Endocytosis of nerve growth factor by ‘differentiated’ PC12 cells studied by quantitative ultrastructural autoradiography

Brain Research, 310(1984) 223-234

223

Elsevier BRE 10318

Endocytosis of Nerve Growth Factor by 'Differentiated' PC12 Cells Studied by Quantitative Ultrastructural Autoradiography ANNA STIEBER. WILLIAM F. HICKEY, RUTH HOGUE-ANGELETFI and NICHOLAS K. GONATAS

Division of Neuropathology, Department of Pathology and Laboratory Medicine, Universi(v of Pennsyh'ania School olMedicine, Philadelphia, PA 19104 (U.S.A.) (Accepted February 21st, 1984)

Key words: nerve growth factor - - PC 12 - - pheochromocytoma - - e ndocytosis - - quantitative autoradiography

The endocytosis of [IZSI]nerve growth factor (NGF) by rat pheochromocytoma cells (PC12 line), previously exposed to the growth factor ('differentiated' or 'primed' cells), was studied by ultrastructural quantitative autoradiography. Cells previously grown in the presence of NGF were incubated at 37 °C with [125I]NGF for periods of up to 24 h. Under these culture conditions, PC12 cells have a rich network of neurites. At the commencement of the experiment, after incubation of cells with [IZSI]NGF for l rain at room temperature, the plasma membranes of perikarya and processes showed similar levels of labeling by [125I]NGF of 0.186 _+ 0.03 grains/,um and 0.152 _+ 0.013 grains/ um respectively. The density of grains per micron of plasma membrane of perikarya reached a plateau between 15 rain to 2 h of incubation of cells at 37 °C with [125I]NGF (0.58 _+ 0.15 grains/~um and 0.65 _+ 0.18 grains/~um, respectively). The endocytosis of [tzSI]NGF in perikarya of cells incubated for 6 h at 37 °C was studied by the "mask' analysis method of Salpeter et al. -~2.At this time, the greatest amount of endocytosis was observed, corresponding to 28.4e4 of total grain counts. The following optimized computed source densities, or relative specific activities _+ standard errors of measurement (S.E.M.), were obtained: plasma membrane, 16.52 _+ 0.86; multivesicular bodies, 9.58 _+ 2.84; endosomes, 5.00 _+ 0.97: smooth vesicles and tubules, 1.66 _+ 0.38; lysosomes, 1.13 + 0.20; mitochondria, 0.46 + 0.10; nuclear membranes or envelopes, 0.32 + 0.14; nuclei, 0.06 _+ II.01: the Golgi apparatus, 0.08 _+ 0.06: and other cytoplasmic elements 0.07 + 0.03. Our findings indicate that smooth vesicles and tubules, endosomes, multivesicular bodies and lysosomes are part of the pathway(sl of endocytosis of NGF. while all other cytoplasmic and nuclear elements, including the nuclear membrane, are not. The heavy plasma membrane labeling of NGF and the relatively low degree of its endocytosis are consistent with the hypothesis that the NGF action is mediated through plasma membrane activated second messenger(s). INTRODUCTION

graphic study of the distribution of [12q]NGF in cells not p r e v i o u s l y e x p o s e d to N G F , ( ' u n p r i m e d ' or 'un-

Studies on the m e c h a n i s m

of action of n e r v e

d i f f e r e n t i a t e d ' ) , we f o u n d that [~2-Sl]NGF occupies

growth factor ( N G F ) are consistent with two princi-

sites on plasma m e m b r a n e s and that it slowly under-

pal hypotheses: (a) N G F binds to plasma m e m b r a n e

goes endocytosis in lysosomesl. Since the PC12 nu-

receptors and acts t h r o u g h a h i t h e r t o u n k n o w n sec-

clei did not s h o w labeling by [1251]NGF, the h y p o t h e -

ond m e s s e n g e r ; (b) N G F has m u l t i p l e c y t o p l a s m i c

sis that N G F is directly i n v o l v e d in the initiation of

and nuclear sites of actions m e d i a t i n g b o t h transcriptional and p o s t - t r a n s c r i p t i o n a l e v e n t s < l 1,14,24.

transcriptional e v e n t s has not b e e n s u p p o r t e d by this study 1.

As originally d e s c r i b e d by G r e e n e and Tischler,

In the p r e s e n t q u a n t i t a t i v e ultrastructural autora-

cloned rat p h e o c h r o m o c y t o m a cells (PC12) have re-

diographic study, we have used an i m p r o v e d m e t h o d

ceptors to N G F and r e s p o n d to the factor in a reversi-

for the c o m p u t a t i o n of grain density distributions to e x a m i n e the binding and e n d o c y t o s i s of [~25I] N G F by

ble fashion by arrest of cell division and the s p r o u t i n g of neuritesl0,~. In a recent q u a n t i t a t i v e ultrastructural a u t o r a d i o -

PC12 cells p r e v i o u s l y e x p o s e d to N G F , ('primed" or ' d i f f e r e n t i a t e d ' PC I2 cells) e~. Particular attention has

Correspondence: A. Stieber, Division of Neuropathology, University of Pennsylvania School of Medicine, Philadelphia, PA 19104,

U.S.A.

000(>8993/84/$03.00 © 1984 Elsevier Science Publishers B.V.

224 been given to the labeling of nuclei, nuclear envelopes, multivesicular bodies, and endosomes, a cytoplasmic organelle which probably plays an important role in receptor mediated endocytosis of various ligands 23. The results of this study will be compared with those obtained with 'unprimed' PC12 cells and with cultured sympathetic neurons, and they will be discussed in the context of the existing hypotheses concerning the mechanism of action of N G F 1,2,4,<11A< 18,24.

MATERIALS AND METHODS

Cell culture The PC12 pheochromocytoma cell line was obtained from Lloyd Greene and cultured according to a previously published methodl,5,12; PC12 cells were placed on collagen coated A C L A R plastic strips and maintained in sterile medium containing 3 /~g/ml NGF for 6 days, at which time an elaborate network of processes was observed ~2.

Preparation and iodination of nerve growth factor Mouse 2.5 S N G F was prepared from adult male submandibular glands by the method of Bocchini and AngelettP. Iodination of N G F and removal of N G F aggregates was performed according to the proce"dure detailed in our previous study I. The specific activity of [125I]NGF was 1200 cpm/fmol. As reported previously, the [125I]NGF obtained with lactoperoxidase-glucose oxidase immobilized on Sepharose enzymobeads (BioRad) could be completely competed off PC12 cells by excess of unlabeled N G F 1.

Electron microscopic autoradiography and quantitative analysis Cells were fixed with 2.5% glutaraldehyde, 1% formaldehyde (from paraformaldehyde), in 0.2 M sodium cacodylate buffer, pH 7.35, containing 0.002% CaC12, for 30 min at room temperature. Cells were post-fixed in 1% osmium tetroxide and 1.5% potassium ferrocyanide, dehydrated in ethanol, embedded in Araldite and processed for autoradiography according to the method of Salpeter as adopted in our laboratoryL9,19-22. Sections, about 80-100 nm thick, were coated with a monolayer of Ilford L4 nuclear emulsion. Sections of PC12 cells not exposed to

[125I]NGF {controls) and secnons oI test cells were mounted on the same collodion-carbon coated slides Autoradiograms were developed with Elon-ascorbic acid according to described methods ~:t A half distance (HD) of 90 nm ~ as used: grain density distributions were calculated according to Salpeter with the aid of a programmed minicomputer (MINC-11. Digital), interfaced with an 'Orthoplex Coordinate Sensor" (digitizer}, by Ladd Research Industries Burlington. VT. Software in use is R T l l and F O R T R A N . Detailed account.~ of the program for the calculation of normalized ~rain density histograms and application of this method of computation to ultrastructural autoradiographv have been published 1,1~'. Using the grain densitx of the first band straddling the plasma membrane, i.;:. 0.5 HD inside and 0.5 H D outside the plasma membrane, the densities are normalized, that is, expressed as proportions of the first band density. In our previous autoradiographic study ot [12511NGF uptake b~ cells not previously exposed to NGF. "unprimed PC12 cells, we used the probability circle method for the analysis ot the distribution of [leSt]NGF over the various cellular organelles or compartments j, As displayed in Tables Ill and IV of this published stud,,, a wide range of values was obtained for a number of cytoplasmic compartments including mitochondria, the Golgl apparatus, smooth endoplasmic reuculum and "other'} The method of computation of grain density distributions we have adopted in the present study eliminates false labeling from scattering of ~adioactivitv from adjacent heavily labeled sources ~-e. Briefly, this method of analysis involves the comparison between actual grain distributions in autoradiographs and randomly generated grain distributions obtained from the same autoradiograms and with identical criteria for grain sources and compartments The comparison between actual and generated gram distributions ' . . . leads to a determination of the most likelv levels of radioactivity in the tissue':: In complex cases. such as this analysis, the "best fit" between observed and generated grain density distributions can be estimated with the help of a computer z:. A computer program was adopted according to Salpeter et al. e: For the generation of grain density distributions we used the 6 masks (overlays) for ~:~l trom the study bv Salpeter el al., which were copied onto transparent plastic at a final magnification of *~ ~L(I{/{I.

225 For the matrix analysis, the following 10 source compartments were used: lysosome (LYS), multivesicular body (MVB), Golgi apparatus (GOLGI). nuclear envelope (NM), mitochondrion (MIT), smooth vesicles and tubules (SMOOTH V and T), plasma membrane (PM), other cytoplasmic (OTHER CYT), and nucleus (NUC), membrane bound cytoplasmic vacuoles. 0.25- I um in diameter, containing a few vesicles but otherwise appearing empty were identified as endosomes (ENDO) -'3. The matrix contained 16 grain compartments: NM, 3 HD outside NM (NM 3 HD OUT), 3 HD inside NM (NM 3 HD IN), inside nucleus over 3 HD from NM (OTHER NUC), compact cisternae of the Golgi apparatus (GOLGi), Golgi associated vesicles (GOLGI AS VES) which are clusters of vesicles at the trans aspect of the Golgi apparatus within a 1 ,urn radius from the innermost cisterna of the Golgi apparatus, endosome (ENDO) which describes a membrane bound. empty appearing vacuole containing none to several

small vesicles (Figs. 1 and 2), 1 HD outside endosome (ENDO 1 HD OUT), multivesicular body (MVB) which describes a typical MVB with dense matrix and several vesicles (Fig. 2), 1 HD outside MVB (MVB 1 HD OUT), lysosomc (LYS). 1 HD outside LYS (LYS 1 HD OLT), mitochondrion (MIT), smooth vesicles and tubules (SMOOTH V and T), PM. 5 HD inside and outside the plasma membrane (PM + 5 HD), and other cytoplasmic constituents (OTHER CYT). All areas of the Golgi apparatus and of the Golgi associated ~esicles were outlined by a marking pen so that identical regions of these two compartments were used for the analysis of actual and generated grain cotnpartments. Grain compartments were classified according to the following 5 priority groups: (1) PM. PM _+ 5 HD: (2) NM, GOLGI, GOLGI AS VES, ENDO, MVB, LYS, SMOOTH V AND T: (3) NM 3 HD IN, NM 3 HD OUT, ENDO 1 HD OUT, MVB 1 HD OUT, LYS 1 HD OUT: (4) O T H E R NUC. MIT: (5) OTH-

Fig. 1. "Primed' PC12 cell incubated with [I:51]NGFfor 6 h at 37 °C. Note heavy labeling of plasma membrane,, ol adiaccnt two cclls: G: Golgi apparatus. Autoradiogramexposure. 14weeks. × 28.000.

226 E R CYT (see Table V). The highest priority is 1 and the lowest 5. A grain falling within two compartments of different priorities was ascribed to the compartment of the higher priority. If a grain fell in two compartments of the same priority, each compartment was given 0.5 credit. Autoradiograms were exposed for 14 or 23 weeks. For quantitative comparisons, all exposures were normalized to 14 week values using the decay table for 1251. Lengths of plasma membranes of perikarya and processes were calculated with the digitizer using an appropriate program. Lengths of plasma membranes were also measured with a cartographer's wheel; essentially similar values were obtained with the two methods.

Experimentalprotocol Cells exposed to N G F for 6 days were rinsed by dipping 3 times in cold (4 °C) sterile culture medium without NGF. Subsequently cells were incubated in 4 ml sterile culture medium without N G F at 37 °C for 1 h. These two treatments are sufficient to remove ex-

ogenous N G F 7. At that time cells were m o v e d into sterile petri dishes containing [lzSI]NGF in culture medium. Each petri dish contained one A C L A R strip, 1 x 1 cm, bearing PC12cells, and 2 m l o f medium. All petri dishes contained [125t]NGF at a concentration of 1,944,000 cpm/ml, or 1 nM. One petri dish was left for 1 rain at the temperature of the laboratory before rinsing and fixing for autoradiography. The rest of the 4 petri dishes were incubated in full culture medium for 15 min, 2, 6 and 24 h before rinsing and fixation. The experiment was terminated by dipping the cover slips for 30 s in 5 successive dishes containing 3 ml each of R P M I at 4 °C. Each wash was monitored for radioactivity and rinsing was stopped when the last one or two washes gave low counts: the actual counts per min per ml obtained for the last wash were 308 (1 rain), 564 (15 min), 654 (2 h), 618 (6 h) and 595 (24 h). To avoid damage and/or detachment of cells from A C L A R strips, the radioactivity of the plastic strip was not counted. In view of the considerable variation of densily of perikarya and network of processes among the various plastic

Fig. 2. 'Primed' PC12 cell incubated with [t25I]NGFfor 6 h at 37 °C. E: endosome: M: multivesicular body. d: dark core vesicles. Autoradiogram exposure, 12 days. x 27,000.

227 PM

strips, we did not attempt to measure the radioactivity or cell protein contents in sister cultures.

OUT. IN RESULTS 1.0

Plasma membrane binding and endocytosis

~_lZSl~NGF-PCI2 + 15 min 37°C

in Figs. 1 and 2 some common properties of PC12 cells exposed to N G F are illustrated. These features are: a prominent Golgi apparatus, numerous membrane bound vacuoles which appear empty or contain one or two small vesicles, hitherto to be referred to as endosomes, multivesicular bodies, lysosomes, and dense core vesicles. In addition we observed numerous processes, usually in bundles of 2 to 5, and usualIv containing large n u m b e r s of dark core vesicles.

0.5

These processes were adjacent to perikarya or at a considerable distance from cell bodies. In Figs. 3 - 7 , grain density distributions in cell bodies, normalized to 1 at the plasma m e m b r a n e , are displayed; Fig. 3 can serve as a baseline for subsequent experiments conducted to explore the degree of endocytosis of [~2q]NGF.

Comparisons

between

histograms

, 2

0

2

4

6

" " " 20~

in

Figs. 3 - 7 are consistent with the conclusion that there is a low level but significant endocytosis of N G F

Fig. 4. Histogram of grain density distributions in 'primed' PC12 cells incubated with []2~I]NGFfor 15 rain at 37 °C. For details, see legend of Fig. 3.

PM "IIN

PM

OU.TI,N

1.0[I251]NGF_PCI2+ I min 2 0 ° C

1.0 [i2 51] NG F- PCI2 + 2 hrs 37 ° C

0.5 0.5

3 2

O

2

4

6

..........

~(~

Fig. 3. Histogram of grain density distributions normalized to one at the plasma membrane (PM). IN, within the cytoplasm: OUT, outside the cell. Divisions of the abscissa are in increments of 2 half distances (HD). HD is 90 nm. "Primed' PC12 cells were incubated with [t25I]NGFfor 1 rain at 20 °C.

52

O

2

4

6

............

2 0 ~jl

Fig. 5. Histogram of grain density distributions in "primed' PC12 cells incubated with [t2511NGFfor 2 h at 37 °C. For details see legend of Fig. 3.

228

PM OUT__I ~

PM OUT IN

1.0

1.0

~ZS~NGF- PCI2+

O 'I]NGF-PC Z+

6 hrs 37* C

24 hrs 37'*C

0.5

0.5

62

0 2 4 6

20~ ~b

Fig. 6. Histogram of grain density distributions in primed' PC12 cells incubated with [1~-5I]NGF for 6 h at 37 °C. For details see legend of Fig. 3.

72

0

2

4

6

-

20~

Fig 7. Histogram of grain density distributions in "primed' PC12 cells incubated with [125I]NGFfor 24 h at 37 °C. For details see legend of Fig. 3.

quantitative impression of the degree of endocytosis of [12SI]NGF, the percentages of grains inside cells

ma m e m b r a n e . This interpretation is consistent with the universal curve of radioactivity scatter from a linear source by Satpeter 20,=1. At this point of the study.

over 4.5 H D or over 5.5 H D were calculated after the

we compared the endocytosis of [~-'SI]NGF by "un-

in 'primed' PC12 perikarya. In order to gain a more

background and the values for the 1 min exposure

primed' cells using the raw data presented in our pre-

were subtracted (Fig. 3). These percentages are pre-

vious

sented in Table 1. The results indicate that only a small percentage, about 25%, of radioactivity is asso-

Tables I and II show that despite some minimal dif-

ciated with intracellular NGF. The histograms are

studies. "unprimed' cells internalize a much greater

consistent with the conservative interpretation that all extracellular grains up to 3.5 H D and intracellular grains up to 4.5 or 5.5 H D from the plasma membrane represent scatter from the heavily labeled plas-

proportion of N G F than primed PC 12 cells.

paper.

Table It 1.

Compartsons

between

ferences in the ume intervals and HD used in these

B i n d i n g o f [125I]NGF on perikarTa a n d processes.

In Table I l l we summarize the n u m b e r of grains per um of plasma m e m b r a n e of perikarya and processes

TABLE I Numbers and percentages of grains from []eSl]NGF internalized by primed' PC12 Lells

Total number of grains include all intraceUular grains and all extracellular grains within 3.5 HD from the plasma membrane. Time~temperature

Total number of grains

Over 4.5 HD inside cell

c)~

Over 5.5 HD inside cell

%

Corrected over 4.5 HD ~q4 J

Corrected over 5.5 HD ( 97

1 min, 15 min, 2 h, 6 h, 24 h,

319.0 803.0 1947.2 1833.2 1103.4

41.3 193.1 755.8 756,8 428.8

12.9 24.0 3g.8 41 3 38.9

40.9 168.g 7t 1.5 7(12.5 398.4

12.~ 21.0 36.5 38.3 36.1

~ 11. l 25.9 28.4 26.0

u 8.2 23.25.5 23.3

20 °C 37 °C 37 °C 37 °C 37 °C

229 TABLE II Number,s and percentages of grains from [1:YI]NGF internalized by 'unprimed' PC12 cells

4 ,~ 4.

. [

i

Total number of grains include all intracellular grains and all extracellular grains within 3 [ ID from the plasma membrane. Tim e temperature

2 h, 4 °C 2 h, 37 °(" 6 h, 37 °C 26 h, 37 o(-

To tal number ql'grain,s

Over 5 HD inaide cell

770 1802 1480 2373

156 660 768 367

after e x p o s u r e of A C L A R

~}

2t).3 36.6 51.9 57.6

Corrected orer 5 ftD insMe cell f c~) 0 18.(} 34.2 40.4

strips for 1 min at r o o m

t e m p e r a t u r e . W e o b t a i n e d similar levels of labeling by [I:SI]NGF of p e r i k a r y a and processes. Using the

'I

kR,

~g~7

Fig. 8. Number of grains per/*m of perikaryal plasma membrane length in "primed' PCI2 cells incubated with [125I]NGF for 1 min at 20 °C (0 time) and 15 min, 2, 6 and 24 h, at 37 °C. Based on the following counts of plasma membrane lengths and grains: 0 time, 1670.86,um, 283 grains: 15 min, 1748.81/*m, 598 grains; 2 h, 1817.80 um, 1116 grains: 6 h, 1523.01 urn, 997 grains; 24 h, 1479.88/¢m, 625 grains.

same a p p r o a c h as d e s c r i b e d in the f o o t n o t e in T a b l e IIl, we calculated the n u m b e r of grains per length of

zontal line. A t no time during the e x p e r i m e n t was

plasma m e m b r a n e of p e r i k a r y a for the entire experi-

t h e r e a histogram pattern e v e n r e m o t e l y suggestive

m e n t , Fig. 8. A c c o r d i n g to this curve, the density of

of labeling of the n u c l e a r e n v e l o p e .

grains per plasma m e m b r a n e length r e a c h e s a plateau b e t w e e n 15 rain and 2 h and declines s o m e w h a t later at 24 h.

Endocytosis o f / Iz~I ] N G F We analyzed the intracellular grain density distributions and the r a d i o a c t i v e source c o m p a r t m e n t s in

Labeling o f the nuclear envelope (nuclear m e m b r a n e ) G r a i n density distributions w e r e calculated for the area within 3 H D f r o m the n u c l e a r m e m b r a n e on the cytoplasmic side ( N M - O U T )

cells e x p o s e d to [1251]NGF for 6 h at 37 °C (Figs. 6 and 8). A t this t i m e the binding of [I:51]NGF on plasma

membranes

has

already

reached

a

plateau

and across the entire

nucleus ( N M - I N ) . T h e entire p e r i m e t e r of the nucle-

NM

ar m e m b r a n e was used and c y t o p l a s m i c areas including heavily labeled sources w e r e not e x c l u d e d f r o m the grain c o m p u t a t i o n s . Values f r o m 0.5 H D inside to 0.5 H D outside the n u c l e a r m e m b r a n e w e r e nor-

4.2

OUT IN

20 115

i ~rZSI~NGFI mmPCI2+20°C 21~(

NM

IN 15 min 37°C

malized to one. Results are s u m m a r i z e d in Figs. 9 11 in which the b a c k g r o u n d is indicated by the horiTABLE I11

I0

I.C

Den~ily ol [1-'sI/NGF on pla,sma membranes ()/perikao, a and processe~ aj?er incubamm qf cells fi~r 1 rnin at room temperature For calculation of plasma membrane allocated grains, all grains within 4 liD from plasma membranes were counted (see Fig. 3). A 0.5 credit to perikarya and processes was given if grain center was within 4 HD from both. S.E.M., standard error ol n l e a s t l r e i ] l c l l t ,

Cells Processcs

(;rain.s

Micra

Grains~urn S.E.M.

312.5

1683.5

0.186

+ 0.030

11 74 S

7021.1

0.152

+ 0.013

0.5

__ t 92 0 2 4 ~

0.5 2 0 2 4

Fig. 9. Histograms of grain density distributions normalized to one at the nuclear membrane (NM) or envelope, in 'primed" PCI2 celts exposed to [125I]NGF for 1 rain at 20 °C and 15 rain at 37 °C. Horizontal line above abscissa marks the background. IN, inside nucleus: OUT, inside cytoplasm.

230 NM

z°I OUTNM

,,...1

1 02

o

2

4

~x

Fig. 10. Histograms of grain density distributions, normalized to one at the nuclear membrane (NM), in 'primed' PC12 cells exposed to [~25I]NGFfor 2 and 6 h, at 37 °C. For details see legend of Fig. 9. (Fig. 8); also, maximal endocytosis of [nSI]NGF has occurred (Table I). In Table IV we present the actual matrix and in Table V we summarize the results of the ' m a s k ' analysis p e r f o r m e d according to Salpeter et al. 22. The low total Z2 value (7.0512) indicates that the c o m p u t e r

o

o o o ~

o ~ o o ~ o

~ o o

oo~

i

'2

Fig. 11. Histogram of grain density distributions normalized to one at the nuclear membrane (NM) in 'primed' PC12 cells exposed to [~25I]NGFfor 24 h, at 37 °C. For details see legend of Fig. 9.

ooo~

.~

PM+ 5 HD

grains 533 Generated grains (from mask) 383.77 Computed grain distribution from optical source density 533.03 Z-"components 0.000001 Relative compartment sizes (%) 9.80 Relative grain distribution from computed grains (%) 45.23 Relative label ( ~:~ grains/ 17~ area) 4.62

Actual

11.32.

20

146

24.83

I).94

3.73

2.11

11.56

48.34

13.69

2.36

1.24

1.16

0.94

NM3HI) IN

8

NM

Grain compartments

T o t a l z 2 - 7.11512: P

Computer analvs'is of data

TABLE V

0.35

1.45

4.16

2.114

17.119

163

23

NM3HD 0 UT

0.21

4.72

22.46

0.13

55.66

879.06

53

OTHER NUC.

1.22

4.77

3.9111

0.14

56.21

152.80

59

SMOOTtt V and T

24.5

25.86

2.15

2.67

1.24

1.96

2.19

1.12

0.009 0.07

31.47

7.(I1/

0.84

0.12

0.12

9.92

4.84

11

LYS1 MVB HD OUT

48.67 43.77

32

I.YS

5.92

/).71

0.12

0.09

8.38

4.68

7.5

7.95

1.51

0.19

11.17

17.75

7.46

16

MVB 1 END() HD O U T

4.09

0.94

111.23

11.32

ll.12

9.13

13

0.53

1.00

1.87

/I.{)113

11.8

73.17

12

ENDO I GOLGI HD OUT

1/.51

{I.311

0.59

0.63

3.48

23

2

(;OLGI A S VES.

312

OTHER CYT.

11.88

4.14

4.68

0.005

48.84

1t.59

26.25

44.54

0.02

309.38

183.1.14 1743.20

49

MIT

r~ %0

232 generated 'best-fit' between actual and generated grains is compatible with meaningful interpretations of the results. The relative labels (% grains over % area) extrapolated from relative compartment sizes (%) and relative grain distribution from computed grains (%J show that endosomes, multivesicular bodies, lysosomes, plasma membranes and smooth vesicles and tubules show significant radioactivity; the nuclear membrane (envelope), nucleus, the Golgi apparatus and associated vesicles, mitochondria and 'other' show low and probably insignificant labeling. In Table VI, the optimized computed source density, a valid indicator of the relative specific activity of each compartment is displayed. Five principal sources of radioactivity are identified: smooth vesicles and tubules (1.66 + 0.38), lysosomes (1.13 ± 0.20), endosomes (5.00 + 0.97), multivesicular bodies (9.58 _+ 2.84) and the plasma membrane (16.52 ± 0.86). The nuclear membranes, nuclei, mitochondria, the Golgi apparatus, and 'other' cytoplasmic compartments are not labeled by [125I]NGF. The optimized source densities of multivesicular bodies and endosomes were not significantly different using Student's t-test. Lysosomes and smooth vesicles and tubules were both found to be different from endosomes using the same test. The difference between the optimized source densities of nuclear membrane and 'other' cytoplasmic elements was not significant using Student's t-test. DISCUSSION The noradrenergic clonal line of rat adrenal pheochromocytoma (PC12) has been useful in studies of the mechanism of action of NGF m,u. It has been shown by biochemical methods involving cell 'fractionation that these cells have high affinity recep-

tors to NGF on their plasma and nuclear mere= branes 2a. The PCI2 cells respond to NGF by apparent morphological and biochemical Mifferentiation'm, 24. After exposure to NGF, PCI2 cells cease dividing and form long neurites I~ We have shown that 'unprinied' PCI2 cells ha~e an elaborate network of smooth endoptasmic reticulum and of Golgi apparatus-complex with acid phosphalase and thiamine pyrophosphatasc m several cisternaeS. We have also shown that these cells after prim, ing with NGF, display increased localization of internalized wheat germ agglutinin ~WGA), c0vatently linked with the marker enzyme horseradish perox~ idase (HRp)I-L in one or two cisternae at the trans aspect of the Golgi apparatus; furth:ermore, in these NGF primed cells the Gotgi apparatus increases in area by' 7()q>} over the area occupied by the Golgi apparatus in PCI2 cells not previously primed b~ NGF 1:. In our previous ultrastructural autoradiographic study of the endocytosis of [125I]NGF by ~unprimed' PC12 cells, we presented evidence and discussed the tow non-specific binding and low levels of degradation of the iodinated preparations of NGF used in these and in the present experiments ~. According to this published stud},, in "unprilned" PC12 ceils. [1251]NGF undergoes endocytosis quite slowly and it localizes in lysosomes and probably m smooth vesb cles ~. The results of the present quantitative autoradiographic study conducted with a more precise method for the computation of intracellular labeling by [12q]NGF are as follows: multivesicular bodies and endosomes were 5-9.5 times more heavily labeled than lysosomes: lysosomes and smooth vesMes and tubules showed similar low but significant levels of labeling by [12SI]NGF; nuclei, the nuclear envelope (computed by the "matrix' method and by histo-

TABLE Vl Source compartments NM

LYS

MVB

ENDO

G O L G I SMOOTtt MIT V and T

.%~UC

OTHER ('YT.

Optimized computed source density 16,52

0.32

1.13

9.58

5.00

0.08

1.66

[!.46

0.06

1.07

Standard error ranges

0.14

0.20

2.84

[).97

(I.06

0.38

o. 10

00t

~{

PM

0.86

233 grams), mitochondria and 'other' cytoplasmic compartments were not labeled (see Table VI and Figs. 9 - l l ) . The computation of endocytosis of [1251]NGF was done after cells were exposed for 6 h at 37 °C to NGF, at which time the number of plasma membrane binding sites has already reached a plateau (Fig. 8), and the amount of internalized NGF was probably at a steady state (Figs. 3-6). At this time, the Golgi apparatus and Golgi associated vesicles did not show any labeling by NGF. Recently, a great deal of attention has been paid to pathways of adsorptive and receptor mediated endocytosis s. Some of the ligands studied have been found in cisternae of the Golgi apparatus s,9,12. It has been proposed that the transport route through the Golgi apparatus is important for membrane repair and recycling. The absence of labeling of the Golgi apparatus by NGF indicates that the endocytosis of ligands, such as lectins, in the Golgi apparatus is not a general property of adsorptive or receptor mediated endocvtosisS.9,13,17. In our previous qualitative light microscopic autoradiographic study of binding of [I:5I]NGF by cultured sympathetic neurons, both perikarya and processes showed labeling~3. These results are consistent with the present quantitative analysis of [~251]NGF over perikarya and processes which showed approximately similar numbers of NGF binding sites (Table III). Our results obtained from PC12 cells are essentially similar to those of Claude et al. obtained from cultured rat sympathetic neurons, and of Rohrer et al. with PC12 cells
Our results are different from those reported by Bernd and Greene who reported labeling by NGF of nuclear membranes of PC12 cells:, The most likely explanation of this significant difference between the findings of Bernd and Greene and ours is to be found in the different methods of computation of grain density distributions used in these two studies 2. Bernd and Greene in order to eliminate grain scatter to the nuclear envelope from adjacent heavily labeled lysosomes and plasma membranes, computed grain density distributions excluding grains within 3 um of plasma membranes or within 3 um of plasma membranes and lysosomes:. In our computations the entire perimeter of the nuclear membrane was examined, We employed a method for the computation of grain density distributions which corrects for false labeling from adjacent heavily labeled sources. We believe that our non-selective approach is more valid and that the nuclear membrane does not show any labeling, since neither the histograms of grain density distributions (Figs. 9-11) nor the "mask" analysis (Table VI) showed any significant levels of labeling. The observed heavy labeling by [I:51]NGF of multivesicular bodies coincide with the findings by McKanna et al. with ferritin labeled epidermal growth factor (EGF) 15. McKanna et al. have suggested that in multivesicular bodies, intact EGF is still associated with its receptor, while in lysosomes EGF is being degraded 15. Our studies do not allow this distinction: however, the hypothesis of McKanna et al. is interesting and merits a future exploration. Another cytoplasmic organelle which has received attention recently because of its suggested role in receptor mediated endocytosis is the endocytic vacuolc or endosome (Fig. 2) 17-:3. In this, presumably pre-lysosomal compartment, the receptor is dissociated from the ligand; subsequently the receptor is presumably retrieved for its reinsertion in the plasma membrane, while the ligand is advanced to the lysosomes for degradation. In PC12 cells, endosomes showed levels of NGF labeling similar to those found in multivesicular bodies (Table VI). It is not clear from our experiments with NGF and from other studies whether the ligand is transported from endosomes to lysosomes17,23. The relationship between thc endocytosis of [1251]NGF and its mechanism of action is not clear e4 Translocations of NGF through the various endocytic

234 c o m p a r t m e n t s m a y be l i n k e d to its m e c h a n i s m of ac-

s o m e s and l v s o s o m e s , which are c o m p a r t m e n t s for li-

tion. O n the o t h e r h a n d , N G F e n d o c y t o s i s m a y be re-

g a n d p r o c e s s i n g and d e g r a d a t i o n .

lated only to the d o w n r e g u l a t i o n of cell surface rec e p t o r s to N G F and their e v e n t u a l d e g r a d a t i o n in ty-

ACKN OWLEDGEMENTS

s o s o m e s 24. O u r study is c o n s i s t e n t with t h e h y p o t h e sis that t h e N G F action is m e d i a t e d solely t h r o u g h a plasma membrane

T h e a u t h o r s wish to e x p r e s s t h e i r g r a t i t u d e to Dr.

activated second messenger(s).

M i r i a m S a l p e t e r for h e r n u m e r o u s c o n s u l t a t i o n s o n

This c o n c l u s i o n is b a s e d o n the findings of a h e a v y

the application of h e r m e t h o d s to this study. O u r

plasma membrane

thanks to Ms E l l e n S c h w a r t z for typing t h e m a n u -

labeling t h r o u g h o u t the e x p e r i -

m e n t and of a r e l a t i v e l y low d e g r e e of N G F e n d o c y -

script.

tosis which i n v o l v e d m u l t i v e s i c u l a r bodies, e n d o -

NS07064 and NS13201

REFERENCES

and morphometric methods. J. Neurocvtol.. 12 t19831 751-756. 13 Kim, S. U., Angeletti-Hogue. R. and Gonatas, N. K.. Localization of nerve growth factor receptors in sympathetic neurons cultured in vitro. Brain Research. 108 (1979) 602-608. 14 Levi-Montalcini. R.. Developmental neurobiology and the natural history of nerve growth factor Ann. Rev. Neurosci. 5 (19821 341-362. 15 McKanna. J. A.. Haigler. H, T, and Cohen. S., Hormone receptor topology and dynamics: morphological analysis using ferritin-labeled epidermal growth factor. Proc. nat. Acad. Sci. U.S.A.. 76 (1979) 5689-5693 16 Merten. C., Stieber, A.. Grand, E.. Gonatas. J. and Gonatas. N. K.. The use of a minicomputer for quantitative ultrastructural autoradiography, L Histochem. Qvtochem., 29 (19811 585-587. 17 Pastan. I. H. and Witlingham. M ('.. Receptor-mediated endocytosis of hormones in cultured cells, Ann. Rev. Physiol.. 43/1981) 239-250. 18 Rohrer. H,. Sharer. T., Korsching, S, and Thoenen. H.. Internalization of nerve growth factor by pheochromocytoma PC12 cells: absence of transfer to the nucleus. J. Neurosci.. 2 (1982) 687-697. 19 Salpeter, M. M. and Bachmann, L., Autoradiography with the electron microscope. A procedure for improving resolution, sensitivity and contrasts. J. Cell Biol.. 22 (1964) 469-477. 20 Salpeter, M, M,, Bachmann. L. and Salpeter. E E., Resolution in electron microscope autoradiography, J. Cell Biol. 41 (1969) 1-20. 21 Salpeter. M. M and Bachmann, L. Autoradiography. In M. A. Hayat (Ed.), Principles and Techniques f o r Electromtcroscopy, Van Nostrand Reinhold, New York, 1973. pp 221-278. 22 Salpeter. M, M.. McHenry. F. A. and Salpeter. E. E.. Resolution in electron microscope autoradiography. V. Application to analysis of autoradiographs. J. Cell BioL. 76 1978) 127-145. 23 Tycko. B. and Maxfield. F. F., Rapid acidification of endocytic vesicles containing a-2 macroglobutin. Cell. 28 [ t982~ 643-651. 24 Yankner B. A. and Shooter. B. M., Thc biology and mechanism of action of nerve growth factor, Ann. Rev Biochem.. 51 (19821845-~fiR

1 Angeletti-Hogue, R., Stieber, A. and Gonatas, N. K., Endocytosis of nerve growth factor by PC12 cells studied by quantitative ultrastructural autoradiography, Brain Research, 241 (1982) 145-156. 2 Bernd, P. and Greene, L. A., Electron microscopic radioautographic localization of iodinated nerve growth factor bound to and internalized by PC12 cells, J. Neurosci, 3 (1983) 631-643. 3 Bocchini, V. and Angeletti, P. U., The nerve growth factor: purification as a 30,000 molecular weight protein, Proc. nat. Acad. Sci. U.S.A., 64 (1969) 787-794. 4 Bradshaw, A. A., Nerve growth factor, Ann. Rev. Biochem., 47 (1978) 91-216. 5 Burchanowski, B. J., Hogue-Angeletti, R., Stieber, A., Gonatas, J. and Gonatas, N. K., The ultrastructure and cytochemistry of PC12 cells with special reference to the Golgi apparatus and GERL, J. Neurocytol., 11 (1982) 323-333. 6 Claude, P., Hawrot, E., Dunis, D. A. and Campenot, R.. Binding, internalization and retrograde transport of 125I-nerve growth factor in cultured rat sympathetic neurons, J. Neurosci., 2 (1982) 431-442. 7 Frazier, W. A., Boyd, L. F. and Bradshaw, R. A., Properties of the specific binding of [125I]Nerve Growth Factor to responsive peripheral neurons, J. biol. Chem., 294 (1974) 5513-5519. 8 Gonatas, N. K., The role of the neuronal Golgi apparatus in a centripetal membrane vesicular traffic, J. Neuropath. exp. Neurol., 41 (1982)6-17. 9 Gonatas, N. K., Stieber, A., Gonatas, J., Mommoi, T. and Fishman, P. H., Endocytosis of exogenous GM t ganglioside and cholera toxin by neuroblastoma cells, Molec. cell. Biol., 3 (1983) 91-101. 10 Greene, L. A. and Tischler, A. S., Establishment of a noradrenergic clonal line of rat adrenal pheochromocytoma cells, Proc. nat. Acad. Sci. U.S.A., 72 (1976) 2424-2428. 11 Greene, L. A. and Shooter, E. M., The nerve growth factor: biochemistry synthesis and mechanism of action, Ann. Rev. Neurosci., 3 (1980) 353-402. 12 Hickey, W. F., Stieber, A., Angeletti-Hogue, R., Gonatas, J. and Gonatas, N.K., Nerve growth factor induced changes in the Golgi apparatus of PC-12 rat pheochromocytoma cells as studied by ligand endocytosis, cytochemical

Supported

by

USPHS

Grants

NS05572.