T11..: .JollltNAI. OF I NV I-:STIC:ATI\' f: DtmMATOt ()(;r

Cop~· ri ghl @ 1~)72 bv The Williams & \Vilkin~ Cn. Vnl. fi8. No.2

fJrint('c/in U.S.A.

THE PRODUCTIO OF HYAL URON IDASE (HYAL RONATE LYASE) BY CORYNEBA CTERIUM ACNES*

S. MADLl PUHVEL, PH.D . AND RO ALD M . RE ISNE R, M .D.

ABSTRA CT

This paper presents evidence of the product ion of hyaluronidase by Corynebacterium acnes. In a study of seven ra ndom isola tes of C. acnes from les ions of cystic ac ne a nd from norma l skin , hya luronidase activity ra nged from less than one turbidity reducting uni t (TRU) to 72 TRU per mg of extracellular prote in. These findin gs raise the interesting pos­sib ili ty that C. acnes hya luronidase may play a role in the deve lopment of acne by contrib­uting to the in!la mmatory phase through a n increased permeabi li ty of t he folli cu lar epithe­lium to free fatty acids and other irritants.

Corynebacterium acnes is the predomin a nt bacterium found in sebaceous gla nd -rich areas of norma l huma n skin (1). It is a lso one of the two bacteria charac terist ica lly found in lesions of acne vul garis (2). Numerous investiga tions have char­acterized many of the biochemica l a nd phys iolog­ica l prope rti es of th is organism (3, 4, 5, 6), how­ever, t he produ ct ion of hya luron idase (hya lu­ronate lyase) by C. acnes has only been reported once prev iously in a brief technical note de­scribing a new screening m ethod for detecting hyaluronidase producing organis ms (7).

M ATERIALS AND METHODS

Five isolates of C. acnes were obtained fro m les ions of acne vul ga ri s in patients with cystic acne attending t he Acne linic a t the University of Ca lifornia Center for Hea lth Sciences, Los Angeles, Ca li fornia . Two isolates o f C. acnes were obtain ed fro m fac ia l skin of indi v idua ls without ac ne. The source of isolation of each s train is li sted in Table I. All isolates conform ed to the accepted characte ri stics of this s pecies (i.e., a nae robic, catalase pos it ive, lique fi ed ge latin) a nd a ll were sens itive to C. acnes bacteriophage.

For production of hya luronidase, ce ll s were inocu ­lated from log phase cultures into 85% sa line solutions in dialys is bags immersed in b rain hea rt infus ion broth (D ifco) suppl emen ted with 1% d ex trose. Anaerobios is was achi eved by constant bubbling of a mixture of 90% N and 10% CO, in to t he dia lys is bags. In cubation was carried out at 37° C for 72 hours. The bacteria l ce lls were re moved from the supernates by hi gh s peed cen ­tr ifugat ion for 20 minutes. Two liter batches of culture supernates were dia lysed against distilled water for 72 hours at 4° C. Following this the protein in the super­nates was precipi tated by dia lysis agai nst four volum es of saturated a mmonium sulfate so lu t ion . The precipi­tated extrace llula r proteins (ECP) were sepa rated by high s peed centrifu gat ion , red issolved in distilled water a nd dia lysed aga ins t di s till ed wa ter to e liminate the a mm onium su lfate. The prote in solu tions were concen-

Received July 16, 1971 ; acce pted for publication Oc­tober 12, 197 1.

This wo rk was supported in pa rt by gra nts from the E lliot a nd Ruth H a ndl er Fou ndation a nd the M ax Factor Memoria l Fund.

*Fro m t he Depa rt m ents of M edic ine, Divi s ions of Dermatology UCLA Cen te r for t he Hea lth Sc iences, Los Angeles, Ca lifornia 90024 a nd Ha rbor Gene ra l Hospita l, Torra nce, a lifornia 90509.

66

trated by lyophili zat ion a nd th e lyophili zed ECP was used for hya luronidase assays. Viscosim etri c a nd turbi­dimetri c assays we re used to d e monstrate hya lu ro ni dase act ivity .

Viscosim etric assay. The viscosimetri c assay is based on the redu ct ion of v iscos ity of hya luroni c acid substra te by the act ion of hya luronidase. FOI' a ll vis­cosity determinat ions a n Ostwa ld viscos im e te r in a thermo-regu lated cons ta nt temperature glass wate r bath at 25° C was used . The flow t im e of di s till ed water in the Ostwald v iscosim eter used was 58.7 seconds. Potas­s ium hyaluronate (G ra d e 1 11-P from human umbilical cord , lot number 30C-2180, Sigma Chemi ca ls, St. Louis, Mi ssou ri) wa used as the substrate t hroughout. Prelim i­na ry ex pe rime nts m easuring t he /l ow time o f' vary ing concentra tions of hya luronic ac id (HA) solu tions in 0.1 M c itra t e butTer established that 0.1 % HA concentra­tions gave fl ow times close to 100 seconds. S ince fr e­quent repeated read ings were to be ta ken in subsequent experim ents, this concen tration was cons ide red optima l. M ore concentra ted HA solu tions ha d increas ingly longer Jlow t im es. 0.4% HA solu t ion for exa mple, had a fl ow time of 317.4 seconds. All fl ow times were determined using a Heuer stop-watch which gave rea dings to the nearest one t enth of a second .

The optim a l pH of C. acnes hya luronidase ac tivity in the viscosim etri c assays was determined by measuring the redu ct ion of flow tim es of 0.1 % HA solu tions in 0.1 M cit rate a nd phosphate buffe rs ra ngin g in pH from 4.0 to 8 .0 in the presence of 250 ~rg of C. acn.es ECP per ml of substrate. 0.4 % HA solutions were pre pared in each buffer tested. 10 m g lyoph ili zed C. acne~ ECP was dis­solved per ml of di st ill ed wa te r. One ml of 0.4 % HA so­lu tion was added to 2.9 ml o f' co rrespondin g buffer a nd 0.1 ml of en zy me solu tion added at ze ro t im e. Flow times were meas ured after 15 minutes at 25° C. The irii­tia l v iscosity of t he HA solution without enzy m e was determined by measuring the llow time of l ml of 0.4% HA mixed with 2.9 ml of correspondin g butTer a nd 0.1 ml of distill ed water.

The effect of heat on C. acnes hya lu ron idase ac tivi ty was determined by heat ing 10 mg of e nzy m e per ml of' citra te butTer at va ry in g temperatu res for 5 minutes. T he hyluronidase activity of t he hea ted sampl es was determined by test ing the e ffect of 0.1 ml of' each on the llow tim e of 1.0 ml of 0.4% HA mixed with 2.9 ml of 0.1 M c itrate buffe r, pH 5.6.

F or est ima tion o f' hya luronidase activity o f' t he dif­fe rent C. acn es isola tes, 10 mg o f C. acnes ECP was dis­so lved in l ml of c itrate butTer (0.1 M , pH 5.6) a nd a series of 10-f'old di lu tions were m a de. Flow tim e deter-

C. acnes-HYALURONIDASE 67

TABLE I

R eduction in. spec ific v iscosity* of 0. / % HA in the

presence of 25 !lg of C. acn.es ECP per ml of substrate

c. V i ~co~ it y ( 1~ rcdul' t ion) acne~

stra in Sou r('C Gm in . IO min . l!l min .

no.

6 normal skin 0.13 (90) 0.09 (93) 0.08 (93)

18 .. 0.09 (93) 0.07 (94) 0.06 (95)

Sa cystic acne 1.09 (16) 10 1 (22) 0.93 (27)

38 .. 0.71 (45) 0.57 (56) 0.49 (62)

79 .. 0.48 (63) 0.:34 (73) 0.27 (77)

142 .. 0.70 (46) 0.53 (59) 0.43 (67)

143 .. 0.62 (52) 0.48 (63) 0.40 (69)

*specifi c viscosity t.,.,,.,,,./t , urrcr I where t is fl ow time in seconds viscosity of 0. 1% HA without enzyme

was 1.30

minations of 3.9 ml of 0. \% HA solu t ions in citrat e buffe r were determined a t 5, 10 and 15 minutes afte r the a dditi on of 0. 1 m I of each d iluti on of ECP. The fl ow ti me of3.9 ml ofO. I% HA soluti on with 0.1 ml ofbutTer without enzy me was determ ined as the control for ini ­tia l v iscos ity of the HA solu tion. In a ll viscosity dete r­mina ti ons the stock HA solu t ion was sto red at 4° duri ng the experiment. Mai ntenance oft he HA at room tem pera ture for severa l hours resulted in a slight spon­taneous decrease in viscos it y. T he reduction in specific viscosity was calculated using th e fo rmula:

s pec ifi c vi scos ity

wh ere t ,

t,,

t., - - 1, t,,

fl ow time of the sa m ple and

fl ow time of the buffer (8).

Tu.rbidim etric a .~say. An alternat ive procedure to demonstrate hya luronidase ac itvity is the turbi d imetri c assay which is based on the formation of n coll oidal sus­pens ion when HA is mixed with se rum a lbumin at an acid pH . Oepolymeri zation of HA by hyaluroni lase prio r to the addition of a lbumin red uces the turb idity of the HA -albumin mi xutre since only HA units of 6000 to 8000 molecular weight or more, react with the ac id a l­bum in to produce colloid form ation (9). The reduction in t urbidity can be measured spectrophoto metri ca lly. One turbidity reduc in g unit (TRU) is defined as the amount of enzy me which in 30 minut es at 37° C reduces the turbidity of 0. 2 mg of HA solut ion to that of 0. 1 mg of HA solution.

0.1 M phos pha te butTer, pH 5.3 in 0.15 M NaC I was used as t he d ilu en t for both the !-lA and the ECP solu­tion. Bovi ne serum a lbumin (Fract ion V, Nutritiona l Biochemi ca ls, Ohi o) was prepa red us sugges t ed by Dor fman (9), by dissolvi ng 10 g of a lbumin , 32.6 g of sodium acetate, 45 .6 ml of glac ial acetic ac id in \0 li ters of water. The p!-1 of the 0. 1% a lbum in solu t ion was brought to 3.72 with concentrated HCI. HA standards we re prepared by addin g 0.1 to 0.6 ml of a solution of 0.4 mg H A/tn l to a seri es of dup lica te colorimeter tubes. and addi ng butTer to 1.0 mi. Blanks contai ned 1.0 ml buffer. The tubes we re heated in a boiling water bath for 5 minutes, and cooled for 5 minutes in an ice bath. ext 4.0 ml of ac id a lbumin solution was added and the tubes

ge ntly mixed . Aller 10 minutes the tu rbidi ty was deter­mined in a Coleman J r. spec trophotometer at 600 m11.

Milligra ms of HA substra te were plotted aga inst optical dens ity read in gs and a stra ight line standard curve was obta ined.

F'or the assay of the unknown s, 0.5 ml of a solu t ion of 0.4 mg H A/m l were added to a series of duplicate color­imeter tubes. C. acn es ECP was dissolved in phosphate buffer in concentrat ions of l mg/m l and a seri es of 2-fold diluti ons prepa rerl . 0.5 ml of each enzy me d ilu t ion was added to each tube of substrate and incubation was carried out for exactly 30 minutes at 37° C. The reac­tion was stopped by heati ng the tubes in a boiling water bath for 5 minu tes, followed by cooling in an ice bath for 5 minutes. 4.0 m l of acid a lbumin solu tion was then added to each tube, and a fter 10 minu tes. the turbidi ty of each t ube was measured in th e sa me way as described above for the standa rds. A co mplete seri es of sta ndards was inclu ded for each seri es of unknowns tested. The HA solu tions and ECP solutions were prepared fresh on the day that they were to be used. T he optica l densit ies of t he unknowns were plotted on the standard curve and the TRU/mg of ECP ca lcu la ted.

RESULTS

T he optima l pH for C. acnes hya luronidase ac­t ivity was between pH 4.8 a nd 6.4. Hyaluronidase activ ity was inhibi ted at pH 4.0 and 8.0. T he re­duction in fl ow t imes of HA solu tion s in t he pres­en ce of C. acn.es ECP at diffe ren t pHs is shown in Tab le II.

C. acn.es h yaluronidase was heat labi le and could be 90% inact ivated by heating at 65° C for 5 minutes, a nd completely inact ivated by heating at 80° C for 5 minutes. Storage at 4° C for seven days d id not affect th e hya luronidase activ ity. P ro longed storage at 4° C was n ot attempted.

T he reduction of v iscos ity of HA by C. acnes ECP was proport io na l to t he con cen tration of ECP/ml of substrate. Figures 1 a nd 2 demon­strate t h e reduct ion of v iscosity of 0.1 % HA by diffe rent con centrat ion s o f ECP from C. acnes stra in 38 a nd C. acnes t ra in 6. The spec ific v is­cosity of 0.1% HA so lution prior to the add ition of en zy m e was 1. 30. As shown in Figure 2, addition of 25 llg of C. acn.es 6 ECP pe r ml of sub t rate,

TABLE II R educt ion in fl ow time of 0.1 % hyaluronic acid in

different buffers after 15 minutes a( 1:ncubation

with 250 11# of C. acnes 79 ECP/ ml of substrate

Flow time Flow time Hcrluction 0. 1 M bu ffer pH of HA with · of HA wilh in now

(lUI Cll 7.YillC (' 11 7. ~' 111 (' time (~ecs . )

Citrate 4.0 100.5 100.6 -.. 4.6 toa.8 83.:3 20.5 .. 4.8 99.8 72.2 27.6 .. 5.0 97.2 69.4 27.8 .. 5.6 98.3 70.0 28.3 .. 6.0 102. 1 71. 1 31.0

Phosphat e 6.4 95.4 67.4 28.0 .. 7.0 95. 1 74.2 20.9 .. 7.5 98.5 85 .2 13.3 .. 8.0 97.6 94.5 3.1

68 THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

reduced the viscosity by 93% in 15 minutes ; addi­tion of 2.5 p.g, by 83% in 15 minutes; and addition of 0.25 p. g, by 44% in 15 minutes.

As can be seen from Table I and from a com­parison of Figures 1 and 2, there was grea t varia­tion in the a mount of hya luronidase activity ex­hibited by the different C. acnes isola tes. The lowest activity was present in C. acnes 8a, iso­lat ed from a lesion of cystic acne. 25 iJ.g of C. acnes Sa ECP per ml of substra te reduced the

_ viscosity of 0.1% HA by only 27% in 15 minutes. By contras t , the highest act ivity was exhibited by C. acnes 6 and 18, both isolated from normal skin . 25 p.g of ECP from these strains per ml of substrate reduced the viscosity of 0.1% HA by 90 and 93% respectively in 5 minu tes.

F igu re 3 demons trates the reduction of vis­cosity of HA by 25 iJ. g of C. acnes 142 ECP per ml of substrate as a funct ion of t ime of incubation . The ini t ial phase of the reac tion is extremely rapid, with 46 % reduction in vi scos ity produced in the firs t 5 minu tes . Incubation of the substra te

100

>-1-u:; 0 u (/)

> 0.: (/) 60 ::: z 2 4 0 1-u ::J 0 LLI 0:: 20 il

------ 250 1-'- 9 ECP/ml

5 10

T IME (mi nu l es )

25 1-'-9 ECP/ ml

2 .5 1-'-9 ECP/m l

0.25 1-'-9 ECP/ml 0.025 1-'-9 ECP/ml

15

F IG. 1. Reduction in spec ific viscosity of 0.1 % HA so­lu t ion by diffe ren t concentra tions of C. acnes 38 E CP .

100

>- ------. 25 J.LQiml

'= (/)

0 80 u (/)

> 0.:

60 (/)

~

z 2 1-

40 u ::J 0 w 0:: 20 il

5 ' 10 ' 15'

TI ME (minules )

F tG. 2. Reduction in specifi c viscosity of 0.1% HA solu tion by di fferent concentrations of C. acn.es 6 E CP .

beyond 60 minutes did not decrease the visco~ity further.

Results of the turbidim etric assays of hyaluron­idase activity a re listed in Tabl e III. The TRU values ranged from less than 1 T R U/mg of E CP to 72 TRU/mg of ECP. There was good correla­tion between the results of the viscosimetric and the turbidimetric assays. C. acnes 6 and 18 had the highest TRU va lues, while C. acnes 8a had an ac tivi ty of less than one TRU/mg of ECP.

DISC USSION

Hya luronid ases belong to th e muco polysac­charidase group of enzy mes. Numerous bacteria in cluding pneumococci, s taphylococc i, clos tridia, proteus and f1 avobacteria produce this enzyme (10) . C. acnes hyaluronidase a ppears to be s imilar to other bacterial hya luronidases in that the op­timal pH of its activity is in the acid range and it is heat labile.

All methods of assaying hya luronidase activity

100

>-1-

(/)

0 80 u (/)

> 0.:

60 <f)

z

z Q 4 0 1-u :;) 0 w 0:: 20 ;f.

20 40 60 T IME (minutes )

FIG. 3. Reduction in specific viscosity of 0. 1% HA solu t ions by 25 llg of C. acnes 142 E CP per ml of sub­s t ra te over a period of 60 minutes.

TABLE Ul

Hyaluronidase activ ity of C. acnes isolates expressed in. turbidity reduct in.g units (TR U) / mg of C. a etJeS ECP

C. H CI1 Cl' !:inur<.:c of TIW/m~ struin no. h:ulal ion t1 1' ECP

6 norma l s kin 72 18 64 Sa cys ti c acne less than 1

38 4 79 13

l42 4 143 4

C. acneS-HYALURONIDASE 69

are based on the reduction of hyaluronic acid to small subuni ts. In contrast to testicular hyaluron­idases, which are heat stable and whose major end products are tetrasaccharides, bacterial hyaluroni ­dases are heat labile and have as their major end produ ct disacchar ides conta ining 4 , 5-desaturated uronides (10).

The levels of hyaluronidase activity of the iso­lates in the presen t study ranged from very weakly active to moderately active. Both the vis­cosimetr ic and turbidim etri c assays are widely used accepted techniques for measuring hya luron­idase activity (11, 12). All hyaluronidase assays are affected by the physico-chemical properties of the substrate since no standardized hyaluron ic acid is available. Resu lts within any one labora­tory may be interna lly consistent, but it is diffi ­cult to accurately compare bacterial hyaluroni ­dase activity determined in one laboratory to that determined in another. Comparisons of activity are further confused by numerous different uni ts wh ich have been used to express enzyme activi ty. Furthermore, publications which have used iden­tical uni ts to express hya luronidase activ ity (e.g . TRU) have often expressed activity in terms of non-comparable parameters such as ml of cul ture supernate (13), hours of incubation (14), or mg of prote in at various stages of enzyme purification (15). Thus, it is diffi cul t to compare the resul ts of the assays in the present paper with those of others. For example, Abramson and Freidman (15), in a study on the purification of taphylo­coccal hyaluronidase from what they described as an active hyaluronidase produc in g stra in of Staphylococcus aureus found hya luronidase ac­tivity in the range of 1.1 TRU/mg of protein. Their protein was obtain ed by lyophili zing the precipitate obtained by 100% a mmonium sulfate saturation of cu lture supernates . T he cul ture medium used was very similar to the one used in the present paper, and consisted of saline which had been dialysed against brain heart infus ion broth supplemented with dextrose. The enzyme was then further concentrated more than 4000 fold by equen t ia l a mmonium sulfate fract ionation and Sephadex gel filtration , so that the fina l puri­fied protein preparation had an activi ty of 4380 TRU/mg. Thus the stage of purification of the enzyme s ignificant ly affects fina l reported ac­tivity.

The role of microbial hya luronidases in disease has not been definit ely established . Even though hyaluronidase production is a characteri stic of many pathogenic bacteria, known non-pathogenic organisms can a lso produce t his enzyme. Several studies which have attempted to correlate hyalu­ronidase prod uction in vitro with invasiveness in vivo, have resul ted in negative findings (16, 17). Yet in the case of staphylococci , for exa mple, there is an a lmost 94% correlat ion between the production of coagulase and the production of hya luronidase. Coagulase negative staphylococci do not produce thi s enzyme (18).

In certain disease conditions it appear that bacterial hyaluronidases, even from normally sap­rophytic bacterias may have a significant effect on the pathogenesis of the disease. A prime ex­ample of this is in the pathogenesis of gingiv itis. Studies have suggested that the hyaluronidases from the microbial flora of dental plaques may affect t he permeability of gingiva l epithelium and be a contributing factor in the development of gingivit is (19). In thi s conn ection, Schultz-Haudt et a l have demonstrated that epithelial cel l bridges are· reduced and intercellular cement sub­stance is removed by the application of testicular hyaluronidase preparations to gingiva l epithelium (20). Thilander has ob erved the widening of in­tercellular spaces of gingival epithelium after applications of hyaluronidase in vivo to the gin ­giva (21).

T he demonstration of hyaluronidase production by C. acnes in vitro may provide new insights into the effect of thi s organism in vivo. C. acnes is numerically the predominant bacterium in the sebaceous follicles of normal human skin . To­gether with S. epidermidis it can also he routinely isolated from lesions of acne vulgaris. T he role C. acnes in the pathogenesis of acne has been inves ­t igated in tensively during the past decade. Re­cen t stud ies have suggested that li pases deri ved from C. acnes in the sebaceous follic les are pri­marily responsible for the hydrolysis of trig lycer­ides of '"virgin " sebum to release free fatty acids which are known to be the major inflammation producing substances in sebum (22, 23, 24).

Hya lu ronic acid , t he substrate of hyaluronidase activity is known to be the prime constituent of the grou nd ubstance of the dermis (25). It has also been shown in human gingival epithel ium that the intercellular cementing substance con­sists in part of hya luron ic acid which is removable in vitro by hya luronidase (26). While the composi ­t ion of the intercellular cement substance in human pilosebaceous follicular epithelium has not yet been elucidated, it is conceivable that the breakdown of this intercellular substance by C. acnes, hya luronidase might be a factor in in ­creasing the permeability of the fo lli cular epithe­lium and fac ili tating the diffusion of irri tating free fatty ac ids into perifollicular areas of the dermis .

In the present study of only seven isolates of C. acnes it was interesting that the two most active hya luronidase producing strai ns were both iso­lated from normal skin and some of the strains from cystic acne were relatively poor hyaluroni ­dase producers under the cond itions of the present experiment.

Further studies are in progress to determine whether there are statistically significant differ­ences in the production of hyaluron idase by iso­lates of C. acnes from both obstructive and in ­flammatory ac ne lesions as compared to isolates from normal skin, and what role C. acnes hyalu­ronidase may play in the pathogenesis of acne.

70 TI-lE JOURNAL OF INVESTIGATI VE DERMATOLOGY

We gratefully ack now led ge the va luable techni ca l as­s is ta nce of Mr. Alvin T rotter.

REFERE CES

1. Ma rp les, M. J .: The Ecology of Human Sh in. Cha rl es C. Thomas, S pringfi eld, Illinois. 1963.

2. Shehadeh, N. H. a nd K li gma n, A. M.: Bacteri ology of ac ne vu lga ri s. Arch. Derm. , 88: 779, 1.960.

3. Moss, C. W., Dowe ll , V. R. , Jr. , Lewis, V. J . a nd Sc hekter, M. ·A.: Cu ltura l cha racte rist ics a nd fatty ac id co mposition of Corynebacterium acnes. J . Bacterial. , 94: 1300, 1967 .

4. Pu hvel, S.M.: Cha racte ri zation of Coty n.ebacterium acn.es. J . Ge n. Mi crobial. . 50:3 13, 1968.

5. Zierclt, C. H., Webster, C. a nd Rude, W. S.: Study of the a naerobic cory nebacteria. In t. J. Syste m. Bacte ri ol. , /8: 33, 1968.

6. Voss , J . G.: Diffe rent iation of two groups of Cory ne­bacterium acnes. J. Bacte ria l. , 101: 392, 1970.

7. S mith , R. L. a nd Wi llett, N. P :: Rapid plate method for sc reening hva luroniclase a nd chrondroitin sulfatase-p rod ucin g microo rga nis ms. App l. M i­crob iol, 16: 1434, 1968.

8. Hult.in , E. a nd Wa nntrop, J .: Viscos imetric deter­minat ion of ce ll ulase act ivity in the intes t ine of the sea urchin : Reaction mecha nis m a nd equi li b­rium consta nt for cullu lase sta bi li zat ion with ca l­cium . Acta Chern. Sca ncl ., 20: 2667, 1966.

9. Dorfman, A. : Mu copolysaccha ridases, pp. 166- 173, M ethods in Enzymologv. Eels ., Holowi ck, . P . a nd l<a p la n. N. 0 ., Academ ic Press. N. Y., 1955.

10. Myer, 1<. , 1-l ofTm a n, P . a nd Linker, A.: 1-l ya luruni ­dases, pp. 447- 460, The En zymes, Vol. 4, 2nd eel . Eels., Boye r, P. D., La rcly, 1-1 . a nd My rbac k, 1< ., Academic Press, N. Y., 1960.

ll. Meyer, 1<. a nd Ra pport, M. M.: Hya luronidases. Adva nc. Enzy mol. , 13: 199, .1.952.

12. Rogers, H. J .: 1-l ya luronidases, pp. 242- 245, Bio­chem ists' Handbooh. Eds. , Long, C. , King, E. J . a nd S perry, W. M., E. a nd F. N. Spon, Ltd ., London , 196 l.

13. Schul tz-Ha udt, S. D. a nd Scherp, H. W.: Produ c­tion of' hya luronidase a nd beta -glu curonidase by viridans streptococci isolated from gingival crev-

ices . J . Denta l Res., 34: 924, 1955 . 14. J esse n, 0 ., Fa ber, V. , Rosenda l, K. a nd Erik en, D.

R.: Some propert ies of Staphy lococcus au reus possibly rela ted to pathogeni city . Pa ri 1: A study of 446 stra in s fro m different ty pes of hu man infection. Acta Path. Mi crobi al. Sca n. , 47: 316, 1959.

15. Abra mson, C. a nd Friedma n, H.: Staphy lococcal hyaluronate lyase: Purifi cat ion a nd characteri­zat ion . J. Bacteriol. , 96: 886. 1968.

16. Crow ley, N.: Hya luro nidase producti on by haemo­lyt ic st reptococc i of hum a n origin . J. Pa th. Bac­te ria l. , 56: 27, 1944.

17. Humphrey, J . H.: Hya luronidase production by Pn eum ococc i. J . Pa th . Bacte ri a l. 56: 273, 1944.

18. Schwabacher, H., Cun li ffe, A. C. , Wi ll ia ms, R. E. 0 . a nd Ha rper, Y. J. : Hyaluronidase prod uction by s ta phylococc i. Brit. J . Path ., 26: 124, 1945.

19. Sode r, P . 0. and Nord , C. E.: Determination of hy­a lu ronidase act ivity in denta l plaqu e materia l. J. Periodont. Res., 4: 208, 1960.

20. Schultz- Haudt, S. D., Dewa r, M. and Bibby, B. G. : Effec ts of hya lu ronidase on huma n gingival epi­thelium . Science, 11 7: 653, 1953.

2 l. Thi lander, 1-1 .: T he effect of leukocytic enzy me ac­t ivity on the s t ructure of gingiva l poc ket epithe­liu m in ma n. Acta OdonL Sca nd ., 21: 432, 1964.

22. Marples, R. R. , Downing, D. T. and Kligman, A. M .: Control of free fatty ac id · in human surface li pids by Corynebacterium acnes. J . Invest. Derm ., 56: 127, 1971.

23. Kellu m, R. E.: Acne vu lga ri s, studi es in pathogen­esis: 11elat ive irritancy of free fatty ac id from to •• · Arch. Derm. , 97: 722, 1968.

24. Strauss, J . S . a nd Pochi, P. E.: ln t racuta neou injec tion of sebum a nd comedones. Histological observations. Arch. Derm ., .92: 443, 1965.

25. Schubert , M .: Inte rce llul a r mac romo lec u les con ­ta ining polysaccha rides, Connective Ti ssue: In ­tercellular m.a.cromolecules. Proceed ings of a sy mposium ~ponso r ed by T he New York Hea rt Associat ion, Litll e, Brown a nd Co. Bos ton, 1964.

26. Thonard, J . . a nd Scherp, H. W. : Cha rac teri zation of hum a n gin giva l epitheli a l in terce ll ular ce­ment ing substance. J . Denta l Res., 38: 7 \l , 1959.