300

BBA 51613

REUTILIZATION OF PHOSPHATIDYLCHOLINE ANALOGUES BY THE PULMONARY SURFACTANT SYSTEM

THE LACK OF SPECIFICITY

HARRIS JACOBS, ALAN JOBE, MACHIKO IKEGAMI, DEBBIE MILLER and SALLY JONES

Perinutal Research Lahorutories, Harbor-UCLA Medrcal Center, UCLA School of Medicrne, 1000 We.yt Curson Street. Tomme. CA

90509 (U.S.A.)

(Received July 26th, 1983)

(Revised manuscript received December 30th, 1983)

Ke,, words: Lung surjactant; Phosphatidylchohne; Phospholipid rrnalog

Five specific “C-labelled analogues of 1,2-dipalmitoyl-sn-glycero-3-phosphocholine, (L-a-DPPC) including I-palmitoyl-sn-glycero-3-phosphocholine, (L-a-IysoPC) and 2,3-dipalmitoyl-sn-glycero-I-phosphocholine (the D isomer of DPPC) were individually mixed with L-(Y-[~HJDPPC and unlabelled natural surfactant isolated from 3-day-old rabbits. The mixtures were injected intratracheally into 3-day-old rabbits which were then killed at preset times up to 72 h after injection. Phosphatidylcholine was isolated from the alveolar wash and from lamellar body fraction from each rabbit and was analyzed for the ratio of 3H-to-‘4C counts/min. These ratios were plotted against the time the rabbits were killed to determine whether a difference existed in the rates of reutilization of the analogues relative to L-a-DPPC. By 60 h L-a-IysoPC was reutilized at 42% of the efficiency of the L-a-[ 3H]DPPC with which it was injected. That part of the L-a-lyso which was reutilized had been converted to [ “C]phosphatidylcholine. Each of the other four analogues was reutilized by the surfactant system as efficiently as L-a-DPPC. These results are most consistent with a process of bulk uptake of surfactant from the alveolar space by the Type II cell with subsequent processing for resecretion which involves minimal specificity for molecular structure.

Introduction

Several recent studies have partially defined the

kinetics of secretion of the pulmonary surfactant

complex [l-6]. These studies have concentrated on phosphatidylcholine, the major surface-active component of surfactant, which is synthesized in the microsomes of Type II cells and is packaged into lamellar bodies for secretion. Jacobs et al. [3] demonstrated that lamellar body phosphati-

Abbreviations: L-u-DPPC. 1,2-dipalmitoyl-sn-glycero-3-phos-

phocholine; L-a-lysoPC, I-palmitoyl-sn-glycerp-3-phosphocho- line.

dylcholine is the strict precursor for the surfactant

phosphatidylcholine found in the alveolar space.

Hallman et al. [4] and Glatz et al. [5] gave radio-

labelled surfactant intratracheally to adult rabbits

and newborn lambs, respectively, and found evi- dence of reutilization of the labelled surfactant. The extent of surfactant phosphatidylcholine re- utilization in 3-day-old rabbits was subsequently quantified and found to exceed 90% [6]. It has also been demonstrated that surfactant phosphati- dylcholine is reutilized possibly entirely as intact molecules rather than by partial degradation with resynthesis of component parts into new phos- phatidylcholine molecules [4,6]. The mechanism

0005.2760/84/$03.00 0 1984 Else&x Science Publishers B.V.

301

and pathways of surfactant phosphatidylcholine

reutilization are as yet undefined. About 60% of the phosphatidylcholine mole-

cules of surfactant have palmitic acid esterified to

the numbers 1 and 2 carbons of the glycerol

backbone [7]. While at least one other fatty acid is

found in the remaining 40%, all phosphatidylcho-

line molecules of surfactant have the phosphate

and polar head group attached to the number 3

carbon of glycerol. The pathway responsible for

reutilization of surfactant phosphatidylcholine has

not been tested for its specificity for any of the

above characteristics of the molecule. Thus in the

present study we utilized analogues of the natu-

rally occurring 1,2-dipalmitoyl-sn-glycero-3-phos-

phocholine (L-(r-DPPC) to test specific characteris- tics of the molecule for their role in allowing

recognition and reutilization of phosphatidylcho-

line by the surfactant system. This was done by

mixing each analogue individually with L-a-DPPC,

injecting the mixtures intratracheally into 3-day-

old rabbits and directly comparing the relative

rates of reutilization of the analogue with L-cy-

DPPC.

Materials and Methods

Preparation ,of analogues. One analogue, L-l-

[ pahnitoyl-‘4C?]%z-glycero-3-phosphocholine (L-C+

1ysoPC) (57 mCi/mmol), was purchased from New England Nuclear. The other radiolabelled ana-

logues of L-(Y-DPPC were synthesized and each contained [ methyl-‘4 Clcholine in the head group.

Three of the analogues were synthesized by the

method of Stoffel et al. [8]. These authors describe a nonenzymatic procedure for the synthesis of

[ methyl-‘4C]phosphatidylcholine starting from un- labelled N, N-dimethylphosphatidylethanolamine

and [ “C]methyl iodide. The reaction is run in the dark, under anhydrous conditions, and at room

temperature. Methanol is used as the solvent.

Sodium hydroxide is used as a catalyst. After 2 h,

the reaction mixture is acidified with 2 M HCl and the lipids are extracted by adding CHCl, and water. The starting lipid, the small amount of lysophospholipid formed and the desired product (the phosphatidylcholine) are all easily separated by thin-layer chromatography on silica gel H TLC plates. The phosphatidylcholine contains one

[14C]methyl group in the choline head group. Using

this method we prepared 1,2-distearoyl-sn-glycero- 3-phosphocholine (~-a-DSPC), 1,2-dipalmityl-sn-

glycero-3-phosphocholine (L-a-DPPC ether), and

1,3-palmitoyl-sn-glycero-2-phosphocholine (/?-

DPPC) by starting with the appropriate N,N-di-

methylethanolamine precursors (Calbiochem-Behr-

ing Corp. or Serdary Research Labs). ~-a-DSPC

contains two 18-carbon stearic acids replacing the

normal 16-carbon palmitic acids, L-a-DPPC ether

has ether linkages joining the glycerol to the 16-

carbon aliphatic side-chains, and p-DPPC has the

polar head group attached to the number 2 carbon

of glycerol.

One analogue was synthesized according to the method of Smith et al. [9]. This procedure is also

non-enzymatic and starts with phosphatidy- lethanolamine as the precursor rather than N, N-

dimethylphosphatidylethanolamine. Silver carbon-

ate on diatomaceous earth is the catalyst used in

the reaction and [‘4C]methyl iodide is again the

methyl donor. The reaction is run at 37°C in the

dark, and under anhydrous conditions. The solvent is a 1 : 1 (v/v) mixture of tetrahydrofuran and

acetonitrile. Phosphatidylcholine is formed by the

addition of three methyl groups to phosphatidy-

lethanolamine. Again the products of the reaction

are easily separated by thin-layer chromatography.

Using this method we prepared 2,3-dipalmitoyl-

sn-glycero-1-phosphocholine (D-DPPC) from 2,3-

dipalmitoyl-sn-glycero-1-phosphoethanolamine

(Serdary Research Labs).

For each analogue synthesized, we used

[14C]methyl iodide from either New England Corp.

(10 mCi/mmol) or from ICN (50 mCi/mmol). Each radiolabelled analoque was purified by

thin-layer chromatography on silica gel H TLC plates and identified by running unlabelled analo-

gous purchased from Calbiochem-Behring Corp. in a parallel lane. Each analogue except L-a-1ysoPC

co-chromatographed with L-a-DPPC in the solvent

system used.

By using phospholipase A, (EC 3.1.1.4) (Sigma), we demonstrated that the synthesized L-(Y-DPPC ether, D-DPPC and /?-DPPC had the expected structures. Phospholipase A, will degrade both /3-DPPC and L-a-DPPC to lysophosphatidylcho- line and free fatty acid [lo]. If rearrangement of the starting molecule occurred in the course of

302

synthesizing p-DPPC, then there would be equal

probabilities of producing phosphatidylcholine

with the phosphorylcholine attached to the num-

ber 3 carbon of glycerol (L-a-DPPC) and the num-

ber 1 carbon of glycerol (D-DPPC). But D-DPPC

is resistant to phospholipase A 2 [l I]. We mixed

the presumed P-[r4C]DPPC with t-cw-[methyl-

‘HJDPPC (New England Nuclear) and treated the

mixture with phospholipase A, (method described

under Phospholipase treatment). When the reac-

tion products were separated by thin-layer chro- matography, 93% of the “H radioactivity and 94%

of the r4C radioactivity were associated with lysophosphatidylcholine. There was no evidence

that any D-DPPC had been formed, indicating

that the synthesized molecule was indeed /I-DPPC.

Similarly, since no rearrangement occurred during

the synthesis of P-DPPC, there is no reason to

assume that any rearrangement occurred in the

synthesis of L-(r-DSPC which was prepared by the

same method.

L-(Y-DPPC ether, like D-DPPC. is resistant to

phospholipase A,. We mixed L-a-[methyl- “H]DPPC (New England Nuclear) with the pre-

sumed L-~-[‘~C]DPPC ether and in a separate vial

with the presumed D-[r4C]DPPC. Both mixtures were treated with phospholipase A, as above and

the reaction products were separated by thin-layer

chromatography. In both cases, over 90% of the

’ H radioactivity was associated with lyso- phosphatidylcholine, while the 14C radioactivity

remained associated with phosphatidylcholine. Iqjection solution. Five separate injection solu-

tions were prepared. Each “C-radiolabelled ana-

logue was mixed with t,-a-[meth,vLiH]DPPC (27 Ci/mmol) which was purchased from New Eng-

land Nuclear. The mixtures were sonicated into

solution using a Fisher Sonic Dismembrator as

before [6]. Unlabelled surfactant was isolated from 3-day-old rabbits by centrifugation of an alveolar wash collected from healthy animals at 8000 X R over 0.7 M sucrose. The surfactant, which layered at the interface, was pelleted at 48 000 X g and was mixed with each of the five solutions [6]. These solutions were adjusted with distilled water and lactated ringer’s solution to a final concentration of 50% lactated Ringer’s, When injected in- tratracheally into adult rabbits 50% lactated Ringer’s has been shown to be well tolerated [12].

In a previous study, we demonstrated that syn-

thetic L-(u-DPPC mixed with natural surfactant

and injected intratracheally into 3-day-old rabbits

functions metabolically like the endogenously pro-

duced surfactant phosphatidylcholine [6]. Phos-

phatidylcholine was isolated from a sample of

each injection solution by thin-layer chromatog-

raphy. The final concentration of phosphati-

dylcholine in each solution (including the added

analogue and the added L-(Y-[ ‘H]DPPC) was such

that each rabbit received 0.25 pmol of phosphati-

dylcholine as determined by phosphatidylcholine

phosphate [13]. This was equivalent to about 4% of

that present in the alveolar space from endogenous surfactant production [3]. Of the 0.25 pmol of

phosphatidylcholine injected, approx. 0.01 pmol was the D-~-[‘~C]DPPC analogue and about 10. ’

pmol was L+[~H]DPPC. Each rabbit received

from 20 000 to 95 000 counts/mix1 as a l4 C-labelled

analogue of L-(Y-DPPC and from 45 000 to 130000

counts/min as L-cu-[“H]DPPC. There was always

more L-(Y-]~H]DPPC than “C-labelled analogue

with the ratio of ‘H to “C counts/min varying

from 1.2 to 2.6. Animuls. 150 3-day-old rabbits weighing 67.7 _t

0.9 g were removed from their litters on the day of

injection and studied in five groups of 30 rabbits

each. Rabbits in each group were injected in-

tratracheally as before with one of the five radio-

active solutions [6]. Briefly, a midline incision was

made in the neck, the trachea was isolated, and 0.2

ml of one of the solutions was injected distally

through a 30-gauge needle while occluding the

trachea proximally with fine forceps. Three rabbits in each of the five groups were killed at one of IO

preset times from 2 h to 72 h after injection by an

intraperitoneal injection of pentobarbitol followed by exanguination. All rabbits were breathing nor-

mally within 1 min of injection and all were healthy

at the time of killing. Fruction isolation. After each rabbit had been

killed the trachea was cannulated and the surfac- tant was recovered from the alveoli by five successive alveolar washes with saline. The five washes from each rabbit were combined and are referred to collectively as the alveolar wash. This procedure recovers more than 90% of the surfac- tant which can be recovered by alveolar wash 171. The washed lung of each rabbit killed from 15 h

303

after injection to 72 h after injection was homog-

enized in 0.32 M sucrose buffer containing 0.01 M

Tris-HCl, 0.001 M CaCl,, 0.001 M MgSO,, 0.0001 M EDTA and 0.15 M NaCl at pH 7.4 [ll]. 0.5 ml

of the homogenate was saved and the remainder

used to isolate a lamellar body fraction by dif-

ferential and sucrose density gradient centrifuga-

tion as before [3,6,14]. Phospholipid analysis. An aliquot of the alveolar

wash, the lung homogenate, and the lamellar body

fraction from each rabbit were extracted according

to Bligh and Dyer [15]. The lipid phase of each

extract was plated on handmade silica gel H thin-

layer chromatography plates. These plates were

run in one dimension in chloroform/

methanol/acetic acid/water (65 : 25 : 8 : 4, v/v) to

isolate phosphatidylcholine [3]. The phosphati-

dylcholine spot was assayed for radioactivity by

counting in Aquasol- scintillation fluid (New En-

gland Nuclear). As stated earlier, L-cy-1ysoPC does

not co-chromatograph with L-a-DPPC. From each

fraction from animals injected with L-cu-lysoPC, we isolated phosphatidylcholine and lysophosphati-

dylcholine. We then assayed the lysophosphati-

dylcholine spot for radioactivity as was done for

the phosphatidylcholine spot. 3H and 14C

counts/min were corrected for cross-channel con-

tamination based on values obtained from count-

ing pure standards of each isotope. No quenching

was observed with the quantities of phosphati-

dylcholine used. All results in each fraction are

expressed as total counts/mm, the ratio of [3H]-

to [‘4C]phosphatidylcholine counts/min or the

ratio of [3H]phosphatidylcholine to [‘“Cl

lysophosphatidylcholine counts/min. Phospholipase treatment. Lyophilized phos-

pholipase A, from Crotalus adamanteus venom

(EC 3.1.1.4) was purchased from Sigma Chemical Company. Analogues of L-(u-DPPC prior to injec-

tion and phosphatidylcholine isolated from al- veolar washes of rabbits given L-a-lysoPC, D-

DPPC and L-a-DPPC ether were tested for their sensitivity to this enzyme. The alveolar washes

were from rabbits killed 48-60 h after injection. In each case, a total of 0.1-0.5 pmol of phosphati- dylcholine was dried under N, in a 20-ml flask. To each phosphatidylcholine sample we added 1.5 ml buffer (1 mM CaCl,, 10 mM Tris, pH 8) 4 ml of diethyl ether (Mallinchrodt Chemical Co.) and 5

units of phospholipase A,. The reaction was al-

lowed to proceed for 1.5 h at room temperature, after which the ether phase was dried under N, at

50°C [16]. The lysophosphatidylcholine, free fatty

acids and residual phosphatidylcholine were ex-

tracted with 4.5 ml of 2 : 1 CHCl,/CH,OH (v/v).

The chloroform layer was dried under N, at 50°C

and plated on a silica gel H thin-layer chromatog-

raphy plate which was run in the same solvent

system given above. The lysophosphatidylcholine

and phosphatidylcholine spots were assayed for

radioactivity by counting in Aquasol- scintilla-

tion fluid (New England Nuclear). For the phos-

pholipase A *-digested phosphatidylcholine from

animals injected with L-a-lysoPC, we also collected

the free fatty acid spot and assayed it for radioac-

tivity as above.

All comparisons were by two-tailed Student’s t-test.

Results

After injecting 3-day-old rabbits intratracheally with labelled L-(Y-DPPC, the change with time of

total alveolar wash phosphatidylcholine counts/ min is best described as a linear sum of exponen-

tials [6]. Similar equations could be generated after

injecting analogues of L-a-DPPC into rabbits.

However, there is no simple method for directly

comparing such complex curves. Injecting rabbits

with L-a-[3H]DPPC mixed individually with each

14C-labelled analogue allows for a direct compari-

son between the natural molecule and the ana-

logues by simply observing the change with time in

the ratio of 3H to 14C counts/min in phosphati-

dylcholine isolated from the alveolar wash and lamellar body fractions [6]. Thus all results from

which our conclusions are drawn, are expressed as the change with time of this ratio in each fraction.

The mean total recovered [ 3H]phosphati- dylcholine in the alveolar wash versus time after

injection for rabbits given a mixture of L-cr- [ ‘H]DPPC and L-(Y-[‘4C]DPPC ether is shown in

Fig. 1A as a semilog plot with the linear least- squares fit to the means. While a sum of exponen- tials would produce a closer fit to the data, the intent is to demonstrate that there is a statistically significant decrease in total recovered counts/min versus time similar to that found before [6].

I I 0 30 60

Hour,

Fig. 1. Total ‘H counts/min and the ratio of ‘H to 14C

counts/min in the alveolar wash. (A) Semilog plot of total

]‘H]phosphatidylcholine counts/min versus time after injec-

tion for rabbits given L-~-[~H]DPPC and L-~-[‘~C]DPPC ether.

Each point represents the mean of data from three rabbits

killed at that time. The error bars are the SE. and the line was

determined by least-squares regression on the log of the means.

The slope of the line is different from zero (P < 0.01). (B) The

mean ratio of ‘H to “‘C counts/min in phosphatidylcholine

isolated from the alveolar washes for rabbits given L-W

[ ‘HJDPPC and L-~-[‘~C]DPPC ether killed at the indicated

times is plotted against the time of death. Each point represents

the mean of data from three rabbits killed at that time. Error

bars are +_S.E. Error bars not shown fall within the data point.

(C) Same as B except that rabbits were given L-~-[~H]DPPC

with /?-[‘4C]DPPC.

The ratios of [ 3H]phosphatidylcholine counts/

min to [‘4C]phosphatidylcholine counts/min in

the alveolar washes from the rabbits killed at each

time for each analogue were averaged and plotted

against the time of sacrifice. The plots for L-W

DPPC ether and /3-DPPC are shown in Fig. 1B

and C. The line shown for each analogue was

determined by least-squares regression on the

means. Neither of the lines shown in Fig. 1B or C nor the regression lines for the same plots for

rabbits given L-a-DSPC or D-DPPC (not shown) were statistically different from zero (P > 0.05). Thus for these four analogues, while the total recovered ‘H and 14C counts/min decreased in alveolar wash phosphatidylcholine, the ratio did not change with time. Also the use of ratios rather

than total ‘H and 14C counts/min cancelled out much of the variability from animal to animal. By using ratios at each time, the standard error of the mean was much smaller (expressed as a percent of

i;;; 0 30 60

HOURS

Fig. 2. Ratio of ‘H to “‘C counts/min in the lung homogenate.

The mean ratios (*SE.) of ‘H to 14C counts/min in phos-

phatidylcholine isolated from the lung homogenate of rabbits

injected with D-DPPC (0) or L-a-DPPC ether (m) were plotted

against the time of death. The lines shown were determined by

linear least-squares regression on the means.

the mean), which implies a more accurate mea-

surement. The ratio of ‘H to 14C counts/min in phos-

phatidylcholine isolated from the lamellar body

fractions were averaged for the rabbits killed at

each time for each analogue and plotted against

the time of death. For the same four analogues

mentioned above, the lines determined by least- squares regression on the means had slopes which

were not statistically different from zero (P > 0.05) (data not shown).

For each of these four analogues, the ratio of

[’ Hlphosphatidylcholine coun ts/min to r4c1

04 I

0 30 60 HOURS

Fig. 3. Ratio of ‘H to 14C counts/min from rabbits given

L-~[‘~C]lysoPC and L-U-[ ‘HIDPPC. The mean ratios of ‘H to

14C counts/min in the phosphatidylcholine from alveolar

washes were plotted against the time of death (A). The mean

ratios of ‘H to 14C counts/min in the phosphatidylchohne

from lamellar bodies were plotted against the time of death (A).

All standard errors were less than 1.7.

305

phosphatidylcholine counts/min in the alveolar

wash was compared to the same ratio in the lamel- lar bodies. Again no difference was found at any

time. The absence of a slope to the regression lines

for the mean ratios in the alveolar wash and lamellar bodies and the lack of any difference in

the ratio of [‘HI- to [‘4C]phosphatidylcholine counts/min in the alveolar wash and lamellar

bodies implies that each of these four analogues

was handled by the pulmonary surfactant system in a manner indistinguishable from the L-ol-DPPC

with which it was mixed.

The recovery of L-(Y-[~H]DPPC in the alveolar washes at 72 h from rabbits injected with these

four analogues varied from 10% to 30% of the

original injected dose. In a previous experiment,

the recovery of labelled natural surfactant phos-

phatidylcholine in the alveolar wash at 72 h com-

pared to the original total injected dose was 22%

[16]. The recovery of these four analogues in the

alveolar wash at 72 h was the same as that for

L-~-[~H]DPPC in the same rabbits. No 3H or 14C

counts/min were detectable in any other lipid of surfactant.

When the phosphatidylcholine isolated from al- veolar washes of rabbits given D-DPPC or L-cu-

DPPC ether were treated with phospholipase A,,

there was no evidence that in vivo rearrangement

of the analogues had occurred. In both cases the

L-(Y-[ 3 H]DPPC was converted to L-LX-[ 3 H]lysoPC,

while over 90% of the [‘4C]phosphatidylcholine

radioactivity was resistant to hydrolysis. Thus these

analogues were not rearranged by the pulmonary

surfactant system. Since L-c~-DSPC and p-DPPC had kinetics similar to those of D-DPPC and L-ol-

DPPC ether it is unlikely that either molecule

could have undergone rearrangement, as the extra

step would almost certainly effect the percent of [‘4C]phosphatidylcholine counts recovered com-

pared to L-LY-[~H]DPPC and/or the observed rate of transport from the alveolar wash to the lamellar bodies and back.

Synthetic L-a-DPPC mixed with natural sur-

factant and injected intratracheally into 3-day-old rabbits is metabolically indistinguishable from the endogenously produced surfactant phosphati- dylcholine [6]. Thus, the pulmonary surfactant sys- tem is able to take up these analogues, package them into lamellar bodies as part of surfactant and

secrete them into the alveolar spaces as if they

were naturally occurring forms of phosphati- dylcholine.

When radiolabelled L-a-DPPC is injected in-

tratracheally into 3-day-old rabbits, there is a de-

crease with time in the total phosphatidylcholine

counts/min that can be recovered from the lung

homogenate as well as from the alveolar wash [6].

For each analogue, the change with time in the

amount of the analogue which could be recovered

from the lung homogenate was compared to the recovery of the added L-(Y-DPPC by comparing the

rate of change of the ratio of 3H to 14C counts/min

in lung homogenate phosphatidylcholine with time.

For each analogue, the mean of the ratio of ‘H to

14C counts/mm in phosphatidylcholine isolated

from the lung homogenate at each time of death

was determined. The slope of the lines determined

by linear least-squares fit to the means were com-

pared to zero. These slopes were not statistically

different from zero for L-a-DSPC, /3-DPPC and

D-DPPC. The one exception was that for the L-LX-

DPPC ether analogue. For this analogue the ratio

of counts/min from L-~-[~H]DPPC to L-(Y-

[‘4C]DPPC .ether decreased linearly with time in

the lung homogenate, implying a more rapid loss of the L-~-[~H]DPPC than the L-~-[‘~C]DPPC

.ether. The regression lines and mean values from

the lung homogenates of rabbits injected with

D-DPPC and L-a-DPPC ether are shown in Fig. 2.

When the total 14C counts/min in the lung homo-

genate + alveolar wash from the ether analogue were plotted against time after injection, the slope

of the line was not statistically different from zero

(data not shown). Thus the lung was incapable of

clearing or degrading the L-(Y-[‘~C]DPPC ether

which was removed from the pulmonary surfac-

tant system.

The results obtained from rabbits injected with L-cy-[14C]lysoPC were different. For rabbits in-

jected with this molecule, we measured 3H counts/min and 14C counts/min both in phos-

phatidylcholine and in lysophosphatidylcholine in the alveolar wash and in lamellar bodies. By the time of the first alveolar wash sample at 2 h after injection, there was almost no detectable radioac- tivity in lysophosphatidylcholine. Thus a plot of the ratio of [ 3H]phosphatidylcholine specific activ- ity to [‘4C]lysophosphatidylcholine specific activ-

306

ity in the alveolar washes versus the time of death

rapidly approached infinity (not shown). Note that

this plot is equivalent to the plots of ‘H to 14C

counts/min in alveolar wash phosphatidylcholine

shown in Fig. 1 for rabbits injected with the other

analogues, all of which co-chromatographed with

L-a-DPPC. As 14C radioactivity decreased in

lysophosphatidylcholine, it increased in phos-

phatidylcholine, indicating that at least part of the

L-cw-[‘4C]lysoPC was converted into [‘4C]phos-

phatidylcholine. This was manifest as a progres-

sive decrease in the ratio of “H to 14C counts/min found in alveolar wash phosphatidylcholine (Fig.

3). Where this conversion took place could not be

determined. Recovery of L-a-[‘4C]lysoPC as [ “C]phosphatidylcholine in the alveolar wash 60 h

after injection accounted for only 15% of the origi-

nal injected radioactivity. Recovery of the simulta-

neously injected L-(Y-[~H]DPPC in the alveolar

washes from these rabbits was about 35% of the

original injected dose 60 h after injection. The alveolar washes from rabbits injected with

L-(Y-[‘~C]~~SOPC contained no 14C radioactivity as L-a-1ysoPC at the later times of death. Phosphati-

dylcholine from the washes from animals killed at

60 h was isolated and hydrolysed to free fatty acid

and lysophosphatidylcholine by phospholipase A 2. Of the total phosphatidylcholine radioactivity

which was subjected to phospholipase A,, over 90% was recovered after separating the reaction

products on thin-layer chromatography plates. Of

this recovered radioactivity, 72% was associated

with lysophosphatidylcholine, while 28% was asso-

ciated with free fatty acids.

Discussion

The Type II pneumocyte in the 3-day-old rabbit is very efficient at reutilizing secreted surfactant

phosphatidylcholine [6]. While the amount of re- utilization exceeds 90%, there is a small percentage of phosphatidylcholine molecules which are not reutilized. Surfactant has many different compo- nents and at least two others besides phosphati- dylcholine, specifically phosphatidylinositol and phosphatidylglycerol, are reutilized [4]. If each of the various components of surfactant were inter- nalized by the Type II cell via a specific carrier, then analogues of any specific component should

be differentially internalized when compared to

the natural form of that component, depending on

the specificity of the transport process. The ana-

logues we used were chosen to test the specificity

of reutilization as it related to specific parts of the

phosphatidylcholine molecule. As indicated, the

results of phospholipase A, treatment of the ana-

logues demonstrated that they had the expected

structures.

The time course of the experiment was based on

two previous studies. One study demonstrated that

the turnover time of surfactant phosphatidylcho- line was about 10 h in 3-day-old rabbits [3]. The

present experiment extended over 60-72 h which covered six to seven turnover times of alveolar

phosphatidylcholine. Any discrepancy in reutiliza- tion of the analogues relative to the naturally

occurring form of phosphatidylcholine would be

demonstrated by a non-zero slope for the regres-

sion line relating the ratio of alveolar wash [3H]-

to [‘4C]phosphatidylcholine counts/min to time

after injection. In the other previous study, 3-day-

old rabbits were injected intratracheally with ra-

diolabelled surfactant. There was a rapid drop in

total counts/min recovered in the alveolar wash over the first 8-10 h during which time the radio-

labelled surfactant flowed predominately into the

Type II cells [6]. Since the first time in the present experiment when animals were killed was at 2 h

after injection, a differential internalization of the

L-a-DPPC compared to the analogue would be

easily detected. The data obtained also indicated

that no reorganization of the analogue occurred

for L-c~-DPPC ether or for D-DPPC and strongly support the conclusion that no reorganization oc-

curred for L-(u-DSPC or for ,B-DPPC. There clearly

was reorganization of the L-a-1ysoPC molecule. The data in the present experiment indicate that

the mechanism(s) of internalization and reutiliza-

tion of surfactant phosphatidylcholine are rela- tively nonspecific. The significant difference in the rate or reutilization of L-a-IysoPC compared to L-a-DPPC indicates that there is some specificity, while the lack of any difference between L-CX-DPPC and the other analogues indicates how limited this specificity is. It seems that almost any analogue of phosphatidylcholine can be internalized and that reutilization will occur without alteration as long as the glycerol backbone contains no free hydroxyl

307

groups. The four analogues which were handled by

the surfactant system as if they were the naturally

occurring L-a-DPPC have structural differences when compared to the natural molecule which are

worth emphasizing. The L-(u-DSPC analogue only

differed from the L-a-DPPC in that the fatty acids

which were esterified to the numbers 1 and 2

carbons of glycerol were 18-carbon stearic acids

instead of 16-carbon palmitic acids. The dominant phosphatidylcholine species of surfactant is L-(Y-

DPPC. However, other fatty acids besides palmi- tate are found esterified to the glycerol of phos-

phatidylcholine. Hence, equivalent reutilization of

L-a-DSPC and L-(Y-DPPC was not surprising and

indicated that there was not a high selectivity for

the dominant species.

L-a-DPPC ether is also very similar to L-(Y- DPPC. The ether linkages between the glycerol

backbone and the aliphatic side-chains replace the

normal ester linkages. This changes the three-di-

mensional configuration slightly [17], but more

importantly makes the molecule resistant to hy-

drolysis by the lung (see Results). The equivalent

reutilization of this molecule compared to L-(Y-

DPPC by the surfactant system supports previous claims that surfactant phosphatidylcholine is re-

utilized possibly entirely as intact molecules [4,6].

P-DPPC is an analogue which does not occur

naturally in biological systems. It has the polar head group attached to the number 2 carbon of

glycerol, but can be hydrolyzed to a 1ysoPC mole- cule and free palmitate by phospholipase A, [lo].

The Type II cell also reutilized the D isomer of DPPC as well as L-a-DPPC ether with an ef-

ficiency to its reutilization of the naturally occur-

ring L-a-DPPC. Thus if a specific carrier is in-

volved in reutilization, it must clearly have a broad

range of configurations which it will recognize.

Our interpretation of these results is that the mechanism of uptake of surfactant from the al-

veolar surface is some form of bulk uptake not

unlike endocytosis. This process probably internal- izes surfactant as macromolecular aggregrates, as suggested in a study by Oyarzun et al. [12]. These investigators injected radiolabelled liposomes of L- or D-DPPC into the alveoli of adult rabbits and measured the rate of clearance from the airways. They found no difference in the rate of clearance of the two isomers, but did not demonstrate that

the aggregrated L-DPPC which they injected was

metabolized in the same way as endogenously synthesized phosphatidylcholine.

Bulk uptake of surfactant by the Type II cell

would also be consistent with the work of Wil-

liams [18], who demonstrated by electron micros-

copy that the membrane of the Type II cell was

reutilized. She used a lectin which binds to (Y-

galactose residues on apical plasma membranes of Type II cells. The lectin was covalently linked to

ferritin. Shortly after injecting this material into

the peripheral airways of adult rats, the material was found by electron microscopy in small, api-

tally located vesicles and in multivesicular bodies

of Type II cells. At later times, it was found in the non-lamellar portion of lamellar bodies. Kuhn [19]

also reported what he called pinocytotic vacuoles

near the surface of the Type II cell. Possibly these

vesicles transport alveolar surfactant back into the

Type II cell for processing, which may give rise to the multivesicular bodies and ultimately lamellar

bodies. Since the L-cy-lysoPC was not inert within

the surfactant system and since L-a-1ysoPC was

altered in the course of reutilization, there exists

some screening mechanism in the reutilization

pathway. It it not clear whether this process occurs

only after internalization or both before and after internalization of surfactant. Either would be con-

sistent with our results. For example, a screening

process which occurs prior to internalization and

is less than 100% efficient would allow for the

uptake of some of the L-cu-1ysoPC. Once inside the

cell this molecule might be preferentially acylated

to L-a-phophatidylcholine, which then enters the

surfactant pathway. Similarly, if screening only

occurs inside the cell, some of the L-a-IysoPC

might be shunted into pathways of complete de- gradation, while other molecules might be acylated

to L-a-phosphatidylcholine. Independent of where the screening occurred, a large fraction of the

injected L-a-[‘4C]lysoPC was removed from the

system. Some of the [‘4C]palmitate from these molecules was found acylated to the number 2 carbon of glycerol in phosphatidylcholine. If this occurred only by a lyso-lyso exchange giving phos- phatidylcholine and glycerol 3-phosphorylcholine, then any L-cy-1ysoPC molecule would have an equal probability of being the donor or recipient of the fatty acid. Thus after the [‘4C]phosphatidylcholine

308

from the alveolar washes of animals killed at 60 h

had been treated with phospholipase AZ. there

should have been roughly equal amounts of r4C

counts/min as L-cu-1ysoPC and as free fatty acid.

The fact that 72% of the radioactivity was found in

the l-position and only 28% in the 2-position

suggests that some of the L-ar-[‘4C]lysoPC lost

from the system was degraded into component

parts and that the free [‘4C]palmitate was re-

utilized (either intact or possibly after some altera-

tion). The relative contributions of the two posible

pathways cannot be determined from the available

data. The only analogue which remained within the

confines of the lung longer than L-c~-DPPC was L-a-DPPC ether. This was demonstrated by a de-

crease with time in the ratio of [“I-I]- to [‘4C]phos-

phatidylcholine counts/min in the lung homo- genate. When, for rabbits receiving L-~-[‘~C]DPPC

ether, the mean total [‘4C]phosphatidylcholin~

counts/min in the lung homogenate plus the al- veolar wash was plotted against the time of

sacrifice, there was no detectable drop from 15 h

to 72 h after injection. The rate of reutilization of

L-LY-DPPC ether as surfactant matched that for

L+DPPC. Thus the excess i4C radioactivity in

the lung homogenate probably represents a faihne

of degradation by the lung of that fraction of the ether analoque removed from the surfactant sys-

tem. This probably represents the small fraction of

surfactant phosphatidylcholine which is not nor- mally reutilized [6]. The other analogues of DPPC

which we tested did not show this effect, which

suggests but does not prove that the initial de- gradative process for surfactant L-a-DPPC is non-

enzymatic. It would be reasonable to expect that if

the first step were enzymatic, then the D-DPPC

would also have been resistant. Furthermore, a non-enzymatic acid hydrolysis of the fatty acids

linked to the glycerol backbone would more effec- tively degrade L-(w-DPPC and the analogues other than I.-a-DPPC ether and would account for these findings. The lack of degradation of L-LU-DPPC ether by the lung makes it a potentially useful probe in studies of site and mechanism of normal surfactant phosphatidyicholine clearance.

Once inside the Type II cell, each of the four analogues reutilizes as L-ar-DPPC was incorpo- rated into lamellar bodies. In fact, as expected for

equivalent reutilization of the analogue and I.-cr-

DPPC, the ratio of ‘H to 14C counts/min for each

of these analogues was the same in the alveolar

wash as it was in lamellar bodies. The experiments

presented here were not designed to determine

whether the intracellular mechanism which packages reutilized phosphatidylcholjne into

lamellar bodies requires a specific phosphati-

dylchohne carrier. They do show that whatever the

packaging mechanism is, it is very nonspecific in

terms of the actual structure of the phosphati-

dylcholine molecule which will be accepted.

The actual steps leading to the formation of

multivesicular bodies and lamellar bodies are as

yet undefined. The data presented here suggest that much of the processing occurs non-enzymati-

calfy. This would be consistent with the concept that the various components of the lamellar body

interact to form a structure with the lowest total

free energy. The ability of lamelfar bodies to

survive inside the cell with a composition signifi-

cantly different from the surrounding cytosol sup-

ports this concept, since a thermodynamically un-

stable structure would require energy input to

maintain its form. Similarly, the fact that lamellar

bodies can be isolated intact from homogenized

lungs by differential and sucrose gradient centrifu-

gation indicates that they require no energy input from the Type II cell to maintain their integrity.

Surfactant in the lamellar body reaches the al- veolar space by exocytosis [20,21]. The surfactant

then forms a monolayer covering the alveolar surface. The critical factor which allows the lamel- lar body to maintain its form inside the Type II

cell (and during isolation) may well be the limiting membrane which is left in the cell membrane after

exocytosis.

This work was supported by NIH Grant HD- 11932 from Child Health and Development, De- partment of Health and Human Services, by Re- search Career Development Award HD-HL-00252 to A.J. and NIH Research Service Award HL- 06544 to H.J.

References

1 Baritusso, A.G., Magoon, M.W.. Goerke J. and Clements, J.A. (1981) Biochim. Biophys. Acta 666, 382-393

309

2 Young, S.L., Kremers. S.A.. Apple, J.S.. Crapo. J.D. and

Brumley. G.W. (1981) J. Appl. Physiol. 51. 248-253

3 Jacobs, H.. Jobe, A., Ikegami, M. and Jones, S. (1982) J.

Biol. Chem. 257, 1805-1810

4 Hallman, M., Epstein B.L. and Cluck, L. (1981) J. Clin.

Invest. 68, 742-751

5 Glatz, T., Ikegami. M. and Jobe. A. (1982) Pediatr. Res. 16.

711-715

6 Jacobs, H., Jobe. A., Ikegami. M. and Conaway, D. (1983)

J. Biol. Chem. 258, 4159-4165

7 Toshima, N. and Akino, T. (1972) Tohoku J. Exp. Med.

108. 253-263

8 Stoffel, W., LeKim D. and Tschung. T.S. (1971) 2. Physiol.

Chem. 352, 1058-1064

9 Smith, G.A.. Montecucco, C. and Bennett, J.P. (1978) Lipids 13, 92-94

IO De Hass, G.H. and Van Deenen. L.L.M. (1964) Biochim.

Biophys. Acta 84, 469-471

11 Brockerhoff, H. and Jensen, R.G. (1974) Lipolytic En-

zymes, p. 217, Academic Press. New York

12 Oyarzun, M.J., Clement& J.A. and Baritusso. B., (1980)

Am. Rev. Resp. Dis. 121, 7099721

13 Bartlett, G.R. (1959) J. Biol. Chem. 234, 466-468

14 Jobe, A. (1977) Biochim. Biophys. Acta 489, MO-453

15 Bligh, E.G. and Dyer, W.J. (1959) Can. J. Biochem. Physiol.

37.911-917

16 Jobe, A.H. (1979) B&him. Biophys. Acta 574, 268-279

17 Morrison, R.T. and Boyd, R.N. (1966) Organic Chemistry,

pp. 16, 560, 662, Allyn and Bacon, Boston

18 Williams, M.C. (1983) Am. Rev. Resp. Dis. 127, 271

19 Kuhn, C. (1968) Am. J. Pathol. 53, 809-833

20 Stratton, C. (1978) Cell. Tiss. Res. 193. 219-229

21 Bensch, K., Schaffer, K. and Avery, M.E. (1964) Science

145,1318-1319