6 DISCUSSION

↓93

The present investigation elucidates the significant role of electrostatic interactions in maintaining the stability of the non-fusogenic structure of HA protein of influenza virus (X-31 strain) at neutral pH. The precursor HA protein is cleaved by host proteases which results in the formation of HA1 and HA2 domains. The HA protein at this stage is believed to be metastable. Low pH triggers a conformational change of the metastable HA into a thermodynamically low energy state. This conformational change exposes the otherwise deeply embedded hydrophobic fusion peptide initiating fusion with host target membrane. Exposure of fusion peptide from the HA2 domain is preceded by dissociation of the globular head domains of HA1 region. Site directed mutagenesis of various amino acids in HA2 domain highlights the importance of respective residues in the conformational process and thereby in the fusion (Gething et al., 1986; Qiao et al., 1999; Han et al., 1999; Cross et al., 2001; Gruenke et al., 2002). Though the dissociation of HA1 domains was proved to be essential in the whole process, the destabilising forces are supposed to be due to enhanced protonation of the titratable amino acids on the HA1 domain (Daniels et al., 1985).

↓94

Huang et al. (2002) pointed out the importance of electrostatic interactions for the stability and destabilisation of HA trimer. The scope of the present research is to determine the contribution of charged residues (ion pairs) to the stability of the cleaved HA protein in its non-fusogenic state. For this purpose, amino acids were selectively mutated, and the effect of mutations on conformational change was analysed. Target of the mutations were those amino acids, which form ion pairs (Fig. 5.1.1-5.1.5). Mutagenesis was aimed, either to destabilise or stabilise the protein (breaking or making salt bridges, respectively). The present work is unique and novel, as it demonstrates the relevance of salt bridges at the interface of HA1 and HA2 for the stability of the HA ectodomain. At the same time it is of relevance to prove that disruption is accompanied with the conformational change of HA. It is for the first time, that the important amino acids (involving salt bridges) which are responsible for the stability of the non-fusogenic HA protein has been identified. The results denote that the conformational transition of the HA ectodomain in different influenza strains could be directly linked to the mentioned salt bridges.

6.1  Choice of mutation sites

All the mutations chosen for the study involved the residues either at the intra-monomer or at the inter-monomer interface. The interface between HA1 and HA2 regions (intra-monomer) gain importance as the HA2 in this region undergoes “loop to helix” transformation (55-75 residues). The region between the residues 64-72 of HA2 is in direct contact with the globular domain of HA1. This region (64-72) shows more than 70 % conservation among all the influenza A subtypes. The residues 85–90, 104–115, 265–270, and 299-307 form the interface region on the HA1 side. At least 20 out of 33 residues within these regions are conserved to a great extent (color index >7, see Fig.5) among the entire influenza A family. Daniels et al. (1985) analysing various naturally occurring mutants surmised that substitutions usually occurred either at interface region of HA1 and HA2 or at the amino-terminus of HA2. Furthermore, the authors quoted that an extensive breakage of intra and inter subunit contacts could possibly lead to conformational change of HA. In accordance with this hypothesis, Huang et al. (2002) proposed that protonation of solvent exposed negatively charged amino acids, disrupt electrostatic interactions between residues (i.e. salt bridges). This disruption causes a conformational change leading to a fusion active state of the HA. To prove this model experimentally, the present study lays focus on the interface region between HA1 and HA2, especially salt bridges. A series of mutants (between HA1 and HA2 of a monomer) were constructed either to destabilise or to stabilise the HA protein (section 5.1.1-5.1.4). Similarly, this study also identified inter-monomer interactions contributing to the stability of the HA protein. In particular, inter-monomer interactions between the three HA1 monomers were studied in the distal region, because the initial step in the process of the conformational change is linked to changes in this region. The role of ion pairs for inter-monomer interactions and for the stability of the ectodomain was determined by targeting salt bridges in the distal region of HA2 monomers (section 5.1.5).

All the selected amino acids for mutagenesis were checked for their homology among influenza A virus strains. Figure. 5.2 show that most of the mutated residues were conserved ranging from average to a higher degree. The most conserved residues involving mutants were Arg 109, Ser 110, Lys 299, Tyr 308 of HA1 chain. In addition, the sequence alignment of the HA2 domain showed that Glu 74 and Ile 77 were conserved to a higher degree. Further, the tetrad salt bridge involving Arg 109, Glu 89, Arg 269 of HA1 chain with Glu 67 of HA2 chain has most of the residues conserved to a higher degree. Arg 269 is the only exception with low conservation, while negative charge at position 67 of HA2 is conserved to approximately 70 %. This shows the evolutionary importance of this salt bridge among all the members of influenza A virus.

6.2 Mutants construction and expression in mammalian cells

↓95

A prerequisite for assessing the fusion activities of HA expressing cells, is good surface expression of the mutant HA or HA-wt proteins. In connection with this, the surface expression of the HA-wt, and the mutant proteins was adjusted by varying the amount of lipofectin and plasmid DNA. Figs. 5.8.1a-5.8.1f shows that all the mutants were expressed at comparable levels in relation to HA-wt, with the exception of R109E. The failure of surface expression of R109E could be attributed to failure in transport of the mutant protein from Golgi complex to the cell surface (section 5.6, and Fig.5.6). The ability of all the mutants (except R109E) to show cell surface expression and further the formation of HA1 and HA2 by TPCK trypsin treatment showed that mutants formed successful trimers. This is confirmed by chemical crosslinking with DSP (section 5.7 and Fig.5.7). The mutants R109G, R269E, T212E-N216R, and T212E-N216E also showed trimer formation with DSP reagent (results not shown). Even the ability of R109E mutant protein to form trimer underlines the fact that trimerisation is a phenomenon occurring in the ER (Copeland et al., 1988; Boulay et al., 1988). The inability of the surface expression of R109E mutant protein has been discussed in subsequent sections.

6.3 Proteinase K assay revealed difference in conformational changes

The active form of the HA protein undergoes an irreversible conformational change to the fusogenic form depending on the pH milieu. The conformational changes can be probed by the proteinase K assay (section 1.4). The low pH conformational change of the HA is an example of a proton sensitive conformational switch involving Bohr protons (Daniels et al., 1985). The threshold pH for conformation change varies among different strains of influenza A virus. Therefore, this proteinase K assay is very crucial to ascertain whether the desired mutant has a stabilising or destabilising effect on the conformational change.

Figure 5.8 show that HA-wt is sensitive to proteinase K at pH ≤ 5.2. On comparison, it was found that all the destabilising mutants showed a higher pH threshold for the conformational change (see fig 5.8.1a-5.8.1f and table 1). It implies that these destabilising mutants lowered the energy barrier requirements for the conformational change. A close look at the Gly and Glu substitutions at the same position, for instance R269G and R269E, respectively, show that Glu substitutions resulted in a conformational change at a higher pH than those with Gly. This confirms that Glu substitutions generated repulsion forces due to Glu-Glu (-ve to –ve) interactions and lowered the energy barrier requirements for the respective mutants. These results indicate that substitutions lowering the required energy could be due to reduced interaction (or increased repulsion) between the subunit interfaces. In addition, increased repulsion (or lowering the required energy) lowers the concentration of protons required to trigger the conformational change.

↓96

Thus the mutations show a decreased pH (proton) requirement for the conformational change in the following order -

Normal ion pair (HA-wt) > no ion pairs (Gly substitutions) > repulsion ion pairs (Glu substitutions).

The stabilising mutations (Fig.5.8.2a-5.8.2c) on the other hand showed no conformational change upon proteinase K treatment till pH 5.0. The only exception being S110D (Fig.5.8.2c), which was intended to be a stabilising mutation but turned out to be destabilising in nature, on proteinase K treatment. Very likely, the reason is that the interacting residue His 64 with its aromatic ring failed to form an ion pair. Another reason could be that the pair Ser 110 – His 64 is in the neighbourhood of the tetrad salt bridge involving Arg 109 and Glu 67. The introduction of S110D would have perturbed the geometry of the tetrad salt bridge. Indeed, the occurrence of a salt bridge involving His residue is not common. Kumar and Nussinov (1999) studying a database of 222 non-equivalent salt bridges from 36 monomeric proteins accounted for only 4 salt bridges involving His residue. The experiments were repeated for three times and the protein concentrations were recorded as percentage taking the sample treated with pH 7.0 fusion buffer as 100 %. Accordingly the concentrations of protein bands seen on the fluorograms (i.e.,quantity or % of the protein resistant to proteinase K at each of pH analysed), were plotted on a graph using standard error as a statistical tool and taking sample size as three.

↓97

Fig.6.3: Non proteolysed HA content from both the stabilising and de-stabilising mutants after proteinase K assay.

6.4 Relation of destabilising mutants with natural variants

The mutations in the interface region of HA1 and HA2 indicated the significance of charged residues in the conformational transition of HA protein. Inspired by these results, an attempt was made to look for the reports of the natural variants harbouring these mutations and show higher pH requirements for fusion activity. It was observed that natural variants of influenza viruses (X-31 and Weybridge strains) show fusion at higher pH than the HA-wt (Daniels et al. 1985). These variants were specifically selected for their resistance to amantadine hydrochloride, which is known to increase the pH of intracellular vesicles. The amino acid sequence analysis revealed that 15 of the 17 mutants from X-31 strain showed substitutions in the HA2 region. On the contrary, 50% of the mutants from Weybridge strain showed substitutions in HA1 region. In both cases, charged residues were substituted to an extent of 75 %. Some of the substitutions were electrostatically destabilising for the salt bridges. Mutations were reported to be either in the interface region (HA1-HA2 or HA2-HA2) or in the amino terminal peptide of HA2.

Similarly Proesch et al. (1990), Hoffman et al. (1997), and Staschke et al. (1998) reported natural variants of various subtypes of influenza virus to be resistant to anti-viral drugs (norakin, tetra-butyl hydroxyquinone, and methyl-O-methyl-7-ketopodocarpate respectively). Amino acid sequences of the respective variants revealed substitutions in either amino terminal peptide of HA2 or in the interface region of HA1 and HA2. All the mutants were reported to have elevated pH of fusion.

↓98

The reports on natural variants showed several substitutions including those of charged residues, but there is no report on salt bridge destabilisation in the interface region. In other words, the charged amino acids (especially tetrad salt bridge) that were considered for the study were not substituted in any of the natural mutants. This signifies the importance of these amino acids towards the stability of the HA protein. Further, this indicates that the complex salt bridge (in the interface) could have been conserved in all the subtypes of influenza virus. The occurrence of a similar salt bridge in the interface region for H2 subtype will be discussed in the subsequent sections.

6.5 Significance of Arg 109 and tetrad salt bridge in maintaining the stability of the HA ectodomain

The Arg 109 is conserved mostly among all the influenza A viruses and also in influenza B viruses. Only incidence of its mutation was reported in H2 subtype where Arg 109 was substituted with Lys. In other words, a positive charged residue at position 109 (numbering according to X-31 strain) is conserved. As deduced from the 3D structure of X-31 subtype (PDB: 1HGD), the Arg 109 forms a salt bridge network with Glu 89 of HA1 on one side and Glu 67 of HA2 on the other side. In addition, Arg 269 is involved in this network forming a tetrad network. It may be stated that this tetrad salt bridge is especially important in maintaining the stability of the non-fusogenic conformation as it connects different parts of the intact protein. The figure below shows the distance in Å between the residues involved in the salt bridge network.

Musafia et al. (1995) studying 1105 salt bridges from 94 proteins concluded that complex salt bridges comprise about 70 % of inter subunit interactions. The significance of this tetrad salt bridge is increased by the presence of Arg, which is an abundantly found residue in inter subunit interactions (Musafia et al., 1995). Further the same authors report that Arg is abundant and constitutes nearly 43.3 % in all the complex salt bridges. Report of Xu et al. (1997) suggests that the hydrogen bonds across the interfaces are predominantly oxygen-nitrogen type. Taking into account all these considerations, it is quiet apparent that this complex salt bridge is ideally placed at the interface with perfect geometry.

↓99

The failure of the expression of R109E mutant protein could be explained by disturbance caused as a consequence of the repulsion generated by substituted Glu with that of Glu 89 of HA1 and Glu 67 of HA2.

Fig. 6.5: The tetrad salt bridge showing the distance in Å between the residues.

The glycosylation assay (Fig.5.6; lanes10-15; and section 5.6) showed that R109E mutant to be sensitive to EndoH. It may be deduced from the result that R109E could have been retained either in ER or cis-Golgi compartments. Since the immunoprecipitation assays showed that N2 antibody detected the protein, it may be said that R109E mutant successfully formed trimer. This gives an indirect hint that R109E passed through ER and possibly held up in cis-Golgi compartment. The N2 antibody is specific for neutral pH and trimer HA (Copeland et al., 1986). Its inability to recognise monomers (Copeland et al., 1986) emphasises that the mutant R109E has formed trimer. In addition, trimerisation is a phenomenon occurring in the ER organelle of the cell (Copeland et al., 1988). It is also a prerequisite for efficient transport of the protein from ER (Boulay et al., 1988). The importance of this tetrad salt bridge in stability of HA protein is highlighted by the identification of a second ligand binding site of concave pocket nature at the interface of HA1 and HA2 (Sauter et al., 1992).

6.6 Relation of stabilising mutants with H2N2 (A/JPN/305/57)

↓100

The HA protein of A/JPN/305/57 (H2 subtype) was found to be very stable compared to A/PR 8/34 (H1 subtype) and X-31 of H3 subtype. Even at pH 5.0 and 37°C, inactivation is very slow when compared to X-31 and A/PR 8/34 (Korte et al., 1999). In order to gain insight into the stability aspects of this H2 subtype, the amino acid sequence was aligned with X-31 sequence. The HA2 sequence in both the strains is highly conserved compared to HA1 region. The structure of HA of A/JPN/305/57 strain was determined at neutral and acidic pH by cryo electron microscopy (Böttcher et al., 1999). At neutral pH, the structure of HA was found to be similar to the 3D structure of X-31 strain. At acidic pH (4.9), the structure showed a continuous central cavity retaining the intact trimer structure. Based on the amino acid sequence of A/JPN/305/57 and taking 3D crystal structure as reference, the formation of salt bridges for H2 subtype was probed. Based on the homology of the HA2 amino acid sequence, it was assumed that H2 subtype may have similar HA2 structure like that of X-31 strain. Apparently HA protein would have similar interface region and thereby similar HA1 structure. In addition, Böttcher et al. (1999) reported that both the structures of A/JPN/305/57 and X-31 strains are similar, other than for a few minor differences. The first step was to verify whether all the charged residues were conserved in the H2 subtype. Based on the sequence alignment, the interface at the HA1 region of H2 subtype shows additionally 17 charged residues between 74-84, 96-125, 256-274, and 280-295. On the other hand, the number of charged residues on the interface of HA2 region is relatively same. Further, the complex salt bridge in the interface region of X-31 is conserved in the H2 subtype. The amino acids forming this salt bridge are Glu 89, Lys 109, and Lys 269 of HA1 (numbering based on X-31 strain), and Glu 69 of the HA2. In the absence of Glu 67 on the HA2 and Lys 299 on the HA1, it may be assumed that Glu 69 would form similar tetrad salt bridge in the interface region. In X-31 strain, tetrad salt bridge involves Glu 89, Arg 109, and Arg 269.

Table. 6.6a: Tetrad salt bridge formation in A/JPN/305/57 (H2 subtype)

Residues involved in salt bridge formation

HA1

HA2

Arg 109, Arg 269, Glu 89

Glu 67

Lys 299

Glu 69

A/JPN/305/57

Lys 101 (109*), Arg 264 (269*), Glu 80 (89*)

Glu 69

*Numbering based on X-31 strain, following sequence homology of A/JPN/305/57 with X-31.

Further, the trimeric shape of the HA of the A/JPN/305/57 (H2 subtype) is very stable even at pH 5.0 (mentioned earlier). The low pH usually affects the charged residues, and further the salt bridges. From on the above results, it is very much evident that salt bridges could destabilise or stabilise the HA structure. The sequence homology and subsequent alignment of the A/JPN/305/57 sequence onto X-31 3D structure identified potential salt bridges. The focus was more at the interface regions (inter and intra monomers) especially in the distal region of the HA structure. Our intention was to find the potential salt bridges responsible for the stability of the low pH structure (continuous central cavity) of the A/JPN/305/57. At least 7-8 additional possible salt bridges were identified in H2 subtypes, when compared to X-31 strain, and are summarised as shown in the table 6.6a and 6.6b.

↓101

Table. 6.6b: The different types of additional salt bridges in A/JPN/305/57 (H2 subtype) when compared to X-31 strain.

Residues involved in salt bridge formation

HA1

HA2

Between HA1 and HA2 (i.e., A-B; C-D; E-F).

His 102 (110*)

Glu 64

Within HA1

(i.e., A or C or E)

Lys 159 (165*)

Glu 240 (246*)

Between HA1 and HA2 (i.e., A-F; C-B; E-D).

Glu 98 (106*) or Glu 99 (107*)

Arg 75

Arg 206 (212*)

Glu 210 (216*)

Asp 181 (187*)

Arg 195 (201*)

Between HA2 monomers

(i.e., B-D-F)

Arg 83 - Glu 85

*Numbering based on X-31 strain, following sequence homology of A/JPN/305/57 with X-31.

Based on this analysis, it may be surmised that A/JPN/305/57 has three rows of inter-monomer salt bridges connecting all the monomers (either A-C-E or B-D-F) at three different positions (Fig 6.6). In X-31, such an inter-monomer linkage is only found between HA2 monomers (i.e., B-D-F). In addition to these inter-monomer linkages, there is one more salt bridge (Glu 98 or 99 [106* or 107*] with Arg 75 of HA2) involving HA1 and HA2 monomers in the hinge region giving it an extra stability to the HA of A/JPN/305/57 strain. This salt bridge has the potential to form a complex salt bridge in the hinge region with another salt bridge (Glu 74-Arg 76) forming a ring like network between B-D-F chains of HA2 monomers. At least one of these salt bridges i.e., T212E-N216R was generated in X-31 strain (discussed in section 6.7). The mutant was found to be stable till pH 5.0 and showed no fusion activity.

Fig. 6.6: Salt bridge networks in A/JPN/305/57 strain.

The figure shows the amino acid residues of A/JPN/305/57 strain superimposed on 3D structure of X-31 strain implicating a three-layer ring like salt bridge network. HA of A/JPN/305/57 strain has been reported to be very stable and its trimer structure is retained even at pH 5.0.

6.7 Introduction of a salt bridge in the distal region stabilises the HA trimer of X-31 strain

↓102

Locking of the distal region of the HA trimer impaired the low pH induced conformational changes and also the fusion activity (Godley et al., 1992; Kemble et al., 1992). Based on this observation, an attempt was made to introduce salt bridges in the same positions viz., T212E and N216R in a HA1 monomer (Fig.5.1.3; section 5.1.3). This double mutation was located at the HA1-HA1 interfaces and was expected to lock the trimer with a ring like complex salt bridge. Probing the conformational change by the proteinase k assay showed that the mutant is resistant to conformational changes till pH 5.0 (Fig.5.8.4). This indicates that the introduction of an ion pair in the distal region enhances the stability of the trimer. On the contrary, introduction of repulsion charges (Fig. 5.8.3) in the same position showed similar activity as that of HA-wt. This result is comparable to that of the disulfide mutation of Godley et al. (1992) and Kemble et al. (1992), where locking of the HA1 trimers abolished both the conformational change and fusion activity. It was surprising and interesting to note the stability conferred by this salt bridge to the HA trimer. Sequence alignment of A/JPN/305/57 against X-31 showed that this salt bridge exists between Arg 206 and Glu 210 in H2 subtype of A/JPN/305/57.The stability of the distal region conferred by a salt bridge would not be the same as that of disulfide bonding. However, the introduction of an additional salt bridge could have easily raised the stability by 0.4 pH units. This is confirmed by fusion assays (discussed in section 6.8). No fusion activity was observed for pH ≥ 5.0 upon incubation for 5 min at 37°C. It may be possible that either longer incubation periods or a pH < 5, could dissociate these stable salt bridges leading eventually to a conformational change. This mutant clearly demonstrates that dissociation of the globular domains is the foremost and most essential step towards a fusogenic conformation of HA. As discussed earlier, this salt bridge seems to be existent in A/JPN/305/57 strain and it may be concluded that this is one of the important factors for the stability and slow inactivation of HA of A/JPN/305/57 at pH 5.0.

6.8 Fusion assays

Membrane fusion i.e., lipid mixing upon acidification is essential to show the functional activity of all the mutants. All the destabilising mutants (except T212E-N216E) showed comparable fusion activities to that of HA-wt, but the extent of activity differed between the mutants. R109E did not show any fusion activity as was expected due to its failure to show surface expression. The fusion assay was performed at a series of pH 5.0, 5.4, 5.6, and 7.0, respectively, in order to see if there is any correlation between the pH threshold of the conformational change assessed by proteinase k assay (conformational assay) and fusion assay. For all of the destabilising mutants (except T212E-N216E) the pH threshold was 5.4 and 5.6, respectively. Thus a difference of ±0.2 pH units was observed between the proteinase k conformation assay and the fusion assay.

Fusion activity was quantitatively higher for R269E than R269G, as it was evident from conformational assay, while the former requires less energy to gain fusion competent structure and hence more fusion activity. On the contrary, the K299G mutant showed relatively more fusion activity than K299G. Further, the K299G mutant showed fusion activity even at pH 5.6, while the conformational assay shows sensitivity only below pH 5.4. Similarly, T212E-N216E mutant though was sensitive to conformational change below pH 5.2, showed no fusion activity even at pH 5.0. The stabilising mutants did not show any fusion at any of the pH, and was in agreement with proteinase K assay.

↓103

In the present state of investigation, no conclusive explanation was possible for this difference. One reason could be that the pH dependence of the exposure of the fusion peptide is different from that of the complete exposure of the hinge region of HA1, which requires a more acidic pH. Another reason could rely in the specific conditions of the assays. While the fusion assay is done in the presence of the target membrane, the proteinase K assay does not involve interaction with the target membrane. It may be that the conformational change is supported by interaction with the target membrane.

Thus to state, the fusion assay confirmed the conformational changes based on proteinase K assay and thereby the fusion abilities of destabilising and stabilising mutants.

6.9 The conformational change also requires interaction of the acidic solvent with the HA2 monomers

There is a notion that acidic pH is required only for the dissociation of globular head domains of HA1. The report of Chen et al. (1995) that HA2 domain in the absence of HA1 is capable of attaining fusion competent structure, even at neutral pH gives credence to this notion. To ascertain this, salt bridge network (E74-R76) connecting all the three HA2 domains (Fig.5.1.5) was targeted. This salt bridge is present at the interface of HA2 monomers in the distal region, and moreover in the neighbourhood of the tetrad salt bridge of the hinge region. Hence it was assumed that this salt bridge will have similar effect on the conformational transition of the HA protein like that of tetrad salt network in the hinge region. To investigate the contribtution of this salt bridge to the stability of HA, Glu 74 and Arg 76 were substituted with E74C-R76C.

↓104

If the Cys residues formed disulfide bond connecting all the three HA2 monomers, the proteinase K assay profile of the mutant HA protein would probably resemble HA-wt. Then it may be deduced that conformational changes in HA2 are independent of low pH. In other words, the acid pH resistant covalent linkage should not change the pH threshold of the cys mutant HA protein. On the contrary, E74C-R76C failed to show good surface expression compared to HA-wt. A very little quantity of the expressed HA is cleaved by TPCK trypsin indicating that the majority of the HA remained in the cell (Fig 5.8.5; lane 1). The distance from the Arg 329 (trypsin cleavage site; Fig 1.2.1) rules out the possibility of trypsin cleavage site being affected in this mutant. In addition, Glu 74 and Arg 76 are totally covered by the globular head domains of the HA1 and they unlikely affect the trypsin cleavages site. Based on these assumptions, the mutations of Glu 74 and Arg 76 should not affect the surface expression of the mutant HA protein. The lack of the cell surface expression indicates the role of these charged residues in the surface expression of the protein. However, the formation of a disulfide linkage for E74C-R76C mutant was confirmed from the fig. 5.9, lane 2 (section 5.9). The conformational assay using proteinase K showed that the E74C-R76C mutant protein was relatively resistant to conformational changes (Fig 5.8.5a).

Moreover, the study investigated, if the lack of surface expression of E74C-R76C has been due to the absence of E74-R76 salt bridge or if the presence of Cys has changed the conformation of the HA. It was envisaged that retaining E74-R76 salt bridge, and incorporating a Cys in the neighbourhood (I77C), would determine if Cys has misfolded the HA protein. If the resultant I77C has lacked surface expression then, it would be the presence of Cys in the distal region of HA2 that has affected the structure of HA.

In contrast to the expectations, the I77C mutant has surface expression and proteinase K assay profile very similar to HA-wt. It was evident from Fig.5.1.5a, that a disulfide bond at I77C, if formed would only link any of the two HA2 monomers. Therefore, the purpose of I77C mutant was primarily not to lock the HA trimer. Accordingly, the autoradiogram from fig.5.9 (lane 4) does not seem to show any indication of disulfide linkage.

↓105

From these disulfide linkage experiments, it can be concluded that

  1. The salt bridge network connecting the HA2 monomers cannot be substituted. The HA gene sequence alignment of all the members of influenza A virus indicate that E74 is highly conserved. On the other hand, R76 is conserved to approximately 75 % and the absence of R76 is compensated in these strains by R75. This only shows that this salt bridge is highly conserved in all the members of influenza A virus.
  2. The importance of this salt bridge could be surmised in bringing all the three HA monomers together, to form a trimer. Though other interactions between the monomers are equally important in trimer formation, E74-R76 is the only interaction in some strains (eg., X-31) connecting all the monomers. Therefore it may be proposed that substitution of this salt linkage would weaken the trimer formation.
  3. Further, at low pH the E74-R76 salt bridge also breaks facilitating the dissociation of the ectodomain. This facilitates the entry of the solvent into regions of the HA2 trimer shielded from solvent at neutral pH. In the process, the contact of the solvent with the hydrophilic cavity (discussed in 1.2.1) would force the hydrophobic fusion peptide out of the cavity.
  4. It may also be concluded (personal communication with Huang and Herrmann) that interaction of solvent with the loop region of HA2 triggers the formation long helix forming the extended trimer coiled coil structure and, by that, the release of the fusion peptide from the hydrophilic cavity.

6.10 Model for the role of salt bridges for HA conformation and its stability

It is very much evident from the present investigation that salt bridges (ion pairs) play a vital role in low pH regulated protein conformational changes of HA protein. Similar views were reported by Wedekind et al. (2001) for exotoxin A of Psuedomonas aeruginosa. The electrostatic free energy contribution to protein stability does not correlate with the number of ion pairs in a protein, but rather depends on the exact location and geometrical constraints of the ion pairs.

↓106

Protein interfaces are generally more hydrophilic than the protein interiors. The residue composition of most protein-protein interfaces appears to be more similar to that of protein surfaces (Xu et al., 1997). In addition, the interfaces tend to form more hydrogen bonds and salt bridges than protein interiors (Sheinerman et al., 2000, Tsai et al., 1997) and thus are the major contributors to the electrostatic interactions. It is rare to observe a single large hydrophobic patch on the interface (Larsen et al., 1998). Rather, the hydrophobic residues are scattered over the entire surface and form small patches interspersed with polar and charged residues. Xu et al. (1997) further point out that the polar nature of the protein interfaces allows larger electrostatic stabilisation with reference to the interior of the proteins. This sort of stability is exemplified in thermophilic proteins (Musafia et al., 1995).

From the present research, it could be inferred that, mainly interfacial salt bridges are responsible for the stability of the non-fusogenic HA protein at neutral pH. At low pH, the negatively charged residues become protonated initiating the dissociation of the trimer. This facilitates the entry of solvent into otherwise tightly packed trimer. Upon expoure to solvent, the salt bridges at the interface of HA1-HA2 (tetrad salt network) and that of HA2-HA2 (E74-R76) become disrupted. The present study showed that the tetrad salt bridge connecting HA1 and HA2 subunits is vital for maintaining the stability of HA trimer in its non-fusogenic structure. More so, the high degree of conservation of ARG at 109 position in the influenza A family shows the importance of this residue for the stability of the HA trimer. In addition, the presence of a conserved negative residue at position 67of HA2 in all influenza A members makes it ideal for the formation of interfacial salt bridge. In the process, the solvent exposure also disrupts the ring salt network connecting the three HA2 monomers exposing the interiors of the stem region. It may be surmised that in the process, the hydrophilic pocket protecting (harbouring) the fusion peptide is destroyed, thus forcing the expulsion of hydrophobic fusion peptide. At the same time the disruption of tetrad salt bridge in the hinge region transforms the loop region of HA2 into helix (Fig.6.10). Thus it may be stated, that the protonation resulting from the exposure of the HA trimer to the solvent, leads to dramatic and irreversible conformational changes in the HA protein.

.
Fig.6.10: Predicted model for the conformational changes at low pH of the ectodomain of the influenza virus HA.

Shown in the picture are two rows A and B. Row A shows the HA protein from the top view and row B shows the HA protein from lateral view.


© Die inhaltliche Zusammenstellung und Aufmachung dieser Publikation sowie die elektronische Verarbeitung sind urheberrechtlich geschützt. Jede Verwertung, die nicht ausdrücklich vom Urheberrechtsgesetz zugelassen ist, bedarf der vorherigen Zustimmung. Das gilt insbesondere für die Vervielfältigung, die Bearbeitung und Einspeicherung und Verarbeitung in elektronische Systeme.
DiML DTD Version 4.0Zertifizierter Dokumentenserver
der Humboldt-Universität zu Berlin
HTML-Version erstellt am:
14.09.2006