CA1328841C - Method for the microbiological production of non-antigenic hyaluronic acid - Google Patents

Abstract

A METHOD FOR THE MICROBIOLOGICAL PRODUCTION
OF NON-ANTIGENIC HYALURONIC ACID
ABSTRACT OF THE DISCLOSURE
The present disclosure is concerned with the production of high molecular weight hyaluronic acid suitable for medicinal administration to mammals without provoking an immune response from microbiological fermentation. The cultures may be prepared from specially developed strains of hyaluronic acid generating bacteria obtained by passaging in serologically negative host animal blood. The cultures are kept in log phase growth for an extended period by appropriate temperature, pH and glucose content adjustments. If the cultured strain is not hyaluronidase negative the hyaluronidase activity is inhibited. The hyaluronic acid is precipitated from the culture by the sequential addition of an anionic surfactant and then a cationic surfactant and extended from the precipitate with a high molarity aqueous calcium ion solution. The isolated aqueous hyaluronic acid solution may then be purified by passage through a nitrocellulose filter. Its pyrogenicity may be alleviated by treatment with a strong acid washed activated carbon.

Description

- 13~8~

Mo-2789-ClP
A METHOD FOR THE MICROBIOLOGICAL PRODUCTION
OF NON-ANTIGENIC HYALURONIC ACID
FIELD OF THE INVENTION
This disclosure is concerned generally with the preparation, purification and use of hyaluronic acid and its salts and specifically with the preparation of hyaluronic acid from a microbiological source.
BACKGROUND OF THE INVENTION
Hyaluronic acid is a naturally occurring high o molecular weight polysaccharide typically recovered as its sodium salt having an empirical formula of (C14 H20 N Na 11)n -~
where n > 1000. The general structure of hyaluronic acid is illustrated ;n Merck Index, N;nth Ed. (3rd pr;nt;ng, 1978), at page 624. It ;s well known that hyaluronic ac;d and ;ts salts, hereafter collectively referred to as HA, can be obta;ned from at least three sources: human umb;l;cal cords, rooster combs and certain bacterial cultures such as group A and C hemolyt;c streptococci. However, certain d;sadvantages are associated with the former two sources (e.g. relatively low y;elds, contaminat;on w;th chondro;tin sulfate, and labor ;ntensive ~ -processing and purification steps). -- ' g~

Mo-2789-CIP -Bl -: -Since HA is found in aqueous and vltrecus humor of eyes and the synovial fl~id of mammalian j oints, there has been considerable interest in obtaining p~ ified HA for use as a fluid replacement to co~~ect 5 pathological conditions in the eye and in joints. The preparation of HA from rooster combs and h~man ~mbilical cords and its use in eye and joint applications is described in U.S. Patent 4,141,973 to E. A. Balazs.
That patent also provides a detailed review of the 10 technical lite ature describing the isolation, characterization and uses of HA.
U.S. Patent 4,303,676, also to E. A. Balazs, describes cosmetic formulations containing sodium hy21uronate fractions in various molecular weight ranges 1~ made from rooster combs. U.S. Patent 4,328,803 to L. G.
Pape discloses the use of an ultrapure hyaluronic acid salt in eye su~gery. The HA product used was a sodium hvaluronate salt a~ailable under the registered trademark ~YLARTIL from Pharmacia, Inc. and obtained in 20 commercial quantities from rooster combs.
Because the medical applications of ~A require that the HA be in3ected into a mammalian body (e.g. as a fluid replacement), it is very important that the ~ ~
injected products be as pure as possible to avoid -2~ reactivity problems. This importance of purity is described in U.S. Patent 4,141,973 which desc-ibes an ultrapure HA prod~ct prepared from rooster combs or, alternatively, from h~man umbilical cords. In addition to purity, it appears that control of molecular weight 30 of an HA product is very important (e.g. the 4,141,g73 patent suggests an average molecular weight of at least 750,000 daltons and U.S. Patent 4,303,676 suggests h2~ing two distinct fractions of contro'led molecular weight, one low and one high). Although there is a ~o-2789-CIP

,, 132~8~1 description of a high molecular weight (1,200,000 daltons) HA preparation of very high purity (i.e. less -~
than 0.05 % protein) in a paper by swann, Arch. Opthal.
88, pp. 544 - 8 (1972), we are unaware of any description of an HA product having the following advantages: (1) derivable from a microbiological source at relatively low costs, in high yields, and with low reactivity upon injection; (2) having a desirably high and closely controlled average molecular weight: and (3) being substantially free of protein and nucleic acid impurities.
The microbiological production of hyaluronic acid is well known in the literature. A rather extensive discussion is found in "The Biosynthesis of Hyaluronate by Group A Streptococci", a 1955 doctoral thesis on file at The University of Minnesota. Japanese Published Patent Application 83-56692 teaches greatly enhancing the production of hyaluronic acid from StrePtococcus zooepidemicus and streptococcus equi cultures by adding high levels of glucose to a protein (yeast extract) containing culture and continuously aerating while shaking. Although a 1976 paper of Kjem in Acta. Pathol.
Microbial Scand. is acknowledged as teaching microbiological production of hyaluronic acid no explicit mention of a chemical defined medium (CDM) or a protein free medium is made. There is no indication that its technology could be applied to such a specialized medium. U.S. Patent No. 2,975,104 to Warren teaches a techni~ue for increasing the acceptable incubation time of streptococci in producing hyaluronic acid by the use of a particular medium which contains a hyaluronidase inhibitor. The growth of Group A ~
streptococcal strains in a chemically defined medium -~ - -(CDM) with the production of a hyaluronic acid capsule Mo-2789-CIP

;~" '' '-, ', : . , - . :, : ., - . , . ,, ,. : . . ,, , ,, ,, : -:,,,, . : : , . :, ,, , .::: . , . .:, 8 ~ ~ :

is disclosed in "Growth Characteristics of Group A
Streptococci in a New Chemically Defined Medium" by Van de Rijn and Kessler at pages 444 to 448 of Volume 27, Number 2 (February 1980) of Infection and Immunity. The 5 recovery of hyaluronic acid from a somewhat different CDM
by heat killing the culture, filtering the medium, ~-precipitating with ethanol, centrifuging the precipitate, suspending the precipitate in an aqueous sodium chloride solution, reprecipitating with cetyl pryidinium chloride, o centrifuging the reprecipitate, dissolving this reprecipitate in an aqueous sodium chloride solution and finally precipitating once again with ethanol is described in "Isolation of Hyaluronic Acid from Cultures of Streptococci in a Chemically Defined Medium" by Kjems 15 and Lebech at pages 162 to 164 of Section B, Volume 84 (1976) of Acta Path. Microbiol. Scand. This procedure has some significant disadvantages. The heat killing of the microorganisms will result in the hyaluronic acid's becoming unnecessarily contaminated with proteins, 20 nucleic acids and other internal cell components which are difficult to separate from the hyaluronic acid and can provoke immune reactions in mammals. Thus, some of the benefit of utilizing a chemically defined medium to avoid protein contamination and achieve non-antigenicity is unnecessarily lost as compared to the contamination suffered merely from the natural death of cells in a grow.ing culture. Furthermore, this CDM itself has limited utility because it will not support the growth of most streptococcal strains. The recovery of hyaluronic acid of a mean molecular weight of about 55,000 daltons from an anaerobically grown culture of Streptococuss pvoqenes inactivated with trichloroacetic acid by filtering using a 0.22 micrometer pore size, -Mo-2789-CIP

~7~ : . , ." ,~" :, i 3 2 ~

diafiltering with a filter ha~ing a nominal retention of 30,000 daltons molecular weight until the conduct-vi~y of the filt~ate is 0.5 mega-ohms (believed to mean inverse ohms per centimeter times 104), precipitating S with ethanol, resuspending in an aqueous sodium chloride solution, precipitating with CETAB (belie~ed to be hexadecyl trimethyl ammoni~m bromide), coarse filtering, resuspending in an aqueous sodium chlo~ide solution and diafiltering with a filter ha~ing a nominal molecular 10 weight retention of 30,000 daltons until the conductivity is "0.5 mega-ohms" is described in U.S.
Patent No. 4,517,295 to Bracke and Thacke-. This procedure also involves the troublesome killing of the cult~re cells. In this case, the trichloroacetic acid 15 not only kills these cells ~ith the resultant contamination problem, b~t it also is relied on for an -effective separation of the cells from the broth.
Furthermore, this broth is only semi-defined and contains casein derived proteins which may also 20 contaminate the obtained hyaluronic acid.
Non-antigenic hyaluronic acid is also discussed in the literature. A process of obtaining hyaluronic ,~
acid free from proteins, antigens and pyrogens by treating an aqueous alkaline suspension to denature the 25 protein, subjecting the suspension under appropriate condi~ions to proteolytic ferments, removing amino acids and mineral salts with ion exchangers, and acidifying to pH 3 to 4 to precipitate the remaining impurities with some of the hyaluronic acid is described in U.S. Patent 30 Number 3,396,081 to Billek. A reportedly ultrapure hyaluronic acid suitable fo~ injection Lnto the human eye or animal joints and obtained by an involved extraction procedure including a five day chloroform extraction under mixing is described in U.S. Patent No.
3~ 4,141,973 to Balazs.
Mo-2789-CIP

i3288~1 A useful procedure for the isolation of hyal~ronic acid capsules from streptococci cult~res by incubating the bacteria, ~Thich had already been isolated by cetrifugation and washed with saline, with sodi~m 5 dodecyl sulfate in a saline suspension until the capsule was released, centrifuging to ~eco~er the supernatant, filtering the supernatant through a 0.22 micrometer pore filter, precipitating the hvaluronic acid with hexadecyltrimethyl amnonium bromide, recover;ng the 10 precipitate by centrifugation, redissolving the precipitate in 2M calci~m chloride and clarifying the solution by cen~rifugation, precipitating with ethanol, redissolving ~th water followed by the addition of sodium chloride and clarifying the solution by 15 centrifugation, and repeating the ethanol precipitation five times is described in "Streptococcal Hyaluronic Acid: Proposed Mechanism of Degradation and Loss of Synthesis During Stationary Phase" by Van de Rijn at pages 1059 to 1065 of volume 156, number 3 (December 20 1983) of the Journal of Bacte-iology. The use of sodi~m dodecyl sulfate to separate another virulence factos, M-protein, from Stre~tococcus equi is discussed in U.S.
Patent No. 4,582,798 to Bsown, Bryant and Lewis. These latter two developments became publicly available after 25 many of the developments with which the present disclosure is concerned.
There is a need for an efficient and inexpensive procedure to obtain high molecular weight hyaluronic acid suitable for injection into mammals. A
30 microbiological fesmentation which gives high yields of a relatively uncontaminated high molecular weight hyaluronic acid and a recovery procedure which does not create un~ecessary purification problems and does not adversely affect the molecula~ weight of the hyaluronic 35 acid are both desirable and would meet this need.
Mo-2789-CI~

: ::` 132~

BRIEF DESCRIPTION OF FIGURES
Figures 1-4 are graphs showing HPLC determined molecular weight distributions (retention times) of HA
made from four microbiological fermentations in 5 accordance with the disclosures herein.
Figures 5~7 are graphs showing HPLC determined molecular weight distributions (retention times) of three commercially available prior art HA preparations.
Figures 8-9 are graphs comparing the efficacies lo as a joint fluid replacement of the HA products of this disclosure with controls and/or, a commercially available product.
Figure 10 is a graph showing the use of the product of this disclosure as a fluid replacement in an 15 equine joint diseased by unspecified cause.
SUMMARY OF THE INVENTION
A procedure for culturing suitable micro-organisms to give high yields of high molecular weight antigen-free hyaluronic acid by growing these organisms 20 in a chemically defined medium free of protein, keeping them in log phase growth by appropriate pH control and addition of dextrose (glucose) for an extended period, and isolating the hyaluronic acid without killing the microorganisms has been developed. The microorganisms 25 are suitable if they produce extracellular hyaluronic acid capsules and either do not produce extracellular hyaluronidase or are grown in a medium treated by heat or an appropriate chemical agent to inhibit the activity of hyaluronidase. A procedure for separating hyaluronic ~-30 acid capsules from the cells which produced them by treatment with an anionic surfactant has also been developed. Also developed is procedure of isolating hyaluronic acid from the bacterial culture which produced it by sequentail treatment with an anionic Mo-2789-CIP ~c ~;., , -. ~, . .

13288~1 s~r,actant 2nd then a cationic s~rfactant. Additionally developed is a procedure for isolating hyaluronic acid from anionic/cationic surfactant complexes by dissolving it in a hi~h molarity aqueo~s solution of a calci~m ion 5 such as calcium chloride. Another development is the reduction of the protein content of an aqueous hyaluronic acid solution by passin~ it through a nitro cellulose filter.--An additional development is the removal of pvrogens from high molecular weight 10 hyaluronic acid by treating an aaueous solution of it with strong acid washed activated carbon. A preferred procedure is to combine some or all of the above developments with dissolution in water and precipitation in a lower alcohol, such as ethanol or isopropanol, to 15 obtain high yields of a pyrogen free very pure high r molecular weight hyaluronic acid. An especially preferred procedure is to apply these developments in such a way as to obtain a high molecular weight hyaluronic acid which as an approximately l weight 20 percent aqueous solution has a protein content of less than about 1.25 mg/ml, prefera~ly 0.10 mg/ml, ànd a nucleic acid content of less than about 0.045 mgiml, ~-preferably 0.005 mg/ml, as determined by W absorbance at 280 and 257 nanometers, respectively. It is further 25 preferred to apply these developments so as to obtain hyaluronic acid which displays a single significant substantially symmetrical HPLC retention peak with ~
retention times representati~e of molecular weights F
between about l,lO0,000 and 4,000,000 daltons and with 30 at least-98Z of the distribution lying within this , single peak. It is particularly preferred to utilize the extraction and purification developments to obtain a hyaluronic acid which as an appro~imately 1 weight percent aqueous solution has an am:no acid content by Mo-2789-CIP

~328~
orthophthalaldehvde fl~iorescence (which inherently it involves the hydrolysis of any protein present back to its constit~ent amino acidic) o less than about 0.4 mg/ml and a nucleic acid content by ethidium bromide t;~
5 fluorescence of less than about 0.06 mg/ml and which has a standardized ~PLC determiined average molecular weight of at least abo~t l,100,000 daltons.
A hyaluronic acid which has a high molecular ---weight, a narrow molecular weigh~ distribution and a ~ -10 very high p~irity has also been developed. The process developments have enabled the recovery and purification of hyaluronic acid without the significant loss of -`
molecular weight and without subs~antial broadening of ~
the molecular weight distribution. A preferred ;
15 hyaluronic acid s pvrogen free, has a single ~
substantially symmetrical HPLC retention peak lying - -between retention times representative of molecular weights between about 1,100,000 and 4,000,000 daltons i with about 98~ of the distribution encompassed by this -~
20 peak and has a W absorbance determined protein content ' of less than about 1.25 mglml and a W absorbance ~'?;,, determined nucleic acid content of less than about 0.04 mg/ml, both determined on a one weight percent aqueous solution. ç
Finally, a techniaue for creating strains of hyal~ronic acid generating bactesia espPcially suitable -for the production of hyaluronic acid by passaging it x through the blood of an animal which is susceptible to the bacteria but has not developed an immune response to ~, 30 the bacteria has been developed. A preferred procedure is to passage a Streptococcus equi strain through horse blood which gives no evidence, such as a detectable antibody level, of prior exposure to this bacterium. A ~;
strain was developed in accordance with this technique ~-Mo-2789-CIP ~

. . ..
..

:, -: .,,, ., ,. , . : . . ~., : .. .. :. . : ;.. ,. . . . . : . :

, ~_ 1,, i3~8~

and is on deposit w th the American T~pe Culture , Collection (ATCC) under number 39,506. ~-DETAILED DESCRIPTION OF T~E INV~NTION
The maintenance of hyaluronic a~id generating 5 mic~oorganisms in the log or eY~ponential growth phase for extended periods is achieved by careful pH and temperatllre control and peri~dically supplementing the cult~re with additional dextrose. Neither shaking nor ~
con~inuous aeration are required. This procedure would, ;-10 of course, be f~tile unless either the particular ~~
microorganism was hyluronidase negative, i.e. it did not generate extracellular hyluronidase, or the activity of ~'r ' any such extracellular hyaluronidase had been inhibited by heat treatment or a chemical inhibitor such as those 15 taught in U.S. Patent No. 2,975,104 to Warren. It is preferred to conduct this growth in a chemically derined media such as that taught in "Growth Characteristics of -Group A Streptococci in a New Chemically Defined Medium" :
by Van De Rijn and Kessler at pages 444 to 448 of Yolume ;
20 27, N~mber 2 (February 1980) of Infection and ImmNnitY ~`
so as to avoid the presence of extraneous protein. This simplifies the s~bsequent p~rification since only the proteins released fr~m the microorganism being cult~red then need to be removed. The temperature and pH of the ~--2~ cult~re should be maintained in ~he range known to be conducive to the growth of the particular microorganism being cultured. In the case of Groups A and C -;
streptococci, which are preferred microorganisms, the tempe-ature is advantageously around 37~C and the pH
30 should be maintained in the range of about 7 and slightly above. In the case of the particularly preferred Group C streptococci and the most preferred Stre~tococcus eoui the pH is preferably maintained between about 7.0 and 7.2 and the temperature is Mo-2789-CIP

`- 13288~
1 1 ~ r prefe~ably abo~t 37~C. Culturin~ fo~ between about 24 and 120 hours ~nder ~hese condi.ions has been found to be ad~antageous. The pH control may be continuous or it may be intermittent. In the latter case the p~ may be 5 readjusted each ti~e dextrose is added. For the Group C
strep.ococci i~cluding the Streptococcus eaui it is prefe-zed to Teadi~st the pH to about 7.6 at each de~trose addition.
Sufficient dextrose should be added to the 10 microo~~anism on 2 timely enough regimen to avoid complete depletion of the dextrose from the culture medium. The mic~oo-ganism utilizes the dextrose in synthes.~ing its hyzluronic acid capsule and th~s - continuously cor.sumes dextrose from the medium. In the -15 case of the streptococci, especially the Group C and most especially the Stre~tococcus eaui addition on no more than a twenty four hour interval is preferred.
With these bacteria it is preferred to add approximately -1 weight percent o dextrose, based on the total weight 20 of the culture including medium. An especially ~-prefe_red procedure ~rth these bacteria is to add 1 weight percent dextsose every twenty four hours but to -also add an additional 0.5 weight percent of dextrose between the larger additions, preferably about s;xteen -2S hours after the 1.0 percent addition. -~
The recove-y of hyaluronic acid is enhanced if the pH cont~ol is d7scontinued s~fficiently before it is to be harvested to cause a significant pH drop. A drop i-to a value of between about 6.5 and 6.8 is preferred. A
30 desirable ~rop can be obtained by ceasing pH control about 12 hours before harvest although ceasing berween about 6 and 12 hours before harvest has been ound suitable. Such ~ pIocedure facilitates centrifugation and increases the yield of hyaluronic acid.
~ " ' ,~: ' ' Mo-2789-CIP

.;, , .

~328~1 rne ability to ma;n.ain the microorganis~ in ,~
log phase growth is surp-ising and makes it possible to continuouslv culture the microorganism. In s~ch a ;:
procedure once some of the n~trients other than the 5 rapid~ consumed dextrose become exhausted, the depleted medi~m may be separated from the cells of the microor~anism and replaced ~-ith a freshly constit~ted medium. Al,ernatively, the depleted nutrients can be added to the medium in app-opriate amounts. However, a.
10 some point, .he waste products and the hyaluronic acid not attached to the cells need to be extracted from the culture. One convenient technique is simply to filter "
the medium using a pore size which retains the cells.
Then the cells can be pro~ided with fresh medium and 15 hyaluronic acid can be ext.acted from the depleted ~`-medium.
In preparing the initial medium and, in the case of con~inuous culturing, in preparing a replacement medium it is preferred to sterilize the medium to avoid 20 the presence of undesirable microorganisms. A
convenient technique for such a sterilization is to pass the medium through a fine pore filter, such as a 0.22 micr~meter pore filter. -In initiating the culture,-it is also -25 convenient to use an inocu'a.ion seed grown on the same -CDM as the growth medium and to use between about l and 5 volume percent Oc inoculate. If the CDM described by Van De Ri~n and Kessler, Supra, is used it is also convenient to add the cysteine and bica_bonate of soda 30 to the medium just befo,e inoculation. These two :
nutrients have a tendency to fo~m salts which can precipitate out of the mediu~.
The separation of hyaluronic acid, especially in czpsular form, from the ce~ls of the microorganism Mo-2789-CIP

., , , , . ,: , ,., . . : . , , .. ~ , ..... ., - .. .

which generated it ~s conveniently a~fected by ~.
inc~bation with an anionic s~rfactanL. Among the suitable surfactants are those compo~nds which are available in sufficient puriLy to use in such a 5 fermentation bath and which carry one or more sulfate or sulphonate groups. Among the particularly preferred surfactants are sodi~m dodecyl sulfate and dioctyl sodium s~lfos~ccinate, with the former being especially preferred. The amount of sur~actant required is 10 dependent on the hyal~ronic acid content of the cultu-e and is generally at least about 0.01 weight percent based on the weight of the culture. In the case of Group C streptococci and sodi~m dodec~l sulfate at a 0.01 wt. % le~el, an incubation time at 37~C of about 15 -15 minutes has been found to be beneficial. In the case of microorganisms force grown, i.e. kept in log phase, for an extended period the increased content of hyaluronic acid requires a higher content of su-factant. ~or - --cultures containing between about 1 and 2.5 grams of 20 hyaluronic acid per 10 liters a surfactant content in the neighborhood of 0.1 weight percent is usually found suitable. A suitable range for such high yield culture is between about 0.05 and 0.25 weight percent, based on the weight of the culture including the medium. The 2~ separated hyaluronic acid can then be isolated from the culture by any suitable means including filtration to remove the larger cells and diafiltration to remove the -lower molecular weight species. A particularly preferred technique is to precipitate the hyaluronic 30 acid and anionic surfactant by the addition of a cationic surfactant. :
The isolation of hyaluronic acid from the medium in which it is grown can be effected by the sequential addition o~ an anionic surfactant followed by Mo-2789-CIP

-" 132~841 the addition of a cationic surfactant. The suitable types and amounts of the anionic surfactant are discussed hereinabove. The cationic surfactant is preferably an ammonium salt and more pr~ferably a quaternary ammonium salt. Especially preferred are quaternary ammonium salts with four aliphatic substituents particularly those in which at least one of the substituents is a long hydrocarbon chain. Among these long chain substituted aliphatic quaternary ammonium salts hexadecyl trimethyl lo ammonium bromide is particularly preferred. A preferred technique for precipitating the hyaluronic acid is to cross titrate the anionic and cationic surfactants in samples of the culture until a heavy readily separable flock is obtained. A convenient manner to accomplish this crosstitration is to prepare numerous one milliliter samples of culture at each of several levels of anionic surfactant, for example at levels between 0.05 and 0.25 weight percent in 0.05 percent steps, and then add various amounts of the cationic surfactant to the samples -for each level of anionic surfactant, for example amounts between lO and 100 microliters of a 10% solution may be added in lO microliter steps. As a double check after the appropriate amount of anionic surfactant has been added to the total culture a sample may be titrated 25 against the cationic surfactant to confirm the appropriate level to be added.
The anionic surfactant is thoroughly mixed and incubated with the culture before the addition of the cationic surfactant. In general, a mixing time of at 30 least about fifteen minutes is preferred with a mixing time of at least thirty minutes being more preferred, especially with culture volumes in the neighborhood of 2000 liters or more.

Mo-2789-CIP

.

i328841 - .5 -It has been fo~nd advantageous to use a 10 percent aq~eo~s solution of hexadecyl trimethyl ammonilm bromide to precipitate the hyaluronic acid from a ten liter culture to which at least about 0.01 weight 5 percent of sodium dodecyl sulfate had been added and mixed fo- at least fifteen minutes. It was found that between about 100 and 400 millilite-s of this solution was suitable and that at least abo~t one hour sho~ld be allowed ~or maximum floc formation.
The preclpitate formed by the sequential addition of anionic and cationic surfactants may be separated from the balance of the culture by anv common liquid solid separation technique. Among the convenient ~~
and pre e-red techniques are fil,ration and 15 centrifugation. It is preferred to cool the entire medium below the growth temperature and preferably below about 10C and to store it for an extended period, in the case of 2000 liter or larg~s cult~res for in excess OI si~teen hours, before effecting this separation. The . ~-20 most pre,erred separation technique is centrifugation.
Utilizing such a procedure resulted in a supernatant almost completely free of hyaluronic acid.
Hyaluronic acid may be separated from an anionic/cationic surfactant complex by dissolution in a 25 high molarity aqueous solution of a calcium ion. A
prefe-red solvent is a 2M solution of calcium chloride.
It is p-eferred to use a reduced volume of solvent compared to the original culture and between about 5 and 10 volume percent has been found suitable although 30 between about 10 and 40 volume-percen, is preferred.
The solution is preferably ca- ied out at temperatures between about 4 and 30C with temperatures between about 4 and lODC being preferred for 2 period of at least 6 hours w~th periods of be ween about 1 and 10 days being Mo-2789-CIP

r ,?
132g841 preferred fo~ original cultures of 2000 liters o, greater.
The dissolved hyaluronic acid can now be isolated from the s~rfactant complex and other materials 5 which precipitated with the su~factant complex ~y any typical solid/liquid separation technique including filtration or centrifugation. Centrifugation is particularly preferred.
The hyal~ronic acid can then be purified by a 10 va-iety of techniques which involve precipitation in a lower alcohol followed by resolution in water.
Particularly suitable alcohols a,e ethanol and isopropanol.
The protein and nucleic acid content of the 15 hyaluronic acid can be significantly reduced by passing an aqueous solution of it through a nitrocellulose filter. It is pseferred to conduct at least one and preferably several ethanol precipitations and subsequent water resolutions before effecting this filtration step.
20 This will reduce the protein load on the filter matrix and reduce the probability that it will become saturated with protein. The pore siz~ of the filter is not critical but is somewhat dependent on the viscosity of the solution being filtered. Pore sizes ~om 0.22 25 micrometers to 2~ micrometers m~y readily be utilized ~th pore sizes of 8 micrometers and greater being preferred for the more viscous solutions to minimize the back pressure. With a single thickness of a standard nitrocellulose filter it is preferred to utilize no more 30 than about 20 mg/cm2 of protein per unit area of filter with a ratio of no more than about 10 mg/cm2 being more preferred and a ratio of no moFe than about 2 mg/cm2 being especia~ly preferred. ~lthough nitrocellulose is known to ha~e affinity for n~cleic acids it is -o-2789-CIP

. .. .. : , -. , .. - , . .. .. - . .: , . , . - . .. . . . . . . . .

.~, , - , - . , . , . , ........... , : : ,~ :...... .. . . .

.. ~ ` - ! , . ...

~32~8~1 s~rpris ng that it is able to extract both nucleic acids and protein from an aq~eo~s solu~ion of hyaluronic acid, which itsel~ is known to have an affinity for these materials. There does not appear to be any basis for 5 predicting that the eauilib~ium between water and nitrocellulose, let alone water con~ainin~ an a~tractant for protein, would lead to substantial extraction of protein. In this reg2-d treating a 4.5 liter solu,ion by ~assage thro~gh an approximately 670 cm 10 nitrocellulose filte- ~esulted in an about 78% red~ction in protein from 0.028 ~. % to 0.006 wt. ~. Such an ~-equilibrium pa-tition;ng was totally unpredictable.
Further filte, thicknesses proportionall~
;ncrease the protein ~e~ unit area of filter which can 15 be remo~ed. For inst2nce with two thicknesses, it ls prefe-red to utilize no more than aboùt 15 mg/cm of protein per unit filte~ area.
Hyaluronic acid may be freed of pyrogens by contact as an aqueous solution with acid washed 20 activated carbon. The pyrogenic agent is evidently independent from the protein and nucleic acids which may provoke an im~une response in a m~mal. A pyrogenic response has been detected in rabbits iniected with hyaluronic acid which has a very low amino acid content 25 (b~ the orthophthalaldehyde fluorescence which hydroiyzes any proteins present and detects them as their constituent amino acids) and a very low ethidi~m bromide fluorescence detected nucleic acid content.
This ,esponse was observed using the established -~
30 protocol of injecting into the ear vein of a frequently handled ~abbit and observing any resultant increase in body te~perature. A r~se of greater than 0.6C o. a total rise of greater than 1.2C from this rabbit and two additionally injécted rabbits is an indicatior. of Yo-2789-CIP
''~ ' 13288~1 pyro~enicitv. When the same batch of hval~ronic acid is tested according to the same protocol after exposing it to a strong acid washed carbon no pyrogenic indication was observed. It is prefe~red to ~se s~fficient 5 activated carbon to give 2000 pre:Eerably 10,000 and most 1, preferably 20,000 m of s~face area per liter of aq~eous 1 weight percent sol~tion being treated. An especially preferred technique is to utilize a fairly high s~rface area ac~ivated carbon of more than abo~t 10 500 m2/g and to ~til-ze in excess of abo~t 10 grams pe-liter of one weight percen. aqueous hyal~ronic acid solution.
The acid washed active carbon treatment can be either continuous or batch. In the former case, 15 treatment may conveniently be effected by circulation th~ough a cartridge. In elther case treatment can be repeated until negat;ve pyrogenicity is achieved. In view o. the minimal Emo~nt of pyrogen present, ~ypically in the nanograms per milliliter range, the concern is 20 not saturation o the adsorbent but rather adequate , contact with the pyrogen.
The aqueous hyal~ronic acid solution obtained by the use of the calcium ion treatment of the ' anionic/cationic surfacta~t precipitant may be further 25 p~rified by repeated prec;pitations ;nto a lower alcohol, such as ethanol or isopropanol and resolutions in water. In a prefer~ed procedure the supernata~t from the calci~m ion ~reatment (preferably a 2M aqueous calcium chloride solution) is extracted with 2 volumes 30 of a suitable alcohol, preferably 9~% ethanol or 97Z
isopropanol and cen rifuged or sieve filtered after at least one hour. Then this precipitate is solubilized overnight at between abo~t 4 and 10C in deionized, distilled wate. uSing between abou~,l/lO and 1/20 of the Mo-2789-CIP

- 19 - 13288~1 original vol~e and the precipitate is removed by sieve filtration or centrifugation. This is followed ~y the addition of s~fficient sodium chloride to ~ive a one percent solution by weight, i.e. 10 grams per liter of 5 solution. This solution is then extracted with two volumes of an appropriate alcohol to reprecipitate the hyaluronic acid (actually at this point more precisely sodium hyaluronate). The precipitate is then isolated by either sieve filtration or centrif~gation. The 10 resolution in water, addition of sufficient salt (sodi~m chloride) to give 1% (10 grams per liter) and precipitation with two volumes of an appropriate alcohol are continued until the aqueo~s solution is clear using ;ncreasingly smaller volumes of water with each 1~ resolution and consequently s~aller volumes of alcohol (twice the volume of the solution being extracted).
This typically requires four additional alcohol p~ecipitations. An outline of this proced~re is as -:
follows:
OUILINE O~ PROCESS ~OR EXTRACTION -OF BACTERIAL H~ALURONIC ACID -1. Grow Streptococcus-orga~ism 2. 1 ml/l SLS 10~.
3. 10 - 40 ml/l.
Hexadecyltrimethylammo~ium bromide 10~.
4. Collect ppt.
. Solu~ilize in 2M CaC12.

6. Collect supernate.

7. 2 Vol. alcohol (ppt. HA, some nucleic acids, some protein).

8. Collect ppt.

9. Solubilize ppt. in DI-H2O.

10. Discard undissol~ed ppt.

11. Collect supernate.

Y.o-2789-CIP

- 20 - 132884~

12. ~ ~aCl Vol. alcohol (ppt. HA).

13. Collect ppt.
5 14. Solubilize in DI-H2O.
15. Discard ppt.
Collect supernate.
l~. 1% NaCl 2 Vol. alcohol (ppt. HA).
17. Collect ppt.
18. Sol~bilize in DI-H20.
19. Discard ppt.
Collect supernate.
15 20. Filter - protein binding type (e.g., nitro-cellulose) (remove some of the minimal protein remaining.
21. l~ NaCl 2 Vol. alcohol ~ppt. HA).
22. Collect ppt.
23. Solubilize 0.15M phosphate burfered saline pX 7.2.
24. ~djust to lZ HA by spectrophotometric assay. -25. Ste_ilize with 0.1Z betapropiolactone 4 - 10 C 24 - 48 hours.
26. Hydrolvze betapropiolactone 37 C 24 - 48 hours.
27. Fill syringes.
The ~Lnal steps of product pTeparation may involve 30 washi~g with g5% EtOH and 99% acetone followed by drying unde_ vacuum. The dried ~A is resuspended in 0.15M
sodiu~ phosphate buffer to a concentration of 1.0~.
This m2y be filter sterilized throu~h a final 0.45 mic~o~eters nitrocellulose type f~lter and/or sterilized 35 in fir.al bulk form with 0.1% betap opiolactone. The betapIopiolactone sterilization is conducted at 4 C for 24 - 48 hou~s followed by hydrolization of the Mo-27~9-CIP

13288~

betapropiolactone at 37~C for 24 - 48 hours. The final product contains 10 mg/ml HA in 0.15M sodium phosphate buffer. When these steps are followed, HA of the highest purity is obtained in high yield (> 99.90% HA).
A more preferred procedure to isolate sodium hyaluronate from the supernatant of the calcium ion treatment involves the use of diafiltration, and reverse precipitation with a winding device. This supernatant may be diafiltered against a membrane which allows species with molecular weights less than about 1~0,000 to pass through. The water to the feed side of the membrane may then be cut off and water flow on the pass side of the membrane continued until the volume is between about 10 and 40% of the initial volume. This typically gives a hyaluronic acid (or calcium hyaluronate) concentration of between about 0.4 and 0.7 weight percent. This is followed by three ethanol precipitations with a winder apparatus and resolutions utilizing the procedure described in copending Canadian Patent Application Serial Number 521,989 filed November 3, 1986 by Karen K. Brown et al. In particular, the -sodium chloride content of this diafiltered concentrate is adjusted to ten grams per liter and it is fed to three volumes ot 95% ethanol in which a device with vertical fingers is slowly rotating. The precipitating sodium hyaluronate attaches itself to these moving "fingers". The precise configuration and speed of rotation of this "winder" is adapted to the particular vessel used and the batch size but a cage like general configuration is preferred. The ethanol is then drained from the vessel and the precipitated hyaluronate is redissolved by the addition of water to the vessel. The winder should preferably be rotated for at least about 30 minutes after the last of the diafiltered hyaluronate solution is added and may also be rotated after the water addition to aid in Mo-2789-CIP

~/
~ .

13288~

in resolution. The third resolution may conveniently be in the final desired product formulation buffer such as an aqueous 0.15M sodium phosphate solution. In such a case it is advisable to assay the hyaluronate after the second resolution.
The fairly pure sodium hyaluronate solution may now be further purified and sterilized. In this regard, it is convenient to drop the viscosity of the solution in accordance with copending Canadian Patent Application Serial No. 521,989 filed November 3, 1986 by Karen K. Brown et al. In particular, o it may be heat treated, preferably at a temperature in excess of about 50C for in excess of about 72 hours in an open vessel, or forced through a filter with a pore size less than about one micrometer, until the 37C viscosity is less than about 250 centistokes. At this point, the solution may be conven;ently tested for pyrogenicity in rabbits and, if a positive result is obtained, appropriately treated with strong acid washed activated carbon. Finally, the now low viscosity pyrogen free solution may be filtered through a 0.45 or 0.22 micrometer pore nitrocellulose filter for both sterilization and final protein and nucleic acid reduction.
The former procedure was used to obtain sodium hyaluronate for further evaluation. The material from four different fermenters was initially evaluated after solubilization in the 2M calcium chloride. One of these lots was then further purified in accordance with this procedure and evaluations made at various stages.
The typical 10 1 fermenter of Streptococcus equi produces 5 9 to 7 g dry weight of cells and 1.0 g to 2.5 9 dry weight of HA. Yield is therefore between 14.3% and 50% (w/w).
Yields of HA from extraction of rooster combs as in U.S.

Mo-2789-CIP

~ . ,.

- 132~8~1 Patent 4,141,973 are reportedly around 0.079%. -It should be noted that a latter-stage filtration through a suitable protein-binding filter (for example a nitrocellulose filter) is necessary in order to remove reactivity of the final product HA. Other types of filters (plain cellulose and cellulose acetate) do not adequately remove reactivity as observed in the horse joint injection test. It is thought that this step removes the minute quantity of reactive proteinaceous material remaining in the HA.
The purity of this bacterial-derived HA was determined via a colorimetric protein assay, U.V.
spectrophotometry, HPLC, and slab gel electrophoresis. Initial experiments involved quantitation of protein contamination as measured via the BIO M D* Protein Assay. This method can 5 detect levels of protein as low as 200 ug/ml. Table I lists the results of testing aqueous 1.0% solutions of hyaluronic acid extracted from four different fermenters of Streptococcus equi.
TA~LE I
BIO RAD* Protein Assay Results Concentration SampleO.D. at 595 nm of Protein Ferm 10.00,0.000 < 200 ug/ml Ferm 20.00,0.010 < 200 ug/ml Ferm 30.00,0.005 < 200 ug/ml Ferm 40.00,0.005 < 200 ug/ml According to such data, the protein content of a 1.0%
bacterial-derived HA solution may be as high as 0.02%.
A second method of determining protein, peptide, and/or amino acid content is UV absorption at 280 nanometers.
A known concentration of bovine serum albumin was used as a control. Table II reports these *Trademark Mo-2789-CIP
"'' ~, . i ~ ''i .,. . , . .,, . ~ ~ .. .

~32~8~1 results. Als~ reported are the W absorptions of some f comparative sol~tions at 257 nm. Absorption at 257 nm ~epresents contamination with n~cleotides or nucleic acid such as DNA and ~'A. It is noted that '., 5 spectrophotome~ric absorption at 280 nm detects more protein contamination than the Bio RAD assay. The l.OZ
sol~tions of bacterial-derived HA contain at most O.lZX
contaminants which absorb at 280 nm. Since these ~ame !, solutions contain almost no nucleic acid contamination, 10 ~he purity is in the range of at least 99.88X. In this _espect it is notable that amino acid analysis of s;milarly extracted HA indicated the presence of ~ 0.04%
protein. This would mean that the HA purity is as high as 99.96%. This is compared with the p~-ity of 1~ commercially available rooster comb derived XA
(HYLARTIL~ available from Pharmacia or HyalovetT~
available from Trans Bussan) which, according to o~r tests, have purities in the range of 99.78Z to 99.86X
respectively.

~o-2789-CIP

:
- 25 - 1 3 2 ~
TABLE I I
PROTEIN AND NUCLEIC ACID CONTAMINATION OF 1. 0%
HYALURONIC ACID AS MEASURED BY W SPECTROPHOTOMETRY
W Absorbence Concentration Nucleic O.D. at O.D. at Protein Acid Sample 280 nm 257 nm mg/ml ua/ml Ferm 1 0.45 - - 0.66 - -Ferm 2 0.83 - - 1.22 - -10 Ferm 3 0.75 - - 1.10 - -Ferm 4 0.47 - - 0.69 - -Miles Labs (Rooster Comb) low purity1.16 1.31 1.70 48.5 15 Pharmacia HYLARTIL~ 0.91 1.63 1.33 60.3 Trans Bussan Hyalovet TM1.27 >2.0 1.86 >74<300 1/30 dil. = 0.27 Further studies on purity were conducted with the bacterial-derived HA. Effectiveness of two alcohol purification processes were followed spectrophotometrically at 280 nm, 257 nm, and 195 nm.
Absorbance at 195 nm represents the actual absorbance of HA and is linearly related to concentration of HA. Table III shows optical density results whereas Table IV
converts all readings to concentrations of protein, nucleic acid and HA. These two more purified lots yielded bacterial-derived HA which was 99.99% pure using -30 either 95% ethanol (ETOH) or 99% isopropyl alcohol -extraction. ~

' ~' ' -:. ' Mo-2789-CIP ~ -:: - ,- - .

s . ':

- 26- ~`~

:: -L e~ ~ C~.l C~l O ~D CO CO OD , ~ V~ C~> O -- _ ~ E _ :~ Ln ~,_ .~', I c- O l.D O
O r~ O 1~
L-J O -- -- -- - .
r.'"' . ^' :' '' ~J C~ L --~ CD
C~ Z Q~ O
--W ~ ~ ~'7 O O O O _:
ZoC~ L al E _ ~ ~7 ~) 1~ 1~
J C: C . IS) ,.. .
C C 1:~ N ,_ _ _:~ ~ ~ O ri' O I C~ ,~
~ ZozO ~_ O O O C~
C ~ '.~ . -:~ N C~ O '`
L ~t C~ t O
C : ~ E o O O O
O C:~ , ~
O N . -.
: ~ ~ ~

~o ; ~ ' ',-~ O 0: 0 0 ' `:~ o o o oaQ ..
C~J N t~
~: : ~ C L --C : ' Mo-278g-CIP - "~
~, :, .' .

, : . . ~.: ' : ~ : - :'.'~ ' ,'' ~

~ I L ~ O 00 ) C 11~ ~
L C
v~ _ ,-c ~ _ c~c~
z _ .
~z ,, ~c . _ D ~ O
~ Z C ~ N C0 ~D
--C~ ~ C--E ~
~ ~ .
~ D ~ _ ~ ~D O C~
~ 1~1 ZO L~.J ~
~ ~ .
:: 0~:

~ L-l C U O C:l ~ ~ ~ E ~-~ ~0 : ~ ~ ~
O _ N c~J~ N ) . ~
~ O O 0 0 ~:
'~ 0~ ' ~
'~.' '' :~, ~' Co Co tC~ _ ~: O ~5 ~ ~ ~ ON
O o c~
~ ON ON ON O
_ ~ _ = = ~ ~
t ~ ~ N ~ ~ ~
~ ~Mo-2789-CI~P

- 28 - 1 3 2 8 8 ~ 1 d Slab gel electrophoresis was used to further r analvze the various 1.0~ hyal~ronic acid preparations listed in Table II for nucleic acids. Such a technique can differentiate DNA from RNA. A O.~Z low endosmosis 5 agarose containing 2 ug/ml e~hidi~m bromide was used in conjunction with short w2ve length W light in order to visualize the nucl~ic acids after electrophoresis. DNA
being of-a m~ch larger molecular weight remains near the origin whereas RNA migrates with the buffer front.
10 Twenty-five ul samples were electrophoresed 18 hours at 90 volts in a Canalco Slab Gel apparatus. Results indicated no detec,able nucleic acids in any of our fo~r preparations or the HYLARTIL~ product. The Hyalovet product showed a significant amount of nucleic acid in 15 the form of RNA.
High Performance Liquid Chromatography (HPLC) was used to analyze the molecular weight of the various ~A preparations. This is a newer a~d more accurate method than viscometry as mentioned in the Patent No.
20 4,141,973. A Waters Micro Bondagel/E-High ~ column was used for molecular weight determination. In this column it was impossible to run aqu~ous standards in the molecular weight range required along with test samples in order to determine accurate molecular weights.
25 However, relative molecular weights based upon retention times on the column were dete mined. Theoretically, with this procedure, the earlier the time of peak detection the higher ~he molecular weight. The col~mn had a retention time of 10 min. with minimum and maximum 30 molecular weight capabilities between 15,000 2nd -7,000,000 daltons. Figures 1 - 4 show the 'HPLC tracings for our first fou~ fermenter preparations. Miles Hyaluron, Ph macia HYLARTIL, and Trans Bussan Hyalovet are shown in Figuses 5 - 7. Ihe retention tLmes of the ~o-2789-CIP

. . " ~, . ,, . . . . . . -.. . . .-.
x ~Y ~ ~ q r~ i?~ .h~P:.. q.. ,.. ~.q,~.:i~.. I r~

- ~ -13288~

peaks and shoulders were determined and relative molec~lar wei~hts calculated based on a linear relationship between molecular weight (15,000 to 7,000,000 daltons) and retention time (0 - 10 minutes).
5 Such relative molec~lar weights are listed in Table V.
As can be seen in Figures 1 - 4, sodium hyal~ronate obtained according to this disclosure gives an essentially single, substantially symme. ical high molecular weight (avg. higher than about 2,000,000 10 dalton) HPLC distribution peak. In Figures 1 - 4 at least about 9~Z of the HA molecules are within the single peaks shown. Such close control of high molecular weight distribution is not shown in existing commercial products as illus~rated in Figures 5 - 7.
15 Figure 5 (Prior Art ~1) illustrates the HPLC tracing for the Miles Hyaluron HA product. Figure ~ (Prior Art ~2) represents the Hyalovet product and Figure 7 (Prior Art ~3) represents the HYLARTIL~ product. It is though~ -that this close control of final molecular weight -ange 20 may be due to the simplicity of the extraction procedure requiring minimal shear-producing steps as well as to the lack of hyaluronidase which could degrade high molecular weight HA. -. .

Mo-2789-CIP

_ 30 _ ~ 32 88 4 TABLE V
Relative Molecular Wei~hts of Hyal~ronic Acid Moieties in Various Pre~arations Relative Molecular wei~ in Millions 5 Sample Range Avera~e Fermenter 12.2 - 3.9 2.8 Fermentes 22.5 - 4.0 3.8 Fermenter 31.7 - 2.8 2.4 Fermenter 41.1 - 3.9 2.6 :~
lO Hvaluron0.015 - 3.0 0.015 HYALOVET~ 0.010 - 3.7 .015, l.8, 3.7 HYLARTIL~< O.OlO - 3.8 1.9 As noted above the relative molecular weight range of HA moieties fo~nd in bacterial-derived HA is 15 narrow with the majority (98~) measu-ing between about 1,100,000 or 2,200,000 and 4,000,000 daltons. Via the same method, Hyalovet contains three distinct moleculaT
weight moieties of 2,700,000; 1,700,000 and 300,000 daltons. Finally, the HYLARTIL~ product contained an 20 array of EA molecular weights from ~10,000 to 3,700,000 -~
daltons. As shown, the ~Y~ RTIL~ product contains the widest variation of molecular weight sizes.
Fr~m the various analytical tests described here;n, it has been determined tha~ HA extracted from 25 bacteria ~ia a simple method is purer than three commercial products made from either rooster combs o-~mbilical co~ds. The latter are produced via a complex process which is inefficient yielding only 0.079~ ~A.
This is compared with H~ extracted from streptococci 30 which can reach yields as high as 50% w/w.
Joint Fluid Re~lacement ~
Hyaluronic acid prepared from bacteria as ~~ -described herein has been tested for reactivity in tibiotarsal and radial-caspal joints of horses. rne Mo-2789-CIP
'''^'`"` ' - . r~ i - 1 3 ~

following cl~nieal index test was devised in order to measure reactivity of HA prepara.ions post ,~
intra-articula- injection of horses.
The test protocol is as follnws:
1. Assess normal movement of joint to be iniected. Assign læmeness indices from 0 to 5 according to the following definitions.
Lameness Index 0 = No lameness 1 = S;i~ht lameness - moderate ,~
2 = Noticeable lameness - moderate 3 = Obvious lameness 4 = Severe lameness - reluctant to move o ;
bea_ weight 5 = Cannot bear weight. If down, animal is unable to rise.
2. Sedate horse (e.g. with Rompun~ sedative).
3. -Shave hair around the joint area to be in3ected. -:~
4. Determine Swelling Observation Index according to the following definitions.
Swellin~ Obser~ation Index 0 = No Swelling ~-Nothing obvious - palpable fluid 2~ 2 ~ Slightly noticeable - palpable fluid ~-3 c~Noticeable s~elling of entire joint `r':4-~ Severe swelling at injection site Severe swelling involving more than the joint alone. i~
5.~ With cloth tape measure, measure joint ;i;
ciscumference ~mmiediately anterior to the anterior --~
aspect of the third metatarsal (tibiotarsal joint) or imnediately distal to the protuberance of the accessory ca:rp~l bone that~is aro~nd the~rad~ial carpal bone 35 (c2:~al joint).
Mo-2789-CIP

~' - 32 - 1328~1 The exact circ.u~ference o~ the joint (in ~
millimeters) before and after iniection is recorded. A t.-"
difference between the circumference each day post in~ection and the original circumference is calculated. ',7 5 If the difference is greater than 1.0 cm, the exact measurement is added t~ the other twc, index values in order to determine the clinical index.
6. Remcive joint fluid tl.0 - 2.0 cc) prior to `
in,ection with a 3.0 cc syrin~e with a 20 - 22 ga. X 1" ;
10 needle.
7. Inject joint with 2.0 cc of a 1 prepa.ation of hyaluronic acid being evaluated for reactivity. ~or this injection, use a 3.0 cc syringe with the same needle ~exchanging sy_inges only) as used 15 to remove joint fluid. This is done so as to reduce trauma to the joint as m~ch as possible. .
8. Apply digital pressure to the injection site for 1 to 3 minutes after injection. This is done to pre~ent backflow of the very viscous HA. ~
9. Observations and measurements are made for .s -four consecutive days post injection, then on day 7. -~
10. The Clinical Index (CI) is calculated as ~-follows: ~ -Total Lameness Index (TLI) - Sum of Daily ~, Lameness Indices Total Swelling Indeg (TSI) = Sum of Daily -Swelling Observations . Sum of Joint Circumference Measurements Greater Than 1.O
; ~ :
cm.
CI = TLI + TSI -11. Interpretation :
Joint injection alone causes trauma with development of some swelling and lameness. -This was proven by evaluating numerous Mo-2789-CIP -, ,J
_ 33 _ 1 3 28 ~ 4 joints injected with phosphate buffered saline (PBS) and some joints in which only fluid was removed. CIs were calculated on these traumatized joints. They varied from 0 to 18.7 a~fon~ 56 joints. However, the average CIs in the three separate studies of traumatized joints sho~ing these wide individ~al ~ariations were 0.7, 5.3, and 3.4. It is thus suggested that an average i`.
CI value of 6.0 or less in at least lO
joints could be expected from injection trauma. On this basis, we have assigned a 10 - joint average CI value of 6.0 or less ~,A' ~ as acceptable in the horse joint reactivity .
1~ test for evaluation of HA preparations.
Any product showing a lO - joint average of CI of '6.0 is unacceptable. Using these ,~
criteria, several HA prepasations were tested. Results are shown in Table VI. i TABLE Vl ;
Evaluation of Hyaluronic Acid Prep æations , by the Horse Joint Reactivi tv Test :~
No. of Average AcceptabilitY '' Pre~aration ioints CI of Preparation 25 Microbiological 14 4.6 Acceptable Source HA Filtered ~ ~` th~ro~gh~itrocellu- 13 ~ Acceptable "
-~ lose --, . ,~
Microbiological 11 11.4 Unacceptable ' 30 Source HA Non- ~
filtered 6 17.4 Unacceptable --;.
Prior Art Purified;and 10 6.2 Unacceptable Filtered ---The microbiological source HA listed in Table -~
. VI was that obtained from fermenters 1 through 4 as Mo-2789-CIP
~ ` ~'.'.
.
-. :

: : .

1~288~1 described previo~sly. It is noteworthy that this ~, material is acceptable for joint inj~ction arter nitrocell~lose filtration b~t not prior to s~ch `.
filtration. On the other hand, Prior Art ~' (see Fig~re ;
5 5) is on the borderline of being ~nacceptable even after nitrocellulose ~iltration. Evidently, the ~eactive proteinaceo~s load in the Prior Art 1,1 prepa,ation is too great to be removed via the protein binding filtration step.
The same Clinical Index can be used to evaluate efficacy of treatment of diseased joints wi-.h HA. In this test system clinical symptoms are ind~ced in joints wi~h intraarticular injection of complete o. incomplete Fre~nd's adjuvant. This adjuvant produces rirst an 15 acute and then a chronic pathology of the joint characterized by extreme lameness and swelling which does not appear to reverse itself within two months.
Some such efficacy studies have been conducted on the bacterial-derived HA.
Experiments were designed to eval~ate the effect of semoving some of the joint fl~id ,rom adj~vant -~
induced pathological joints and replacing it with bacterial-derived HA. This was done with both acute joints (HA injection within three days of the Fre~nd's 25 Co~plete Adj~vant injection) and with chro~ic joints (HA
injected within 12 to 34 days of Freund's ;~jection).
Clinical Index evaluation was beg~n the day of adj~vant injection and continued for fo~r days following the HA
injection a~te which weekly observations were made for 30 three weeks. Figures 8 and 9 display the ~es~lts over 30 day periods.
Fig~re 8 represents the acute situation. The zero day readings were all adj~sted to seven on the relative index scale so that comparisons could be better o-2789-ClP

_ 35 _ 13288~1 ~is~alized. Zero day repr~sents three days post Fre~n~'s injection in the acute joints. Figure 8 then portr~ys the change in Clinical Index for the first 30 days post injection with HA from fermenters 1 and 2 and ~
5 Prior Art ~1 after f~rther purification. These res~lts are compared with similar adi~vant injected joints left untreated (control). It is notable that the control horses contin~ally worsen thro~gh day 4 post Fre~nd's injection before showing some improvement on their own.
10 Howeve., this improvement does not ~each the starting level by day th ee and by day four appears to be plateauing. This is the t~pical picture for induction of chronic pathology. A significant improve~ent in ';~
acute symptoms is observed after injection of HA.
The chronic situation is represented by Figure -9. Horses which ha~e been injected with ~reund's Complete Adj~vant 12 - 34 days prior to HA treatment can serve as their own controls since these horses had been `-stable for at least seven days prior to day zero. The 20 control line represents these control index levels.
Again, immediate clinical impro~ement is noted after -treatment with HA. Longer term observation of these ~r, horses has indicated that the improvement tends to plateau. Therefore, it is expected that more than one .
25 treatment may be necessary.
One horse entered the study with a diseased '~
io:nt of unspecified cause. As indicated in ~ig~re 10, two injections of bacterial-derived HA were a~ministered r~
to this joint 30 days apart. In this case, the zero day --30 readings were adjusted to 10 in order to display the `:
complete treatment response. Immediate improvement was i~-noted after both the first and second injections of HA.
The improvement after the first injection was followed by some return of pathology as indicated by s~elling Yo-2789-CIP

r ~ f ~L 3 2 ~3 8 ~ 1 only. After the secon~ HA injection joint swelling was eliminated and has not ret~rned within three months post r treatment.
From the data presented herein it is obvio~s 5 that ~ltra p~re bacterial-derived HA is nonreactive in horse joints and displays efficacy for reversing lameness and/or swelling in diseased joints. Since bacterial-derived HA as described herein is p~rer than any commercially available product, incl~ding those used 10 in opthalmalogical trea~ments, it is highly probabl~
that these HA preparations could also be ~sed to replace vitreous h~mor of the eve during surgery. They sho~ld be able to substitute for any other use applied to the rooster comb or umbilical cord HA.
The invention is further illustrated, b~t is not intended to be l;m; ted by the following examples in which all parts and percentages are by weight unless otherwise specified.

The viruience of a Stre~tococcus equi and therefore its hyaluronic acid production, which is a virulence factor, was enhanced by passaging throu~h serologically negative horse blood. In particular, so~e pus from an abscess of a horse showing the symptoms of a 25 Stre~tococc~s eaui infection, which was later established by differential s~gar testing and staining to contain this microorganism, was inc~bated with some serologically negative horse blood (blood having no detectable levels of antibody against this 30 microorganism) and the CDM described in the Van de Rijn and ~essler article, s~pra, for three ho~rs at 37C. A
small aliquot was then placed on a blood agar plate (smsll petsi dish ~ith solid agar and small amount of medi~m containing blood) o~ernight at 37~C. The large Mo-2789-CIP

13288~1 mociod colony was then transferred into a fresh mixt~re ~, of serologically ne~ative horse blood and CDM for three ho~rs at 37C. After a total of four inc~bations in horse blood/CDM the final isolated strain was grown in j~st CDM and a sample was submitted to the ATCC and received deposit n~mbe~ 39,506. This strain was s~bstantially ~ore virulent than the initial nat~rally occ~rring strain.

The strain developed in Example 1 was ~sed to grow cult~res for 22 and 64 ho~rs and the hyal~ronic acid generated was harvested and isolated. In partic~lar these c~lt~res were,grown in a stirred vented vessel with air feed lines which s;mply compensate for 15 the oxygen cons~med by the microorganisms b~t do not "aerate" by vigorously adding air or oxygen in the same CDM as ~sed in Example 1 at 37C with the pH initially set at 7.6 and thereafter controlled at a value of abo~t 7.2. In the 64 ho~r c~lture 1 weight percent of 20 dextrose, based on the total weight of c~lt~e, was added every twenty-fo~r ho~rs. For both c~lt~res the p control was terminated abo~t sixteen ho~rs before harvest.
Both c~ltures Temained in the log or 25 exponential growth phase for at least the period of pH
control. This was appaTent from the pH controller ' recosd which indicated additions approximately every '~ five min~tes and from optical density meas~rements which increased during the entire cu}turing peri~d as iolloas:

. :

, Mo-2789-CIP

- 38 - 132~8~1 .i ITEM 22 Hour C~lture 64 Ho~r C~lture 6' In HoursOPtical Density Optical Density }
3 0.13 0.07 6 0 ~ 17 0 ~ 12 r 22 0~18 0~22 ~ -28 0.22 0.22 33 0.235 -38.5 0.5 -64 0.675 (at 1:10 dil~tion) The la~ in the sixty fo~r hour culture was attri~uted to its exhausting the glucose in its medium before the next addition. It was s~rprising that a streptococcal 15 organism could be kept in log phase in a CDM for an ';
extended period. The only p-ior reported extended log phase growth for this genus of microorganisms appears to be an aerated-Shaker Culture in a protein rich medium which is reported in Japanese Published Patent 20 Application ~6692-83.
Sodium hyaluronate was then harvested and isolated from each culture by the first techniq~e describe~ hereinabove in~olving ethanol precipitation followed by centrifugation and redissolution in water. ~ -25~However, in this case five precipitations with ethanol were utilized. The ~dry cell weight of each culture was '' estimated by designating one centrifugation pellet as a --standard and then estimating the dry cell weight content ~--fs the pellet weight. The d-y cell content and 30 hyaluronic acid yield were as follows:
.' ' . "',, . '.' ' ~'.
~ Mo-2789-CIP
,,~ ~" .,-,.

`,:

':

~3288~1 Yield Hyaluronic Based on Acid Dry Cell ~.
Culture Drv Cell WeightObtained Wei~ht 5 2~ Hour 2.9 grams/10 11.1 g/10 1 37.9%
64 Hour 3.79 ~rams/10 12.06 g/10 154.4Z
Thus, the longer term culture is substantiall~ more .
efficient in producing hvaluronic acid. However, subsequent experiments indicated that culturin~ for 10 bevond 72 hours results in only ma-ginal increases in ~ield evidently because nutrients other than glucose became depleted from the CDM.

.:
Hyaluronic acid was harvested and isolated from 15 2 one hundred liter culture prepai-ed in a manner similar to .hat described in Example 2. However, it was 2nalyzed for protein content by W absorption at 280 na~ometers and nucleic acid content by UV absorption at .
260 nanometers immediately before and immediately after 20 passage t~rough a single thickness 1.2 micrometer pore ~-293 mm diameter nitrocellulose filter. The 4.5 liters !
OI aqueous sodium hvaluronate solution being treated had the following pre and post filtration contents:
Pre Fil- Post ~ t~
tration tration Percent ContaminateContent Content Reduction Protein0.28 mg/ml 0.06 mg/ml 78% ; ~:
Nucleic acids 0.09 mg/ml 0.023 mg/ml 74%
The protein load on the filter was about 1.88 ~
30 mg/cm2 and it remo~ed approximatelv 1.48 mg/cm2 of protein to effect 2 protein reduction of 78%. After one .
further ethanol precipitation and resolution this lot was found to contain 20.08 grams of sodium hyaluronate for a yield of about two grams pel ten liters of :~.
3~ o~i~inal culture. - -Mo-2789-CIP

_ 40 _ 1 3 2 8 8 ~ 1 The degree of contaminate removal appeared to be related to the contaminate load on the nitrocellulose filter. In a similar experiment 7.5 liters of an aqueous sodium hyaluronate solution were passed through a similar filter. The protein reduction appeared to be about 43%
and the nucleic acid reduction seemed to be about 29% but the protein load was about 4.25 mg/cm2. The absolute removal was about 1.85 mg/cm2 in reducing the lo concentration from 0.38 to 0.215 mg/ml. The absolute reduction in nucleic acid content was also similar between the two experiments with 0.045 mg/cm2 removed from the 4.5 liter batch and 0.052 mg/cm2 removed from the 7.5 liter batch.

An aqueous approximately one weight percent sodium hyaluronate solution displaying pyrogenicity in the rabbit test was rendered non-pyrogenic by treatment -with strong acid washed activated carbon. The solution was contacted with 29 grams per liter Gelman 12011 a strong acid washed activated carbon with a surface area of about 700 m2/g (thus giving 20,000 m2/liter of surface ;:
area) for one and sixty minutes. A reference solution of lipopolysaccaride (LPS), the commonly suspected pyrogenic agent, was similarly treated. The LPS content of both was evaluated by Limulus amoebacyte lysate (LAL) analysis with the following results:

Post Post Initial 1 minute 60 minute ContentTreatmentTreatment H~ solution 15.5 ng/ml12.08 ng/ml 8.12 ng/ml LPS solution 30 ng/ml 13.35 ng/ml 12.75 ng/ml The results with the hyaluronate solution was 35 confirmed by testing of similarly treated samples in the rabbit test.

Mo-2789-CIP

~r ~, ..

- ~ . ; . . . . - . . I . . .. .. .. ..

~ f ~;
1328~1 The ability of ac~ivated carbon to extract LPS
from an aqueous sodium hyal~ronate sol~tion appeared to be uniq~e. Although other materials such as ion exchange resins and affinity binding colf~mns marketed 5 for LPS adsorption extracted LPS from a water solution they were unable to extract from an aqueous hyaluronate solution.

Hyal~ronic acid was prepared using the st-ain 10 of microorganism prepared in accordznce with Example 1 -~
by culturing a five hundred liter batch for forty-eight hours in accordance with the procedures of Example 2 except that a second glucose addition of 0.5 weight percent hased on the total culture weight was made 15 sixteen hours after each one percent addition by isolating and purifying the product by the second preferred procedure utilizing the "winder" which is discussed hereinabo~e. The procedure included the especially preferred features of lowering the ~iscosity 20 by heat treatment, and filter sterilizing bv passage through a 0.2 micrometer pore filter as well as a subsequent pyrogen~treatment by passage at between 37 ,:
and 56C through a Gelman 12011 strong acid washed acti~ated carbon filter pro~iding 20,000 m s~rface area 25 per liter of solution being treated. The final aqueous solution had a sodium hyaluronate cor.tent of 1.19 ~eight percent. This hyaluronic acid had a FPLC (fast prot~in liquid chromatography) dete-mined average molecular weight of 1.88 x 106 daltons ~th a narrow essentially 30 symmetrical single distribution peak, a 37DC visc~sity of 147 cSt, a nucleic acid content of less than 0.003 mg/ml by ethidium bromide fluorescence and a total afino acid content of less than 0.005 mg/ml by orthophthalaldehyde fluorescence. lt caused no readily Mo-2789-CIP

~ d '.; S, - 42 - 13288~1 apparent antigenic reaction when injected into horses and ~ave a negative indication in the rabbit pyrogen test.
Although the invention has been described in 5 detail in the foregoing .or the purpose of illustration, it is to be ~nderstood that such de~ail is solely for that purpose and that variations can be made therein by those skilled in the art witho~t departing from the spirit and scope of the invention except as it mav be 10 limited by the claims.

Mo-2789-CIP

Claims (14)

1. A process for extracting high molecular weight hyaluronic acid from an anionic/cationic surfactant complex comprising dissolving the hyaluronic acid in a high molarity aqueous solution of a calcium ion.

2. A process for the preparation of high molecular weight hyaluronic acid suitable for medicinal administration to mammals without provoking an immune reaction comprising eliminating its pyrogenicity by contacting an aqueous solution of it with a strong acid washed activated carbon.

3. The strain of Streptococcus equi deposited with the ATCC under number 39505.

4. A process for the production of high molecular weight hyaluronic acid suitable for medicinal administration to mammals without provoking an immune reaction comprising a) culturing the ATCC deposit Number 39506 strain of Streptococcus equi in a medium free of extraneous protein, and b) isolating and purifying the generated hyaluronic acid.

5. A process for the production of high molecular weight hyaluronic acid with a narrow molecular weight distribution suitable for medicinal administration to mammals without provoking an immune reaction comprising a) increasing the virulence and therefore the hyaluronic acid generation of an existing strain of Streptococcus equi by passaging it through serologically negative horse blood, b) growing a culture of the strain of Streptococcus equi prepared by said passaging in a medium free of extraneous protein, c) maintaining the bacterium in log phase growth for at least about twenty-four hours by 1) maintaining the pH of the medium between about 7 and 7.2, 2) maintaining the temperature at about 37°C, and 3) feeding at least about one weight percent of glucose, based on the total culture weight, to the medium no less frequently than about every twenty-four hours, d) separating the capsular hyaluronic acid generated from the walls of the cells that generated it by incubating the culture with at least about 0.01 weight percent, based on the total culture weight, of a sulphonate or sulphate based anionic surfactant, e) precipitating the generated hyaluronic acid from the medium containing said anionic surfactant by the addition of an aliphatic quaternary ammonium salt, f) separating the precipitated hyaluronic acid from the anion surfactant and ammonium salt by dissolving it in a high molarity aqueous solution of a calcium ion, g) precipitating the hyaluronic acid from the calcium ion solution by combining the solution with a lower alcohol selected from the group consisting of ethanol and isopropanol, h) repeatedly redissolving the reprecipitated hyaluronic acid in water, adding sufficient sodium chloride to provide at least about ten grams per liter and reprecipitating the hyaluronic acid by combining the solution with ethanol or isopropanol, i) passing an aqueous solution of the repeatedly dissolved and precipitated hyaluronic acid through a nitrocellulose filter, and j) treating an aqueous solution of the repeatedly dissolved and precipitated hyaluronic acid with sufficient strong acid washed activated carbon to remove any pyrogenicity from it as measured by the rabbit pyrogen test.

6. A process for the production of high molecular weight hyaluronic acid suitable for medicinal administration to mammals without provoking an immune reaction comprising a) growing Streptococcus equi in an aqueous chemically defined medium which is free of protein not released by the microorganism in a non-aerated shaker culture, b) inactivating any extracellular hyaluronidase generated by the microorganism, c) maintaining the microorganism in essentially log phase growth for in excess of about twenty-four hours by 1) maintaining the pH in the range of between about 7 and 7.2 by continuous or intermittent addition of base, 2) maintaining the temperature in the range of about 37°C, and 3) adjusting the glucose content of the medium to at least one weight percent at least every twenty-four hours, d) isolating the generated hyaluronic acid from the culture without disrupting the Streptococcus equi cells, and e) purifying the isolated hyaluronic acid without causing any significant molecular weight degradation.

7. The process of Claim 6 in which the Streptococcus equi is the strain deposited under ATCC
number 39506.

8. The process of Claim 6 in which the chemically defined medium is sterilized before inoculation with the Streptococcus equi.

9. The process of Claim 6 in which i) the chemically defined medium is filter sterilized before inoculation, ii) the inoculate is grown on the same chemically defined medium as the growth medium, and iii) the inoculate comprises between about one and five volume percent of the growth medium.

10. The process of Claim 6 in which the pH is permitted or caused to drop to between about 6.5 and 6.8 before isolation of the hyaluronic acid.

11. A process for the production of a hyaluronic acid suitable for medicinal administration to mammals without provoking an immune reaction comprising a) increasing the virulence and hyaluronic acid generation of an existing strain of Streptococcus equi by passaging it through serologically negative horse blood, b) growing a culture of the strain of Streptococcus equi prepared by said passaging in a medium free of extraneous protein, c) maintaining the bacterium in log phase growth for at least about twenty-four hours by 1) maintaining the pH of the medium between about 7 and 7.2, 2) maintaining the temperature at about 37°C, and 3) feeding at least about one weight percent of glucose, based on the total culture weight, to the medium no less frequently than about every twenty-four hours, d) separating the capsular hyaluronic acid generated from the walls of the cells that generated it by incubating the culture with at least about 0.01 weight percent, based on the total culture weight, of a sulphonate or sulphate based anionic surfactant, e) precipitating the generated hyaluronic acid from the medium containing said anionic surfactant by the addition of an aliphatic quaternary ammonium salt, f) separating the precipitated hyaluronic acid from the anionic surfactant and ammonium salt by dissolving it in a high molarity aqueous solution of a calcium ion, g) precipitating the hyaluronic acid from the calcium ion solution by combining the solution with a lower alcohol selected from the group consisting of ethanol and isopropanol, h) repeatedly redissolving the reprecipitated hyaluronic acid in water, adding sufficient sodium chloride to provide at least about ten grams per liter and reprecipitating the hyaluronic acid by combining the solution with ethanol or isopropanol, i) passing an aqueous solution of the repeatedly dissolved and precipitated hyaluronic acid through a nitrocellulose filter, and j) treating an aqueous solution of the repeatedly dissolved and precipitated hyaluronic acid with sufficient strong acid washed activated carbon to remove any pyrogenicity from it as measured by the rabbit pyrogen test in such a manner that a hyaluronic acid is obtained which has 1) a molecular weight distribution in which 98%
of the moieties fall within a single substantially symmetrical peak which lies between retention times representative of molecular weights between about 1,000,000 and 4,000,000 daltons, 2) a UV absorbance determined protein content of less than about 1.25 mg/ml, and 3) a UV absorbance determined nucleic acid content of less than about 0.045mg/ml.

12. The process of Claim 11 in which the hyaluronic acid obtained has 1) a standardized FPLC determined average molecular weight of at least about 1,100,000 daltons, 2) an orthophthalaldehyde fluorescence determined amino acid content of less than about 0.4mg/ml, and 3) an ethidium bromide determined nucleic acid content of less than about 0.06 mg/ml.

13. The process of Claim 11 in which the hyaluronic acid is purified by passaging an aqueous solution of it to a bath of ethanol or isopropanol in which a device with vertical fingers is slowly rotating.

14. The process of Claim 11 in which the final viscosity of the aqueous hyaluronic acid solution obtained is lowered to less than about 250 centistokes at 37°C by either heating it at greater than 50°C in an open vessel or forcing it through a filter with a pore size less than about one micrometer.