WO1997012910A1

WO1997012910A1 - Helicobacter pylori protein

Info

Publication number: WO1997012910A1
Application number: PCT/GB1996/002404
Authority: WO
Inventors: Bow Ho
Original assignee: CORTECS INTERNATIONAL Ltd
Current assignee: CORTECS INTERNATIONAL Ltd
Priority date: 1995-09-29
Filing date: 1996-09-27
Publication date: 1997-04-10
Anticipated expiration: 1998-03-29
Also published as: BR9610953A; NO981407D0; CN1201464A; NO981407L; AU7137896A; JP2000500123A; CA2233328A1; MX9802486A; ZA968185B; KR19990063833A; EP0859788A1; GB9519865D0

Abstract

The present invention provides a novel antigenic protein obtainable from the coccoid form of H. pylori. The use of the antigenic protein in methods of detecting H. pylori, as well as in the preparation of vaccines, is also disclosed. A method of culturing the coccoid form of H. pylori is also provided.

Description

HELICOBACTER PYLORI PROTEIN

The present invention relates to a novel antigen derived from the coccoid form of H. pylori , its general use in medicine, its use in the preparation of vaccines, as well as its use in the detection of the coccoid form and the diagnosis of H. pylori infection, as well as determination of the disease prognosis of a subject.

H. pylori is a Gram negative bacteria that has been strongly implicated in chronic active gastritis and peptic ulcer disease (Marshall et al , Medical Journal of Australia , 142:439-444 (1985); Buck, G.E., Journal of clinical Microbiology, 3:1-12 (1990)) . In in vi tro culture, H. pylori exists in two distinct morphological forms, the culturable spiral form and the non-culturable coccoid form (Marshall eϋ al , Microbios letters, 25:83-88 (1984); Kung, J.S.L., and HO, B., Workshop on Gastroduodenal Pathology and Campylobacter pylori (abstract P9) , edited by F. Megraud and H. Lamouliatte, Bordeaux, France (1988)) . The spiral form of the bacterium does not survive beyond about 2 hrs when exposed to air. Under unfavourable conditions, the spiral form undergoes differentiation into the coccoid form (Vijayakumari and Ho, Acta Gastro-enterologica Belgica, 56:101 (1993) ) .

To date, there has been only a single report of the successful in vi tro transformation of coccoids to spirals, which experiment has proved unrepeatable (Mai et al , Gastroduodenal Pa thology and Campylobacter pylori , pp28-33, edited by F. Megraud and H. Lamouliatte, Elsevier Science Publishers (1989) . Although the formation of coccoids in vi tro could be induced by antibiotics or by deprivation of nutrients (Nilius et al , Zbl . Bakt . 280: 259-272 (1993) , studies of the coccoid form have been hampered due to the lack of information regarding this form, and its role in the life cycle of H. pylori , as well as the lack of any method for obtaining a synchronous culture of this form. Up to now the coccoid form of H.pylori has been regarded effectively as a "dead", non-viable form whose role in the life-cycle of the bacterium is unclear.

The results discussed herein indicate that the coccoid form can indeed exist in a viable form and does have a role in the life cycle of H. pylori , and that this role has implications in the diagnosis, prognosis and treatment of H. pylori infections.

At present, various antigens of the H. pylori spiral form have been identified (see, for example, WO 93/22682) and are used, for example, in diagnostic kits for the detection of H. pylori , e.g. the HELISAL™ test marketed by CORTECS LTD. However, the results described herein indicate that it is essential to be able to detect the coccoid form of the bacterium to ensure accurate and complete (i.e. coccoid only infection) diagnosis. At present, no specific antigens for the coccoid form have been identified.

Thus, in a first aspect, the present invention provides an antigenic protein having a molecular weight of 60 kDa, as determined by native PAGE, obtainable from the coccoid form of H. pylori . In particular, the protein can furhter be characterised in that it has the following N- termmal amino acid sequence:

D-T-H-K-S-E-I-A-H-R-F-N-D-L-G. Given the antigenicity of this protein, and its unique presence in the coccoid form, it is useful as a means of detecting the presence of the coccoid form of H. pylori .

Therefore, in a second aspect, the invention provides the use of the antigen of the invention in the detection of antibodies against H. pylori , and specifically detecting the coccoid form. In addition, the novel antigen of the invention can be used in combination with other antigens, particularly those obtainable from the spiral form of H.pylori to provide more sensitive methods of detecting H.pylori .

In addition, the coccoid antigen of the invention can be used to raise antibodies, which antibodies can then be used to detect the antigen, including the antigen when present as part of the intact coccoid. For example, therefore, the antibodies could be labelled and used as a means of detecting coccoids in tissue samples and the like. Methods of raising antibodies using the antigen are well known to the skilled man, as are means of labelling such antibodies for use in such methods. Thus, a third aspect the present invention provides the use of the coccoid antigen of the invention in the preparation of antibodies, e..g polyclonal antibodies. In a further aspect, the invention provides the use of such antibodies in the detection of the coccoid form of H. pylori which comprises antibodies raised against the coccoid antigen of the invention.

The antigen of the invention also finds use as part of an antigen composition, which may contain antigens against both che spiral and coccoid forms of H. pylori . In a sixth aspect, therefore, the invention provides a composition comprising the antigenic protein of the invention, optionally together with one or more other H. pylori antigens, these one or more other antigens being obtainable from either the spiral or coccoid form of H. pyl ori .

In a seventh aspect, the invention provides a method of detecting the coccoid form of H. pylori , e.g. by detecting antibodies, which includes the step of contacting the antigen of the invention, or the antigen composition of the invention, with a sample. Usually the sample will be a biological sample, eg a blood sample, a urine sample or a saliva sample. The antigen or antigen composition can be brought_ into direct contact with the biological sample. Alternatively, the biological sample can first be treated to render it more suitable, eg by filtration, pH adjustment etc. Examples of suitable methods are those described in UK patent application no. 9422991.1.

In an eighth aspect the present invention provides a method for the diagnosis of H.pylori infection which includes the step of contacting the antigenic protein of the invention with a biological sample obtained from a subject. The antigen can be provided in the form of an antigenic composition as described herein. In general, the diagnostic method of the invention will also include the step of detection of other antigens obtainable from either the spiral or coccoid form of H.pylori . In this way a more sensitive method of diagnosis for H.pylori infection is provided. Suitably, the diagnostic method of the invention is carried out on a sample of blood, a sample of urine or a samoie of saliva. Suitably, the diagnostic method of the invention will be be carried using a test device or test kit, e.g. that used in the HELISAL™ test. In a further aspect, therefore, the present invention provides a kit for use in the diagnosis of H.pylori infection which comprises the antigenic protein of the invention. Generally, the kit of the invention will also include one or more other antigens obtainable from either the spiral or coccoid form of H.pylori .

The identification of the unique antigen of the invention also opens up the possibility of providing a vaccine against H.pylori which will be active against both the spiral and coccoid forms of the bacterium. In a tenth aspect, therefore, the present invention provides a vaccine for the prophylaxis or treatment of H.pylori infections which comprises the antigen of the invention together with one or more adjuvants and/or carriers. In a preferred embodiment of this aspect the vaccine includes one or more antigens derived from the spiral form of H.pylori . Suitably, these additional antigens will include at least one which is unique to the spiral form.

In an alternative embodiment, the vaccine can comprise the coccoid form (either killed or "live") of H. pylori itself since the cells could be taken up in the GI tract and then induce an immune response. Furthermore, forms of H. pylori which exist between the true coccoid and true spiral forms could be used on the basis that they are expressing the novel antigen. In other words, the coccoid, or intermediate, form(s) of H. pylori are used as a vehicle for delivery of the novel antigen to achieve an immune response. In addition, it has been determined that the novel antigen of the invention can be used to detect IgM antibodies produced in children in response to H. pylori infection. In an eleventh aspect, therefore, the present invention provides a method of detecting IgM antibodies against the coccoid form of H. pylori in children, which comprises the step of bringing the antigen of the invention into contact with a biological sample obtained from a child. Suitably, the biological sample will be a blood sample, a urine sample or a saliva sample.

As described herein, there has now been devised a method for culturing the coccoid form of H. pylori so that it is viable. Thus, in a final_aspect, the present invention provides a method of culturing the coccoid form of H. pylori which comprises the step of regularly adding carbon dioxide to a culture medium comprising the spiral form of H. pylori such that conversion to the coccoid form occurs and wherein the coccoid form obtained is viable. Suitably, C0₂ is added at least twice a day and the culture is allowed to run for nine weeks to ensure conversion.

The invention will now be described with reference to the following examples, which should not be construed as limiting the invention. The examples refer to the figures in which:

FIGURE 1: shows a typical growth curve for H. pylori grown in a chemostat with concurrent pH and urease measurements;

FIGURE 2: shows wet preparations of (a) spirals and (b) coccoids seen under phase contrast microscope, magnification xlOOO. Spiral cells were uniformly dense while coccoids were of two types : (A) compact with dense cytoplasm and (B) those with loose cytoplasm like "ghost" cells;

FIGURE 3 : shows a transmission electron micrograph of a coccoid, magnification x 80,000;

FIGURE 4 : shows a transmission electron micrograph of a coccoid with flagella, magnification x 80,000;

FIGURE 5: shows a silver stained SDS-PAGE protein profile. Lanes: 1 high molecular weight marker (Sigma) : 2 low molecular weight marker (Sigma) : 3 spiral NCTC 11637: 4 coccoid NCTC 11637: 5 spiral V₂ : 6 coccoid V₂;

FIGURE 6 : shows modified periodic acid Schiff stained smears of (a) spirals and (b) coccoids, magnification x 1000;

FIGURE 7 : shows modified gram staining of coccoids, wherein in (a) they have been treated with the salivary enzyme ot amylase and in (b) they have not;

FIGURE 8: shows DNA of H.pylori . Lanes: 1 Hindlll cut λ DNA: 2 spiral NCTC 11637: 3 coccoid NCTC 11637: 4 spiral V₂: 5 coccoid V-,;

FIGURE 9 shows the results of modified Albert's stain, magnification x 1000;

FIGURE 10: shows detection of urease enzyme activity on PAGE. Lanes: 1 high molecular weight marker (Pharmacia) : 2 spiral NCTC 11637:

3 coccoid NCTC 11637: 4 spiral V₂: 5 coccoid V₂;

FIGURE 11: shows silver stained native PAGE protein profile. Lanes: 1 high molecular weight marker (Pharmacia) : 2 spiral NCTC 11637: 3 coccoid NCTC 11637: 4 spiral V₂ : 5 coccoid V₂;

FIGURE 12 : shows a western immunoblot under non-denaturing conditions with the coccoid antigen. Lanes: 1 molecular weight marker

(Pharmacia) : A preimmune anti-spiral serum: B anti-spiral serum: C preimmune anti-coccoid serum: D anti-coccoid serum;

FIGURE 13 : shows indirect fluorescent antibody test of coccoids, magnification X1000, wherein it can be seen that, like the spirals, the coccoids fluoresce under ultra violet light, indicating that their surface antigens are similar.

EXAMPLE 1

(a) Bacterial strain and preparation of the differentiated forms of . pylori .

A local H. pylori strain V- isolated from a patient with non-ulcer dyspepsia was used. This strain was initially grown on chocolate blood agar (CBA) to check for purity. The plate culture was then used as inoculum for a 250ml Schott flat-bottomed round bottle containing 30ml BHIH (brain heart infusion supplemented with 10% horse serum and 0.4% yeast extract) , and incubated at 37°C for 72h. This in turn serves as the inoculum for chemostat or batch cultures.

A 1.5L fermenter containing 540ml BHIH was set up as described in Ho and Vijayakumari (Λicrojbios, 76:59-66 U993) . The medium was inoculated with 2x30ml of 3 day old H.pylori cul ture, giving a ratio of 1:10

(inoculum:medium) . Hence, carbon dioxide was supplied twice daily, for five seconds each time, and dissolved oxygen was maintained by a feedback control of the impeller speed of around 35-40 rpm. Samples were withdrawn at time intervals and checked for urease activity, pH, viability, and microscopic examination for morphological changes.

The culture was maintained under these conditions for up to 3 months during which daily monitoring of the cells was continued. When homogenous/synchronous culture was observed, the cells were harvested by centrifugation at 10,000g for 40min. and washed once. The pellet was then used to prepare coccoid antigen using the modified glycine method (Ho, B., and Jiang, B., European Journal of Gastroenterology and Hepatology, 7:121-124 (1995) .

Alternatively, a IL Schott round-bottomed bottle or IL Erlenmeyer flask with a side-arm and fitted with a tight fitting rubber bung, containing 270ml BHIH was used. A "'mm diameter hole was bored so as to accomodate the fitting of a disposable filter unit containing a 0.22μm filter having a diameter of 50mm (e.g. Gelman) . Each 270ml of BHIH was inoculated with 30ml of 3 day old H.pylori culture. Carbon dioxide was supplied twice daily via the 0.22μm filter.

The culture was incubated in a 37°C shaker incubator (New Brunswick) maintained at 90rpm for up to nine weeks and the cells were subsequently harvested by centrifugation at 10,000g for 4Omin. The cell pellet was washed once and antigen prepared as described above.

A 48 hr old culture on chocolate blood agar was used to provide "synchronous" spirals.

(ii) Serum specimens. Sera from 50 patients with gastroduodenal disease and 50 blood donors were kindly provided by A/Prof K M Fock of Toa Payoh Hospital and Dr D Kuperan of National University Hospital, Singapore, respectively.

(iii) ELISA and Western immune-blotting.

Indirect ELISA and Western immunoblotting were performed according to the method of Khia and Ho, Biomed . Letts . 50: 71-78 (1994) .

RESULTS

Figure 1 shows viability, pH and urease specific activity for a typical culture. From this it can be seen that at 9 weeks the culture had become a coccoid culture. This was confirmed by microscopic examination failure of spiral for growth on CBA. The time taken for the spiral form to differentiate into the coccoid form is dependent on the constant supply of Carbon Dioxide. It is also clear that there are two forms of coccoid. One has a dense cytoplasm while the other has a "ghost" -like appearance . This latter form is considered to be non- viable . In contrast to other reports (Nilius et al , infra) , the chemostat culture showed mostly dense coccoids. These coccoids were also harvested and suspended in BHIH supplemented with 20% glycerol. The suspended coccoids were then stored at -80°C until needed.

ELISA.

By ELISA, 37 out of 50 patients (74%) were sero-positive for H. pylori IgG antibodies to both the coccoid and spiral antigens (Table 1) . Of the 50 blood donors, 16

(32%) and 14 (28%) were found to be sero-reactive to IgG antibodies to coccoid and spiral antigens, respectively

(Table 1) . Using spiral antigen as the standard for comparison, the sensitivity and specificity of IgG ELISA to coccoid antigen were 98 and 94%, respectively.

TABLE 1. Serum IgG antibodies to H. pylori coccoid and spiral antigens by ELISA.

Patients Blood donors (n=50) (n=50)

Spiral Positive Negative Positive Negative Coccoid

Positive 36 1 14 2

Negative 1 12 - 34

Sensitivity: 98% Specificity: 94% These results show that in the sample used a 4% higher detection rate was achieved when using both the coccoid specific antigen as well as spiral antigen(s). Clearly, therefore, better rates of diagnosis, and thus treatment will result from the use of the coccoid specific antigen.

Western blotting.

By Western blotting, serum IgG was shown to recognise similar major protein bands in both the coccoid and spiral antigens. The conserved protein bands in both antigens detected were 128, 116, 110, 95, 91, 66, 60, 54, 50 and 33 kDa.

EXAMPLE 2

Bacterial strains.

Two strains of H. pylori were used, the local H. pylori strain V₂ referred to in example 1 above isolated from a patient with non-ulcer dyspepsia and the standard strain NCTC 11637.

Preparation of cul tures .

Coccoid and spiral sultures were prepared as described in example 1. For coccoids, a batch culture was grown as described earlier (Ho and Vijayakumari, infra) . A small aliquot was aseptically removed at time intervals to assess culturability on chocolate blood agar, and the viable count enumerated using the Miles and Misra technique (Miles and Misra, Journal of Hygiene, 38:732- 738 (1938) . The percentage of coccoids was estimated by counting in triplicate the number of spirals to coccoids using a Neubauer bacterial cell counting chamber under a phase contrast microscope. Urease specific activity was measured using the phenol spectrophotometric method of Hamilton-Miller and Gergan, infra , while pH of the culture medium was monitored.

Light and electron microscopic examination of cul tures .

Cultures were observed for motility and morphology using phase contrast microscopy. Air dried smears of cultures were stained using the modified Gram stain in which counter staining with dilute carbol fuschin was carried out for 10 to 30 minutes instead of the usual 1 minute.

In the Periodic acid Schiff staining for polysaccharide, the method of Hotchkiss (Archives in biochemistry, 16:131-141 (1948)) was modified with a longer staining time of 90 minutes compared with the normal 15-45 minutes in fuschin sulphite and a prolonged counter staining for 45 minutes with malachite green.

Laybourn's modification of Albert's stain for volutin granules (Cruickshank, Medical Microbiology : A guide to laboratory diagnosis and control of infection, pp656-657 (1968) ) was further modified for the coccoids with a longer staining time of 45 mins instead of the usual 3-5 mins used for the spirals.

Cells for transmission electron microscopy were fixed in 4% glutaraldehyde, dried on copper grids and negatively stained with 0.5% phosphotungstic acid (pH 6.8) . Grids were viewed using Philips JOEL-JEM-1200EX transmission electron microscope.

Biochemical assays.

The presence of 19 different enzymatic reactions and the biotype were determined using the commercial strip API ZYM kit 2520 (Kung et al , Journal of Medical Microbiology, 29:203-206 (1989) ) . Urease specific activity was determined quantitatively using the phenol spectrophotometric method of Hamilton-Miller and Gergan

(Investigative Urology, 15:327-328 (1979)) while protein content was assayed by the modified Lowry assay

(Schacterle & Pollack, Analytical Biochemistry, 51:654-

655 (1973)) . ATP was quantitated using the bioluminescence assay kit (Bio-Orbit, Finland) and the polysaccharide content measured by the L-cysteine sulphuric acid assay as described by Chaplin and Kennedy

( Carbohydrate Analysis : a practical approach, ppl-2. Edited by M.F. Chaplin and J.F. Kennedy, Oxford: IRL press

(1986)) .

DNA extraction and inicroassay.

The DΝA of both forms were extracted according to the procedure of Clayton et al ( Infection and Immuni ty, 57:623-629 (1988)) and electrophoresed on a 1% agarose gel . Total DΝA content per cell was assayed according to the method by Kapuscinski and Skoczylas (Analytical Biochemistry, 83:252-257 (1977)) .

Protein profile and Western immunoblotting.

Protein profiles were elucidated by polyacrylamide gel electrophoresis (PAGE) according to the method of Laemmli ( Nature , 227:680-685 (1970)) . In the non-denaturing native PAGE, 30 μg total protein of whole cell preparations were electropnoresed on a 6% separating gel and 5% stacking gel. In the sodium dodecyl sulphate (SDS) denaturing PAGE, the same amount of protein was electrophoresed on a 10% separating gel and 5% stacking gel. Relative molecular weight was determined with reference proteins run under the respective electrophoretic conditions. Both types of gels were visualised by silver staining according to the procedure of Sambrock et al (Molecular Cloning: A Laboratory Manual, 2nd edn. Cold Spring Harbour, New York:Cold Spring Harbour Laboratory (1989)) . In addition, the protein bands of both the native and SDS PAGE were electro- transblotted onto Immobilon P (Millipore) membrane using a modification of the method of Towbin et ai ( PNAS USA 76: 4350-4354 (1979) ) .

Antibodies raised in rabbits against either the spiral or the coccoid were used as probes to identify the specific and immunogenic proteins in both forms.

Haemagrg-lutina ion and -haemagg-lutiπation inhibi tion assay. A slight modification of the microtitration plate assay of Morgan et al ( Journal of Clinical Microbiology 29: 395-397 (1991)) was carried out with 20 μl of 2% v/v red blood cells (human or rabbit) which were added to 25 μl of bacterial culture containing a range of 10⁷-10¹² cells/ml in individual microtitration wells. Each mixture was incubated in quadruplicates at 4°C overnight before the haemagglutination patterns were read. The haemagglutination inhibition assay was performed with bacteria pretreated with 1 mg ml^"1 protease (pronase E, Sigma) at 37°C for 60 minutes or heated at 60°C for 10 minutes. Similarly, the red blood cells were pretreated with 4.0 μg ml^"1 Neuraminidase (Sigma) or 1 mg ml^"1 protease at 37°C for 60 minutes before haemagglutination assay.

Indirect fluorescent antibody test .

Smears of coccoids and spirals were dried on glass slides (HTC, Wellcome) before being treated with human sera positive for IgG antibodies against H. pylori . After 30 minutes' incubation, the smears were washed three times for ten minutes each with phosphate buffered saline (PBS - pH 7.6) . Smears were then treated with 1:40 dilution of goat anti-human IgG conjugated FITC (Wellcome) for another 30 minutes. Smears were washed four times for fifteen minutes each with PBS, mounted with buffered glycerol and viewed under ultra violet light using the Reichert-Jung fluorescence microscope.

RESULTS AMP DISCUSSION

"Synchronous" culture of coccoids .

A "synchronous" culture of coccoids has been successfully prepared in a chemostat environment. Figure 1 shows a typical growth curve for H. pylori in a chemostat. Growth in the first two weeks was similar to that described by Ho and Vijayakumari ( infra) . The late stationary phase showed a gradual decrease in viable counts to 10^s CFU/ml within the next two weeks. Subsequently, the declining death phase continued linearly for the following five weeks. Throughout the approximately nine week culturing period, the percentage of coccoids was shown to be inversely proportional to spirals.

The pH of culture medium decreased from neutral to 6.58 in the first 3 days and remained at 6.53 +/- 0.13 in the stationary phase for the next four weeks. It then increased to a maximum of 6.98 on the 7th week, after which the pH was stable at 6.84 +/- 0.02 in the following two weeks. Catrenich & Makin { Scandinavian Journal of Gastroenterology, 6(suppl. 181) : 58-64 (1991)) reported a similar inversely proportional relationship between pH and viable counts and postulated that the loss of viability and conversion to coccoids was due to basic pH endogenously produced by deaminase activity. However, in our study, as the pH of the culture medium remained within the tolerable range for the growth of H. pylori (pH 6.5-7.5) as reported by Marshall et al ( infra) and Dick (Ann . Rev. Microbiology, 44: 249-269 (1990)), the conversion of spirals to coccoids was most probably due to nutrient depletion or metabolite inhibition rather than pH changes .

The average urease specific activity (USA) as 4.90 +/- 4.42 U mg^"1 protein in the exponential phase and increased to 11.58 +/- 10.03 U mg^"1 protein in the stationary phase. During this period, a trimodal peak was observed at the 18th, 156th and a highest peak at the 333rd hour respectively. The USA then decreased linearly with the declining death phase. At the end of the culture period, USA was 0.18+/- 0.03 U mg^"1 protein. This indicates that urease activity is valuable for the actively reproducing spirals and decreases with the formation of coccoids. Marshall et al ( Gastroenterology, 99:697-702 (1990)) reported that the primary function of urease activity of H. pylori in vi tro is to protect the bacterium against acidity. As such, the increase in USA in the stationary phase may be an adaptive response of the spirals to the increasing acid pH due to metabolism. Concomitantly, as the pH increases in the declining phase, the USA is decreased.

Mi croscopy .

Under phase contrast microscopy, the spirals appeared as curved or S-shaped rods with uniform contrast in cytoplasmic density (fig 2a) . The coccoids on the other hand were circular and consisted of two types: one type was shown to be compact with dense cytoplasm, while the other type was with loose cytoplasm and had the appearance of "ghost" cells (Fig 2b) . It was not possible to separate the two types of coccoids using sucrose density gradient centrifugation. In transmission electron microscopy, our observations were similar to earlier reports by Marshall et al (1984 infra) where the spirals were shown to be curved rods with cell dimensions of 0.3-0.5 μ x 1.0-3.0 μ and possessing tufts of 1-6 flagella ending in bulbous tips. In contrast, the coccoids were circular with diameters ranging from 200- 300 nm and had intact cell membranes (Fig 3) .

Unlike the spirals, which had characteristic darting motility, the coccoids were non-motile when observed under the phase contrast microscope. Transmission electron microscopy, on the other hand, showed the presence of flagella in some coccoids (Fig 4) . This could either mean that the flagella is a remnant of the spirals (Marshall et al , 1984 infra) or that the coccoids actually possess flagella but they are inactive due to the dormant state or the lack of energy to drive it.

Suerbaum et al { Journal of Bacteriology, 175:3278-3288

(1993) having cloned the flagellin genes, observed that mutation in the major flagellin gene resulted in non- motile and a flagellated H. pylori while mutation in the minor flagellin gene produced normally flagellated and motile bacterium. In this study, the denaturing SDS PAGE protein profile showed that the 58 kDa minor flagellin protein is present in equal intensities in the spirals and coccoids, but the major 52 kDa flagellin protein is reduced in intensity in the coccoids as compared to the spirals (Fig 5) . Since the minor flagellin is located proximal to the flagella hook, a structure required for flagella attachment (Kostrzynska et al , Journal of Bacteriology, 173:937-946 (1991)) , the findings could explain the intact flagella on the coccoids but absence of motility.

On modified Gram staining, the spirals appeared Gram negative with counter staining for 10 minutes while the coccoids remained weakly Gram negative even after counter staining for 30 minutes. In the modified Periodic Acid Schiff staining, spirals stained green (Fig 6a) but the coccoids were bright red, indicating the presence of polysaccharides on their cell wall (Fig 6b) . This reaction was sustained by the fuzzy halo of approximately 50-60 nm radius around the coccoids as observed on transmission electron microscopy (Fig 3) . This polysaccharide layer which contributes > 50% of the coccoid cell component could account for the poor staining of the coccoids in the modified Gram staining reaction. Consistently, treatment of the coccoids with salivary enzyme ct amylase before staining to digest the polysaccharide content of the coccoids greatly improved the Gram reaction (Fig 7) . Moreover, the polysaccharide content of the coccoids was 10 x more than the spirals (Table 2) . It has been shown that the presence of a polysaccharide coat could help the survival of bacteria under adverse environmental conditions. The hygroscopic layer could act as a cellular buffer controlling gaseous exchange and preventing excessive absorption or loss of fluid that could lead to cell lysis and death (Wilkinson, Bacteriological Review, 22:46-73 (1958)) . TABLE 2 : Constituents of Spirals and Coccoids

♦Units in μmol/cell

Thus, the coccoids could possibly survive outside the human body with protection offered by the thick polysaccharide layer from atmospheric oxygen tension as well as the unfavourable environment. Similar observations were rendered for Campylobacter jejuni by Rollins and Colwell (Applied and Environmental Microbiology, 52:531-538 (1985)) where an increase in viscosity of the culture suspension was noted as it underwent transition from the spirals to coccoids. They suggested that the production of an extracelluar viscous polysaccharide as an adaption to ensure extended survival for C. j ejuni .

In H. pylori cultures, this increase in viscosity was not observed but the granular spiral culture tended to flocculate. Under phase contrast microscopy, the "synchronous" coccoids were seen in large clumps. The polysaccharide layer could function to hold the coccoids together resulting in the floccules.

DNA content.

Intact chromosomal DNA was extractable from coccoids indicating that these forms are probably viable (Fig 8) . total DNA content of spirals was approximately 5.22 x IO^-7 ng/cell, while that of coccoids was 3.13 x IO^"7 ng/cell (Table 2) . The decreased DNA in the coccoids as compared to the spirals could be accounted for by the population of coccoids having loose and leaky cytoplasm as observed under phase contrast microscopy (Fig 2b) . On the other hand, possession of only half the total amount of DNA in the coccoids as compared to the spirals, could also indicate the dormancy or survival strategy of this differentiated form.

Novitsky and Morita (Applied and Environmental Microbiology, 33: 635-641 (1977)) reported a 48% reduction in DNA content of the marine vibria ANT 300 under starvation survival. They suggested that this could be a strategy for conservation of energy resulting in the degradation of extraneous or partially replicated DNA.

Biochemical characteristics .

The spirals were oxidase and catalase positive but the coccoids did not show any visible reaction in these qualitative tests (Table 3) . In the API ZYM test, both the spirals and the coccoids showed similar enzyme profiles belonging to Biotype II as described by Kung et ai (-Journal of Medical Microbiology, 29:203-206 (1989)) . Interestingly, equal amounts of acid and alkaline phosphatase activities were observed in the coccoids and spirals (Table 3) . The presence of these enzymes in the dormant coccoids could function in the transport of inorganic phosphates and generation of energy source in the form of ATP. The coccoids contained ATP but 100 x less than in the spirals (Table 2) . This signifies that the coccoids are a viable but dormant form. Similarly, a 99% reduction in endogenous respiration was exhibited by the marine vibrio ANT 300 as part of its survival strategy under long term nutrient starvation (Novitsky & Morita, Applied and Environmental Microbiology, 32:617- 622 (1976) ) .

TABLE 3 : Enzyme profile of spirals and coccoids

ENZYME SPIRALS COCCOIDS

Oxidase + -

Catalase + -

Urease + w 3.61U/mg protein 0.18U/mg protein

Acid + (≥40nM) + (≥40nM) phosphatase

Alkaline + (≥40nM) + (≥40nM) phosphatase

Napthol AS- + (20nM) + (20nM) Bl phospho- hydrolase

Leucine + (≥40nM) + (5nM) arylamidase

+ = positive - = negative W = weak positive

In addition, recent work has elucidated the formation of intracytoplasmic polyphosphates in spirals of H. pylori under adverse conditions (Bode et al , Journal of General Microbiology, 139:3029-3033 (1993) ; Caselli et al , Gut, 34:1507-1509 (1993)) . These structures represent a reservoir of stored energy and phosphorus and may be regarded as an alternative energy source when ATP is in short supply. Volutin granules (polyphosphates) were observed in the coccoids by Albert's stain (Fig 9) and could constitute a survival strategy. Furthermore, the presence of the phosphohydrolase enzyme could play a role in the metabolism of these polyphosphates.

The mean urease specific activity of coccoids was twenty times less than that of the spirals having an activity of 0.18 +/- 0.03 U mg^"1 protein as compared to 3.61 +/- 0.52 U mg ^'l protein in the spirals. The low urease actvity in the coccoids could either be due to the preformed enzymes left in the coccoids or that the dormant coccoids do not require as much urease enzyme activity as the actively reproducing spirals. This could also be the reason why the rapid urease test on coccoids was negative even after 48 hours and that the urease activity band could not be detected by the sensitive staining on PAGE by Shaik's method (Shaik et ai, Analytical Biochemistry, 103:140-143

(1980)) as shown in Fig 10. Nevertheless, the urease subunits A and B of 29 and 66 kD respectively seem conserved in the cocoids as shown on the denaturing SDS PAGE (Fig 5) . It is interesting to note that equal band intensities of these two urease subunits were observed in the spirals and coccoids in spite of the 20 times lower urease activity in the coccoids.

The dominant presence of only the urease A and B subunits in the coccoids could explain the decreased urease activity as urease subunits C and D which are also essential for urease activity in H. pylori (Labigne et al , Journal of Bacteriology, 173:1920-1931 (1991)) were absent.

Protein profile .

It was also observed as shown in the SDS-PAGE (Fig 5) that there was a decrease in the number and intensity of various protein bands of molecular weight < 14 > 200 kD in the coccoids indicating the dormant, inactive state of this form. This is further strengthened by the fact that the total protein content per cell in the coccoids was only half that found in the spirals (Table 2) . Reeve et al ( Journal of Bacteriology, 160:1041-1046 (1984)) demonstrated that protein degradation was necessary for starvation survival of Escherichia coli and Salmonella typhimurium as part of energy conservation. On the other hand, conservation of the possible urease subunits A and B as well as the presence of flagellar proteins could mean that the urease enzyme and flagella are necessary for the conversion of the coccoids to spirals and subsequent survival of the spirals in the gastric environment.

In the native PAGE, the presence of three novel proteins of 955, 871 and 661 kDa proteins in trace amount and a distinct band of 60 kDa were observed (Fig 11) . Appearance of the novel protein bands would appear to indicate that the coccoids have a purpose and do not simply represent a degenerative form of H. pylori . In addition, western immunoblotting identified the 60 kDa protein to be specific and immunogenic for the coccoids (Fig 12) .

Hae-maggrlutination and haemagglutination inhibition assay.

Majority of strains of H. pylori are known to agglutinate various red blood cells including human and rabbit (Morgan et al , 1991 infra , Taylor et al , Journal of Medical Microbiology, 37:299-303 (1992)) . In this study, both the spirals and coccoids agglutinated human red blood cells equally well at bacterial concentrations of > 10⁸ cells. With the rabbit red blood cells, the spirals agglutinated at a minimum of IO⁸ cells but the coccoids required IO⁹ cells to produce visible haemagglutination. Haemagglutination was inhibited by heat and protease treatment of the red blood cells. This indicated that the haemagglutinin in the coccoid, similar to that observed by Huang et ai (FEMS Microbiology Letters, 56: 109-112 (1988)) in the spirals, is a protein while the receptor is not a protein but sialic acid. Hence, the haemagglutinating property of the spirals is retained in the coccoid as was also observed by Wadstrom et ai (European Journal of Gastroenterology and Hepatology, 5(βuppl.2) :S12-S15 (19931).

Indirect fluorescent antibody test. The coccoids were shown to fluoresce under ultra violet light (Fig 13) indicating that their surface proteins are intact and similar to spirals. This property could be utilised to detect the specific presence and survival of H. pylori in the environment and may help to elucidate the route of transmission. Similar properties have been employed to study the survival of viable but non- culturable forms of Salmonella enteri tides (Roszak et al , Canadian Journal of Microbiology 30: 334-338 (1983)) and Vibrio cholerae as well as E. coli (Xu et al , Microbiol . Ecology 8: 313-323 (1982)) in the environment.

From the results it would appear that the coccoid form can exist in a viable form, contrary to what was believed previously. It has intact DNA, ATP enzyme activities, presence of novel and conserved protein and the presence of a thick polysaccharide coat to protect it under adverse conditions in the environment.

The fastidious nature of the spirals in both nutritional and physiological requirements inhibits their survival in the environment (Marshall et al, 1984 infra ; Dick, 1990 infra) . The microaerophilic spirals are sensitive to oxygen and have been reported to become non-viable if exposed to air for more than 2 hours (Soltesz et al , Journal of Clinical Microbiology, 30:1453-1456 (1992)) . Thus, the coccoids, as an alternative morphological state, might be the link in the cell cycle and possibly in the mode of transmission of H. pylori . Similar viable but non-culturable differentiated forms have been observed in the cell cycle of other microorganisms like Myxococcus xanthus (White et al , Journal of Bacteriology, 95:2186-2197 (1968)), marine vibrio ANT 300 (Novitsky & Morita, 1977 infra) and ArthroJbacter crystallopoietes (Boylen & Ensign, Journal of Bacteriology, 103: 569-577 (1970) ) and shown to play an integral role in their survival strategy under adverse conditions.

Many reports have cited the detection of H. pylori in water by non-culturable methods like the ³H thymidine uptake studies (Shahamat et al^", Klin. Wochens chr. , 67:62- 62 (1989)) and detection by polymerase chain reaction in faeces (Mapstone et al , Lancet, 341:447 (1993)) and in the water supply in Peru (Westblom et al , Acta Enterolgica Belgica, 56(suppl) :47 (1993)) . H. pylori prevalence studies by ³C urea breath test has also been associated with consumption of river water (Klein et al , Lancet, 337:1503-1508 (1991)). These indirect detection procedures do not exclude the possibility of detecting coccoids in the environment as this study shows the presence of DNA in the coccoids. Further, the isolation of H. pylori from human faeces immediately after excretion does not exclude the possibility that the spirals would not survive in the normal atmosphere for greater than 2 hours (Soltesz et al , 1992 infra) owing to their microaerophilic nature and oxygen toxicity (Krieg & Hoffman, Ann . Rev. Microbiol . , 40:107-130 (1989)) . These earlier reports only serve to establish the fact that there exists more than a single form of H. pylori .

Claims

CLAIMS :

1. An antigenic protein having a molecular weight of 60 kDa, as determined by native PAGE, obtainable from the coccoid form of tf. pylori .

2 . An antigenic protein as claimed in claim 1 which has the following N-terminal amino acid sequence:

D-T-H-K-S-E-I-A-H-R-F-N-D-L-G.

3. The use of an antigenic protein as defined in claim 1 or claim 2, in the detection of antibodies against tf. pylori .

4. The use as claimed in claim 3 wherein the detection of antibodies is used to detect the presence of the coccoid form of H.pylori .

5. The use as claimed in claim 3 wherein the antigenic protein from the coccoid form of H.pylori is used in combination with one or more other antigens obtainable from either the spiral or coccoid form of H.pylori .

6. The use of an antigenic protein as defined in claim 1 or claim 2 in the preparation of antibodies.

7. The use as claimed in claim 6, wherein the antibodies are polyclonal antibodies.

8. The use of antibodies as defined in claim 6 or claim 7 in the detection of the coccoid form of tf. pylori .

9. A kit for use in a method as defined in claim 8, comprising antibodies as defined in claim 7 or claim 8.

10. A composition comprising an antigenic protein as defined in claim 1 or claim 2, optionally together with one or more other tf. pylori antigens.

11. A composition as claimed in claim 10 wherein the optionally one or more other antigens are obtained from either the spiral or coccoid form of H.pylori , or both.

12. A method for the detection of tf. pylori which includes the step of contacting an antigenic protein as defined in claim 1 or claim 2 with a sample.

13. A method as claimed in claim 12 which is for the detection of the coccoid form of tf. pylori .

14. A method as claimed in claim 12 or claim 13 wherein the antigenic protein is in the form of a composition as defined in claim 10 or claim 11.

15. A method as claimed in any one of claims 12 to 14 wherein the sample is a biological sample.

16. A method as claimed in claim 15 wherein the biological sample is a sample of blood, a sample of urine or a sample of saliva obtained from a subject.

17. A method for the diagnosis of H. pylori infection which includes the step of contacting an antigenic protein as defined in claim 1 or claim 2, with a biological sample obtained from a subject.

18. A method as claimed in claim 17 wherein the antigenic protein is in the form of a composition as defined in claim 10 or claim 11.

19. A method as claimed in any one of claims 12 to 18 which also includes the step of detection of other antigens obtainable from either the spiral or coccoid form of H.pylori .

20. A method as claimed in claim 19, wherein the biological sample is a sample of blood, a sample or urine or a sample of saliva.

21. A kit for use in the diagnosis of H.pylori infection which comprises an antigenic protein as defined in claim 1 or claim 2.

22. A kit as claimed in claim 21 which further comprises one or more other antigens obtainable from either the spiral or coccoid form of H.pylori .

23. A vaccine for the prophylaxis or treatment of H.pylori infection which comprises an antigenic protein as defined in claim 1 or claim 2, together with one or more adjuvants and/or carriers.

24. A vaccine as claimed in claim 23 which further comprises one or more other antigens obtainable from either the spiral or coccoid form of H.pylori .

25. A vaccine as claimed in claim 23 or claim 24, wherein the antigenic protein is provided in the form of the coccoid form of tf. pylori .

26. A vaccine as claimed in claim 23 or claim 24 wherein the antigenic protein is provided in the form of an intermediate form of tf. pylori between the coccoid and spiral forms, which carries the antigen.

27. A method of detecting IgM antibodies against the coccoid form of tf. pylori in a child, which comprises the step of bringing an antigenic protein as defined in claim 1 or claim 2 into contact with a biological sample obtained from said child.

28. A method of culturing the coccoid form of tf. pylori , which comprises the step of regularly adding carbon dioxide to a culture medium containing the spiral form of tf. pylori such that conversion to the coccoid form occurs, and wherein the coccoid form obtained is viable.

29. A method as claimed in claim 28, wherein carbon dioxide is supplied twice daily for a period up to 3 months.