Attachment of Nontypable Haemophilus ... - Wiley Online Library

18 downloads 0 Views 1MB Size Report
sparfloxacin against. Pseudomonas aeruginosa pneumoniae in neutropenic mice. Antimicrob. Agents. Chemother. 36: 1352-1357. 15 ) Porras,. C.S.,. Svanborg,.
Microbiol. Immunol., 42 (10), 697 - 702, 1998

Attachment of Nontypable Haemophilus to Human Pharyngeal Epithelial Cells by a Ganglioside Receptor Kenji Kawakami*, Kamruddin Ahmed, Yoshiaki Utsunomiya, Kazunori Oishi, and Tsuyoshi Nagatake

influenzae Mediated

Naoto Rikitomi, Akihiro Hori,

Department of Internal Medicine, Institute of Tropical Medicine, Nagasaki Japan

University, Nagasaki, Nagasaki 852-8102,

Received June 3, 1998. Accepted July 6, 1998 Abstract:

Nontypable

infections

and

receptor nificant

Haemophilus

has

the

to which decrease

attachment

NTHI

not

hand,

decrease

the

centration glioside of

D1b

NTHI

the

receptors

Key

words:

in

attachment, in

and

D2

treatment of

ng/ml.

Treatment

Other cells

on

human

a dose-dependent of

NTHI

Haemophilus

isolates

of

with

Ml,

of NTHI

as

manner.

This

Ganglioside,

D3,

inhibited a decrease

monoclonal

indicates

epithelial

epithelial

important

colominic

T lb the

acid

resulted

of the

acid.

The acid,

Only

treatment

the

attachment.

decreased

glycoconjugate,

with On

M1

significantly after

also

cell in a sig-

receptor.

asialoganglioside

in attachment

sialic

epithelial

N-acetyl-neuraminic

attachment

antibody

that

cells

decreased and

respiratory

the

component

of fucose, and

of human

characterized

significantly

M3,

D2

showed

study

pharyngeal

an

acid

M2,

GD2

pathogens

We

1,000 ƒÊg/ml

12.5 ƒÊg/ml

ganglioside

anti-human

influenzae,

acid

with

of

major

cells.

acetyl-salicylic

Only

the

of pharyngeal

sialic

pretreated

gangliosides

NTHI.

of

epithelial

pretreatment

suggesting NTHI

is one

pharyngeal

at a concentration

with

attachment

of 12.5 D2.

(NTHI)

to

N-acetyl-galactosamine,

D1a,

other

attach

Neuraminidase

decreased

N-acetyl-glucosamine,

the

influenzae to

bind.

in NTHI

was

gangliosides

ability

did at

treatment the

not

a con-

with

gan-

attachment

GD2,

is one

of

cells.

Attachment,

Receptor

The attachment of bacteria to pharyngeal epithelial cells is prerequisite for the establishment of respiratory infection. The host epithelial cell expresses the receptor on its surface, to which the bacteria attach by its adhesin (8). Haemophilus influenzae (H. influenzae) is one of the major pathogens of respiratory infections and meningitis. The adherence of this organism to the epithelium is a critical step in the process of colonization and infection caused by nontypable H. influenzae (NTHI) (13). To understand the pathogenesis, several important studies have been done to reveal the attachment mechanism of this bacteria with respiratory epithelial cells. These studies clearly indicated that the receptor of H. influenzae resides in the sequence of gangliosides (6, 9, 10, 22). Although several studies have shown that H. influenzae causing respiratory infections are mostly nontypable (1, 11, 12, 22), till now, only one study was done to evaluate ganglioside interactions with NTHI (6). That study

showed the receptor as a minor ganglioside. Gangliosides are important constituents of the surface membrane of most cells of mammalian tissues. It is generally assumed that the gangliosides occur on the outer side of the cellular plasma membrane (23). It is recognized that gangliosides are important receptor sites for many bacteria and toxins. Interestingly, another study identifled proteoglycan as a receptor for NTHI (13). Therefore, to solve confusion, this study was done for identify the receptor of NTHI for pharyngeal epithelial cells.

Address correspondence to Dr. Kenji Kawakami, Depart-

Abbreviations: GD2, diganglioside 2; IgG3, Immunoglobulin G3; NANA, N-acetyl-neuraminicacid; NTHI, nontypable Haemo-

*

ment of Internal ki University,

Medicine, 1-12-4

Institute

Sakamoto,

of Tropical Nagasaki,

Medicine, Nagasaki

Materials and Methods Bacteria. NTHI isolated from the sputum of patients with respiratory infection were used in this study. Serotypes of H. influenzae strains were determined by type-specific antisera (Difco Laboratories, Detroit, Mich., U.S.A.). H-95-135 was the isolate mainly used in this

Nagasa-

852-8523,

philus influenzae; PB, phosphate buffer; PBS, phosphate-buffered saline; TLC, thin-layer chromatography.

Japan.

697

698

K. KAWAKAMI

experiment.

Other

H-95-120. sion

All

(BHI)

eysville,

broth Md.,

Dickinson tries,

and

Osaka,

on

ing

5%

in

a 5%

Co.),

for

BHI

agar

(Becton

Dickinson

with

were

to

0.75%

washings

hr

were pH

centrifuging

at

in

The 80 •~

were

for

10

min

of

bacteria

cells

at a concentration

with

neuraminidase

gens,

Sigma

water

bath,

room

of

Chemical and

Bacteria with

and

mmol

phos-

washed

three

at

tem-

room

Pure

D2,

min

M1,

D3

and

were

T1b and

a water

cells

treated

with

Corporation,

30

min.

1D3,

bohydrate, quantitative vitalex,

used

1 •~

108

CFU/ml

(purified

from

a

Effects At

10

min.

PB

at

of

the

chain

of

and

of

the

than

glioside

(2)

on

NAD,

BHI

and

cells

were

after

neuraminidase,

31.7•} treated

respectively.

μg/ml. The gangliosides D1a

30

a 30%

at

D1b

104 cells/ml

IgG3

However

mouse

and C

antibody

of

IgG3

resulted

of

M1

also

tested

with

the

lowest

250

T

b

Gangliosides at con(Table

did

not

M2,

Only

500

were

attachment

gangliosides

dose

a

1).

decrease

M3,

ganglioside in

effective

and

mixture T1b.

attachment

gan-

decrease

respectively

and

M1

decreased

the

64-75% 125,

and

125 ƒÊg/ml,

asialoganglioside

manner

in

decreased

and

caused

except

gangliosides

D1b,

sugars

acid)

adherence

of

the

D1a,

12.5

We

nificantly

no

N-acetyl-glucosamine,

concentrations

gangliosides

attachment.

isotype

which

significantly

of

Attachment

1,000 ƒÊg/ml,

acid,

of

Ml,

and

on and

decrease

constituents

Sigma

500

D2,

D3

D2

sig-

dose-dependent being

12.5

ng/ml.

on

Adherence

lipopolysaccharide immunotype-1

with

was agar

were

epithelial

0.001)

acetyl-salicylic

adherence

D1a,

Effects of

and

of

of

mixture,

acid,

37

treatment NTHI

Data

pharyngeal

control

colominic

more

at

Fisher

After

done

if P values

(P•ƒ

Carbohydrates

NANA,

for

C

Different

mix-

incubated

monoclonal

to

4.3%

the

5 units/ml

N-acetyl-galactosamine,

control.

bacteria.

culturing

of

brain;

5 •~

of

and

of

a concentration

37

of

antibody

aeruginosa

viability

and

(fucose,

of

M3,

concentrations

Tokyo)

was

Attachment

NTHI

22.1 •}

0.001)

were

N-acetyl-

M2,

bovine

in

on of

to

(P•ƒ 20

at

in

colominic

suspended

bacteria

(ANOVA).

significant

50 two

average.

analysis

variance

Neuraminidase

decreased

with

N-acetyl-neuraminic

M1,

an

a

on

sample,

attached as

at

on Products

each

Statistical

attachment

10.8%

PB

ganglioside fucose,

gangliosides

a mouse

as of

hemin

min

with g for

acid,

the O-side

Pseudomonas was

10

80 •~

of

monoclonal

against

Viability

for

Industries),

different

GD2

(14)

of

The

treated

at

at a concentration

(Seikagaku

isotype

of

For

statistically

10 min

0.05.

centrations

anti-human

from

than

Effects

bath.

Epithelial

for

considered

for

bacteria

counting

of

min. three

collected

Southern

presented

analysis

30 of

perfrin-

times

N-acetyl-galactosamine,

Co.),

in

C

concentrations

asialoganglioside

Chemical

37

centrifuging

Chemical

were

Clostridium

for

was

g

attached

counted.

made

cells/ml,

Results

Epithelial

cells/ml

three

acetyl-salicylic

(NANA),

D1b,

at

a concentration

glucosamine, acid

104

cells.

from

washed

different

(Wako

V Co.)

then by

at

epithelial

5 •~

(type

temperature

treated

and

The

pharyngeal

a total

were

at

male.

perature.

ture

were

cells

Treatment

(Shandon

England).

7.2,

C for

80 •~

cells

cytospin

analysis.

one-way

by

done

pH

5 X 104

at 37

at

Finally,

were

Statistical

kept

centrifuging

attachment

with

less

swab

1/15

cells

g

adult

a cotton

collected

7.2.

at 80

epithelial

healthy

with

separated

were

bac-

Bacteria

and of

were by

the

was

(PBS),

CFU/ml,

bacteria

cells

the

108

and

by

specimens

isolate.

Pharyngeal

slide

assay

concentration ratios

Astmoor,

and

a magnification in each

a normal

30

a

of

modification.

at equal

in PBS

successive

CX

operated

at

fimbriation cells.

for 100

with

saline

1 •~

at

temperature.

glass

used

acetate

Tokyo)

scrapped

swab (PB),

mixed

Ltd.,

was

a JEM

scanned

from was

the

by

14

from

grid

uranyl

Ltd.,

the

collected

oropharynx

times

for

of

Unattached

at a concentration

under

randomly

epithelial

buffer

grown

copper

examined

determine

Pharyngeal

from

contain-

attachment

15)

in phosphate-buffered

cells

viability

control.

An (3,

concentration

were

day,

the

assay.

epithelial

cul-

Bacteria

saline

(JEOL

were

phate

was

Co.)

observation.

microscope

cells

NTHI

and

a

next

with

before

suspended

Mo.,

The

compared

described

room

staining

were

Louis,

and

A carbon-coated

Bacteria

The

use.

NAD,

in normal

Specimens

cells

and

microscopic

36,000

Indusadenine

St.

overnight.

was Attachment

as

incubator.

suspended

electron

of

hemin

Cock-

Chemical

Co.,

teria

(Becton

(ƒÀ-nicotinamide

C until

negative

kV.

NAD Chemical

at -40

CFU/ml.

sec.

and

incubator

infu-

Co.,

Isovitalex

Pure

kept

CO,

109

(Wako

and

heart

and

5%

and

were

of

Dickinson

hemin

Isovitalex,

H-94-152

in brain

containing

Japan)

Electron agar

H-93-151,

stocked

(Becton

Sigma

U.S.A.),

were were

U.S.A.)

dinucleotide,

tured

isolates

isolates

ET AL

determined

containing

incubated

each

in

carby

5% a 5%

of

Anti-Human

GD2

Antibody

the

NTHI The

adherence

of

a dose-dependent

Iso-

mouse

CO2

3).

anti-human There

NTHI

manner

was

GD2 no

decrease

was

decreased

after

cells

significantly were

monoclonal in

attachment

treated

antibody

in with (Table

when

cells

RECEPTOR

OF NONTYPABLE

H. INFLUENZAE

699

Table 1. Attachment of H. influenzae to pharyngeal epithelial cells after treatment with different gangliosides

were

treated

with

monoclonal

a control

antibody

of

the

same

isotype

mouse

Electron

(1D3).

stain Effect

of

Strains We three treated to

the

gests glioside

Ganglioside

D2

on

the

Adherence

of

found

a significant isolates

of

with

1.25 ƒÊg/ml without

universality D2

for

under

Viability

control the

stains

NTHI.

decrease

in

when

these

NTHI of

the

D2 (Table

receptor

adherence

strains

ganglioside

ganglioside of

the

as 2).

sequence

Observation

were an

nonfimbriated

electron

as

revealed

by

negative

microscope.

Other

of NTHI

other

Microscopic

All

of

were

pre-

compared This in

After this ty

suggan-

study, compared

CFU/ml.

of

Bacteria

treatment there to

with was the

no control.

different significant The

carbohydrates difference viable

count

used of was

in

viabili1 •~

108

K. KAWAKAMI

700

Table

2. Attachment

epithelial

cells

of three after

strains

treatment

of H.

with

influenzae

to

1.25 ƒÊg/ml

Our

pharyngeal

ganglioside

ET AL

D2

findings

clude

with

that

The

the

GD2

from

in

in

the

originated tor

for

and

the

of

the

to con-

of

were

GD2.

purified

antibody

GD2. does

of

was

Therefore,

We

sequence

us

sequence

monoclonal

species.

in the

permit

assay

ganglioside

different lies

in

human

function

NTHI

NTHI

inhibition

against

from

of

lies

the

brain

mouse

possible

strains

receptor

used

bovine

raised

other

not

it is

vary

when

the

recep-

assume GalNAc1ƒÀ1-4

(Neu-

Acƒ¿2-8NeuAcƒ¿2-3). Several

Table 3. Attachment of H. influenzae to pharyngeal epithelial cells after treatment with mouse-anti-human gangliosides D2 and 1D3

our

features

findings.

receptor

of

could

M.

detect

(8)

may

disturb

and

tured

human

and

unidentified

with

acid

for

H.

have

influenzae

pathogens.

cells

(3).

glioside

in

Ml,

of of

that

the

identify

the

ment,

D2

was

and

D3

gangliosides decrease

binding

decreased

sequence further

share

and

all

D2,

where

bition

was cells

attachment

of

these

anti-human significantly

GD2 a

signifi-

with the and

gan-

occurring

of

sequences 4).

A

observed

for in

antibody dose-dependent

also

by

cells

manner.

of the

is due

to

the

which

receptors;

essentially by

aber-

glycosphingolipid

Another

of

bacteria,

reason

may

be

interaction

NTHI

to

epithelial

is

primarily

to

(1).

This

Apicella

et of

indicates as

that

in,

The

attachment and

NTHI found

well

as

al

(4)

reported

the

H.

influenzae

all

are

other

H.

influenzae

study

H.

in

all the

fresh

clinical

nonfimbriated.

influenzae

are

impor-

infection.

cells which

in infection

respiratory

to epithelial

facwere

influenzae

respiratory

for

do (16).

adherence

were

factors

H.

this

that,

nonfimbriated

of bacteria there

in

during

responsible

of

nonfimbrial used

NTHI

epithelium

proteins

that

nonfimbriated

80%

respiratory

as

The We

tant

process

identified

20).

were

fimbriated

surface

been 13,

attachment

Interestingly,

attachment

have (5,

isolates,

decreased

idenone

identifying

adhesin-receptor

attachment

High-molecular-weight

sputum

epithe-

the

of

and

characterized of

(7).

sets

cells

experiments,

of are

surface

fimbriae.

facilitate

nonfimbriated.

inhiof

cell

Gram-negative

with

ganglioside

attachment

treatment

In

tors

treatment

types cancers

of

receptor

Therefore, in

their

nasopha-

the

researchers in

by

proteins

cultured

organization

several

during

M2,

significant

after

used

and

are

mediated not

attach-

gangliosides

different

of

supported

identify

(13).

role

cells.

even the

by cells

different

at the

there

used

was

The

of

discrepancies

animal

synthesis

that

To

inhibit

rant

of

to

a receptor

these

and

expressed

role

D1b

human

gan-

suggest-

to

with

effect

Furthermore,

in

the

gangliosides.

except

a dose-dependent

with

these

(Table not

gangliosides

observed.

express

all

respectively,

common

D 1b was

possibly

(NeuAcƒ¿2-

elucidate

required

treated

which

pharyngeal

as

as

NTHI

(6).

membrane

attempt

shown between

is further

outer

et al also

for

types

significantly, in

reasons of

of

with

D1a

Noel

different

separately to

Another

by

was

oligosaccharide

specific

the cul-

ganglioside

GM1

NTHI

sialylated

proteoglycans

the

attention

tested

NTHI

of

which

using

it

this

near

bac-

and

interaction

however,

mobility

to

(17).

tified

epithe-

of

steps

TLC

the

it

of

bacteria

cells,

to

while

detect

commonly

washing

epithelial

binding

mucin

is

two-dimensional

essential

(3),

binding

between

NTHI;

the

receptor

treatment

12.5 ƒÊg/ml,

resides

D 1a

We

T1b,

and

in attachment

with

lial

NTHI

minimum

NTHI

M3,

125

receptor

our

gangliosides

of

receptor

on

NTHI

and

on

The

GalNAcƒÀ1-4

after

recep-

respirato-

receptor

19).

NTHI.

treated

With

the

the

aeruginosa,

received

for

D 1 a, D 1b

ganglioside.

attachment

be

attachment we

at concentrations

ing

to

as

other

catarrhalis)

gangliosides

mixture,

gliosides each

as

receptor

decrease

for

Pseudomonas

(9,

is proposed

candidate

only

also

bacteria (M.

Therefore,

cant

of

not

but

proposed

binding

cells

proven b,

case

catarrhalis

epithelial

the

the was

for

Moraxella

3)

been type

In

GalNAci31-4Gal lial

in

ryngeal

Gangliosides

is

gangliosides

NTHI

ry

acid

assay

the

method

not

receptor

many

laryngeal

sialic

sialic

require

By

present

could

to

of

M2

(TLC)

interaction

(21).

the

in TLC

critical

identification

ganglioside

the

TLC

ganglioside

tor

in

that,

the

were

the

where

lactosylceramide,

weak

the

by

receptor

It is explained to

that

methodology

chromatography

(18). teria

our

revealed

catarrhalis,

the

thin-layer

Discussion

of

It was

are

is a complex responsible

RECEPTOR

Table

4. Sequences

of different

OF NONTYPABLE

H. INFLUENZAE

gangliosides

in that process (6). In conclusion, we provided here evidence that the receptor for NTHI lies in the sequence of GD2 on human pharyngeal epithelial cells; however, much still needs to be learned about the factors which determine the successful attachment of H. influenzae in the respiratory tract.

350. 9 )

Krivan,

H.C.,

Roberts,

pathogenic

Acad.

Kubiet,

M.,

We thank Akitoyo Ichinoseand Kiwao Watanabefor their excellenttechnicalassistance. Part of the projectwas supported by Grant-in-Aidfor Encouragementof YoungScientists,1996, Ministryof Education,Science,Sportsand Cultureand Grantfor ScientificResearch (InternationalCooperationon the Preventionand Treatmentof AcuteRespiratoryInfections)fromthe Ministryof Healthand Welfare,Japan.

U.S.A.

and

11 )

Ramphal,

3)

4)

5)

6)

7) 8)

to in

1988.

the

some

Many

carbohydrate

glycolipid.

Proc.

6157-6161. R.

1995.

Adhesion

from

blood

and

lactoferin.

mucins

Murphy,

F.T.,

and

Haemophilus antigens,

and

Infect. 12 )

Dis.

Musher,

of

and

nontypable

sputum

to

Infect.

13 )

14 )

the

immune

human

human

Immun.

63:

Kubitschek,

Pneumonia

Love,

bacterial

Immun.

62:

T.,

K.,

and

and

of

febrile

surface

infection.

Rev.

David,

to

J.,

and

Baughn,

tracheobronchitis

Intern. M.

nontypable

adhesion

due

Med.

99:

444-450.

1994.

High-molec-

Haemophilus

cellular

influenzae

proteoglycans.

Infect.

4028-4033.

Sonoda,

F.,

Matsumoto,

Iwagaki,

K.

lipopolysaccharide

A.,

1992.

against

neutropenic

Kobayashi,

Effects

specific

sparfloxacin in

to

Crennan,

acute Ann.

D.C.,

proteins

mediate

Oishi,

Nontypable aspects,

response

R.K., and

influenzae.

G.J.,

1987.

of clinical

1-15.

Haemophilus

Noel,

A.M.

a review

9:

1983.

Apicella,

influenzae:

M.D.,

E.R.

1) Ahmed, K., Ichinose, A., Dai, T.C., Takahashi, A., Utsunomiya,Y., Kawakami,K., Nagatake,T., and Matsu

2)

85:

influenzae

ular-weight

moto, K. 1993. Neutrophil response to nontypable Hemophilus influenzae in respiratory infections. Microbiol. Immunol. 37: 671-677. Ahmed, K., Kawakami, K., Watanabe, K., Mitsushima, H., Nagatake, T., and Matsumoto, K. 1995. Corynebacteri urn pseudodiphtheriticurn, a respiratory tract pathogen. Clin. Infect. Dis. 20: 41-46. Ahmed, K., Matsumoto, K., Rikitomi, N., and Nagatake, T. 1996. Attachment of Moraxella catarrhalis to pharyngeal epithelial cells is mediated by glycosphingolipid receptor. FEMS Microbiol. Lett. 135: 305-309. Apicella, M.A., Shero, M., Dudas, K.C., Stack, R.R., Klohs, W., LaScolea, L.J., Murphy, T.F., and Mylotte, J.M. 1984. Fimbriation of Haemophilus species isolated from the respiratory tract of the adults. J. Infect. Dis. 150: 40-43. Barenkamp, S.J. 1996. Immunization with high-molecularweight adhesion proteins of nontypableHaemophilus influenzae modifies experimental otitis media in chinchillas. Infect. Immun. 64: 1246-1251. Fakih, M.G., Murphy, T.F., Pattoli, M.A., and Berenson, C.S. 1997. Specific binding of Haemophilus influenzae to minor gangliosides of human respiratory epithelial cells. Infect. Immun. 65: 1695-1700. Hakomori, S. 1992. Functional role of glycosphingolipids in tumor progression. Tohoku J. Exp. Med. 168: 211-222. Karlsson, K.A. 1989. Animal glycosphingolipidas membrane attachment sites for bacteria. Annu. Rev. Biochem. 58: 309-

specifically

V.

899-902.

to

References

Ginsburg,

found

Sci.

Haemophilus

and

bind

GalNAcƒÀ1-Gal

Natl. 10 )

D.D.,

bacteria

sequence

tracheobronchial

-

701

S.,

the

of

antibodies

aeruginosa

Antimicrob.

Nagatake,

combination

monoclonal

Pseudomonas

mice.

of

and

pneumoniae

Agents

Chemother.

36:

1352-1357. 15 )

Porras,

C.S.,

Svanborg,

1985.

Method

for

zae 4: 16 )

to

human

buccal

Read,

R.C.,

epithelial

Wilson,

APR.,

tion

R.,

T., of

cells.

Binding

nontypable mucin.

vitro.

and

Hanson,

Haemophilus

Eur.

L.A. influen-

J. Clin.

Infect.

J.M.,

Immun.

N.,

Microbiol.

Svanborg,

C.

La (Branhamella)

1991.

V.,

P.J.

Dis.

64:

Todd,

1991. with

163:

Murphy, outer

influenzae

Andersson,

Lund,

Cole,

Infect.

between

Haemophilus

Rikitomi,

A., and

influenzae

J.

Bernstein,

1996.

and

in

M.S.,

H.S.

P.K.,

Haemophilus

mucosa

Reddy,

Rutman,

Jeffery,

of nontypable

piratory

18 )

Lagergard,

adherence

310-315.

Brain,

17 )

E.C.,

testing

H.C., Interac-

human

res-

549-558. T.F.,

and

Faden,

membrane

proteins

and

pharyngeal

human

of

1477-1479. B.,

Matsumoto,

Mechanism

catarrhalis.

K.,

of Scand.

Lindstedt,

adherence J. Infect.

R.,

of MoraxelDis.

23:

559-

567. 19 )

Saiman, pili

bind

cystic

L.,

and

to

asialo

fibrosis

Prince, GM1 epithelial

A.

1993.

which cells.

Pseudomonas

aeruginosa

is increased J.

Clin.

on

the

Invest.

surface 92:

1875-

1880. 20

)

St.

Geme

III,

J.W.

1994.

The

HMW1

adhesin

of

nontypable

of

702

K. KAWAKAMI

Haemophilus influenzae recognizes sialiated glycoprotein receptors on cultured human epithelial cells. Infect. Immun. 62: 3881-3889. 21) van Alphen, L., van den Broek, L.G., Blass, L., van Ham, M., and Dankert, J. 1991. Blocking of fimbriae mediated adherence of Haemophilus influenzae by sialyl ganglio-

ET AL

sides. Infect. Immun. 59: 4473-4477. 22) van Alphen, L. 1992. Epidemiology and prevention of respiratory tract infections due to nonencapsulatedHaemophilus influenzae. J. Infect. Dis. 165(Suppl.1): S 177-180. 23) Wiegandt, H. 1992. Gangliosides. Glycolipids New Comp. Biochem. 10: 199-260.