Literature Review - Part 1

1.1 Bacterial heat shock proteins: Their history

Heat shock proteins are proteins which are expressed when a bacterial cell is exposed to high temperatures. They were initially recognized by their increased expression following exposure of the bacterial cells to elevated temperature (Ritossa, 1962; Lindquist, 1986; Lindquist and Craig, 1988); the laboratory of F. C. Neidhardt ‘discovered’ these proteins by coincidence as a result of an accident: a control experiment intended to validate the use of temperature-sensitive mutants in essential genes for study of macromolecule synthesis regulation, which, instead of showing the act of shifting E. coli cells within their normal growth temperature range would not disturb their gene expression severely, demonstrated that dramatic changes (10-50X increase/decrease) occurred in the rates of individual polypeptide synthesis transiently after temperature shifts (either higher or lower). This brought attention to the group of proteins that were induced to high levels (Lemaux, Herendeen, Bloch, Neidhardt, 1978). Earlier on, an unusual mutant was isolated from a background containing a temperature-sensitive nonsense suppressor (Cooper and Ruettinger, 1975), this experiment having nothing to do with what was being studied in Neidhardt’s lab, the details of which will be discussed in the following section.

The most recent findings regarding these heat shock proteins are:

1. Homologous proteins identified in both prokaryotes and eukaryotes suggests an important general function of these heat shock proteins
2. The various proteins can be grouped into a few distinct families with high degree of structural conservation during evolution
3. Not only heat stress induces formation of stress proteins; other conditions bring about the induction of proteins which are identical, similar to those induced by heat

Currently, the major lines of research involve the study of the molecular basis of regulation of gene expression in bacteria and eukaryotes by temperature induction, and the structure and function of heat shock and stress proteins. Molecular cloning, protein purification and functional studies in vitro and in vivo have led to a better understanding on the role in sorting and folding assembly of proteins.

1.2 Types and functions of bacterial heat shock proteins

Up to 1984, 17 heat shock proteins were discovered in E. coli, ranging from 10,000 to 94,000 daltons in molecular size. These proteins have been divided into 4 families: the hsp100 family (80-100 kDa), the hsp 90 family (82-96 kDa), the hsp 70 family (67-76 kDa) and the hsp 60 family (58-65 kDa); there are also other heat shock proteins of smaller molecular weight groups. Certain hsps such as B56.5 and B66.0 are abundant in the cell; others remain undetected until induced by heat. Seven of them have been identified as products of known genes and have been characterized extensively - for example, the hspB56.5 is coded by the gene groEL. The structure of these proteins is conserved during evolution, which implies that it has similar functions in all organisms (Georgopoulos, Ang, Liberek and Zylicz, 1990; Morimoto, Tissiéres and Georgopoulos, 1990). Although called heat shock proteins, these proteins are also synthesized under non-stress conditions, though at reduced rates, and play a fundamental role in normal cell physiology.

1.2.a Localization of bacterial heat shock proteins

There are generally 3 places where bacterial heat shock proteins may be found. Some are associated with the outer membrane; some in cytosolic compartments, and others in the periplasmic space (Hoffman et al., 1990; Ensgraber and Loos, 1992)). Recently it has been discovered that some of these heat shock proteins are also secreted into the culture medium, for example the 66 kDa Salmonella typhimurium heat shock protein (Ensgraber and Loos, 1992), and may contribute to the organisms’s pathogenecity.

1.2.b General functions

These heat shock proteins play an important role in protective and homeostatic mechanisms to cope with physiological and environmental stress at cellular level, and prevent the cells from damaging effects of temperature. Indeed, a wide variety of cells were found to have higher thermotolerance after preincubation at elevated temperatures (Lindquist and Craig, 1988). They are also described as molecular chaperones (Ellis, 1987; Laskey et al, 1998) as they mediate correct folding and assembly of polypeptides (for example, nucleoplasmin is important in the assembly of nucleosomes); however, they are not components of the functional assembled structures (Ellis and van der Vies, 1991; Pelham, 1986). There is also strong evidence accumulating that these heat shock proteins are necessary for acquisition of monomeric and oligomeric protein native structure after synthesis of ribosomes or after transfer across the membrane. Transient exposure of hydrophobic or charged residues during these processes can result in misfolding or aggregation of proteins, and therefore seem to require the presence of protein factors in folding in vivo (Langer and Neupert, 1991). Figure 1 shows the function of heat shock proteins as chaperones. Table 1 shows the heat shock proteins, the genes involved, and these proteins’ possible functions.

(picture not available)

Figure 1. Heat shock proteins as chaperones. A. A model is proposed in which the DnaK and GroEL heat shock proteins bind to unfolded polypeptides within the cell (Rothman, 1989; Ellis, 1990). Substrates for this activity would include new synthesized polypeptides emerging from the ribosome, polypeptides destined for translocation, and globular proteins which have become unfolded following heat shock or other stresses. The heat shock proteins would stabilize the unfolded structure, protecting it from aberrant folding and from non-specific interactions with other intracellular proteins. In response to an appropriate signal – perhaps ATP alone for DnaK, and ATP plus GroES for GroEL – the polypeptide would be released and allowed to fold or assemble into a functional form. B. In response to heat shock, or other forms of environmental stress, increased levels of heat shock proteins are required in order to cope with the accumulation of partially unfolded or denatured proteins in the cell (Lathigra et al., 1991).

1.2.b The hsp70 family

This family of heat shock proteins have been found abundantly in all organisms examined to date. They have a very high degree of evolutionary conservation – between the bacterial members of the hsp70 family, DnaK, and eukaryotic members, the homology is higher than 45% (Langer & Neupert, 1991).

The proteins of this family tightly bind ATP, and seem to feature a weak ATPase activity. However, no consensus sequence for an ATP binding site has been identified so far.

The DnaK family is the only prokaryotic homologue of the hsp70 (Georgopoulos et al. 1990). It was characterized as a heat shock protein encoded in the heat shock regulon of E. coli (Bardwell & Craig, 1984). It is an abundant, constitutively expressed protein with an apparent molecular weight of 70,000. The purified protein (Zylicz & Georgopoulos, 1984) possesses a weak ATPase activity (with a turnover rate of about one ATP per minute) and can be autophosphorylated at threonine residues (Zylicz et al. 1983), and its enzymatic activities are well studied. The ATP hydrolysis and autophosphorylating activity depend on divalent cations (Dalie et al. 1990). The same Ca²⁺ ions that inhibit the ATPase stimulate the autophosphorylation activity, indicating a regulatory role of Ca²⁺ (Cegielska & Georgopoulos, 1989). There may exist a highly conserved calmodulin-like binding domain in various members of the hsp70 family (Stevenson & Calderwood, 1990).

The function of these proteins is to stabilize protein conformations distinct from the stably folded native structures. These seem to be necessary for targeting to and transportation of polypeptides across membranes for assembly into oligomeric structures, or for interactions with other proteins. The shielding of previously buried sequences by hsp70 proteins may reduce the tendency of proteins to aggregate, especially at raised temperatures (Langer & Neupert, 1991.

1.2.c The GroEl/hsp60 family

A heat shock protein with a molecular mass of about 60 kDa has been found in all prokaryotic and eukaryotic cells investigated so far (Langer & Neupert, 1991). Its first member, the GroEL protein was purified from E. coli and characterized several years ago (Hendrix, 1979). There is considerable conservation between the different proteins – about 46%-54% sequence identity (Reading et al., 1989).

These proteins have several features in common. They share sequence similarity and immunological cross-reactivity, and are constitutively expressed, although they can be induced by heat shock. There is also a close similarity in quaternary structure.

All of these proteins possess a weak ATPase activity, which may be important for their function in assisting proteins in the process of acquiring their native structure (Hemmingsen et al., 1988). They ay also act as molecular chaperones. The term “chaperonin” was proposed to define them as a third class of chaperone proteins besides nucleoplasmin and the hsp70 protein family (Hemmingsen et al., 1988; Ellis et al., 1989).

The GroE proteins are the prokaryotic homologues of the hsp60, and belongs to the most abundant proteins in E. coli and several other bacteria. It is encoded in the GroE operon, which is part of the E. coli heat shock regulon. When temperature is elevated from 37°C to 46°C the expression of the encoded proteins is increased four to fivefold (Langer & Neupert, 1991). The transcript of about 2100 nucleotides contains two open reading frames. In addition to the GroEL protein (molecular weight 52,259), the operon also encodes the GroES protein (molecular weight 10,368; Hemmingsen et al., 1988). The GroEL protein possesses a weak ATPase activity (Hendrix, 1979) which can be inhibited by GroES (Chandrasekhar et al., 1986). Partial cosedimentation of purified GroES protein with GroEL in a glycerol gradient also suggests a physical interaction. ATP and MgCl2 are necessary for this (Chandrasekhar et al., 1986).

Recently, a genetic approach in the study of these proteins showed that GroEL and GroES proteins are both necessary for bacterial growth at all temperatures (Fayet et al., 1989). The actual level of GroE proteins can determine the maximal growth temperature (Kusukawa & Yura, 1988), suggesting a more general role of the GroE proteins for cell function. So far, there have been strong evidence that the GroE proteins are involved in DNA replication, and translocation of proteins across membranes.

There is evidence for the involvement of these proteins in the folding and assembly of proteins. In some cases, GroEL-like proteins have been shown to be necessary for the acquisition of the native structure of proteins in the cell. Its complex quaternary structure may aid in the carrying out of these functions. A general role of hsp60 for protein folding in vivo, thus, appears possible. Also, several other proteins were identified which assist protein folding in vivo, (Rothman, 1989), including cis/trans-peptidyl-prolyl-isomerase (Lang et al., 1987). Within the thermodynamic limits chaperonins and these other proteins may assist folding at a kinetic level. One of their main functions may be the prevention of premature folding and aggregation, which is favoured both by the high protein concentration in the cell and by high temperatures (Langer & Neupert, 1991).

1.3 Induction of heat shock proteins

Induction of heat shock proteins occurs primarily at transcription level, and the transcription of heat shock genes is enhanced as a consequence of increased activity of specific transcription factors such as σ³² (the part of the enzyme RNA polymerase holoenzyme that increases the enzymes affinity to promoter regions) in E. coli (Yura, Nagai and Mori, 1993). Some heat shock proteins, particularly hsp70 (DnaK in E. coli) exert negative-feedback control on the synthesis of hsps following initial induction (DiDomenico, Bugaisky and Lindquist, 1982), presumably by interacting with the heat shock transcription factors indirectly or directly (Craig and Gross, 1991). It follows that the structure, function and regulation mechanisms of these heat shock proteins appear to be conserved, at least in some basic forms and reflect rapidly changing cellular requirements for the hsps of most living organisms in nature as well as a need for rapid and fine adjustment of hsp levels to assure optimal growth and survival in the environment (Yura, Nagai and Mori, 1993).

It does not take drastic changes in temperature to induce the production of heat shock proteins – a moderate shift in the temperature from 30°C to 42°C is enough for transient induction, although this heat shock response is exhibited over a wide range of growth temperature in E. coli. This hsp induction occurs immediately after the shock, peaks at about 5 minutes after, and reaches a lower steady-state level in 20-30 minutes after the shock occurs (Yura, Nagai and Mori, 1993). There are basically three groups of response depending on the initial temperature and the temperature shift:

1. Below optimum temperature for growth (eg. 28-33˚C shift): measurable transient induction of heat shock proteins
2. From low temperature to the 35-43˚C range: accelerated synthesis of all cell proteins, including heat shock proteins (an increase in the absolute rate of heat shock protein synthesis, and not just relative to other proteins [Neidhardt, VanBogelen and Lau, 1982; Yamamori, Osawa, Tobe, Ito and Yura, 1982)
3. From low temperatures to 43-47˚C: more pronounced heat shock induction, restricted rate of growth and diminished synthesis of general cellular proteins.

A near-exclusive synthesis of heat shock proteins is brought about by shifts to temperatures that do not permit balanced growth even in very rich media (response #3), and this synthesis continues as long as the cells can make protein (Neidhardt and VanBogelen, unpublished data).

It is found that induction occurs coordinately at the level of transcription as well (Yamamori, Ito, Nakamura and Yura, 1978); Yamamori and Yura, 1980). In 1975, it was discovered (Cooper and Ruettinger, 1975) that a nonsense mutation in the rpoH gene of an E. coli mutant Tsn-K165 affected the synthesis of several major protein essential for growth at high temperature, now known to be heat shock proteins (Neinhardt, VanBogelen and Lau, 1983; Yamamori and Yura, 1982). It was later discovered that this rpoH gene coded for a protein belonging to the class of σ factors that are an important component of the bacterial enzyme RNA polymerase needed for transcription of heat shock genes (Grossman, Erickson and Gross, 1984).

It has been shown that when cells are heat shocked, the synthesis of heat shock proteins is rapidly induced while most pre-existing synthesis is repressed; when the temperature returns to normal, the synthesis of these proteins is gradually repressed and normal synthesis is restored. It has also been shown that hsp70 production in the cells is quantitatively correlated with the degree of stress, and that this level of synthesis is controlled at both the transcriptional and posttranscriptional level through repression of hsp70 mRNA synthesis and destabilization of HSP 70 transcripts (DiDomenico, Bugaisky, Lindquist, 1982):

Production of functional heat shock proteins blocked --> heat shock transcription continues; heat shock mRNAs are stable and accumulate in vast quantities;

Block released --> specific quantity of functional heat shock proteins must accumulate before heat shock transcription is repressed and preexisting heat shock mRNAs are destabilized.

However, not only temperature shifts induce the production of heat shock proteins. Factors other than temperature shift, such as amino acid analogues (Goff and Goldberg, 1985) and alkaline shifts (Taglicht et al, 1987) were also found to induce the same response.

1.4 The heat shock response

When the cells are exposed to temperatures exceeding the optimum for growth, increased amounts of heat shock proteins are synthesized. In E. coli, for example, a shift from any normal growth temperature to 42-46ºC elicits rapid changes in the synthesis of most of the 1000 cellular proteins within a minute and this climaxes at 7 minutes (Neidhardt et al, 1982). At the same time, the cell develops thermotolerance, which is the increased resistance to a higher heat challenge. This phenomenon has been demonstrated in S. typhimurium. It also produces heat shock proteins in response to stresses other than temperature; however, these adaptations (save for the adaptation to starvation and acid shock) do not produce cross-protection to heat (Foster and Spector, 1995).

In E. coli, a subset of heat shock proteins comprising 17-20 proteins (including chaperonins GroEL, GroES, DnaK, DnaJ, GrpE and RpoD, the RNA polymerase σ⁷⁰) forms a regulon whose temperature-dependent induction requires an alternate sigma factor, σ³², which dictates the specificity of the enzyme RNA polymerase to heat shock promoters. As a consequence of elevated temperature, some of the resulting heat shock proteins denature, and this protects them from further degradation or refolding. Others, such as the Lon and Clp proteases are involved in proteolysis. There are also several genes, such as htr, which are involved in high-temperature resistance; however, not all of them are thermally regulated.

The release of preformed σ³² is needed for the thermal induction of the Htp regulon so that more of this sigma factor is available to express the target genes. The concentration of free DnaK or DnaJ available to bind and activate this sigma factor is decreased by the binding of these two gene products to abnormally folded proteins resulting from heat denaturation. The rpoH gene is also subject to transcriptional and posttranscriptional level control. rpoH promoters P1, P4 and P5 are transcribe by E σ⁷⁰; promoter P3 requires σ²⁴, which happens to be the factor also required for heat shock induction of htrA, which is a periplasmic protease as well as for survival at extremely high temperatures. Temperature up-shifts cause transient disruption of the secondary structure capable of blocking rpoH transcript translation; thus permitting higher rates of translation initiation at higher temperatures.

Not all the genes associated with thermotolerance in Salmonella typhi have been identified; it is, however expected that its system is analogous to that of E. coli (Foster and Spector, 1995).

1.5 Bacterial heat shock proteins as virulence factors

It has been shown that proteins which play an essential role in host-parasite interactions are co-ordinately regulated at the transcriptional level, and that their expression is induced by environmental stimuli likely to be encountered within the infected host (Miller et al., 1989). In the case of heat shock proteins, change in temperature signals alteration in expression of virulence determinants.

Studies carried out on the DnaK and GroEL heat shock proteins have shown that both these protein groups perform important functions during normal cell growth of E. coli at all temperatures. It can be argued that as these heat shock proteins are essential in cell “housekeeping”, therefore they cannot be considered as virulence factors, although they may play an additional role during infection (Lathigra et al., 1991).

Bacterial heat shock proteins can stimulate immune responses to both pathogen-specific and autoreactive epitopes, and may thus have a “toxic” effect on the immune system (Lathigra et al., 1991). For example, in the case of chlamydial infection, the GroEL homologue appears to trigger an immune response which is responsible for the tissue damage associated with the disease (Morrison et al., 1989). Thus the heat shock proteins may be considered from an immunological standpoint as virulence factors.

1.6 Bacterial heat shock proteins as antigens of pathogens

It is strange that such highly conserved proteins as heat shock proteins should be the targets of an immune response. Possible explanations include that heat shock proteins

• are abundant cellular proteins
• have conserved epitopes that may prime the host for an immune response to these proteins
• may be preferentially processed for presentation due to either structural or functional features, and
• may be virulent factors. (Shinnick, 1991)

Overall, the immunoreactivity of these proteins might be due to their being major constituents of a pathogen when it is within an antigen-presenting cell, and are thus processed and presented to the host’s immune system.

At an optimal growth temperature of 37°C, E. coli cells have both hsp60 and hsp70, although they account for only about 1.6% and 1.4% of the total protein content respectively. When heat shocked, however, hsp70, hsp60 and GroES rise to more than 15% of total cell protein (Neidhardt et al, 1984).

Not all of these proteins are located intracellularly. Some members of the hsp90, hsp70 and hsp60 families seem to be surface accessible. Not only that, heat shocked cells and other stressed cells also express certain heat shock proteins such as hsp60 on their surfaces or even excrete them (Koga et al., 1989). Because members of the hsp70 family have been suggested to play roles in targeting intracellular proteins for lysosomal degradation (Chiang et al., 1989) and in antigen presentation on the macrophage, maybe the conserved domains of the pathogen’s heat shock proteins will target them to the organelles and proteins involved in antigen processing and presentation and thus facilitate their processing. Presentation might also be facilitated by the ability of hsp70 and hsp60 to denature proteins. It is possible that a pathogen’s heat shock proteins, then, might be ‘co-processed’ along with other proteins, thus increasing the contingency of a heat shock protein being presented by the macrophage (Shinnick, 1991).

1.7 Heat shock proteins and the immune response

Bacterial heat shock proteins can cause damage to the human host. This is brought about by eliciting cross-reactive antibodies and T cells. Production of these bacterial heat shock proteins occurs during infection, particularly when the bacteria are inside phagocytes, the evidence being that antibodies against these proteins have been detected in patients who have mounted immune response to these bacterial pathogens. Because the amino acid sequence in these proteins have been conserved in evolution, both the bacterial and human heat shock proteins contain similar enough epitopes to elicit T cells and antibodies that react with both. For the benefit of the human host, an early anti-heat shock response is of protective value, since it limits incipient bacterial infections by killing infected host cells (stressed human cells express these proteins on their surfaces); however, if out of control, this response can become a major cause of tissue damage in the infected area since both human and bacterial cells are targets for antibody binding, complement killing and phagocyte attack, and there is no way of telling the two species of heat shock proteins apart (Salyers and Whitt, 1998; Murray and Young, 1992)

It is strange that of all the antigens expressed by an invading microorganism, it is the molecular chaperones belonging to highly conserved heat shock protein families that are prominent immune targets (Young et al., 1993). For example, it has been reported that as many as one in five of mycobacteria-reactive T cells in immunized mice may be directed towards the bacterial 60 kDa heat shock protein (hsp60 or chaperonin) (Kaufmann et al., 1987). The fact that some chaperones make up a major proportion (1-2%) of the total cell protein content, with the potential for further induction under stress conditions, could make an important quantitative contribution to their immunogenecity. Furthermore, it has been suggested that there may be an additional inherent bias introduced during development of the immune repertoire which favours recognition of conserved, ‘self-like’ proteins (Cohen & Young, 1991).

In the case of several pathogens, however, the humoral immune response to these heat shock proteins is directed predominantly towards non-conserved epitopes (Shinnick, 1991). This has been proposed as a parasite survival strategy, whereby these epitopes might serve as a diversion of the host’s immune response away from conserved epitopes, which may be regions required for functional activity. However, some heat shock proteins in some proteins seem to be targets of a protective immune response, indicating that this is not a general survival strategy.