The effects of serum and of hyaluronic acid derivatives on the action of hyaluronidase

Telephone: +44 (0)1582 763133 Web: http://www.rothamsted.ac.uk/

Rothamsted Research is a Company Limited by Guarantee Registered Office: as above. Registered in England No. 2393175. Registered Charity No. 802038. VAT No. 197 4201 51. Founded in 1843 by John Bennet Lawes.

Rothamsted Repository Download

A - Papers appearing in refereed journals

Hadidian, Z. and Pirie, N. W. 1948. The effects of serum and of

hyaluronic acid derivatives on the action of hyaluronidase. Biochemical

Journal. 42 (2), pp. 266-274.

The publisher's version can be accessed at:

•

https://dx.doi.org/10.1042/bj0420266

The output can be accessed at:

https://repository.rothamsted.ac.uk/item/8wywz

.

© 1 January 1948, Portland Press Ltd.

I948

The Effects of Serum

and of Hyaluronic Acid Derivatives

on

the

Action

of

Hyaluronidase

By Z. HADIDIAN AND N. W. PIRIE*, The WorcesterFoundation for Experimental Biology, Shrew8bury, Massachusetts, and the Department of Physiology, Tufts MedicalSchool

(Received 26 June 1947)

In thepreceding paper(Hadidian&Pirie, 1948) the viscosity of hyaluronic acidpreparations inadefined environment was used as a criterion in selecting standard preparations. In this paper the loss of viscosity by such a preparation has been used to follow the course of hyaluronidase action. This method has the primary advantage that loss of viscosity appearstobe the firstchange that can be recognized during the enzymic decomposition of hyaluronic acid. Factors influencing this in vitro changearemorelikely,therefore,toaffect the action ofhyaluronidase in vivo than those influencing the later stages of the decomposition of the molecule. It is clear that severalenzymesareinvolved in the hydrolysis of hyaluronic acid to reducing sugars. ThusHahn (1945)haseffectedapartial separation of the enzyme responsible foroligosaccharide pro-duction from thatresponsible for monosaccharide production; and Rogers (1946b) finds differences between theproducts of the reactioncatalyzed by enzymes fromdifferentsources. Heinterprets this

asevidence that the enzymes studied by Hahn are present indifferentratios in thesesourcesand points

outthatit isimprobablethateither enzymebrings about the reduction inviscosity.

Several inhibitors are known for the action of hyaluronidase on hyaluronic acid, whether this is measuredbyfall inviscosity,mucin-clot prevention

or appearance of acetylglucosamine. McClean & Hale(1941) andHobby, Dawson, Meyer& Chaffee (1941)found that theseraofanimals, that had been injected withhyaluronidase from different sources, inhibited the action of the enzyme used for immuni-zation butdidnotinhibit the action of other hyaluro-nidases. Later,however, McClean(1942)mentioned briefly an inhibitor present in normal serum, and this substance has also beeninvestigated by Haas (1946). Heparin, gastric mucin, chondroitin sul-phuric acid and partly depolymerized hyaluronic

acidalso inhibit theenzyme. McClean (1942) sug-gests that theseglucosamine-containingsubstances

act competitively because of their structural similarity tohyaluronic acid. Workingonthis sug-gestionwehave madesomeinhibitorsby modifying hyaluronic acid. We also record some observations

onthe inhibitorpresentinserum.

* Onleave ofabsence from Rothamsted Experimental Station,Harpenden,Herts.

EXPERIMENTAL

General

Viscositymeasurements weremadebythetechnique used in the preceding paper(Hadidian&Pirie, 1948). The time taken for the viscosityto fall half way from the initial to thepresumed final value has been takenas ameasureof the rateof enzyme action.This measurement of'half-time' is conventional in studies on this enzyme and has been adopted forconvenience, although it isclear that it has

noparticular merit. It does not, forexample, correspond to theloss of viscosityby half of thehyaluronicacidpresent butto asmallerproportionthanthis; andto aproportion that varies inversely with the original viscosity of the hyaluronic acid. This is due to the non-linearity of the relation between the concentration and the viscosity of

ahyaluronic acidsolution,andtothenon-linearityof the relation between the axial ratios ofelongatedparticles and the viscosftyof their solutions. The enzymeaction weare studyingcanbepicturedin two extremeways.The enzyme may attack aparticle ofhyaluronic acid andproceed to

breaklinkagesinthatparticleuntiltheproductshavelittle

or noeffectonthe viscosityof the solution;thehalf-time wouldcorrespondtoastatewhere less than halfthe particles have beenaffected because of the firstnon-linearity. On the otherhand, itmay breakall theparticlestosomeextent before anyoneparticlehas beensofarreducedastohave littleeffectontheviscosity of thesolution;because of the secondnon-linearity the half-time wouldcorrespondtothe breaking of less than half the relevantlinkages.

Some confusionis alsointroduced into the use of

half-time measurements instudyinginhibitionreactionsbythe viscosityof the inhibitors. In manycasestheviscosity of theinhibitorisunaffectedbytheenzyme; the finalvaluefor the endof theaction cantherefore be takenastheviscosity of the inhibitor solutioninthe saltmixtureused. Insome

cases, however, the sum of the increments in relative viscosity, duetohyaluronicacid and inhibitorseparately,is

less than theobservedincrements when theyarepresent

together.Thiseffect,whenitoccurs,issmall andaslongas

no significance isattachedtodifferencesin half-time less than30%,itisunlikelytoalter theconclusions.

Measurementsof therate ofloss ofviscosityhave been

madeon anextensiverange ofhyaluronicacidpreparations

underidenticalconditionsandwiththesameenzyme pre-paration. The half-times have varied by afactor often.

Part of this variation is due to differences in the initial

viscositiesof thepreparations. Ifdifferencesinviscosityare

due todifferencesintheaveragelengthof theparticlesin

a preparation,thehalf-timecorrespondstothebreakingof a smaller number oflinkageswhen theparticlesare long

usthat umbilical cord may contain substances that activate or inhibit the action of hyaluronidase on hyaluronic acid, and that these may have been removed to varying degrees during the purification. The fact that viscous preparations are the most uniform in their half-timeis in favour of this interpretation. We have tested many of the products

dis-cardedduring the purification and have found, in agreement with other workers, that fractions of very low viscosity are inhibitors. No fraction giving activation has been con-sistently preparable.

Preparation of

inhibitor8

Nitration of hyaluronic

acid

and

8ubstances

related

to it.

Dry hyaluronic acid, with the loose, open structure it has

afterdrying by sublimation, is cooled in a freezing mixture to about -100. Sixty parts of fuming

HNO3

(sp.gr. 1.59), also cooled to the same temperature, are added and the mixture is maintained at -100 for 15 min. with occasional

stirring. All the hyaluronic acid dissolves in a few minutes. Ice (300-400 parts) is now added with assiduous stirring. The temperature falls still further and a white curd separates from abrilliantblue fluid. As soon as most of the ice has

meltedand all pockets of undiluted

HNO3

in the curd have been broken up with a rod, the mixture is centrifuged for afew minutes and the

solid

is washed 3-4 times on the centrifuge with ice water. It is kneaded carefully each time toget rid of as much

HNO3

as possible, and the washing is continued until the wash water can be neutralized by a drop or two of 0- 1 -NaOH. The

solid

is now suspended in

dis-tilledwater and brought into solution by adding sufficient NaOH to bringthe pH to 6-7. This pale yellow, viscous fluid will be referred to as acid-insoluble nitrate.

All the supernatant fluids are mixed and kept cold. They are neutralized with 5N-NaOH with vigorous stirring and

coolingsothat the temperature never rises above 200 during the neutralization. The fluid, at pH 6, is

concentrated

on a hot plate in a draught of air until

NaNO3

begins to

crystallize out. It isthen dialyzed for 10-20 hr. to remove most of the nitrate, concentratedagain,and redialyzed. At this stage thesolutionfrom 100 mg. of hyaluronic acid has avolume of 2-3 ml. The solution remaining after thorough

dialysiswill be referred to as acid-soluble nitrate. The ratio of acid-soluble to acid-insoluble nitrate varies from preparation to preparation. When hyaluronic acid of low viscosity (e.g. with a relative viscosity of 1-4 for a1 g./l.

solution measured in the presence of salt) is used as the starting material there is

little

acid-insoluble product. When even less viscous fractions are used there is no acid-insoluble material at all. It seemed possible, therefore, that

all

the acid-soluble material was derived from the less viscous components of an

initially

non-homogeneous hyaluronic acid. It isunlikely that this is the sole origin; for if the acid-insoluble nitrate is dried by

sublimation

and nitrated again by themethoddescribed, a further quantity of soluble nitrate is formed. Some degradation of the molecule during nitration is to be expected, and this explanation of the origin of some of the soluble nitrate is borne out by the increased proportion of product in that form if the nitration is continued for longer than 15 min. The sum of the weights of nitrate in both forms is about equal to the starting weight of hyaluronic acid and, when highly viscous hyaluronic acid is used, 85%is in the acid-insoluble form.

Nitration always leads to some fall in the viscosity of hyaluronic acid. The extent of this fall is

variable,

but in

isabout half that of the

parent

hyaluronic

acid.

The

acid-soluble

nitrate

is

much

less

viscous.

Nitrated

hyaluronic

acid,

of both

types,

can besaltedout from neutralsolution

by

(NH4)2SO4.

With the

acid-insoluble

product,

pre-cipitation

begins

at about one tenthof

saturation

and is

complete

at one third of saturation. Themorereadily

pre-cipitated

fractions

are the more

viscous,

and they are

precipitated

fully

by

lower

concentrations

of

(NH4)2SO4

than the less

readily

precipitated

fractions.

It is clear,

therefore,

thatthenitrateis anon-homogeneous mixture, butit is not certain whether thisreflects

an

original

non-homogeneity

of the

parent

hyaluronic

acid

or

is

due to destruction

during

thenitration.

Precipitation

ofthe acid-soluble nitrates and of nitratesderived

from

hyaluronicacid

preparations

of

very

lowinitial

viscosity

does notstartuntil

one third saturation

by

(NH4)2SO4.

Salting out

does not therefore seem to be a useful

alternative

to

neutralization,

evaporation

and

dialysis

in the

preparation

of acid-soluble nitrates

because,

although

acid-insoluble nitrate is

pre-cipitated

by

saturated

NaNO,,,

the soluble

nitrates

are not, and the

high

nitrate concentration confuses

(NH4)2SO4

precipitation.

Hyaluronic

acid is not salted out

by

(NH4)2SO4

from

solutions

containing

no other

component,

but,

asone of the methods

suggested

for the

purification

shows,

it can be

salted out if

pyridine

or some forms of

protein

are also

present.

The nitrates behave

similarly,

for,

although

esterifi-cation of

the

hydroxyl groups

hasmade

the

nitrates

suffi-ciently hydrophobic

to be salted

out,

they'an

be

salted

out

at lower concentrations of

(NH4)2SO4

in the presence of

pyridine.

There

is no reason to doubt that the

hyaluronic

acidhas

nitrated as other

polysaccharides

do

(cf.

Speiser & Eddy,

1946);

for the

dry product

burnsin aflashandthe nitrogen

content

(Dumas)

hasrisenfrom

3-4

to

8-6 %.

Noestimate

other

than the measurement of

nitrogen

contenthas been made of the number of nitrate

groups

introduced. If

hyaluronic

acid has the structure

generally

assigned

toit, a

chain of

alternate

glucuronic

acid and

acetylglucosamine

residues

connected by glucoside

linkages,

there arefourfree

hydroxyl

groups

for each

repeating

disaccharideunit. The

sodium

salt of the tetranitrate that could be madefromsuch

a

substance wouldcontain12

%

of

nitrogen

andthatofthe

dinitrate

8-55%.

Esterification of the

hydroxyl

groupsin

this

way

does

not

apparently

affecttheN-acetylgroup,for

the

acetyl

content

falls

from

10-11

to7-8%.Thevalue to

be

expected

for thedinitrate is 8-7

%.

Controlexperiments show that neithernitrous nornitricacids, inamounts

com-parable

to those

present

ina

hydrolysate,

areestimated as

volatile acid after

hydrolysis

and

distillation

by the

pro-cedures used

on these

samples.

On the other handdirect

estimation,

by

the Elson &

Morgan

(1933)

method, on the

nitrated

material showsan

apparent

disappearance ofthe

glucosamine.

This result is due to the action of the HNO8

liberated

during hydrolysis.

The samedisappearancecan be

brought

about

by hydrolyzing

normal hyaluronic acid in

the

presence

of

a

quarter

ofits

weight

of

NaNO3,

anamount

corresponding

to four nitrate

groups.

Furthermore,

de-esterification,

by

the method to bedescribedlater,restores

the

apparent

glucosamine

content,

and hydrolysis in the

presence

of

a

reducing

agent

suchas HI

protects

the

glucos-amine to

a

great

extent from destruction. Fuming

HNO3

probably

acts to some extent as an oxidizing as well as

28Z.

HADIDIAN

AND N. W.

PIRIE

the glucosamineisgoingon to anextent atall comparable

to thatfound by Skanse & Sandblad (1943)toaccompany peroxideoxidation.

Acetylation ofhyaluronicacid

Like some other acidpolysaccharides (e.g. alginicacid, pectinand agar)hyaluronicacid doesnotacetylatereadily. Aceticanhydrideinneutraloralkaline aqueoussolution,or

in the presence ofpyridine under anhydrous conditions,

failed to bringaboutsignificantacetylationin afew hours at room temperature or on slight warming. Two recent

methods are,however, effective;inonethe acidisswollen withaceticacid andH2S04isusedascatalyst(Chamberlain,

Cunningham&Speakman, 1946),and in the other the acid is swollen with formamide andacetylated withacetic

an-hydride andpyridine(Carson&Maclay,1946).Theproducts

of these two acetylations differ somewhat in their pro-perties.

(1) Hyaluronic acidisswollen withaceticacidby adding

glacial acetic acidinathinstreamto a3-10g./l. solution. Mixingiscompleteuntil about 2 vol. have beenadded,after which precipitation begins and is substantially complete

with4 or5 vol. Themixture iscentrifugedandthesolid

washed with acetic acid and usedimmediately. This pre-cipitation doesnotleadtoanysignificantfractionationof the hyaluronic acid, for only about 3% of the starting

materialisfoundinthe solution when this isevaporatedto

dryness. This soluble fraction has the normalnitrogenand glucosamine content, -but its viscosity is low. The pre-cipitate,onthe otherhand,ifdialyzed immediately, gives

asolution with theoriginalviscosity.We donot,therefore,

agree with the statement made by many workers on

hyaluronic acid thatprecipitationwithacetic.acidleadsto

afallinviscosity; butourobservations arein agreement

with thoseofMeyer,Smyth&Dawson(1939)whoprepared

arelativelyviscousproduct fromsynovialfluid

by

acetic acidprecipitation.

The fibrous clot ofhyaluronicacid ispressedasfree from acetic acidaspossibleandaddedto200 times itsweightof

a mixture of 1 part of conc. H2S04, 90 parts of acetic

anhydride and 120iparts of benzene. There islittle or no

solution of the hyaluronic acid, and, after a suitable

interval,themixtureiscentrifugedand the solidsuspended

in wateranddialyzed. Ifacetylationisallowedto

proceed

forafewminutesonly,theproducthasanacetylcontent of 16-20%;after40 min.at220 this risesto25-33%and isnot

increasedbylonger acetylation atthat temperature.

By

following the method of Chamberlain et al. (1946) and raising thetemperatureto600 for 15min.,after

acetylation

hasproceeded for1hr.at 250, the acetylcontent can be raisedto36%. Acetyl determinationsweremade

by

the method usedinthe

preceding

paper

(Hadidian

&Pirie,

1948)

and include both0-acetyl and N-acetyl. The theoretical value for the tri-O-acetyl derivative ofa

polysaccharide

built up from equal quantitiesofglucuronic acid and

N-acetyl glucosamine is 34% and for the tetra-0-acetyl

derivative 394%. It would therefore appearthat,asin the

case ofnitration, esterification remainsincomplete.

Pro-ductsthat have been

acetylated

forafew hoursonlyare

stillsolublein wateroverthe whole

pH

range, but those that have beenacetylatedatroomtemperaturefor many

hours, orhave beenheatedto 600during acetylation, are notsoluble onthe acid side ofneutrality. There is a

pro-gressive fall in viscosity during acetylation. Products

acetylated for a few minutes only retain 70-80% of the original viscosity increment, whereas the tri-O-acetates retain only10%or less.

(2) Hyaluronicacid that has been dried bysublimation

does not change its appearance when left for a few hours at

200in formamide, but at400itswellssomewhat and at500

turns in 1 hr. to a viscous dough when suspended in ten times its weight of formamide. This dough can be mixed with an equal volume ofpyridine by stirring for a few minutes. The plastic, rather than fluid, mixture is now mixed with half its volume of acetic anhydride, which causes an initial shrinkage, but is incorporated into the mixture in a few minutes. As in the other. method of acetylation the properties ofthe final product depend on the duration of acetylation and the temperature at which it proceeds. The mixture is stirred with three times its weight of ice; when it is brought to pH 2 with 5w-HCI

a soft curd separates. Thefluid is dialyzed. If acetylation has proceeded for 1 hr. at

200

this acid-soluble product has an acetyl content of14-16%, while after 3 hr. at300 it rises to 30%. In neither case has it any significant viscosity. The curd is not soluble at neutrality in the absence of salt and is only partly soluble in the presence of salt, but it swells slowly to give a suspension of gelatinous particles, which dissolve- at pH 9-10. The solutions have only one fifth of the viscosity of the parent hyaluronic acid. All these acetylated fractions coagulate with strong acid and, like the nitrates, can be precipitated with

(NH4)2S04.

De-esterification. Thenitrates and acetates of hyaluronic

acid can be partly esterified without extensive de-gradation by exposure to alkali in the presence of ethanol. Solutions of either ester containing 3-10 g./l. are mixed with 10 vol. of ethanol. With the esters, unlike with hyaluronic acid itself,noprecipitate separates even on the addition of a salt such aspotassium acetate. The addition of ethanolic KOH at 220 leads to the appearance of an opalescence, and later, afteratimeinterval depending on theamount of alkali added, ofaprecipitate. For example,

an ethanolic solution containing 0-8 g./l. of a moderately viscous nitrate (rel. visc. 2-3 at 1 g./l. in salt) gave an immediate precipitatewith 7 g./l. KOH; but with 2-5 g./l. precipitation took 2hr., and with 0 7 g./l. precipitation occurredonly on the addition of some salt. Similarly with theacetatethereisimmediate precipitation with 1-2 g./l. KOH, but with less than this amount precipitation is delayed. Under these conditions de-esterification remains incomplete as determined from the nitrogen and acetyl content of these products. The immediate purpose of the work, the making of substances that inhibit hyaluronidase is, however, satisfied; and conditions that would lead to complete de-esterification have not been investigated.

Factors

affectinq

the

hyaluronidase-hyaluronic

acid reaction

Effectofsaltconcentrationonthe rate ofreaction. In studying the effect of possible inhibitors and

acti-vators on the hyaluronidase-hyaluronic acid re-action, it would first be necessary to establish con-ditions of saltconcentration under which such work

can be carried out with the least possible inter-ference from salt effects. Results of Robertson, Ropes&Bauer(1940), McClean&Hale (1941), and

optimmum salt concentration for thereduction of the viscosity of hyaluronic acid by hyaluronidase, but the optima theyfind vary with their different

sub-strateand enzyme preparations. We have tried the effect on this reaction of several salts, previously studied for their viscosity-reducing activity, with results shownin Fig. 1. These reactions were carried

oNaCl

1-4 oKCI

oNH4C1

1-2 0

/M0CI2

QCaCI2

1-0 0

SodiuIm phosphIate

eNa2SO4

0-8

0.6

0-4

0-2

0-1 0-2 0-3

Molarityof salt

Fig. 1. Variationsintherateof hydrolysis ofhyaluronic

acid by hyaluronidase with varying concentrations of differentsalts. Temp. 25°; pH 6-9±0-10; 12-5mg./l. bull-testis enzyme; 0-3g./l. solution of a medium viscous

hyaluronic acid preparation. In the experiments with

NaCl, KCI,NH4C1and

Na1SO4

thepHwasmaintained withphosphatebuffer, and in those withMgCl1 andCaCl2

with borate buffer. Inallcasesthemolar ratio of saltto

buffer was 100. In expressing the relative rates, the

reciprocal ofthe half-time in 0-125M-NaCl is taken as

unity.

outatpH 7-0,using salt solutions buffered with 1 % of phosphate or borate buffer. 'Of necessity, pH control in low salt concentrations was not quite adequate, but frequent checks showed that variation

was not morethan 0-2 unit; this mayproduce as

much as 20% variation in the rate of reaction. From theresults shown and from othersondifferent substrates, three groups are distinguishable:

CaCl2

and MgCl2, effective in very low concentrations; NaCl,KC1 and

NH4CL,

justaseffective but inhigher concentrations; Na2SO4 and Na2HPO4, relatively ineffective. Sodiumacetateis intermediate between the chloride andphosphate. Lineweaver & Ballou (1945), in their study ofthe effect of salt

concen-trationontheactivity ofpectinesterase,find similar variations in the effectiveness of various salts in activating the enzymic reaction, but with that enzyme there is the samemaximum activation at

with McClean & Hale (1941) that the chloride ion is an important activating agent. However, the difference in the effectiveness of the chlorides cannot be attributed to Cl- alone. Expressed as normality of C1-, assuming complete ionization, the optima with Mg++, Ca++,

Na4,

NH+ and K+ are at 0-05, 0-06, 0-1 and 0-15 respectively. Furthermore, in the presence of0-05M-phosphate, maximumactivationis obtained with 0-05m-NaCl. Two factors seem to be involved; general salt concentration and the presence of an activating ion, such asCl-.ThepH optimum in the presence of0-05m-NaCl and0-05M-phosphate is at 6-3; this value is intermediate between the optima of 6-8 and 5-0 found by McClean (1943) in 0 017M-and 0-167M-NaCl respectively. Our experiments have been carried out at pH 7 as a compromise between the pH optimum and the pHatwhich the enzymeand inhibitors areused in in vivo tests that willbe describedelsewhere.

Estimationof inhibitoryactivity. Sincenomethod exists for direct estimation ofhyaluronidase, the course of the reactionbetweenit andaninhibitor has to befollowed by the indirect methodof determining theamount ofhyaluroni-dase that is made unavailable for the hyaluronic acid-hyaluronidase reactioninthepresenceof the inhibitor or afterthe hyaluronidase has beenexposed to the action of theinhibitor foragivenperiod of time. If, then,oneworks in a region where there is a linear relationship between hyaluronidase concentration and therateof loss of viscosity by hyaluronic acid, one can justifiably consider the rate as adirectmeasureof theamountof activehyaluronidase

inthe system.The rangeoverwhich sucharelationexists

varieswithdifferent preparations ofhyaluronicacid tested

atthesameconcentration.Withhighlyviscouspreparations andwith thepartiallypurified enzymepreparationthatwe

use,deviation fromlinearity isslightupto anenzyme

con-centration of 8mg./l. With less viscous preparations this range may extend upto50mg./l. Thefollowingmethods

areusedinestimatinginhibitory action: in instances where thelengthof exposure of the enzymetothe inhibitorisof little or no consequence, the enzyme is introduced into amixturecontaining both inhibitor andsubstrate,and the subsequent viscosity changesfollowed;where theextentof

inhibitionisdependentontheduration of exposure of the enzyme to the inhibitor, as in the case of inhibitionby

serum,the enzyme and the inhibitorareincubatedtogether

underthe proper salt andpH conditions foragivenlength

of time, then the substrateisintroducedinto themixture,

and the course of the reaction followed. The extent of inhibitionisdeterminedfrom theratioofthehalf-timeof the reactionwith inhibitortothat without. Thusathreefold inhibition would be obtained ifahalf-time of600sec. is

obtainedwith inhibitorascompared with200 sec.without. Thereciprocal of this ratio is the fraction of the initial enzyme concentration thatisactive. Partiallypurified

bull-testisenzyme(Schering)wasusedinalltheseexperiments. Inhibitionbyhyaluronicacidesters. Nitrates and

acetates that have beenmade from viscous

hyal-uronic acid retain muchof their initial

viscosity.

The

Z.

HADIDIAN AND N. W. PIRIE

tested in the usual way. The viscosity of nitrates is notreduced by hyaluronidase and the viscosity of acetatesoftenfalls at first and then stays constant. We look on acetates behaving in this way as

mixtures of material that can be attacked and material that is not attacked and doesnotinhibit thereaction. The stability of these hyaluronic acid esters towards the enzyme contrasts with the sus-ceptibility to attack of the p-aminobenzyl ether (Humphrey, 1943). In Humphrey's preparations, however, less than one ether linkage seems to be present for each glucosamine in the hyaluronic acid, and no indication is given of the extent to which hydrolysis, which was recognized by the appearance ofN-acetylglucosamine, took place.

withnitric esters aremoresystematic; all of them inhibit. Attempts to get less nitration by using

asmaller ratioofnitric acidtohyaluronicacidorby allowingthenitration toproceedforashortertime have led to less active products. The acid-soluble fractionsfrom the nitration of viscous materialand theacid-soluble nitrates of non-viscous materialare

strongerthan theacid insoluble nitrates. By suit-ably controlled removal of nitrate groups with alkali and ethanol the inhibitory power is sub-stantially increased; but it is clear from the table that the intensity of the alkalitreatmentis of im-portanceand that toodrastic a treatment leadsto aweaker inhibitor. Forcomparison the results of experiments withnon-viscous hyaluronic acidand

Table 1. Effects of various inhibitors on the hyaluronidase-hyaluronic acid reaction

Descriptionofinhibitor

Viscoushyaluronic acid acetylated by sulphuric method for 60 min.; 33 % acetyl Above product deacetylated for 15 min. with 2 g. KOH/I.; 14% acetyl

Viscoushyaluronic acid acetylated by sulphuric method for 10 min.; 18 % acetyl Viscous hyaluronic acid acetylated by formamide method for various times;

15,25 or30% acetyl

Acid-solubleproduct from nitration of viscous hyaluronic acid

Acid-insolublepart of the samenitration

Acid-soluble product from nitration of viscous hyaluronic acid Acid-insolublepart of the same nitration

Aboveinsoluble product denitrated for 2 hr. with: 0 7 g. KOH/I. 2-5 g.KOH/L.

7-5g.KOH/I. Nitrated non-viscoushyaluronic acid

Thesamenon-viscoushyaluronic acid before nitration Heparin

Concentration

(g./l.)

0.1

0.1 0-1 0-1

0-01 0-1 003

0*03

0005

0004 003

0-01

0*1

0-1

Inhibition

1-5 6 4 Inactive

3-5

7-3

30 2-2 5 7-5 2-2 5 2 2

Inthese experiments2ml. of a0-6 g.l. solution of hyaluronic acid in 0IM-NaCland0-Im-phosphatebuffer pH 7 were mixed with 1ml. ofasolution of theinhibitor atfour times the concentration specified in the second column. 1 ml. ofa50mg./l.solution of hyaluronidase was then added and the flow time in an Ostwald viscosimeter was measured at

variousintervals. Thehalf-time was foundinthe usual way, and the amount of inhibition set out in the third column

isderivedby dividing the half-time foundinthe presence of the inhibitor by that found in a control experiment with enzyme andsubstrate alone.

The inhibition is measured by mixing suitable

amountsof inhibitor withhyaluronic acid and then addingtheenzyme. There isaslight increase in the inhibition if the enzyme and inhibitor are mixed first, butin this instance theextentof inhibition is

notincreased by allowing the enzyme and ester to bein contactformorethan the time necessary for mixing. Some comparable results are collected in

Table 1. Valuesare givenforonly oneconcentration of each inhibitor because the inhibition is pro-portionaltotheconcentration.Thetableshowsthat

acetatesinhibit weakly,ornot atall, and the most active are partially acetylated products the pre-paration of which isamatter ofsomeuncertainty. On partial de-acetylation the inhibition always increases, but conditionsthatwouldlead toregularly repeatableresults havenotbeen found. The results

heparin are also included in the table. Many

non-viscous preparations of varied origin have been tested andnonehas been found withasignificantly greater inhibitory power than the one cited. Only

one sample of heparin was tried; the inhibition found hereiscomparabletothatfoundby McClean (1942) and Rogers (1946a) using different experi-mentalprocedures. Ourtests weremadeat250with hyaluronicacidhavingarelativeviscosityof7-8for

a1 g./l. solution in005M-NaCl and

0.05M-phosphate

bufferpH7,althoughexperiments described inthe next paragraph show thatthere is more complete inhibitionwithsmallerconcentrations ofphosphate. The same sample of dried, partially purified hyaluronidase from bulltestes (Schering) wasused throughoutat12-5mg./l. Similarexperimentshave

been made withlargerand smallerconcentrations of

having a relativeviscosity as low as 2 for the 1 g./l. solution. These variations in the conditions have not significantly affected the amount ofinhibition observed.

Variations in phosphate and NaCl concentration, however, have a marked effect on inhibition by the esters. Several of these, including an acid-soluble nitrate, an acetate, a denitrated nitrate and a de-acetylated acetate,-were tested in0-05M-NaClwith phosphate concentrations varying between 0-013 and024M; and in 0 05m-phosphate with NaCl con-centration varying within thesame range. Within this range there is an increasing inhibitory activity with decreasing salt concentration. None shows a significant inhibitory activity in 005M-NaCl and

024M-phosphate and all show maximum activity in.

005M-NaCl and 0 013M-phosphate; but they all follow different courses between these extremes. Products with identical inhibitory activity under

our standard testing conditions may differ by a factor of 2, at lower phosphate concentrations. Thus the figures given in Table 1 would not necessarily indicate the relative effectiveness of these substances underphysiological conditions.

For convenience of writing, the operations that have been performed on hyaluronic acid in making theseinhibitors have been called acetylation, nitra-tion and de-esterificanitra-tion. It is certain that the bulk of the material is undergoing these changes, for the observed acetyl content can be increased and diminishedas canthe Dumasnitrogencontentand theinflammability of the nitrates. It is less certain that these are the changesprimarily responsible for the inhibition. All of them are associated with changes in viscosity andsopresumably in particle size, and it is possible that this is the significant change rather than theextentof esterification. The difficulty found in getting reproducible results from experiments on acetylation and de-acetylation slightly favours this interpretation. On the other hand, the stabilityof these inhibitors in the presence ofhyaluronidase suggeststhat agraded breakdown of the molecule isnotalonesufficienttomakeagood inhibitor. Furthermore, our attempts to degrade the molecule by othermeans, e.g. acid, alkali and partial enzyme digestion, have resulted in only feebly active products. It is probable that in-hibitors have both anoptimum particle size andan

optimurit degree of substitution. Thehigh activity foundwithnitrated non-viscous products suggests thatthissize may befairlysmall; but active material doesnotdiffuse through cellophan.

Serum as an inhibitor. McClean (1942), in the

course of studying the in vivo decapsulation of streptococci, discovered a factor inthe seraofseveral

specieswhichinhibitsenzyme from different sources, andnoted that they variedgreatlyin theirinhibitory

inhibition is due to the presence in the serum of a polysaccharide which acts by competing with the substrate for the enzyme. Haas (1946), on the basis of some by no means conclusive evidence, has claimed that this inhibitory action of serum is due

tothe presencein it of an enzyme which inactivates hyaluronidase.

Theinhibition of hyaluronidase by serum at 250 proceeds at a measurable rate, as shownby Haas (1946). The method used for itsstudy by Haas, and with somemodification by us, has beentoexposethe enzyme to theaction ofserumforagivenlength of time, then mix it withhyaluronic acid solution con-taining enough phosphate buffer to obtain a final concentrationof 0 05-0-1M-phosphate. The validity of this method restsupon the fact, discovered by Haas andconfirmed by us, that in the presence of such concentrations of phosphate (the exact

con-centration varieswith differentserumsamples) the activity of hyaluronidase in reducing the viscosity ofhyaluronic acid is unaltered byincubationwith

serum. One cannot, however, assume on such

evidence that interaction between hyaluronidase and the serum factorceasesin the presence of phos-phate.

If one introduces phosphate into a mixture of hyaluronidase andserum that hasbeen incubated for a time sufficient to inactivate a part of the hyaluronidase, and thenremovessamplesatvarious intervals for testing their hyaluronidase content, one finds that in low phosphate concentrations (001-0.05M) the inhibition of hyaluronidase will continue, and in high phosphate concentrations

(005-0.1

M)itwill bereversed. However, atabout 0-05Mtherateofchangeineitherdirection is very slow. Since the process ofestimatingthe hyaluroni-dase after the inactivation reaction seldom lasts longer than 30min., this method of studying the inhibitory action ofserum seems valid so long as

significance is not attached to small changes in activity.

Like the reaction between hyaluronidase and hyaluronicacid,the reaction ofdialyzedserumwith hyaluronidase does not proceed in the absence of salt. It is activated by small amounts, with an

optimum between 0-03 and 01m-NaCl. The exact

optimum varies somewhat with the amount of

serumused, but variations within the rangespecified

areslight. In0-2M-NaCltheinactivation of hyaluro-nidasebyserumattheendof 10min. ofincubation is aboutone fifth ofthemaximum. Phosphatewill

Z.

HADIDIAN

AND N.

W. PIRIE

the serum-hyaluronidase reaction is apparently in-hibited, 0*075M-phosphate giving full inhibition. The serum used in these experiments was dialyzed for 4-5 days against several changes of distilled water. For the most part we have used pig serum in this study, but most of the experiments described

Scrtim(nil.)

o004

o0o10

* 0-20

10 20

Timeofincubation (min.)

Fig. 2. Rate ofinactivationof hyaluronidase byvarying

amounts ofserum. 0 04, 01 and 0-2ml. dialyzed pig

serum, pH 7; incubation at 25° with 50 /Ag. oftestis

enzymein2mLof 0-05N-NaClcontaining0.3ml. borate

buffer pH 6-9. 2ml. of 0-6g./l. solution of a -high

viscosity hyaluronic acid in01IM-phosphate buffer and

005M-NaCl added tothe serum-hyaluronidase mixture attheendofincubation and theresidual hyaluronidase

activity determined viscosimetrically. Hyaluronidase half-timewithout inhibition

activity

= half-time with inhibition .

Owing

to the

non-linearity ofthe relationship between enzyme

con-centration and therateof hydrolysis of this particular

hyaluronic acid preparation the actual amount of

hyaluronidase, expressed as fraction of the original

amountremaining active after incubationis aboutathird

morethan theactivityindicated.

._

'.

-5 4a

Ca ._

Go

0

'0

S'4

have been repeated with human, rabbit, chicken or bovine serum. In view ofthestatementofHaas (1946) that serum obtained by clotting is less active than that obtained from blood defibrinated by stirring, all sera used were obtained in the latter manner.

However, in subsequent experiments with sera

obtainedby clotting and by stirring from thesame

specimen of blood,wehave beenrepeatedly unable

to demonstrate under optimal salt conditions any significantdifference between thetwo.

As aresult of these findings, the following pro-cedure was adopted for testing the inhibitory activity of serum, with such modifications aswere

required by varying conditions. The enzyme (50

pg.)

isincubated with 0-05-0 3 ml. ofdialyzed,

neutra-lized serum for 10 min. in 2 ml. of 0.1M-NaCl buffered with0*2 ml. of 0 2M-borate bufferatpH6-9. To this is then added 2 ml. of hyaluronic acid solution, 0 6g./l. in 0OIM-phosphate, and the course of the depolymerization of hyaluronic acid followed vis-cosimetrically.

Having established beforehand the range of enzymeconcentration over which therelationship between hyaluronidase concentration and therate

ofloss of viscosity by thehyaluronicacid prepara-tiontobe used doesnotdeviatetoomarkedlyfrom linearity, one can follow the course of apparent inactivation of the enzyme by the serum. At 25° with 50 ug. of enzyme and 0 10 ml. ofdialyzed pig

serumin 2 ml. ofbuffered 005M-NaCl, onlyabout 50

%

of the enzyme remains active 30 sec. after mixing, and 25

%

after 1 min. (Fig. 2). Maximum inhibition is reached within 10-500 min.depending upon the serum concentration. With high concen-trations ofserumforthenext5-10 hr.noapparent change occurs,buteventually,in everyinstance the hyaluronidase activity begins to return (Fig. 3).

Sertim(ml.)

0-00

O004

o010

*020

5000

Timeofincubation(min.)

Fig.3. Recovery of hyaluronidaseactivityafteritsinhibition by varyingamountsofserum. Conditionsofexperiment

arethesame asspecifiedunder Fig.2, where thecourseof the reaction for the first 35 min. isshown indetail. Serum

incubated under thesameconditions, but without hyaluronidase, had lostnoneofits inhibitoryactivityattheernd

ofthe experiment when testedon afreshenzymesolution.

O.1

.,

I948

recovery of hyaluronidase activity immediately follows itsmaximum inhibition. When there is an

excess, apparently a condition of equilibrium is

reached and is maintained as long as there is an excessofthe serumfactor available. The recovery

process in every instance is slow, and is complete

within72 hr.only withlowconcentrations ofserum.

Ahyaluronidase solution kept under thesame tem-perature,salt and pHconditionsmayloseaquarter

of its activityduringthis period. Serum kept under identical conditions losesvery little, ifany, of its

original activity. In repeated experiments, the final hyaluronidase activitywasnever morethanthatat

the beginning of the experiment, thus making it unlikely that this phenomenon is due to the

pro-duction ofhyaluronidase bycontaminatingbacteria. Experimentswerealso carried outin the presence

of 005M-NaF and 005M-benzoateto prevent any

possibilityofbacterial contamination. Both

inter-fere tosomeextentwithhyaluronidase activity but

thereappearance ofhyaluronidase after inhibition

canbedemonstrated in theirpresence.

In these experiments inactivation and

re-acti-vation proceed inasolution that is not otherwise

altered. By changing the enviroument the

re-activationprocesscanbehastened. Dilution of the mixturewithoutchangein saltconcentrationcauses a15%increasein theratefor fourfold dilution.The samedilutiondoubles the rate ofre-activationif it

is madewithNaClcontaining003M-phosphateand

quadruplesit if made with NaCl containing 0

1M-phosphate.

We find thereactionbetweenserumand

hyaluro-nidasemuch faster than is manifest from the data of Haas(1946).This inpartis due to the fact that Haas

usesno salt inhisreaction mixture exceptborate

buffer and whatever salt there is present in the

serum,inpart to the fact that he hasno determina-tions at less than 5 min., and in part to his

pre-sentation of thecourseof the reactionas anincrease

in the 'activity' ofserum. Thus, according to his

data,0 1ml.chickenserumhasan'activity'of 2 at

the end of 5 min., and 4 at 20 min. The former

correspondstoaninhibition of 65%of the

hyaluro-nidase, andthelatterto80%,i.e.onlyanadditional 15% of the original hyaluronidaseis inhibited be-tween 5 and 20min. How much of the 65% in-hibition at the end of 5min.takesplaceinthefirst

few seconds there isno way ofdetermining from

Haas'sdata,butfromour owndata it is evidentthat

thereaction is slowedabruptly after about 1 min. whenan excessofserumispresent.

Theinhibition ofhyaluronidase byderivatives of

hyaluronicacid resembles theserum inhibition, in

ageneral way, in its dependence on the ionic

en-vironment, although the serum inhibition is

sup-pressed byconcentrations ofphosphateat whichall

Biochem.1948,42

actions differ in two significant respects; ester in-hibition reaches its maximum more quickly than serum inhibition and there is no recovery on pro-longed incubation. It is obvious that the hyaluronic acid esters cannot becalledenzymes, but the ter-mninology most suitable for the serum inhibitor is uncertain. Haas (1946) has called it an enzyme because the reactionproceedsatameasurablerate

andbecause the inhibitor is destroyed at

550,

but these properties, which we confirm, are not

ex-clusively the properties of enzymes and the fact that theinhibited hyaluronidase can recover its activity isastrong argumentagainst theapplicability of the word enzyme to the inhibitor. Essentially, an enzyme is a substancethat combineswithits sub-stratein such a way that, whendissociation occurs, the substrate is altered; the serum inhibitor

com-bines with hyaluronidase and dissociates from it withoutaffecting the only property ofhyaluronidase ofwhich anything is known. Under these circum-stances the

anqlogy

would seem to be with the antigen: antibodyand toxin: antitoxin reactions ratherthan with theenzyme: substrate reactions. The indiffusibility of the inhibitor and its inacti-vation by heating suggest that it is, or that it is associatedwith,aproteinand this is

bore

outby McClean's observation, which we confirm, that it precipitates in theglobulin fraction ofserum. The destruction of the inhibitor on incubation with a

hyaluronidasepreparationcould be dueto aprotease present in the enzyme, but it is alsopossible that it isanothermanifestation ofhyaluronidaseactivity. On this view the inhibitor isasubstrate for hyaluro-nidaseon which the enzymeactsslowly; it is this that distinguishes it from the esters for they are

apparentlynot attackedat all bythe enzyme. In

anattempttoget evidence for thispointof viewwe

have estimated theglucosamine inhydrolysatesof theultrafiltrate fromdialyzedserumthathadbeen incubated with hyaluronidase for sufficiently long

todestroyapartof itsinhibitorypower. No glucos-amine could be detected. This suggests that the inhibitor doesnotcontainglucosamine,thatthis is

not liberated during its destruction, or that the quantityistoosmall for detectionbyourmethods. There is therefore stillnoevidenceonthenatureof theinhibitororthemechanism of its action.

% SUMMARY

1. The effect of various salts on the action of hyaluronidaseonhyaluronicacid has been studied viscosimetrically. There isanoptimum molarityfor eachsalt,but itsvalue and therateof the reactionat

theoptimum vary with different salts.

2. Methods ofpreparingnitratesandacetatesof hyaluronic acidaregiven. These estersinhibit the

274

Z.

IIADIDIAN

AND N.

W. PIRIE

1948

enzymeaction and thisinhibition is compared with that brought about by a constituent of normal serum.

3. The extent of the inhibition by these agents is dependent on the saltconcentration.

4. The 'seruminhibitor, unlike the inhibitors made byesterifying hyaluronic acid, is destroyed by the testicular hyaluronidase used.

Thisinvestigation was aided by a grant from G. D.Searle

andCo.

REFERENCES

Carson,J. F. & Maclay, W. D. (1946). J. Amer. chem. Soc. 68, 1015.

Chamberlain, N. H., Cunningham, G. E. & Speakman, J. B. (1946). Nature,Lond., 158, 553.

Elson, L. A. & Morgan, W. T. J. (1933). Biochem. J. 27, 1824.

Haas, E. (1946). J. biol.Chem. 163, 63.

Hadidian, Z. & Pirie, N. W. (1948). Biochem. J. 42, 260. Hahn, L. (1945). Ark.Kemi Min.Geol.19A, no. 33. Hobby, G. L., Dawson, M. H., Meyer, K. & Chaffee, E.

(1941). J. exp. Med.73, 109.

Humphrey, J. H. (1943). Biochem. J. 37, 460.

Lineweaver, H. &Ballou, G. A. (1945). Arch. Biochem. 6, 273.

McClean, D. (1942). J. Path. Bact. 54, 284. McClean, D. (1943). Biochem. J. 37, 169.

McClean, D. & Hale, C. W. (1941). Biochem. J. 35, 159. Madinaveitia, J. & Quibell, T. H. H. (1941). Biochem. J.

35,456.

Meyer,K.,Smyth, E. M. & Dawson, M. H. (1939). J. biol. Chem. 128, 319.

Robertson, W. B., Ropes, M.W. &]Bauer, W. (1940). J. biol.Chem. 133, 261.

Rogers,H. J. (1946a). Biochem. J. 40, 583. Rogers,H. J. (1946b). Biochem. J. 40, 782.

Skanse, B. & Sandblad, L. (1943). Acta phy8iol. Scand. 6, 37.

Speiser,R. & Eddy,C. R. (1946). J. Amer. chem. Soc.68,287.

The Fate

of Certain

Organic

Acids and Amides in the Rabbit

3. HYDROLYSIS OF AMIDES BY ENZYMES IN VITRO

BY H. G. BRAY, SYBIL P. JAMES, BRENDA E. RYMAN AND W. V. THORPE Departmentof Phy8iology, Medical School, Univer8ity of Birmingham

(Received4July 1947)

In the first two papers of this series itwasshown that the rabbit is able tohydrolyze the carbamyl group in benzamide and phenylacetamide (Bray, Neale & Thorpe, 1946), but that the introduction of a

hydroxyl group in thepositionpara to thecarbamyl renders the latterstable, less than 7

%

hydrolysis occurring in vivo (Bray, Ryman & Thorpe, 1947). Furtherinvestigations are inprogress of theeffect ofhydroxyl groups in other positions and also of amino and nitro groups. This paper reportsaparallel study in which we have examined the activity of various enzymepreparations, especiallythosefrom rabbit liver, in hydrolyzing the carbamyl group.

Similar investigations are those of Gonnermann (1902, 1903), in which the action of several enzyme preparations, including pepsin, trypsin and sheep's liver andkidney, on a number of amides and anilides was studied, of Geddes & Hunter (1928) and of Grassmann&Mayr (1933) who examined the action ofyeastasparaginase on several amides. Thecriterion usedby Gonnermann asanindication ofhydrolysis

was the isolation of the unchanged amide or of productsof its hydrolysis, whilst Geddes & Hunter

estimated the ammonia liberated. Relevant results from theseinvestigations are summarized in Table 1. Geddes & Hunter claim that asparaginase is present in calf liver and also that yeast asparaginase will hydrolyze glutamine. Glutaminases hydrolyzing thisamide but not asparagine have been shown to be present in various tissues, including liver and kidney (Krebs, 1935).

Inthe present work we have determined quan-titatively the degree of hydrolysis by a formol titration method based on that used by Balls & Lineweaver (1939) for assessing the activity of papain with hippurylamide as substrate. We have also compared the stability of the amides used to acidandalkaline hydrolysis.

Material8 METHODS

Amide8. Theamideswhich couldnotbepurchasedwere

prepared by the action of aqueous ammonia upon the

correspoadingacidchloride or uponthemethylesterofthe parentacid. Aminobenzamides wereprepared by the