chain atoms

Hydrophobicity index change

Table 4.8 (ctd.)

(b) Insoluble sjIPNS single mutants

Tyr160 Leu

locations of these substituted residues in sjIPNS, to examine whether the residues are located

on solvent exposed surfaces or in the buried hydrophobic core

Next, the thirteen amino acid substitutions were assessed with regards to whether the change was conserved or non-conserved Conserved substitutions were found in mutant Ala37Cys, Asp112Glu, Ile156Leu and Ala202Pro, which involved replacement by an amino acid residue of the same property Take mutant Ala37Cys for example, the hydrophobic alanine at position 37 was substituted with cysteine, which is also hydrophobic The remaining substitutions involved non-conserved changes For instance, the polar uncharged serine residue

at position 19 was replaced by a highly hydrophobic phenylalanine in mutant Tyr19Phe Apparently, all of the soluble sjIPNS mutants except mutant Asp112Glu have non-conserved amino acid substitutions whereas 3 out of the 5 insoluble sjIPNS mutants (mutant Ala37Cys, Ile156Leu and Ala202Pro) have conserved changes Hence, it seems that the improved solubility effect may be more reliant on a change in the amino acid property at the mutated site

Here, the substituted residues were analyzed purely in terms of certain amino acid properties However, the local environments of the substituted residues also play an important role in modulating the outcomes of the amino acid replacements For example, an increase in residue volume size in a crowded local environment may result in steric strain As such, it is important to include the structural analysis of the interrelated local environments of the mutated sites other than specific amino acid properties such as size and charge

4.4.2 Computational analysis of the biophysical properties of sjIPNS mutants

Wilkinson and Harrison (1991) used six easily calculated parameters derived from the

amino acid composition of 81 proteins to analyze the cause of inclusion body formation in E coli in a statistical investigation They concluded that the fraction of turn-forming residues and the approximate charge average of a protein showed the strongest correlation with formation

of insoluble proteins These parameters, together with the GRAVY (grand average of

Chapter 4 Results

hydropathicity) and pI values of sjIPNS mutants were analyzed in relation to their solubility and the results obtained are presented below

Fraction of turn-forming residues Turns are the more difficult structures to form

during protein folding Using the Chou and Fasman index (Chou and Fasman, 1978), five amino acid residues, aspartate, asparagine, proline, glycine and serine, have been proposed to

be involved in forming the turns of proteins Thus, if a protein contains a high proportion of these residues, the protein may be inclined to form inclusion bodies due to slower folding rate (Wilkinson and Harrison, 1991) The sjIPNS mutants in Table 4.9 are grouped under soluble and insoluble mutants with respect to their solubility at 25ºC The fractions of the turn-forming residues in various mutants were calculated and there was no clear relationship between these values and protein solubility Both decrease (e.g mutant Ser19Phe and Asp112Glu) and increase (e.g mutant Glu120Gly, Glu120Gly/Thr309Val and Glu120Gly/Thr308Ala/ Thr309Val) in the fractions of turn-forming residues were observed amongst the soluble mutants

Approximate charge average Debye-Hückel equation stated that the logarithmic of

protein solubility is proportional to the square of the net protein charge (Tanford, 1958) Hence, protein solubility increases with increasing net charge, positive or negative In Table 4.9, the approximate net charges of the mutants were calculated by subtracting the total number of basic residues (lysine and arginine) from the total number of acidic residues (aspartate and glutamate) and dividing this number by the total number of amino acid residues

in sjIPNS Majority of the soluble mutants showed no changes in their net protein charges compared to wildtype sjIPNS except mutant Glu120Gly, Arg242Thr, Glu120Gly/Thr309Val and Glu120Gly/Thr308Ala/ Thr309Val Thus, it is obvious that the solubility of sjIPNS mutants is also not related to this parameter

GRAVY and pI scores Most of the mutants showed a more positive value for their

GRAVY scores which accord with the prior observation that most of the substitutions involved replacements with a more hydrophobic residue However, the increase in GRAVY scores alone cannot account fully for the improved solubility observed amongst the mutants as substitution

turn-Approximate charge average b

GRAVY score c pI c Expression

level (expressed

as % of total soluble proteins) d

Soluble sjIPNS mutant

a The fractions of turn-forming residues (includes aspartate, asparagine, proline, glycine and serine) in

sjIPNS mutants were expressed as ratios of the sjIPNS protein sequence length (329 amino

acids)

b Approximate charge average was calculated using the difference between the number of basic

(arginine and lysine) and acidic (aspartate and glutamate) residues divided by the total number of

amino acid residues in sjIPNS

c GRAVY and pI of various mutants were predicted from ProtParam tool (website:

http://expasy.hcuge.ch/cgi-bin/ )

d The percentages of soluble sjIPNS mutants expressed at 25ºC were obtained by densitometric

scanning

Chapter 4 Results

of Tyr160 by more hydrophobic phenylalanine and leucine residues respectively improved the solubility of the former mutant but not the latter mutant This indicates that the local environments of the mutated sites play a significant role too The pI values obtained for each mutant did not show any obvious correlation with the expression of soluble protein (Table 4.9)

In the study of human thymidylate synthase (McElroy et al., 1992), it was reported that the

increased solubility observed in some of the mutants was not related to the pI values of these enzymes However, another group noted that the pI values were related to the solubility of

dihydrofolate reductase mutants (Dale et al., 1994) The reason for the varied relationship

between isoelectric points and solubility in different proteins is still unknown

4.4.3 Secondary structure analysis of sjIPNS mutants

King et al (1989) have proposed that partially folded protein intermediates with

distinct secondary structures are responsible for the formation of inclusion bodies In addition,

an analysis of scIPNS random mutants with reduced solubility (Sim and Sim, 1999) revealed that specific amino acid substitutions that will affect secondary structure predictions exert a greater influence on protein solubility than trivial assessment of other biophysical parameters

In the above study, two protein secondary structure prediction programs, neural network

prediction (nnPred) (Kneller et al., 1990) and self-optimized prediction method (SOPM)

(Geourjon and Deléage, 1994) were used for analyzing the secondary structures of scIPNS

mutants Both programs have been validated to predict the secondary structure elements of A nidulans IPNS (alPNS) for which the crystal structure has been elucidated (Roach et al., 1995;

1997) with at least 65% accuracy The results showed that mutants with greater changes in the predicted secondary structures had reduced solubility compared to wildtype enzyme These observations prompted us to make use of a further improved version of the self-optimized prediction method (SOPMA) (Geourjon and Deléage, 1995) to examine the secondary structures of various sjIPNS mutants

The SOPMA predicted secondary structures of the respective sjIPNS mutants were aligned with the SOPMA predicted secondary structure of wildtype sjIPNS and the differences were highlighted (Fig 4.25a) It appeared that sjIPNS mutants had changes in predicted secondary structure profiles that were not confined to specific regions of the protein However,

an interesting phenomenon observed from the alignment is that the majority of the changes in the predicted secondary structures of the various mutants seem to coincide at the same regions (highlighted by the solid bars in Fig 4.25a)

The number of residue sites where the secondary structure elements were altered with respect to the wildtype was enumerated and these values were plotted against the solubility data (Fig 4.25b) However, there seems to be no obvious relationship between the predicted secondary structure perturbations and solubility For instance, mutants possessing less than ten sites altered in the predicted secondary structures were both found to be soluble (showed by the “♦” and “ ♦ ”) and insoluble (showed by the “ ♦ ”) Hence, the improved solubility of the sjIPNS mutants could not be simply explained by secondary structural changes

4.4.4 Tertiary structural analysis of sjIPNS mutants

The results obtained thus far indicated that protein insolubility is not entirely dependent on parameters such as hydrophobicity, charge and identity of the amino acid replaced, neither is it plainly related to the biophysical properties of the mutated proteins, such

as GRAVY and pI values Hence, the next step in understanding the mutational effects on sjIPNS solubility is to analyze the spatial locations and the local environments of the substituted residues in the three-dimensional structure of sjIPNS

sjIPNS 1 CCEECCCCCCCEEEECCCCCCCHHHHHHHHHHHHHHHTTTTEEEEECTTCCHHHHHHHHHHHHHHCCCTTCHHEEEEHCCTTCCCCCCCEEEECTTCCCHHHHEECCTTCCTTCCCCCTT S19F 1 CCEECCCCCCCEEEECCCCCCCCHHHHHHHHHHHHHHTTTTEEEEECTTCCHHHHHHHHHHHHHHCCCTTCHHEEEEHCCTTCCCCCTTEEEECTTCCCHHHHEECCTTCCCCCCCCCCC D112E 1 CCEECCTTCCCEEEECCCCCCCHHHHHHHHHHHHHHHTTTTEEEEECTTCCHHHHHHHHHHHHHHCCCTTCHHHEEEHCCTTCCCCCCCEEEECTTCCCHHHHEECCTTCCTTCCCCCTT E120G 1 CCEECCCCCCCEEEECCCCCCCHHHHHHHHHHHHHHHTTTTEEEEECTTCCHHHHHHHHHHHHHHCCCTTCHHEEEEHCCTTCCCCCCCEEEECTTCCCHHHHEECCTTCCTTCCCCCTT Y160F 1 CCEECCTTCCCEEEECCCCCCCHHHHHHHHHHHHHHHTTTTEEEEECTTCCHHHHHHHHHHHHHHCCCTTCHHEEEEHCCTTCCCCCCCEEEECTTCCCHHHHEECCTTCCTTCCCCCTT R242T 1 CCEECCCCCCCEEEECCCCCCCHHHHHHHHHHHHHHHTTTTEEEEECTTCCHHHHHHHHHHHHHHCCCTTCHHEEEEHCCTTCCCCCCCEEEECTTCCCHHHHEECCTTCCTTCCCCCTT T308A 1 CCEECCTTCCCEEEECCCCCCCHHHHHHHHHHHHHHHTTTTEEEEECTTCCHHHHHHHHHHHHHHCCCTTCHHEEEEHCCTTCCCCCCCEEEECTTCCCHHHHEECCTTCCTTCCCCCTT T309V 1 CCEECCCCCCCEEEECCCCCCCHHHHHHHHHHHHHHHTTTTEEEEECTTCCHHHHHHHHHHHHHHCCCTTCHHEEEEHCCTTCCCCCCCEEEECTTCCCHHHHEECCTTCCTTCCCCCTT T309L 1 CCEECCCCCCCEEEECCCCCCCHHHHHHHHHHHHHHHTTTTEEEEECTTCCHHHHHHHHHHHHHHCCCTTHHHHEEHHCCTTCCCCCCCEEEECTTCCCHHHHEECCTTCCTTCCCCCTT

A 7C 1 CCEECCTTCCCEEEECCCCCCCHHHHHHHHHHHHHHHTTTTEEEEECTTCCHHHHHHHHHHHHHHCCCTTHHHHEEEHCCTTCCCCCCCEEEECTTCCCHHHEEECCTTCCTTCCCCCTT I156L 1 CCEECCTTCCCEEEECCCCCCCHHHHHHHHHHHHHHHTTTTEEEEECTTCCHHHHHHHHHHHHHHCCCTTCHHEEEEHCCTTCCCCCCCEEEECTTCCCHHHHEECCTTCCTTCCCCCTT Y160L 1 CCEECCCCCCCEEEECCCCCCCHHHHHHHHHHHHHHHTTTTEEEEECTTCCHHHHHHHHHHHHHHCCCTTCHHEEEEHCCTTCCCCCCCEEEECTTCCCHHHHEECCTTCCTTCCCCCTT A202P 1 CCEECCCCCCCEEEECCCCCCCHHHHHHHHHHHHHHHTTTTEEEEECTTCCHHHHHHHHHHHHHHCCCTTCHHEEEEHCCTTCCCCCCCEEEECTTCCCHHHHEECCTTCCTTCCCCCTT H260N 1 CCEECCCCCCCEEEECCCCCCCHHHHHHHHHHHHHHHTTTTEEEEECTTCCHHHHHHHHHHHHHHCCCTTCHHEEEEHCCTTCCCCCCCEEEECTTCCCHHHHEECCTTCCTTCCCCCTT

E120G/T309V 1 CCEECCTTCCCEEEECCCCCCCHHHHHHHHHHHHHHHTTTTEEEEECTTCCHHHHHHHHHHHHHHCCCTTCHHEEEEHCCTTCCCCCCCEEEECTTCCCHHHHEECCTTCCTTCCCCCTT T308A/T309V 1 CCEECCCCCCCEEEECCCCCCCCHHHHHHHHHHHHHHTTTTEEEEECTTCCHHHHHHHHHHHHHHCCCTTCHHEEEEHCCTTCCCCCTTEEEECTTCCCHHHHEECCTTCCCCCCCCCCC E120G/T308A/T309V 1 CCEECCCCCCCEEEECCCCCCCCHHHHHHHHHHHHHHTTTTEEEEECTTCCHHHHHHHHHHHHHHCCCTTCHHHEEEHCCTTCCCCCCTEEEECTTCCCHHHHEECCTTCCTTCCCCCCC consensus 1 ****** **************.***********************************************.**.**.********** *************.******** ***** sjIPNS 121 CCCCCEECCCCTTCCTTCHHHHHHHHHHHHHHHHHHHHHHHHHHTCCHHHHHHHHHHHHHHHHEEEEECCCCCCCCCCCCCCTTCEEEEECCCCCHEEEEEEHCCCCCEEEECCCCCEEC S19F 121 CCCCCEECCCCCCCCTTCHHHHHHHHHHHHHHHHHHHHHHHHHHTCCCHHHHHHHHHHHHHHHEEEEECCCCCCCCCCCCCCCCCEEEEEHHCCHHEEEEEECCCCCCEEEECCCCCCCC D112E 121 CCCCCEEECCCTTCCTTHHHHHHHHHHHHHHHHHHHHHHHHHHHTCCHHHHHHHHHHHHHHHHHEEEECCCCCCCCCCCCCCTTCEEEEECCCCCHEEEEEEHCCCCCEEEECCCCCEEC E120G 121 CCCCCEECCCCTTCCTTCCHHHHHHHHHHHHHHHHHHHHHHHHHTCCCHHHHHHHHHHHHHHHHEEEECCCCCCCCCCCCCCTTCEEEEEHHCCHHEEEEEEHCCCCCEEEECCCCCCCC Y160F 121 CCCCCEECCCCTTCCTTCHHHHHHHHHHHHHHHHHHHHHHHHHHTCCHHHHHHHHHHHHHHHHHEEEECCCCCCCCCCCCCCTTCEEEEECCCCCHEEEEEEHCCCCCEEEECCCCCEEC R242T 121 CCCCCEECCCCTTCCTTCCHHHHHHHHHHHHHHHHHHHHHHHHHTCCCHHHHHHHHHHHHHHHHEEEECCCCCCCCCCCCCCTTCEEEEEHHCCHHEEEEEEHCCCCCEEEECCCCCCCC T308A 121 CCCCCEECCCCTTCCTTCHHHHHHHHHHHHHHHHHHHHHHHHHHTCCHHHHHHHHHHHHHHHHHEEEECCCCCCCCCCCCCCTTCEEEEECCCCCHEEEEEEHCCCCCEEEECCCCCEEC T309V 121 CCCCCEECCCCTTCCTTCHHHHHHHHHHHHHHHHHHHHHHHHHHTCCHHHHHHHHHHHHHHHHEEEEECCCCCCCCCCCCCCTTCEEEEECCCCCHEEEEEEHCCCCCEEEECCCCCEEC T309L 121 CCCCCEEECCCTTCCTTCHHHHHHHHHHHHHHHHHHHHHHHHHHTCCCHHHHHHHHHHHHHHHHEEEECCCCCCCCCCCCCCTTCEEEEEHCCCHHEEEEEEHTCCCCEEEECCCCCEEC A3 7C 121 CCCCCEECCCCCCCCTTCHHHHHHHHHHHHHHHHHHHHHHHHHHTCCHHHHHHHHHHHHHHHHEEEEECCCCCCCCCCCCCCTTCEEEEECCCCCHEEEEEEHCCCCCEEEECCCCCEEC I156L 121 CCCCCEECCCCTTCCTTCHHHHHHHHHHHHHHHHHHHHHHHHHHTCCHHHHHHHHHHHHHHHHHEEEECCCCCCCCCCCCCCTTCEEEEECCCCCHEEEEEEHCCCCCEEEECCCCCEEC Y160L 121 CCCCCEECCCCTTCCTTCHHHHHHHHHHHHHHHHHHHHHHHHHHTCCHHHHHHHHHHHHHHHHEEEEECCCCCCCCCCCCCCTTCEEEEECCCCCHEEEEEEHCCCCCEEEECCCCCEEC A202P 121 CCCCCEECCCCTTCCTTCCHHHHHHHHHHHHHHHHHHHHHHHHHTCCCHHHHHHHHHHHHHHHHEEEECCCCCCCCCCCCCCTTCEEEEEHHCCHHEEEEEEHCCCCCEEEECCCCCCCC H260N 121 CCCCCEECCCCTTCCTTCHHHHHHHHHHHHHHHHHHHHHHHHHHTCCHHHHHHHHHHHHHHHHEEEEECCCCCCCCCCCCCCTTCEEEEECCCCCHEEEEEEHCCCCCEEEECCCCCEEC

E120G/T309V 121 CCCCCEECCCCTTCCTTCHHHHHHHHHHHHHHHHHHHHHHHHHHTCCHHHHHHHHHHHHHHHHHEEEECCCCCCCCCCCCCCTTCEEEEECCCCCHEEEEEEHCCCCCEEEECCCCCEEC T308A/T309V 121 CCCCCEECCCCCCCCTTCHHHHHHHHHHHHHHHHHHHHHHHHHHTCCCHHHHHHHHHHHHHHHEEEEECCCCCCCCCCCCCCCCCEEEEEHHCCHHEEEEEECCCCCCEEEECCCCCCCC E120G/T308A/T309V 121 CCCCCEECCCCCCCCTTCHHHHHHHHHHHHHHHHHHHHHHHHHHTCCCHHHHHHHHHHHHHHHEEEEECCCCCCCCCCCCCCCCCEEEEEHHCCHHEEEEEECCCCCCEEEECCCCCEEC consensus 121 *******.*** **** ****************************.***************.****************** ****** **.******* ************* * sjIPNS 241 CCCCCEEEEETTCEEEEECTTCCCCCCCEEEEECCTTCCCEEEECCCCCCEECCCCCCCCCCCCCCCCCCHHHHHHHHHHHHHHTTTCC

The positions where there are changes in the predicted secondary structures of sjIPNS mutants are highlighted

* “E”, β-strand; “H”, α-helix;

“T”, turn and “C’, coil

% of sjIPNS soluble expression

♦soluble sjIPNS single mutants

♦insoluble sjIPNS single mutants

♦soluble sjIPNS double & triple mutants

Chapter 4 Results

4.4.4.1 Modeling of bacterial IPNS tertiary structures

To date, only the crystal structure of A nidulans IPNS (aIPNS) of fungal origin has been elucidated (Roach et al., 1995; 1997; Burzlaff et al., 1999) Due to the lack of a crystal

structure for sjIPNS, homology modeling using SWISS-MODEL program (Guex and Peitsch,

1997; Guex et al., 1999) was thus carried out for sjIPNS to obtain its tertiary protein structure

for the analysis of the spatial locations of the substituted sites in sjIPNS Correspondingly, the tertiary structures of scIPNS, sfIPNS, slIPNS and nIPNS were also modeled using the same program The predicted structures of the bacterial IPNS isozymes were subsequently viewed using the molecular visualization software, SwissPdb Viewer program (SPdbV) All structural analyses were performed using the manipulation tools in SPdbV program as described in

Section 3.2.11.2.3

The predicted secondary and tertiary structures of bacterial IPNS were compared with the experimentally crystallized structures of aIPNS to access the quality of the structures predicted by SWISS-MODEL program Secondary structure alignment analysis was done by manually highlighting the color encoded SWISS-MODEL predicted secondary structures (yellow for strands, red for helices and gray for loops) of bacterial IPNS isozymes onto their corresponding aligned sequence positions The alignment results showed that the secondary structures are vastly similar amongst the bacterial IPNS isozymes as well as between the bacterial IPNS and aIPNS isozymes, with about 99.1 to 96.0% identity (Fig 4.26) For tertiary structure comparison, the predicted structure of each bacterial IPNS enzyme and the aIPNS crystal structure were superimposed onscreen by the SPdbV program and the root mean square (RMS) value which reflects the degree of deviation between two structures was calculated by the program The predicted sjIPNS structure aligned very well with the aIPNS crystal structure whereby the positions of the Cα atoms of the two structures only deviated by a distance of 0.39Å (Fig 4.27a) Similarly high structure resemblance was also observed for the superimposition of predicted scIPNS, sfIPNS, slIPNS and nIPNS structures onto aIPNS crystal