Culture medium for Haemophilus influenzae type B
The subject of the invention is a culture medium for Haemophilus influenzae type b, in which the source of protein nitrogen comprises at least one plant peptone and in which the heme source consists of protoporphyrin DC. The invention also relates to a method for producing polyribosyl ribitol phosphate (PRP) used in the manufacture of a vaccine against Haemophilus influenzae type b meningitis.
The capsule is the major factor of virulence of Haemophilus influenzae typeb strains. It is a polysaccharide consisting of a succession of repeating units of ribosyl ribitol phosphate. The expression polyribosyl ribitol phosphate (PRP) or capsular polysaccharide type b is used interchangeably to denote the Haemophilus influenzae type b capsule.
Haemophilus influenzae type b populations are oftm heterogeneous; capsulated bacteria coexist with noncapsulated bacteria. The noncapsulated bacteria have lost their capacity to express the capsule following genetic mutations which occur spontaneously. According to Hoiseth et al. (Infectious and Immimity (1985), 49: 389-395), the loss of the expression of the capsule occurs at a frequency of 0.1 to 0.3% at each bacterial generation. At the genetic level, the cap locus has been shown to be involved in the expression of the capsule at the surface of these bacteria (KroU, J. S., et al., J. Bacteriol. (1988) 170: 859-864). The capsulated bacteria have a cap locus which contains at least tvo copies of an 18 Kb gene. The noncapsulated bacteria no longer have an 18 Kb gene or only one single copy of this gene. To identify the capsulated bacteria, the test of agglutination on a slide of bacteria in the presence of an anti-PRP antibody is usually used or molecular biology techniques which charactoize the cap locus are used.
Vaccines based on PRP, or PRP covalentiy linked to a carrier protein, are used to prevent Haemophilus influenzae type b infections. To manufacture these vaccines, it is necessary to produce large qxiantities of bacteria in large volteies of culture medium from which PRP is extracted and then purified. Nevertheless, the ease with which the capsulated Haemophilus influenzae type b bacteria revert to noncapsulated forms can constitute a stumbling block for the production of PRP.
For the industrial production of PRP, culture media are generally used which are based on animal peptones which represent the principal source of protein nitirogen

supplemented with yeast extract, glucose, hemin, P-NAD and inorganic salts. By way of example, the production medium described in US 4,459,286 is mentioned.
Because of the risks linked to BSE, it is sought to replace products of animal origin and more particularly products of human or bovine origin with products offering better biological safety.
Carty et al. (in Dev. Indust. Microbiol. 26: 763-767 (1985)) have shown that animal peptones could be replaced by soybean peptone for the production of PRP. The composition par liter of this medium (MP medium) is the following: soybean peptone: 10 g; yeast extract: 10 ml; NaCl: 5 g; K2HPO4: 2.5 g; Na2HP04: 3.3 g; dextrose: 5 g; hemin chloride: 10 mg; NAD: 10 mg.
Takagi et al. (J. Chem. Tech. and Biotech 81: 182-188 (2006) have sought to optimize the composition of the Carty medium (MP medium). They have shown that the PRP concentration in the culture medium could be increased by 70% to reach 0.25 g/1 when the hemin and p-NAD concentration was increased. To increase the production of PRP, it therefore appears to be necessary to increase the cofactor (hemin and P-NAD) concentrations which are necessary for the growth of Haemophilus influenzae type b.
The need therefore still exists to improve the methods for producing PRP, in particular when the culture volumes are large ( 100 liters) while applying the best biological safety conditions.
Accordingly, the subject of the invention is a novel culture mediimi for Haemophilus influenzae type ft, characterized in that the protein nitrogen source is of nonanimal origin and comprises at least one plant peptone and in that the heme source consists of protoporphyrin DC. Such a medium is particularly suitable for tihie industrial production of PRP. It offers greater biological safety since pqjton of animal origin have been replaced by plant peptones. It also responds to contingaicies of production cost since 10 to 20 times less protoporphyrin IX than hemin is required in order to obtain PRP levels in the culture supematants which are industrially exploitable Gevel > 0.2 g/1). The expression "medium for the culture of Haemophilus influenzae serotype b" is understood to mean a medium which encourages the growth of Haemophilus zn/?«enzae 5ero/y/?e Z) and which comprises:
- a source of protein nitrogen,
- a heme source,

- a P-NAD source
- a carbohydrate source,
- a source of vitamins and growth factors, and
- inorganic salts.
The expression "protein nitrogen source" is understood to mean a preparation in which the quantity of amino acids, peptides, polypeptides, peptones and/or proteins represents at least 50% of the dry weight of this composition.
The expression "protein nitrogen source of nonanimal origin" is understood to mean a protein nitrogen source which has a nonanimal origin. As a result, the preparation is not produced fiom animal cells, animal tissues, or animal organs or bodies. It is generally produced fix)m plants, algae, bacteria, yeasts or ftingi.
The culture medium according to the invention denote, firstiy, a liquid medium, however, it may also be in solid form. The solid form is obtained by adding a gelling substance, such as agar which is usually used at a concoitration range of from 10 to 30 g/1, to the liquid medium.
The source of heme according to the invention is represented by the protoporphyrin DC of formula:
CH,

ROOC OOR
where R denotes either H or a salt, preferably an alkali metal salt, in particular a sodium salt.
In the case of the present invention, protoporphyrin DC is not complexed with iron. Up until now, all the media recommended for producing PRP contained, as heme source, either protoporphyrin DC complexed with iron (which is the case for heme), or a protoporphyrin DC complexed with FeCl (which is the case for hemin), or finally and

more rarely a protoporphyrin DC complexed with FeOH (which is the case for hematin). Even if the Haemophilus influenzae serotype b strains possess a ferrochelatase which allows them to convert protoporphyrin IX to a form complexed witti iron (Loeb et al., J, Bacteriology (1995), 177; 3613-3615), it has nevo: been shown that it would be possible to use, as heme source, noncomplexed protoporphyrin K in a culture medium based on plant peptone to produce PRP at a concentration which is industrially exploitable. It is generally considered that a rate of PRP cTf at least 0.1 to 0,2 g/1 is required in the culture supernatant in order to guarantee an industrial production of PRP
Surprisingly, it has been observed that by using a protoporphyrin DC as heme source and a plant pq)tone as protein nitrogen source, the protoporphyrin DC concentrations necessary were 10 to 100 times lower than those which are used when the heme source is a protoporphyrin DC complexed with iron. As shown in example 1, a protoporphyrin DC concentration of between 100 and 200 ]ig/\ is sufBciait to obtain a maximum production of PRP (the concentration in the culture supernatant is » 0.4 g/1) whereas a hemin concentration 10 to 20 times hier is necessary to obtain an equivalent PRP production. Moreover, even at a concentration as low as 10 yLg/\ of protoporphyrin DC, a significant production of PRP observed while it is insignificant with hemin. At 50 ng/1 of protoporphyrin DC, the production of PRP reaches or exceeds 100 |ig/l (which is already a rate which can be used on an industrial scale) whereas it is of the order of 10 jig/1 or even less when the culture medium includes hemin at the same concentration (see figures 1 and 2).
Accordingly, the subject of the invention is:
A culture medium according to the invention in which the protoporphyrin DC concentration is at least 0.01 mg/1, at least 0.02 mg/1, at least 0.03 mg/1, at least 0.04 mg/1 or preferably at least 0.05 mg/1.
Generally, the protoporphyrin DC concentration in the culture medium is betweai 0.1 mg/1 and 5 mg/1, and preferably between 0.1 mg/1 and 2 mg/1. In these concentration ranges there is optimum use of the raw materials to give an optimum production of PRP in the culture superaatants.
The protoporphyrin DC which is suitable for the subject of the invention may be of animal origin and may be produced firom animal (bovine, porcine and the like) tissues. The degree of purity of these preparations is generally at least 80%, preferably at least

90%, and more preferably still at least 95% (weight/weight). Although the contaminants may contain residual quantities of amino acids, of peptides and/or of proteins, the protoporphyrin DC preparations may not be considered as bdng a protein nitix)gen source for the purposes of the invention since the residual quantities of amino acids, of peptides and/or of proteins, optionally present, generally represmt less than 5%, in general less than 1% of the dry weight of the preparations.
Preferably, to ensure greater biological safety, a protoporphyrin DC is used which is free of any contaminant of animal origin. To produce such a protoporphyrin DC, it is possible to use the method of production as described in the French patent application registered under the registration No. 07/02334 and filed on 30/03/2007 using the steps which are described in scheme 2.
In another preferred embodiment, the culture medium according to the invention comprises a protoporphyrin DC free of any contaminant of animal origin.
According to the subject of flie invention, the principal source of protein nitrogen is represented by one or more plant peptones. They are generally in the form of hydrolysates. They are obtained by enzymatic or chemical treatment of the proteins generally extracted from the parts of the plant which have the highest contents of proteins. Preferably, plants are used which have not been genetically modified. When the chemical route is used, one of the methods consists in treating the protein extract with hydrochloric add in the hot state and under pressure. The hydrolysate is then neutralized with sodiimi hydroxide and then freed of solid by-products. When the enzymatic route is used, one of the classic methods consists in treating the protein extract with papain.
Plant peptones are prq)arations containing mainly a mixture of amino acids, and of small peptides whose MW is 1 KD. The peptides whose MW is > IKD generally represent less than 40% of the mixture. One may also, if necessary, use ultrafiltered hydrolysates in order to enrich or select the small size peptides. One may also subject the ultrafiltered hydrolysate to an additional chromatography stqp in order to select the hydrolysate fractions having a PM < 1 KD, or < 500 Daltons or even < 350 Daltons. Plant peptone preparations are thus obtained in which more than 40% of the peptides, more than 50% of the peptides, or even more than 60% of the peptides have a PM < 1 KD, or < 500 Daltons, or even < 350 Daltons. The plant peptones which are suitable

for the subject of the invention, are obtained in particular from potato such as those provided by Organotechnie (plant peptone El or plant peptone ETl), from soybean such as those provided by Organotechnie or Kerry, from cotton (Hy cotton provided by Quest), from rice (Hy rice provided by Keny), from broad bean provided by Solabia, from wheat such as those provided by Organotechnie (wheat peptone El) or Kerry (Hypep™ 4602, Hypep 4601) or from garden pea, in particular the enzyme hydrolysates of garden pea provided by Kerry (HY pea 7404) or oxoid (VG 100) or acid hydrolysates of garden pea provided by Oxoid refwenced under the name "Acid hydrolyzed vegetable peptone". Preferably, the plant pqptone suitable for the subject of the invention is a wheat peptone and more preferably still the plant peptone is a garden pea peptone.
To define the concentrations for using the plant peptone, account is taken of the content of protein nitrogen of the peptone. This content is calculated using fee Kjeldahl method (Lynch JM et al. , J AOAC Int. (1999) 82(6): 1389-98). Usually, the protein nifrogen contents of the plant peptones in accordance with the subject of the invention are between 8% and 15% per gram of peptone (weight/weight). In this range, good results are obtained when the plant peptone concentration in die culture medium according to the invention corresponds to a protein nitrogen concentration ranging from 0.08 to 2.25 g/1 and preferably in a range of concentration ranging from 0.4 to 1.5 g/1.
Accordingly, the subject of the invention is a medium in which the total plant peptone concentration is equivalent to a protein nitrogen conceatration ranging from 0.08 g/1 to 2.25 g/1.
There is used as a source of J3-NAD (also called factor V or p-nicotinamide adenine dinucleotide), a purified preparation of P-NAD itself or a purified preparation containing a derivative of p-NAD chosen from nicotinamide riboside (NR), P-nicotinamide adenine mononucleotide (NMN), or P-nicotinamide adaiine dinucleotide phosphate (NADP). The degree of purity of the preparation is generally at least 80%, preferably at least 90% and still more preferably at least 95%. There is preforably used, in the case of the present invention, a source of P-NAD which is free of protein contaminant of animal ori. These purified preparations are used at a concentration of at least 1 iM. By way of example, p-NAD is used at a concentration ranging from 2 to 50 mg/1 of culture medium.

As a source of carbohydrate, any sugar which is metabolized by Haemophilus influenzae type b, such as fructose, ribose, xylose, fiicose, glycerol or more particularly glucose may be used. Generally, the carbohydrate source has a nonanimal origin and the carbohydrate concentration in the culture medium is at least 10 mM. When glucose is used, its conc«itration in the culture medium is generally betweao 2 to 20 g/1.
The culture medium according to the invention also comprises a source of vitamins and growth factors. There is used to this end a yeast extract which is obtained from the soluble fraction of the product of autolysis of brewer's yeast derived from the culture of Saccharomyces sp. Numerous amino acids and vitamins such as vitamins B5, Bl, B2, B6, PP, H and B12, trace elements and oligo nucleotide derivatives are found in its composition. The commercially available autolytic extracts of yeast produced by Quest, Difco or Solabia are suitable for the subject of the invention.
The concentration of yeast extract in the medium according to the invention is usually within a range of concenfration ranging from 0.2 g/1 to 15 g/1 and preferably win a range of concentration rang from 0.2 g/1 to 10 g/1 and even more advantageously within a range of concentration ranging from 0.2 to 5 g/1. It has been observed that the RPR production by bacteria was better when the concentration of yeast extract was in the concentration range ranging from 0.2 to 5 g/1.
The yeast extract also represents an additional source of protdn nitrogen of nonanimal origin. The contents of protein nitrogoi in the yeast extracts are indeed gaierally from 9 to 11% (weight/weight). To avoid nitrogenous hypercatabolism which may be responsible for the accumulation of toxic wastes during culture, the concentrations of plant peptone(s) and of yeast extract are generally adjusted such that the total protein nitrogen contait in the medium does not exceed 2.5 g/1. Preferably, the concentrations of plant peptone(s) and of yeast extract are adjusted such feat fee total protein nitrogen contait in fee medium according to fee invraition is 0.5 to 2.5 g/1.
The culture medium according to fee invention also comprises inorganic salts. The inorganic salts used are generally in fee form of salt solutions, at least one of which exerts a sufficiait buffering power for fee initial pH of fee medium, before inoculation of bacteria, to be preferably between 6.5 and 7.5 and most preferably between 7 and 7.5. In general, a mixture of monovalent cations such as Na* and/or K*, divalent cations such as Ca" and/or Mg, phosphate anions in HPO4", H2PO4' and/or P04~ form, and S04~

and Cr anions in the form of salt solutions whose molarities may vary witiiin a concentration range ranging from 10' mM to 100 mM is used.
In addition to the components described in the preceding paragraphs, it is clearly imderstood that a culture mediirai according to the invention may incorporate in its composition one or more other inorganic and/or organic components provided that they do not negatively interfere with the production of PRP. Very preferably, components originating from a nonanimal source are introduced. Thus, it is possible according to the invention to add to the medium amino acids produced by chemical synthesis, or by microbiological fennentation, such as tryptophan and/or cystine, inorganic nitrogm generally in the form of salt solutions providing NH4* ions and/or even othar substances such as sodium lactate. These additives are generally used at low concentrations. The amino acid supplement in the culture medium is generally at a concentration 1 mM. Similarly, the ammonium salts and/or the sodium lactate are generally at a concentration 10 mM. Finally, although it is not necessary to supplement the culture medium according to the invention by a supply of iron in the form of ionic iron since it is already present in sufficient quantity in the composition of the plant peptone and of the yeast extract, it is possible, as a precaution, to add to the culture medium a solution of iron salt, in a concentration range which may range from 0.5 to 10 mg/1 in order to avoid any iron deficit which may occur during bacterial growth.
Advantageously, the culture medium according to tiie inv«ition is free of any protein of any polypeptide, peptide and/or amino acid of human or bovine origin or even free of any protein, of any polypeptide and/or amino acid of animal origin, or even more advantageously still, free of any contaminant of animal origjn.
According to a particular embodiment, the subject of the invention is a culture medium which comprises:
- from 0.1 mg/1 to 5 mg/1 of protoporphyrin DC,
- from 2 to 50 mg/1 of p-NAD,
- from 2 to 20 g/I of glucose,
- from 2 to 5 g/1 of a yeast extract,
- a gardai pea peptone equivalent to a protein nitrogen concentration of 0.4 g/1 to 1.5 g/1, and
- a cocktail of inorganic ions comprising Na", NH4 Ca", Mg**, HP04~,

H2P04', SO4"" and CI" ions in the form of salt solutions such that the pH of the medium is between 6.5 and 7.5, preferably between 7.0 and 7.5. By using this medium composition, the development of noncapsulated revCTtant bacteria during culture in liquid medium is prevented. It was indeed noted that by inoculating into this medium a population which initially contains 100% of capsulated bacteria (the genome of the entire bacterial population of a cap locus which contains two copies of the 18 kb gene), a bacterial population which still contains 100% of capsulated bacteria is obtained after a period of culture equivalent to 40 bacterial generations. This mediimi composition, which particularly includes that which is mentioned in example 3.2.2.1 and which is used in example 4, contributes toward improving the PRP yields by exercising a stabilizing role on the population of capsulated bacteria (cf. example 4).
According to another aspect, the subject of the invention is a method for producing polyribosyl ribitol phosphate (PRP) in which:
(i) Haemophilus influenzae serotype b is cultured in a liquid culture
medium according to the invention, (ii) the culture supernatant obtained in (i) is collected, and (iii) the PRP is extracted from the culture supernatant. To produce the PRP according to the method of tiie invention, it is possible to carry out in step (i) one or more successive Haemophilus influenzae serotype b cultures in a liquid medium according to the invention. The successive cultures make it possible to increase the biomass.
To do this, the bacteria obtained from a freeze-dried product or from a frozen product are inoculated into a volume of medium generally not exceeding 1 liter. After one night of culture or when the optical density of the medium is sufficient, this first culture is transferred into a second culture medium which is identical to the first, but whose volume may be up to 10 to 20 times larger. The quantity of bacteria inoculated into the second medium is adjusted such that the initial optical density (OD) of the second culture medium at 600 nm is between 0.2 and 0.4 in order to promote rapid growtii of the bacterial population. This second culture is usually carried out in a fermenter but other types of containers may be used (flasks, spinners, and the Uke). When the culture is carried out in a fennenter, during the duration of the culture a temperature of 37°C ± 1°C, constant stirring, a pressure of 0.1 bar, a p02 of 30% and an

air flow rate of 0.25 volume of gas per volume of medium per minute are usually used. It is witiiin the competence of persons skilled in the art to choose other parameters for this type of culture. At the end of the exponential bacterial growth phase, it is possible to further amplify the biomass by transforing it into another fisrmeater of a larger capacity using the same procedure and so on. The culture volumes obtained may be up to, or even exceed, 1000 liters. The culture(s) is(are) generally carried out according to the batch mode. It is also possible to adopt other modes of culture, in particular the fed-batch mode. In this case, a nutritive carbohydrate supplement is added to the medium during the exponential growth phase so as to prolong bactorial multipUcation and to obtain, at the end of the exponential growth phase, a higher bacterial density. The quantity of carbohydrate added is evaluated as a function of the level of lactate present in the medium at the time of addition.
The supernatant of the last culture is finally collected after inactivation of the bactCTia. The inactivation is conventionally carried out with the aid of a formalin solution at a final concentration of 0.35%-0.37% (v/v). The supernatant is conventionally separated fi-om the bacteria by a centrifugation step. The PRP contained in the resulting suponatant is then extracted and pimfied according to conventional methods well known to persons skilled in the art.
According to another embodiment, the subject of iht invention is a method for producing PRP in which:
(i) Haemophilus influenzae serotype b is cultured on a solid medium, (ii) one or more colonies obtained in (i) are transferred into and cultured
in a liquid culture medium according to the invention, (iii) the culture supernatant obtained in (ii) is collected, and (iv) the PRP is extracted fi:om the culture supernatant. The solid culture medium which can be used in the method of (he invention should be suitable for the culture of Haemophilus influenzae serotype b. It also comprises:
- a source of protein nitrogen,
- a heme source,
- a P-NAD source,
- a carbohydrate source,

- a soiirce of vitamins and growth factors,
- inorganic salts, and
- a gelling substance, usually agar at a concentration of 10 to 30 g/1.
In the method for industrial production of PRP, a preliminary step of culture in solid medium is conventionally used. Usually, the bacteria obtained form a freeze-dried product or from a frozen product are resuspended and then inoculated onto a charcoal-based solid medium supplemented with horse blood. After 16 to 20 hours of culture in an incubator at 37°C under 10% COj, bacterial colonies are collected and amplified in a liquid mediimi. This method has the disadvantage of using a solid medium which contains, as protein nitrogen source, proteins of animal origin. The invaitors have therefore tried to identify compositions of solid media in which the protein nitrogen source is free of any protein of animal origin.
Initially, the inventors have demonstrated that it is possible to use a solid medium in which the yeast exfract was at the same time able to serve as a protein nifrogen source, a source of vitamins and growth factors, the sources of heme, P-NAD, carbohydrate, and inorganic salts having ihs same characteristics as those which have previously been described. A minimum concaitration of 0.05 mg/1 of protoporphyrin DC, 0.1 \iM for the P-NAD source and 0.1 mM for the carbohydrate source are recommended whilst fht concentration of yeast extract in the medium corresponds to a content of protein nitrogen of 0.2 to 1.5 g/1. The colonies obtained are viable even if the growth symbolized by the size of the colonies is not always optimal. They may be directly transferred into a liquid culture medium according to flie invention. The procedure is then carried out as above in order to amplify the culture volumes, and extract and purify tiie PRP.
Preferably, tihe solid culture medium comprises as a source of protein nifrogen at least one peptone of plant origin used in the form of a chemical or enzymatic hydrolysate. It is possible to use in particular, as yeast extract supplement, at least one plant peptone obtained from wheat, cotton, rice, soybean, field bean, potato, garden pea or a mixture of these at a protein nifrogen concenfration which may range in particular from 0.2 g/1 to 2 g/1, the total protein nifrogen concenfration in the solid medium preferably not exceeding 2.5 g/1. By way of example of mixtures of plant peptones which may be used, there may be mentioned a mixture based on soybean, cotton and rice peptones or a mixture based on garden pea, cotton and wheat pq>tones or finally a

mixture based on garden pea and potato. The inventors have indeed noted that, when the culture medium also comprises a plant peptone as source of protein nitrogen, bacterial growth and the viability of the bacteria were better than those observed on a charcoal-based solid medium (charcoal agar) supplemented with boiled or defibrinated horse blood. They are at a maximum when the plant peptone used is a garden pea peptone.
Accordingly, the subject of the invention is also a method for producing PRP in which the protein nitrogen source of the sohd medium is free of protein of animal origin and comprises at least one plant peptone. Preferably the plant peptone is a garden pea peptone.
Advantageously, the solid culture medium according to the invention is free of any protein, of any polypeptide, of any peptide and/or of any ammo acid of human or bovine origin or more advantageously still free of any contaminant of animal origin. In the latter case, the heme source consists of synthetic protoporphyrin DC, the P-NAD and carbohydrate source also being of nonanimal origin, the source of vitamins and of growth factora being provided by a yeast extract and the gelling substance being agar (a product derived from algae).
For this reason, another subject of the invention is a method for the production of PRP according to which the sohd culture medium and the liquid culture medium are free of any contaminant of animal origin.
It has also been sought to optimize the composition of the solid medium so that it is possible to select the bacteria colonies which produce the most PRP. The composition of such a medium should make it possible to obtain:
- good individualization of the colonies;
- good viability of the colonies;
- development and sufficient size of the colonies so as to be able to study their morphology. In order to be able to discriminate between the colonies, it is necessary to have a medium which makes it possible to obtain Haemophilus influenzae type b colonies having a sufficient size at the end of 16-24 hours of culture (about 3 to 5 nmi).
In one of the preferred embodiments of the method for producing PRP according to the invention, the solid medium comprises: - at least 1 mg/1 of p-NAD,

- at least 0.5 mg/l of protoporphyrin DC,
- a plant peptone and a yeast extract in sufficient quantity for the protein nitrogen concentration in the solid medium to be at least 0.2 g/1 and in a proportion such that the ratio between the quantity of plant protein and the quantity of yeast extract in the medium is 0.1 to 9 when the concentration of protein nitrogen of the medium is 0.2 g/1 to 0.8 g/1 and is 1 to 9 when the concentration of protein nitrogen of the medium is > 0.8 g/1,
- a carbohydrate,
- a detoxifying agent, and
- a cocktail of inorganic ions comprising Na K"*", Ca"*, Mg", Fe"*, HP04~, H2PO4", S04'" and CI' ions in the form of salt solutions such that the pH of the culture medium is between 6.5 and 7.5, preferably between 7.0 and 7.5.
The carbohydrate used is preferably a sugar of nonanimal origin, metabolizable by the bacterium, such as fructose, ribose, xylose, fucose, glycerol or, in particular, glucose. Good results were obtained with glucose at a concentration of at least 0.1 g/1. Usually, glucose is used at a concentration of 0.1 g/1 to 20 g/1 and preferably 0.1 g/1 to lOg/l.
The detoxifying agents promote the growth of the bacteria by neutralizing the inhibitory substances which may be present in the agar preparations as reported by Evans J. M et al. (J. Med. Microbiol. Vol 7, pp 305-309, 1974). As detoxifying agent, charcoal, starch, Tween®, polyvinyl alcohol, sodium oleate or sodium dithionite are preferably used. Good results were observed with Tween 80® (sorbitane polyoxyethylene monooleate) used at a concentration of 0.5 to 10 mg/l whereas a solid medium composition without Tween® produces small size colonies which are indistinguishable from the morphological point of view. Small size colonies which are indistinguishable from the morphological point of view are also observed when the P-NAD concentration is less than 1 mg/l, when the protoporphyrin DC concentition is less than 0.5 mg/l or when the glucose concentration is less than 0.1 mg/l.
In order to ensure good growth and good viability of the colonies, the quantities of yeast extract and plant peptone are such that the concentration of total protein nitrogen is at least 0.2 g/1. The ratio between the quantity of plant peptone and the quantity of yeast extract in the medium may vary to a large degree, ranging from 0.1 to 9

as long as the concentration of protein nitrogen in the culture medium does not exceed 0.8 g/1. On the other hand, at a higher concentration, there is poor individualization of the colonies due to poor spreading of the bacterial suspension on the solid medium when this ratio is less than 1.
By inoculating this solid agar-based medium with a heterogeneous population of Haemophilus influenzae serotype b bacteria (i.e. which contains both capsulated and noncapsulated bacteria), after 18 to 24 hours of culture white colonies and gray colonies are observed which are distinguishable by transparency with the aid of a beam of white Ught. The white colonies produce more PRP than the gray colonies. Moreover, the white colonies also produce more PRP than the colonies obtained from a charcoal agar-based solid medium supplemented with horse blood (cf. example 2). This medium composition is considered as a selective medium composition since it makes it possible to sort the colonies which produce the most PRP.
According to an even more preferred embodiment of the method according to the mvention, the solid medium comprises:
from 5 to 50 mg/1 of p-NAD,
from 0.5 to 5 mg/1 of protoporphyrin IX,
from 1 to 10 g/1 of glucose,
from 1 to 10 mg/1 of Tween 80,
from 3 to 4 g/1 of K2HPO4,
from 0.9 to 3 g/1 of KH2PO4,
from 0.5 to 2 g/1 of K2SO4,
from 20 to 500 mg/1 of MgCb,
from 2 to 50 mg/1 of CaCl2-2H20,
from 1 to 5 mg/1 of FeCl3-6H20,
from 4 to 8 g/1 of NaCl,
from 4 to 8 g/1 of a yeast extract, and
from 4 to 8 g/1 of a garden pea peptone such tiiat the ratio between the quantity of
garden pea peptone and the quantity of yeast extract is 1 when the protein
nitrogen concentration of the medium is > 0.8 g/1.
The white colonies obtained from this medium composition produce up to 400 times more PRP than the gray colonies. Their cap locus was studied by digesting tiie

genomic DNA with the aid of the restriction enzymes Smal and KpnI. Pulsed field electrophoresis was tiien performed on the digestion product followed by visualization with the aid of a specific PvuII probe according to the opiating conditions described in example 3. Surprisingly, all the electrophoretic profiles fix)m white colonies contain an electrophoretic band of 45 kb. No 18 kb electrophoretic band is observed. The cap locus of these colonies consequently contains at least two copies of the 18 kb gene, which means that the bacterial population derived from the white colonies is completely capsulated. On the other hand, the electrophoretic profiles fix>m the gray colonies mainly contain an electrophoretic band of 18 kb. This particularly preferred selective medium composition additionally makes it possible to sdect white colonies whose bacterial populations are completely capsulated.
In fact, one of flie additional means for increasing the yields in the PRP production method whidi comprises a preliminary phase of culture on a solid medium consists in transferring into a liquid medium only white colonies which have been obtained from a selective solid medium composition. Preferably, a solid medium composition is used which makes it possible to obtain white colonies essentially consisting of capsulated bacteria.
Accordingly, the subject of the invention is also, in a preferred embodiment, a method for producing PRP in which only the white colonies obtained on a selective solid medium composition are transferred into tiie liquid culture medium.
In a particularly preferred anbodiment, these white colonies are transferred into a liquid culture medium which exercises a stabilinng role on the capsulated bacterial population. In addition to tiie fact that all the culture steps are carried out with media in which the protein nitrogen source is of nonanimal origm, or even with media free of any contaminant of animal origin notable when synthetic protoporphyrin DC is used, this method also makes it possible to optimize the production of PRP when the Haemophilus influenzae serotype b population is heterogeneous and contains both capsulated bacteria and noncapsulated bacteria. The step of culture on a solid medium makes it possible to select the white colonies which contain a population of completely capsulated bacteria. The step of amplifying the biomass in the liquid medium stabilizes, as sera above, the population of capsulated bacteria by preventing the development of noncapsulated revertants. The yields of PRP/litCT of culture whidi are finally obtained are thai at a

maximum (see example 4).
This method can also be used in the production of a population of completely capsulated bacteria. The bacterial population obtained is completely capsulated when the electrophoretic profile of the genomic DNA of this population shows that the cap locus contains at least two copies of the 18 kb gene (cf protocol of example 3) and when the inoculation of an aliquot of tiiis population onto a selective solid medium composition according to the invention produces more than 95% of white colonies, preferably at least 98% of white colonies. After amplification of the bacteria in a stabilizing liquid medium (i.e. which prevents the appearance of noncapsulated revertant bacteria), the bacterial population obtained is preserved by freeze-drying or by freezing (in this case, a freezing agent of nonanimal origin such as glycerol is added to the culture medium). Thus inoculum batches are made containing a homogeneous population of completely capsulated bacteria and which offer an additional guarantee of biological safety since they were obtained using culture media which is free of any contaminant of animal origin. These inoculum batches can serve in tum to produce PRP.
The subject of the invention is therefore:
- A method for producing PRP, in which all of the stages are carried out using media which are free of any contaminant of animal origin.
- A method for producing a population of completely csulated Haemophilus influenzae serotype b bacteria in which;
(i) Haemophilus influenzae serotype b is cultured on a solid medium comprising:
- from 5 to 50 mg/1 of p-NAD,
- from 0.5 to 5 mg/1 of protoporphyrin EX,
- from 1 to 10 g/1 of glucose,
- from 1 to 10 mg/1 of Tween 80,
- from 3 to 4 g/1 of K2HPO4,
- from 0.9 to 3 g/1 of KH2PO4,
- from 0.5 to 2 g/1 of K2SO4,
- from 20 to 500 mg/1 of MgCla,
- from2to50mg/lofCaCl2-2H2O,
- from 1 to 5 mg/1 of FeCl3,-6H20,

- from 4 to 8 g/1 of NaCl,
- from 4 to 8 g/1 ofa yeast extract, and
- from 4 to 8 g/1 of a garden pea peptone such that the ratio between the quantity of garden pea peptone and the quantity of yeast extract is 1 when the protein nitrogen concentration of the medium is > 0.8 g/1;
(ii) one or more white colonies obtained in (i) are transferred into and cultured in a liquid culture medium comprising:
- from 0.1 mg/1 to 5 mg/1 of protoporphyrin IX,
- from 2 to 50 mg/1 of p-NAD,
- from 2 to 20 g/1 of glucose,
- from 2 to 5 g/1 of a yeast extract,
- a garden pea peptone equivalent to a protein nitrogen concentration of 0.4 g/1 to 1.5 g/1, and
- a cocktail of inorganic ions: Na*, NH4*, Ca", Mg, HPO, H2P04", SO4" and CV in the form of salt solutions such that the pH of the medium is between 6.5 and 7.5, preferably between 7.0 and 7.5, and
(iii) the bacterial culture obtained in (ii) is frozen or freeze-dried. Finally
-A method for producing a population of completely capsulated
Haemophilus influenzae serotype b bacteria, in which all the steps are carried
out by means of media free of any contaminant of animal origin.
The subject of the invention is also the use of a homogeneous population of
capsulated Haemophilus influenzae serotype b bacteria obtained according to this
method, for the production of PRP.
The subject of the invention is a vaccine against Haemophilus influenzae type b meningitis comprising PRP obtained from one of the embodiments of the method according to the invention.
The subject of the invention is finally a solid culture medium for Haemophilus influenzae serotype b, the protein nitrogen source of which is of non-animal origin and which comprises:
- at least! mg/1 of p-NAD,
- at least 0.5 mg/1 of protoporphyrin DC,

- a peptone and a yeast extract in a sufficient quantity for the protein
nitrogen concentration in the medium to be at least 0.2 g/1 of protein
nitrogen and in a proportion such that the ratio between the quantity
of plant peptone and the quantity of yeast extract in the medium is 0.1
to 9 when the protein nitrogen concentration of the medium is 0.2 g/1
to 0.8 g/1 and is 1 to 9 when the protein nitrogen concentration of the
medium is > 0.8 g/l,
a carbohydrate,
- a detoxifying agent, and
- a cocktail of inorganic ions: Na*, K Ca**, Mg", Fe HP04~, H2PO4", SO4" and Cr in the form of salt solutions such that the pH of the medium is between 6.5 and 7.5, prefa:ably between 7.0 and 7.5.
Preferably, the solid culture medium comprises:
- from 5 to 50 mg/1 of p-NAD,
- from 0.5 to 5 mg/1 of protoporphyrin DC,
- from 1 to 10 g/1 of glucose,
- from 1 to 10 mg/1 of Tween 80,
- from 3 to 4 g/1 of K2HPO4,
- from 0.9 to 3 g/1 of KH2PO4,
- from 0.5 to 2 g/1 of K2SO4,
- from 20 to 500 mg/1 of MgCh,
- from 2 to 50 mg/1 of CaCl2-2H20,
- from 1 to 5 mg/1 of FeCl3-6H20,
- from 4 to 8 g/1 of NaCl,
- from 4 to 8 g/1 of a yeast extract, and
- from 4 to 8 g/1 of a garden pea peptone such that the ratio between the quantity of garden pea peptone and the quantity of yeast extract is 1 when the protein nitrogen concentration of the medium is > 0.8 g/1.
The present invention will be understood more clearly in tiie light of the following examples which serve to illustrate the invaition without as a result limiting the content thereof
Figure 1 represents the PRP production levels (in mg/1) in a of culture medium in

which the peptones of animal origin have been replaced by garden pea peptone as a function of the concentrations of hanin (—" ♦ - ) or of protoporphyrin DC of animal origin ( M ) or synthetic (- A -) origin (in jig/l).
Figure 2 rq)resents the PRP production levels (in mg/1) in a culture medium in which the peptones of animal origin have been replaced by wheat protein as a function of tiie concentrations of hemin ( —■ ♦ "" ) or of protoporphyrin DC of animal ( M ) or synthetic ( - A -) origin (in g/1).
Figure 3 represents the electrophoretic profiles of various bacterial populations of Haemophilus influenzae serotype b: the band M represents the electrophoretic profile of a heterogeneous stock population with two bands of 18 kb and 45 kb, band F the electrophoretic profile of its daughter population (F) after selection on selective solid medium with a single band of 45 kb; the bands B represait the profiles of various white bacterial colonies (B) with a single band of 45 kb and &e bands G, the various profiles of various gray bacterial colonies (G) with a singje band of 18 kb.
The band MW shows the position of the molecular weight markers (kb).
Example 1: Influence of protoporphyrin DC on the productioii of PRP in a plant peptone-based culture medium
1) Methodology
The production of PRP obtained after 16 hours of bacterial culture in liquid plant peptones-based culture media in which the heme source is either hemin or protoporphyrin DC of animal origin (porcine protoporphyrin) in disodium salt form, or protoporphyrin DC of purely synthetic origin in disodium salt form was compared. The hemin and protoporphyrin DC concentration ranges in the culture media vary from about 0.010 g/1 to about 2 g/1 so as to obtain PRP titration curves as a fimction of the heme source tested and as a function of the plant peptone tested. An enzymatic hydrolysate of garden pea peptone provided by Kerry (Hy pea 7404) and a wheat peptone provided by Organotechnie (19559) were tested at a concentration in the culture medium equivaloit to 0.87 g/1 of protein nitrogen. With reference to tiie current conditions for production of PRP, the production of PRP was also measured in a medium which contains a peptone of animal origin, such as the casein hydrolysate (HAC) supplied by Solabia at a

concentration equivalent to 0.87 g/1 of protein nitrogen in the presence of increasing concentrations of hanin (cf table 3).
1.1) Preparation of the media
1.1.1. Stock solution of P-NAD (Fluka) at 1 g/1 in water which has been ultrafiltCTed and then sterilized by filtration through 0.22 jim.
1.1.2. Stock solution of hemin (Sigma) at 0.25 g/1 in water which has been ultrafiltered, comprising 5 ml of 25% aqueous ammonia (Cooper) to aid dissolution. The stock solution is sterilized by filtration through 0.22 m.
1.1.3. Stock solution of porcine protoporphyrin DC (Sigma) at 0.25 g/1 in ultrafiltered water comprising 5 ml of 25% aqueous ammonia (Cooper) in order to facilitate the dissolution. The stock solution is sterilized by filtration through 0.22 m.
1.1.4. Stock solution of synthetic protoporphyrin DC at 0.25 g/1 in ultrafiltered water comprising 5 ml of 25% aqueous anamonia (Cooper) in order to facilitate the dissolution. The stock solution was also heated to 80°C in a water bath with stirring in order to complete the dissolution before being sterilized by filtration through 0.22 m. The protoporphyrin EX, in disodium salt form, was synthesized according to the mettiod described in the French patent application registered undw the registration No. 07/02334 and filed on 30/03/2007 using the steps which are desoibed in scheme 2.
The stock solutions of hemin and protppoiphyrin DC were checked for their respective contents of active compounds after sterilizing filtration. The protoporphyrin IX and hemin contents were detamined by Reversed Phase Hi Performance Liquid Chromatography (RP-HPLC). The chromatographic chain comprises a twin head pump module allowing the formation of a binary gradient, a programmable automatic injector, a diode array UV detector and a chromatographic column of the type Synergi 4\m, Polar RP-80A (150x4.6) mm, ref 00F-4336-E0,

Phenomenex.
The stock solution of hemin (Sigma) to be checked after filtration is diluted 1/3 in distilled water. The stodc solutions of porcine protoporphyrin IX (Sigma) and of synthetic protoporphyrin DC to. be checked after filtration are diluted 1/4 in distilled water. In parallel, a calibration series is prepared in ammoniated water ranging fi"om 0.025 g/1 to 0.125 g/I of protoporphyrin DC fi:om porcine protoporphyrin DC fi-om Sigma (ref: 25838-5) and a calibration series ranging fixjm 0.050 g/1 to 0.150 g/1 of hemin fim hemin fiom Sigma (ref: H5533-256). The samples to be checked and the various solutions of the calibration series are injected in a volume of 20 1 (for the hemin solutions and samples) and of 5 1 (for the protoporphyrin DC solutions and samples). The initial mobile phase consisting of a volume for volume mixture of acetonitrile and 10 mM KH2P04 pH2.5 is set at a flow rate of 1 ml/min. A discontinuous gradient is then produced fiiom this mobile phase in order to separate the molecules of interest which are detected at a wavelength of 400 mn. After having established the calibration series and on the basis of the surface area of the peaks for die samples to be checked, the hemin and protoporphyrin DC concentrations in the various filtered stock solutions are deduced therefi-om which were 0.295 g/1 for tiie hanin solution, 0.187 g/1 for the porcine protoporphyrin DC solution and 0.249 g/1 for the synthetic protoporphyrin DC solution, respectively. These concentrations were not subsequently adjusted to the target concentration of 0.250 g/1 but these concentrations really present in the stock solutioi of hemin and protoporphyrin DC were used in the analysis of the results. ;
1.1.5. Stock solution of autolytic yeast extract (Solabia) at 125 g/1 in water which has been ultrafiltered and then sterilized through 0.22 nm.
1.1.6. Stock solution of glucose at 465,12 g/1 in water which has beai ultrafiltered and then sterilized by filtration through 0.22 jim.
1.1.7. Enrichment solution
It consists of 40 ml of stock solution of yeast extract, 43 ml of

stock solution of glucose and 5 ml of stock solution of P-NAD. 1,1.8, Basal medium
- Wheat plant peptone (Organotechnie-Ref 19559) or garden pea plant peptone (Kerry-Ref Hypea 7404) in a sufficient quantity to provide the equivalait of 0.95 g of protein nitrogen per liter of basal medium, the quantity of protein nitrogen being assayed according to the kjedhal method,
- 50% sodium lactate in 50% aqueous solution (VWR): 1.8 ml,
- disodium hydrogen phosphate- I2H2O (Budenheim): 31.14 g,
- sodium dihydrogen phosphate-2H20 (Merck): 2.03 g,
- L-cystine (Jera France): 0.07 g,
- 37% HCl (VWR): 0.07 ml,
- L-tryptophan (Jera France): 0.02 g,
- ammonixun sulfate: 1 g,
- magnesium sulfate-7H20; 0.4 g,
- calcium chloride-2H20: 0.02 g/1,
- ultrafiltered water: sufficient quantity for 1 liter.
The basal medium is finally sterilized using an autoclave at 121 °C for 30 minutes.
1.2) Operating protocol
The culture was paformed in 500 ml Erlenmeyer flasks. Into each Erlenmeyer flask, there were introduced 100 ml of basal medium containing either wheat peptone, or garden pea peptone, 8,8 ml of enrichment medium and a variable volume of stock solution of hemin, of porcine protoporphyrin DC or of synthetic protoporpyrin DC such that the theoretical concentrations of hemin (or of protoporphyrin DC) tested in the various culture media is between 10 pg/l and about 2000 iig/l (see tables I and IT). Each Erlenmeyer flask is inoculated with the contents of a frozen product of Haemophilus influenzae serotype b which contains 10 to 10 bacteria/ml at an inoculation rate of 0,2% (VAO- After 16 hours of incubation with stirring at 175 rpm in an incubator at 37°C, the OD of the bacterial suspension obtained and the PRP concentration are measured in each Erlenmeyer flask by collecting a small amount of culture supernatant.

1.3) Assay of PRP
The PRP productivity was determined on the basis of a sandwich-type ELISA assay in duplicate on the culture suponatants.
The ELISA microplates are sensitized ovranight at +4°C by introducing into each well 100 jil of a solution of immunosera from rabbits hyperinununized wilh Haemophilus influenzae type b microbes, which solution is diluted beforehand in a 0.2M carbonate buffer pH 9.6 (dilution «l/2000). After rinsing and saturation of the ELISA microplates, a calibration series is produced in each microplate from a purified solution of PRP at 1 mg/ml in distilled water by produdng successive dilutions in a dilution buffer (PBS/0.05% Tween 20/1% bovine serum albumin). The culture supematants to be assayed are also introduced by also carrying out successive dilutions in the dilution buffer. After another incubation of about 2 hours at 31°C, followed by a phase for rinsing of the microplates, there are introduced into the mioowells 100 .L of a solution of biotinylated rabbit antibodies obtained by treating the sera of rabbits vaccinated with the Haemophilus influenzae type b vaccine conjugated witii the tetanus protein with a biotinylation agent, diluted beforehand in the dilution buffer (dilution «l/500). After incubation for 1 hour at 37''C followed by a rinsing step, fliere are added to each of tiie microwells 100 p,l of a solution of streptavidin coiq)led witii peroxidase (Southern Biotechnology- ref 7100-05) diluted beforehand in the dilution buffer (dilution «1/5000). After incubation for 1 hour at 37°C followed by a rinsing step, there are added to each microwell 100 1 of a visualization solution (solution of ortbo-phoiylenediamine at 0.4 mg/ml in 0.05M phosphate-citrate buffer, pH = 5, supplemmted with 0.3 1 of hydrogen peroxide at 0.03%). After a visualization time of 20 minutes protected from Ught, the reaction is blocked by adding 50 |il/well of 2N H2S04. The microplates are read at 492 and 620 mn (in order to take into account the absorption of the plastic). From the optical density values obtained on the samples tested, the PRP content in the various culture supematants is determined by interpolation by means of flie calibration series.
a
1.4) Results
The results are represented in tables 1 and 2 below and in figures 1 and 2.

Table 2:
Wheat + pordne protoporphyrin IX
Porcine 1 'H ——£-s.
protoporphyrin *" J*" MeanPRPin Standard
IXinng/l J mg/IAssay deviationPRP
Assay 1 2 *
7.48 25.5 ~ 28.3 ~ 26.9 1.98
37.4 136 149 1425 9.19
74.8 429 299 364 91.92
187 452 408 430 31.11
374 397 374 385.5 16.26
748 399 377 388 15.56
1122 446 329 387.5 82.73
1496 464 283 373.5 127.9
Wheat + syntiietic protoporphyrin IX
protoporphyrin j" J*" MeanPRPin Standard
IXinjig/I mg/1 Assay deviationPRP
9.96 33.5 " 27.9 30.7 " 3.96
49.8 317 319 318 1.41
99.6 389 437 413 33.94
249 345 350 34X5 3.54
498 380 409 394.5 20.51
996 401 404 402.5 2.12
1494 358 350 4 5.66
1992 378 430 404 36.77
Wheat+ hemin
Hemining/l j ?*" MeanPRPin Standard
mg/l mg/lAssay deviationPRP
Assay 12
11.8 " 4.98 ~ 5.61 S3 0.45 ~
59 13 nj_ 13 J5 0.49
118 17J 175 1760 0.14
295 353 302 32.75 3.61
590 7L4 73.4 72.4 1.41
1180 327 256 291.5 50.2
1770 487 446 466.5 28.99
2360 487 J04 495.5 12.02


Figure 1 reproduces the results of table I while showing the PRP production curves obtained as a function of the conctration and source of heme used in a medium which contains garden pea peptones. The PRP production curves are equivalent depending on whether a synthetic protoporphyrin DC or a protoporphyrin K of animal origin is used as source of heme. On the other hand, the production of PRP is substantially lower when the medium contains hemin compared with a medium which contains protoporphyrin DC, this being in the entire concentration range tested. Only about 200 |ig/l of protoporphyrin DC have to be used in order to have an optimum PRP production (« 480 mg/ml) while about 2500 jig/l of hemin are required in order to have a i maximum production of PRP. About 12.5 times less im>toporphyrin DC than hanin is tharefore required in a culture medium based on gard«i pea peptone in order to have a maximum production of PRP.
Figure 2 reproduces the results of table 11 while showing the PRP production curves obtained as a function of the concentration and source of heme used in a medium which contains wheat peptones. The PRP production curves are equivalent depending on whether a synthetic protoporphyrin DC or a protoporphyrin DC of animal origin is used as source of heme. On the other hand, the production of PRP is lower when the medium contains hemin compared with a medium which contains protoporphyrin DC at equivalent concentrations, and the lowar the hemin concentration the more clearly this appears. Only about 100 ng/1 of protoporphyrin DC have to be used in order to have an optimum PRP production (« 400 mg/ml) while about 1500 \ig/l of hemin are reqxiired in

order to have an equivalent production of PRP. About 15 times less protoporphyrin DC than hemin is therefore required in a culture medium based on wheat peptone in order to have a maximum production of PRP.
The results of tables I, II and III also show that it is necessary for the hemin concentration in a medium based on ammal peptone and hemin, which is in fact the medium composition recommended for producing PRP, to be 2 to 5 times higher than the concentration of protoporphyrin DC necessary in a medium based on plant peptone such as wheat or garden pea peptone in order to obtain the same concentration of PRP in the culture supernatant. For example, to obtain a concentration of about 400 mg/1 of PRP in a medium based on a hydrolysate of casein and honin, the hemin concentration should be at least 500 ig/1, whereas a protoporphyrin DC concentration of about 100 ng/1 is sufiGcient in a medium based on wheat or garden pea peptone and protoporphyrin DC. The media based on plant peptone and protoporphyrin DC thCTefore appear more advantageous than the current media containing an animal peptone and hemin serving for the production of PRP.
Example 2: Influence of the composition of the solid medium on tlie production of
PRP by the colonies
A freeze-dried material of a homogeneous population of Haemophilus influenzae serotype b bacteria containing about 10 microbes is taken up in 1 ml of Dulbecco PBS buffer (Gibco ref 14040-083). Ten-fold serial dilutions are carried out in this buffer. 50 JAI of each of the dilutions: 10", 10"* and 10"', are collected and inoculated either on Petri dishes containing various selective solid media compositions according to the invention, or on a Petri dish containing a standard solid medium containing a charcoal agar (Difco, Ref 289410) supplemented with 10% (v/v) defibrinated boiled horse blood (BioMrieux, Ref 55832).
After incubating overnight at 37°C in an incubator containing 10% CO2, the colonies are examined by transparency, under a 75 W lamp, Ike Petri ulation was then measured and compared in a third step. 3.1) Characteristics of the stock population 3.1.1: Analysis of the cap locus 3.1.1.1: Reagents Bacterial lysis buffers
-Pett IV buffer: 10 mM Tris-HCl pH 7.4, IM NaCl
-IX Ivsis solution: 6mM Tris-HCL pH 7.4, 1 M NaCI, lOmM EpTA,
0.5% Brij 58,0.2% sarkosyl, 5mg/ml lysozyme, Ip-g/ml Rnase , ..
-ESP solution: lOmM Tris-HCL pH 7.4, 1 mM EDTA. 1% SDS,
1 mg/ml proteinase
-TE solution: 10 mM Tris-HCl pH 7.4,0.1 mM EDTA Enzymatic digestion
-Small (GIBCO -BRL Ref: 15228-018)
IPX digestion buffer 4: (GIBCO BRL, supplied wife fee enzyme) - to be
diluted 10 fold in sterile purified wat«: free of nuclease at fee time of use
-.Ken/: (INVTTROGEN Ref: 155232-036)

IPX digestion buffer 4: (INVITROGEN Ref: 155232-036) - to be diluted 10 fold in sterile purified water firee of nuclease at the time of use Pulsed field electrophoresis buffer
lOX TBE buffer: 890 mM Tris-HCL pH 7.4, 890 mM boric acid, 250 mM EDTA pH 8.0 - to be diluted 20 fold in ultrafiltered water at the time of use PvuII Probe labeled with digoxigenin:
The specific labeled PvuII probe was obtained firom a DNA preparation obtained firom the plasmid pBR322*'pU038 (Department of Pediatrics -University of Oxford - John Radclifife Hospital). 20 |ig of plasmid DNA were digested for 2 hours at 37*C in the presence of 40 units of enzyme pndl (NEBIOLABS Ref. #R0151-S; in a lOX buffer 4 (NEBIOLABS Ref #B7002-S) diluted 10 fold beforehand in sterile water fi-ee of nuclease. The digestion product was then subjected to electrophoresis on agarose gel at 1% weight/volume in the presence of a IX TAB buffer to which 0.25% volume/volume of bn>moph«iol blue, 0.25% volume/volimie of xylene cyanol FF, and 30% volume/volume of glycerol have been added. The 2.1 kb band of intest corresponding to the Pvun DNA firagment is collected at the end of the migration. The DNA is then extracted from the agarose gel by passing over a "Nucleospin" column (Macherey-Nalgel Ref: 740590.250) and then its integrity is checked by spectrophotometric reading at 260 nm. Finally, the PvuII probe is labeled with digoxigenin using the labeling kit "DIG-Chem-Link Labeling and Detection Set" (ROCHE Ref: 1836463). The labeled probe is stored at-20°C. 3.1.1.2: Operating protocol
An ampoule of the stock population is thawed and inoculated onto a Petri dish containing a standard solid medium consisting of charcoal agar (Difco, ref 289410) supplemented with 10% (v/v) defibrinated boiled horse blood (BioM6rieux, ref 55832). After incubating for 18 h at 37° in an incubator containing 10% CO2, the colonies obtained are harvested and suspended in a Pett IV buffer so that tiie OD 680 nm is » 1.8. The bacterial suspension is mixed volume for volume with low-melting point agarose at

2% (v/v) (Ref: BioRad, ref 162-0138), tempered at 50*C and then this mixture is
distributed in plug molds (BioRad Ref: 170-3713) in an amoimt of» 80 jil/plug. Agarose
molds containing iht whole microbe are thus obtained. Eadi plug is placed in 1 ml of IX
lysis solution. After incubating for 6 h at 37*C, this solution is replaced with 1 ml of an
ESP solution. After another incubation ovemigjit at 50*'C, ch plug is washed 3 times
with 4 ml of a TE solution for 30 min. The genomic DNA of flie lysed bacteria whidi is
contained in each plug is thm digested ovemit at 25®C with the aid of 300 jil of a IX
digestion buffer 4 (GIBCO BRL) containing 20 units of enzyme Smal (GIBCO -BRL
Ref: 15228-018) and then washed with 4 ml of a TE solution. The digestion is continued
for 7 hours at 30°C with the aid of 200 jil of a IX buffer 4 (INVITROGEN Ref: 155232-
036) containing 20 units of the enzyme nl (INVITRCX3EN lef: 155232-036) followed
by washing in a TE solution. These two restriction enzyme release the cap locus of the
bacterial genomic DNA. The digested plugs are inserted into a certified agarose gel at
0.8% v/v (BIORAD ref: 162-0138) and thm subjected to pulsed field electrophoresis
carried out in a 0.5X TBE buffo- for 13 hours, using mamtaa of the "Chef mqyper"
type (Biorad) set so that 6 volt/cm, an angle of 120°, a linear {sogrossion, an initial
switch time of 0.9 s and a final switdi time of 11.54 seconds are applied. The gel is
transfored to a positively diarged nylon filter (Rodie R 1209272) by semidry transfer
wi& the aid of tiie ajaratus "Vacugene XL Vacuion blotting System** (Pharmacia)
according to the manufacturer's recommendations. The DNA traosfiared onto the n)don
filter is fixed with UV for 3 min at 312 nm. The filter is fliea pnybiidized for 2 hours at
42°C in '*DIG easy hyb*' buffer (Roche ref: 1585738), and then hybridized ovemigjit at
42*C in "DIG easy hyb** buffer containing 20-50 ng of a specific PvuII probe labeled
with digoxigenin/ml of buffer. This probe specifically rwognizes the Haemophilus
influenzae serotype b cap locus. The filter is then washed twice witii the aid of a low
"stringency" buffer at 65*C followed by washing in a higji ''stringency" buffer. TTie filter
is then visualized with the aid of a lurdnescait sulstrate (CDP -star: Rodi
Ref: 2041677) after having added a solution of alkaline phoq>hatase-labeled
antidigoxigenin antibodies using the kit "Dig-Chem-link labeling and detection Set"
(Roche). The electrophoretic profile obtained is represraited in figure 3. Two bands of
18 kb and 45 kb are observed, which indicates that tiie stracture of the cap loois of the
stock population is heterogaieous. A portion of the population possrases a cap locus

which contains two copies of the 18 kb gene, corresponding to the electrophoretic band of 45 kb, while the other portion possesses a cap locus in a nonduplicated form, corresponding to the electrophoretic band of 18 kb. Consequently, the stock population is a mixture of capsulated and noncapsulated bacteria. This heterogeneity is moreover confirmed using the test for determining the percentage of white colonies obtained after inoculation of the stock population onto a selective solid medium (see example 2).
3.2) Culture of the stock population on selective solid medium: determination of the percentage of bacteria forming white colonies on selective solid medium and deriving of a daughter population essentially consisting of capsulated bacteria.
3.2.1) Detamination of the percentage of bacteria forming white colonies.
The step of culturing on a selective soUd medium and the morphological analysis of the (iolonies obtained are carried out according to the same operating conditions described in example 2. The composition of the selective solid medium corresponds to that of the selective medium A of example 2.
The number of white colonies per 100 colonies visualized is determined. 60% of the colonies are in the form of white colonies, which indeed confirms that the initial stock population is heterogeneous and contains a mixture of capsulated and noncapsulated bacteria.
3.2.2) Selection and characterization of the dauter population
3.2.2.1) Selection of the daughter population
A white colony which was obtained after 18 to 24 hours of culture on the selective solid medium A is inoculated into a tube containing 2 ml of a composition of liquid medium identical to the selective solid medium without Bacto agar. After another incubation of 20 hours at "iTC, with shaking, the contents of the tube are transferred into an Brlenmeyer flask containing 50 ml of a liquid mediimi accordmg to the invention whose composition per liter is as follows:
P-NAD:5mg
protoporphyrin DC: 1 mg
glucose: 20 g
yeast extract: 5 g
garden pea peptone (Hypea 7404 (Quest)):7.42 g

sodium lactate in 60% aqueous solution: 1.49 ml cystine: 0.07 g tryptophan: 0.02 g Na2HP04-12H20:31.14 g NaH2P04-2H20:2.03 g (NH4)2S04:1 g MgS04-7H20:0.4 g CaCl2-2H20: 0.02 g The Erlenmeyer flask is placed in an incubator at 37°C, with shaking. When the OD at 600 nm is close to 2, a volume of glycerol is added such that its final concentration in the bacterial suspension is 20% (v/v). The bacterial suspension is distributed into Nunc tubes in 1 ml before being frozen at -70°C. A daughter bacterial population was thus derived in the form of frozen material, produced from a white colony obtained on a composition of selective solid medium and which is derived from a bacterium of the stock population.
3.2.2.2) Characterization of the daughter bacterial population The analysis of the cap locus of the daughter population was carried out according to the protocol described in paragraph 3.1.1.2. The electrophoretic profile shows a single band of 45 kb, which indicates that the cap locus of iht daughter population is essentially in a duplicated form of the 18kb gene (cf figure 3). Consequenfly, the daughter bacterial population essentially consists of capsulated bacteria. The homogeneity of this population is confirmed by the fact that it also produces 100% of white colonies when it is inoculated onto a selective solid medium.
3.3) Comparison of tiie production of PRP by the bacteria of the stock population and of the daughter population
The contents of an ampoule containing « lO'* bacteria obtained either from the stock population, or from the daughter population, are directly inoculated into an Erlenmeyer flask containing 200 ml of a liquid medium whose composition is that indicated in paragraph 3.2.2.1.
After incubating for 24 h at 37°C +/- PC, with shaking (175 rpm), the culture supernatant is collected, and then the PRP concentration is detCTmined by ELISA

according to the method described in example 1. The same trial was repeated 3 times. The results are presented in the table below. The values indicated represent the mean value for three trials.

Trial 1 Trial 2 Trial 3
Stock frozen material Heterogeneous population 145* 185 116
Daughter frozen material
Homogeneous population of capsulated bacteria 402 447 429
*: results expressed in mg/1
Conclusion: the production of PRP by the daughter population consisting of a homogeneous population of capsulated bacteria is improved « by a factor of 3. Consequently, the mettiod which consists in using a step culture of on a selective solid medium which makes it possible to select white colonies containing 100% of capsulated bacteria improves the PRP yields obtained. This me&od may also be used to constitute a population of completely capsulated bacteria from an initial population which contains a mixture of capsulated and noncapsulated bacteria.
Example 4: Role of the stabilizing culture medium on the bacterial population and
on the production of PRP
The starting bacterial population consists of a population of completdy capsulated bacteria whose cap locus contains at least two copies of the 18 kb gene and which produces 100% of white colonies on a selective solid medium. 4 .1) Operating protocol
The contents of a frozen material containing per ml from 10* to 10'° bacteria obtained from the daughter population selected according to the operating protocol of paragraph 3.2.2.1 are inoculated into a 1 liter fermenter containing 500 ml of a liquid medium whose composition is that indicated in paragraph 3.2.2.1. After incubating for 14 hours at 37°C, with shaking, a volume of the first culture is transferred into a second

1 liter fennenter containing 500 ml of the same liquid medium so as to have an initial OD equal to 0.3. After another incubation of ~ 5 hours under the same conditions (OD value obtained in the region of 4), a volume of the second culture is transferred into a third 1 liter fennenter containing 500 ml of medium so as to have an initial OD equal to 0.3, and then, after incubating for » 3 hours (OD value obtained in the region of 4), the volume is poured into a fourth 1 liter fermenter containing 500 ml of the same liquid medium. This operating protocol is an adaptation to laboratory scale of the steps which are normally carried out for the industrial production of PRP in a fennenter of 13 000 liters.
The numb of bacterial generations is calculated at the end of each culture using the conventional formula N= Log X/XO x 1/Log 2 in whidi X represents the biomass at the end of the culture and XO the biomass at the start of the culture. The number of cumulative bacterial generations obtained at the end of the fourth culture in a 1 liter fermenter in fact corresponds to the number of bacterial generations obtained at the end of the culture in a 13 000 liter fermenter. At the end of each culture, the cap locus was characterized and the percentage of bacteria which form white colonies on selective solid medium was determined according to the methods described in example 3. The results obtained are grouped together in the table below.

The number of cumulative genarations at the aid of flie fourth culture is 24.09 generations.
The successive cultures do not modify the characteristics of the bacterial population, which remains completely capsulated during the successive cultures. The production of

PRP also remains stable throughout the culture at a very high level. The composition of this medium therefore exercises a stabilizing role on the capsulated bacterial population since the characterstics of the bacterial population do not change appreciably during the culture.

Claims
1. A medium for the culture of Haemophilus influenzae serotype b, characterized in that the source of protein nitrogen is of nonanimal origin and comprises at least one plant peptone and wherein the heme source consists of protoporphyrin EX.
2. The medium as claimed in claim 1, in which the protoporphyrin DC concentration is at least 0.01 mg/1.
3. The medium as claimed in claim 2, in which the protoporphyrin IX concentration is between 0.1 mg/1 and 5 mg/1.
4. The medium as claimed in one of claims 1 to 3, in which the plant peptone is a wheat peptone.
5. The medium as claimed in one of claims 1 to 4, in which the plant peptone is a garden pea peptone.
6. The medium as claimed in one of claims 1 to 5, in which the total plant peptone concentration in the culture medium is equivalent to a protein nitrogen concentration of between 0.08 g/1 and 2.25 g/1.
7. The medium as claimed in one of claims 1 to 6, free of any contaminant of animal origin.
8. The liquid culture medium as claimed in one of claims S to 7, which comprises:

- from 0.1 mg/1 to 5 mg/1 of protoporphyrin DC,
- from 2 to 50 mg/1 of p-NAD,
- from 2 to 20 gl of glucose,
- from2to5 g/lofayeastexfract,
- a garden pea peptone equivalent to a protein nitrogen concentration of

between 0.4 g/1 and 1.5 g/ and - a cocktail of inorganic ions comprising Na*, NH4'*", Ca", Mg"*, HPO4"", H2PO4", SO4" and Cr ions, in the form of salt solutions such that the pH of the medium is between 6.5 and 7.5, preferably between 7.0 and 7.5.
9. A method for producing polyribosyl ribitol phosphate (PRP) in which:
(i) Haemophilus influenzae serotype b is cultured in a liquid culture
medium as claimed in one of claims 1 to 8, (ii) the culture supernatant obtained in (i) is collected, and (iii) the PRP is extracted from the culture supernatant.
10. A method for producing polyribosyl ribitol phosphate (PRP) in which:
(i) Haemophilus influenzae serotype b is cultured on a solid culture
medium, (ii) one or more colonies obtained in (i) are transferred into and cultured in
a liquid culture medium as claimed in one of claims 1 to 8, (iii) the culture supernatant obtained in (ii) is collected, and (iv) the PRP is extracted from the culture supernatant.
11. The method as claimed in claim 10, in which the source of protein nitrogen of the solid culture medium is of nonanimal origin and comprises at least one plant peptone.
12. The method as claimed in claim 11, in which the plant peptone is a garden pea peptone.
13. The method as claimed in claim 11 or 12, in which the heme source consists of protoporphyrin IX.
14. The method as claimed in one of claims 10 to 13, in which the solid medium comprises:
- atleastlmg/lofp-NAD,

- at least 0.5 mg/1 of protoporphyrin DC,
- at least one plant peptone and a yeast extract in a sufficient quantity for the protein nitrogen concoitration in the solid medium to be at least 0.2 g/1 and in a proportion such that the ratio between the quantity of plant protein and the quantity of yeast extract in the medium is 0.1 to 9 when the concentration of protein nitrogen of the medium is 0.2 g/1 to 0.8 g1 and is 1 to 9 when the concentration of protein nitrogoi of the medium is > 0.8 g/1,
- a carbohydrate,
- a detoxifying agent, and
- a cocktail of inorganic ions comprising Na", K", Ca**, Mg, Fe HP04~, H2P04', S04' ■ and CI' ions in the form of salt solutipns such that the pH of the medium is between 6.5 and 7.5, preferably between 7.0 and 7.5.
15. The method as claimed in claim 14, in which the solid medium comprises:
- from 5 to 50 mg/1 of p-NAD,
- from 0.5 to 5 mg/l of protoporphyrin DC,
- from 1 to 10 g/1 of ucose,
- fix)m 1 to 10 mg/1 of Tween 80,
- from 3 to 4 g/1 of K2HPO4,
- from 0.9 to 3 g/1 of KH2PO4,
- from 0.5 to 2 g/ml of K2SO4,
- from 20 to 500 mg/I of MgCb,
- from 2 to 50 mg/1 of CaCl2-2H20,
- from 1 to 5 mg/1 of FeCl3-6H20,
- from4to8g/lofNaCl,
- from 4 to 8 g/1 of a yeast extract, and
- from 4 to 8 g/1 of a garden pea peptone such that the ratio between the quantity of garden pea peptone and the quantity of yeast extract is 1 when the protein nitrogen concentration of the medium is > 0.8 gA.

16. The method as claimed in claim 14 or 15, in which only white colonies are transferred intq the liquid culture medium.
17. The method as claimed in one of claims 10 to 16, in which the solid culture medium and the liquid culture medium are free of any contaminant of animal origin.
18. A method for producing a population of completely capsulated Haemophilus influenzae serotype b bacteria in which;
(i) Haemophilus influenzae serotype b is cultured on a solid medium comprising:
- from5to50mg/lofp-NAD,
- from 0.5 to 5 mg/1 of protoporphyrin EX,
- from 1 to 10 g/1 of glucose,
- from 1 to 10 mg/1 of Tween 80,
- from3to4g/lofK2HP04,
- from 0.9 to 3 g/1 of KH2PO4,
- from 0.5 to 2 g/1 of K2SO4,
- from 20 to 500 mg/1 of MgCla,
- from 2 to 50 mg/1 of CaCl2'2H20,
- from 1 to 5 mg/1 of FeCl3,-6H20,
- from 4 to 8 g/1 of NaCl,
- from 4 to 8 g/1 of a yeast extract, and
- from 4 to 8 g/1 of a garden pea peptone such that the ratio betweoi the quantity of garden pea peptone and the quantity of yeast extract is 1 when the protein nitrogen concaitration of the medium is > 0.8 g/1;
(ii) one or more white colonies obtained in (i) are transferred into and cultured in a liquid culture medium comprising:
- from 0.1 to 5 mg/1 of protoporphyrin DC,
- from 2 to 50 mg/1 of p-NAD,
- from 2 to 20 g/1 of glucose,
- .from 2 to 5 g/1 of a yeast extract,
- a garden pea peptone equivalent to a protein nitrogen concentration of

0.4g/ltol.5g/l, and
- a cocktail of inorganic ions comprising Na*, NH4*, Ca**, Mg"*, HP04~,
H2PO4', S04~ and CI' ions in the form of salt solutions such that the pH of the medium is between 6.5 and 7.5, preferably between 7.0 and 7.5; and (iii) the bacterial population obtained in (ii) is frozen or freeze-dried.
19. The method as claimed in claim 18, in which all the stages are carried out using media that are free of any contaminant of animal origin.
20. The use of a population obtained according to the mdhod of claim 18 or 19, for the production of PRP.
21. A vaccine against Haemophilus influenzae type b meningitis, comprising PRP obtained according to the method of one of claims 10 to 19.
22. A solid culture medium for Haemophilus influenzae serotype b, the source of protein nitrogen for which is of nonanimal origin and which comprises:

- at least 1 mg/1 of p-NAD,
- at least 0.5 mg/1 of protoporphyrin DC,
- a plant peptone and a yeast extract in a sufficient quantity for the protein nitrogen concentration in the medium to be at least 0.2 g/1 of protein nitrogen and in a proportion such that the ratio between the quantity of plant peptone and the quantity of yeast extract in the medium is 0.1 to 9 when the protein nitrograi concentration of the medium is 0.2 g/1 to 0.8 g/1 and is 1 to 9 when the protein nitrogen concentration of the medium is > 0.8 g/1,
a carbohydrate,
- a detoxifying agent, and
- a cocktail of inorganic ions comprising Na*, K*, Ca, Mg, Fe, HPO4"', H2PO4", S04~ and CI' in the form of salt solutions such that the pH of the medium is between 6.5 and 7.5, preferably between 7.0

and 7.5.
23. The solid medium as claimed in claim 22, which comprises:
- from5to50mg/lofp-NAD,
- from 0.5 to 5 mg/1 of protoporphyrin DC,
- from 1 to 10 g/1 of glucose,
- from 1 to 10 mg/1 of Tween 80,
- from 3 to 4 g/1 of K2HPO4,
- from0.9to3g/lofKH2PO4,
- from 0.5 to 2 g/1 of K2SO4,
- from 20 to 500 mg/1 of MgCl2,
- from 2 to 50 mg/1 of CaCl2-2H20,
- from 1 to 5 mg/1 of FeCl3-6H20,
- from 4 to 8 g/1 of NaCl,
- from 4 to 8 g/1 of a yeast extract, and
- from 4 to 8 g/1 of a garden pea peptone such that the ratio between the quantity of garden pea peptone and the quantity of yeast extract is 1 when the protein nifrogen concentration of the medium is > 0.8 g/1.