DevelopmentalBrain Research, 18 (1985) 57-78 57 Elsevier

BRD 50147

Effects of Monocular Closure at Different Ages on Deprived and Undeprived Cells

in the Primate Lateral Geniculate Nucleus

M. P. HEADON, J. J. SLOPER, R. W. HIORNS 1 and T. P. S. POWELL

Departments of Human Anatomy and 1Biomathematics, University of Oxford, South Parks Road, Oxford ( U. K.)

(Accepted August 10th, 1984)

Key words: visual deprivation - - lateral geniculate nucleus - - primate - - hypertrophy

This study has examined the effects of monocular visual deprivation on cells in the lateral geniculate nucleus of the primate by com- paring the sizes of cells in deprived and undeprived LGN laminae of experimental rhesus monkeys with those of cells in the corre- sponding laminae of normal animals. A number of conclusions may be drawn from this comparison: (1) monocular visual deprivation has major effects on cells in the undeprived LGN laminae and these vary with age at closure; (2) the initial effect of monocular closure from birth is to cause marked hypertrophy of undeprived parvocellular cells with little shrinkage of the deprived parvocellular cells, whereas late monocular closure (after 2 months of age) causes marked shrinkage of both undeprived and deprived parvocellular cells; (3) following monocular closure at birth, the LGN abnormality continues to evolve until at least 3 months of age, with a marked paral- lel shrinkage affecting both deprived and undeprived parvocellular cells. The initial hypertrophy of the undeprived cells is reversed and the deprived cells become smaller than normal; (4) cells in the monkey LGN are sensitive to visual deprivation until about 1 year of age, much later than previously thought. Visual experience, however, modifies this sensitivity so that the effects of monocular visual deprivation are both qualitatively and quantitatively different at different ages; (5) there are important differences between the sus- ceptibility of cells in the magnocellular and parvoceUular laminae to visual deprivation; and (6) actual shrinkage of cells to markedly below normal size occurs and the smaller size is not simply failure of growth.

INTRODUCTION

This study describes the changes which occur in

cells of the pr imate lateral geniculate nucleus (LGN)

following monocular depr ivat ion by comparing the

sizes of cells in both depr ived and undepr ived L G N

laminae of exper imenta l animals with sizes of cells in

the corresponding laminae from a series of normal

animals. Previous studies in the monkey have meas-

ured changes in depr ived cells by using the unde-

prived cells in the corresponding laminae of the L G N

of the other hemisphere of the same animal as a con-

troiS, 2°,3°-32,34. Compar isons be tween exper imenta l

and normal animals have not been used previously

because it had been thought that there was too much

variabil i ty between individual animals to allow reli-

able comparisons to be made. However , this impres-

sion of variabil i ty between monkeys has been largely

based on sizes of cells in the undepr ived laminae of

exper imental animals. Following measurements of

L G N cells made in a number of normal animals as a

control for the effects of binocular depr ivat ion 9, it be-

came apparent that the variabi l i ty in size of cells from

normal animals was markedly less than that for cells

from undepr ived laminae of exper imenta l animals.

Reliable comparisons could therefore be made be-

tween different animals. Such comparisons have now

shown that undepr ived L G N cells undergo major size

changes in response to visual depr ivat ion, and that

much of the apparent variabil i ty in their size was in

fact part of a series of changes produced in unde-

prived laminae by visual deprivat ion. At certain

times the changes in undepr ived laminae may be the

major effect of depr ivat ion or may mask large

changes in the depr ived laminae. The changes pro-

duced by visual depr ivat ion are more complex, and

the L G N is sensitive to visual depr ivat ion for much

longer than had been previously thought , and this has

Correspondence: M. P. Headon, Department of Human Anatomy, South Parks Road, Oxford OX1 3QX, U.K.

0165-3806/85/$03.30 © 1985 Elsevier Science Publishers B.V. (Biomedical Division)

58

a number of implications for our understanding of

both normal and abnormal deve lopment of the pri-

mate visual system.

The data on L G N cell sizes in normal monkeys

which are used as a baseline in this paper are de-

scribed in detail in the preceding paper 14. Prelimi-

nary reports of some of these findings have appeared previously10-13.

MATERIALS AND METHODS

Under open e ther or Nembuta l anaesthesia , the

lids over one eye were sutured together in 29 infant

TABLE I

Details of monkeys used

Asterisks indicate animals with constant period of closure = 60 days.

Age at Duration of Age at closure closure per fusion (days) (days) (days)

OM236 4 2 6 OM215 0 5 5 OM167 1 l0 11 OM407 0 16 16 OM380 0 21 21 OM399 1 27 28 OM403 0 34 34 OM397 0 41 41 OM405 0 55 55 OM404 0 70 70 OM166 6 84 90 OM210 3 403 406 OM277 1 1852 1853

OM198 32 19 51 OM413 30 58 88* OM267 20 967 987

OM171 58 20 78 OM408 59 61 120" OM200 68 321 389

OM374 100 60 160" OM168 102 60 162" OM194 103 239 342 OM212 114 400 514 OM373 127 124 251

OM193 215 59 274* OM363 275 60 335* OM385 291 62 353* OM370 368 62 430* OM376 547 59 606*

rhesus monkeys (Macaca mulatta) at varying ages

and the infants re turned to their mothers for different

survival periods (Table I). These animals fall into

two major groups. In 13 animals closure was started

early, usually within 48 h of birth, and progressively

longer survival per iods were given. In the remaining

16 animals closure was per formed after varying peri-

ods of normal visual experience. Nine of these mon-

keys were given an approximate ly constant per iod of

closure of about 60 days to give an indication of

changes in sensitivity to the effects of monocular clo-

sure with increasing periods of preceding normal vi-

sion. Details of methods of perfusion, histological

prepara t ion and techniques of measurement are

given in the preceding paper 14 and have been con-

stant throughout the study. The same observer meas-

ured the areas of 50 cells in each lamina of each L G N

and the mean and s tandard deviat ion of these were

calculated. For making comparisons between groups

of animals this mean has been t reated as a single

value for calculation of group means and s tandard

deviations. Standard deviat ions for single animal

measurements (Tables II , IV, VII and IX) thus refer

to the variabil i ty of the 50 cells measured and have

been used for statistical comparisons of depr ived and

undeprived laminae within individual animals. For

the tables relating to groups of animals (Tables III , V

and VIII) the s tandard deviat ions refer to the varia-

tion between means for different animals, and these

have been used for the statistical comparisons made

between groups of animals. The normal data used to

calculate the changes in this paper are from Table IV

of the preceding paperl4, and refer to the whole

group of normal animals. Results for each of the ex-

per imental groups have also been tested against an

appropr ia te ly aged subgroup of the normal animals.

This gave closely similar results, except in the case of

magnocel lular cells of infant animals where both

comparisons are given. All statistical comparisons

have been made using Student ' s t-test.

RESULTS

For each animal the areas of 50 cells were meas-

ured from laminae I - V I of both LGNs. Because

there are important differences between responses of

cells in the magnocel lular laminae (I and II) and par-

vocellular laminae ( I I I - V I ) to monocular closure,

59

8 E

r-

.9,o @

._>

c I,,.

%

0

- 1 0

-20

-30

-40

-50

. . . . . . . Magnocellular , . P a r v o c e l l u l a r

"'/Z

I l i t J / / i a ,¢ / i ,,J

A

%

3O

o C

• 2o

-~ 10 E

c 0

._> - 1 0

- 2 0

x: -30 U

N

05 -4O

¢ \ /

/

\~_ - - _ i ll

• . . . . . • Undepr ived m a g n o c e l t u l a r

o . . . . . o Depr ived m a g n o c e i t u t a r

= _- U n d e p r i v e d parvocettuLar

c ~ D e p r i v e d parvocet tu tar

X " - - -"//

_ - - ~ . . . . . / / . . . . . o.

"y/.

B

I i J L i i / / I I / / L l -500 20 40 60 80 100 400 420 1840 1860

A g e ( d a y s )

Fig. 1. A: graph showing mean shrinkage of cells in the deprived magno- and parvocellular LGN laminae in relation to cells in corre- sponding undeprived laminae of the same animal against age at perfusion for animals with monocular deprivation started in the first few days of life. B: graph showing mean size changes for cells in both deprived and undeprived, magno- and parvocellular laminae in relation to the mean cell sizes for the corresponding laminae of normal animals for the same animals as A.

6O

TABLE II

Mean LGN cells areas (/~m 2) for a monkey with monocular closure from day 1 to day 28

Lamina Undeprived % Change Deprived % Change % Change

Mean + S.D. compared Mean +- S.D. compared deprived vs to normal to normal undeprived

I 357.4+53.7 +22% 284.5+-43.4 II 325.1 +46.4 + 12% 296.7+51.3

III 245.8+38.9 +26% 196.4+25.6 IV 235.8+29.9 +23% 188.0+24.3 V 248.9+30.7 +32% 189.2+ 18.6 VI 228.3+_27.1 +26% 183.9+20.6

- 3 % - 2 0 % * *

+2% - 9%*

0% -20%** - 2 % - 2 0 % * *

+1% -24%** +1% -19%**

* P < 0.01. ** P < 0.001.

results for these sets of laminae have been t rea ted

separately throughout . In general laminae I and II

have behaved similarly, as have the individual parvo-

cellular laminae, and so for graphical presenta t ion

the means of the percentage changes for each set of

magnocellular or parvocel lular laminae have been

used. However , under certain condit ions the laminae

receiving connections from the contra la tera l eye (I,

IV and VI, referred to as the crossed laminae) be-

haved differently from the laminae receiving from

the ipsilateral eye (II, I I I and V, the uncrossed lami-

nae). These differences between crossed and un-

crossed laminae have been presented separately.

M o n o c u l a r c losure s tar ted at bir th

In all animals with early monocular closure, cells in

the depr ived L G N laminae appea red smaller and pal-

er when compared to cells in the adjacent undepr ived

laminae. This was apparent after only 2 days of clo-

sure. These changes were confirmed by measure-

ments of mean cell area in which cells in correspond-

ing deprived and undepr ived laminae were com-

pared. The difference between depr ived and unde-

prived laminae was significant for laminae II and IV

after only 2 days of closure; it increased rapidly for

the first 3-4 weeks, and then increased slowly there-

after (Fig. 1A).

However , when the sizes of cells in depr ived and

undeprived laminae of these same animals were com-

pared with those of normal animals it became clear

that the changes produced by monocular closure

were considerably more complex than this. Closure

of one eye at birth ini t iated a sequence of events

which profoundly affected cells in both depr ived and

undeprived laminae, and cont inued to evolve for at

least 3 months. The changes in cell size differed in di-

TABLE III

Mean LGN cell areas (#m 2) for 7 monkeys with monocular closures from birth to between 11 days and 8 weeks

Lamina Undeprived % Change Deprived % Change % Change Mean +_ S.D. compared Mean + S.D. compared deprived vs

to normal to normal undeprived

I 313.3+20.9 + 7% n.s. 253.1+29.1 -14%** -19% II 308.2+28.1 + 6% n.s. 250.6+29.6 -14%** -19%

III 234.9+16.4 +20%*** 185.7+13.1 - 5% n.s. -21% IV 226.8+9.6 +18%*** 184.1+7.3 - 4%* -19% V 223.7+23.0 +19%*** 170.5+10.9 - 9%** -24% VI 215.7+19.0 +19%*** 168.5+17.2 - 7% n.s. -22%

* P < 0.05. ** P < 0.01.

*** P < 0.001.

TABLE IV

Mean L G N cell areas (t~m 2) for a monkey with monocular closure from day 3 to day 406

61

Lamina Undeprived % Change Deprived % Change % Change Mean + S.D. compared Mean + S.D. compared deprived vs

to normal to normal undeprived

I 342.2+65.1 +16% 260.4+47.4 II 320.1 + 60.1 + 10% 228.0+52.3

III 177.5+23.3 - 9% 119.4+18.4 IV 187.5+24.7 - .2% 122.8+20.5 V 165.0+26.9 - 12% 122.4+18.4 VI 167.5+31.1 - 8% 97.8+16.3

-11% -24%* -22% -29%*

-39 % -33 % * -36 % -35 % * -35% -26%* -46% -42%*

* P < 0 . 0 0 1 .

rection at different times following closure and were

different in the magnocellular (I and II) and parvo-

cellular ( I I I - V I ) laminae.

During the first 4 weeks following closure, cells of the undeprived parvocellular laminae underwent

marked hypertrophy of about 25% compared to nor-

mal animals, and this increase in size was largely

maintained for a further 4 weeks (Fig. 1B). Results from a typical animal in which one eye was closed for

27 days from day 1 are shown in Table II and the

'overall means for all animals closed from birth to be-

tween 11 days and 8 weeks are shown in Table III. During this time there was little significant shrinkage

of cells in the deprived parvocellular laminae. The

difference in size of cells in deprived and undeprived parvocellular laminae following monocular closure at

birth is thus a result almost entirely of hypertrophy of

cells in the undeprived laminae. After the initial period of hypertrophy and plateau

the response of the undeprived cells then reversed

and they shrank almost as rapidly as they had initially

hypertrophied. By 90 days of age they were compar-

able to cells in normal animals (Fig. 1B; Tables IV and V). Over this same period cells in the deprived parvocellular laminae shrank in parallel with the unde-

prived cells, so that the difference between deprived and undeprived cells remained almost constant. Fol-

lowing the initial phase there was thus a marked par-

allel shrinkage of deprived and undeprived parvocel-

lular cells which had virtually no effect on the compa-

rison between deprived and undeprived laminae.

When comparisons were made only between de-

prived and undeprived laminae, the behavior of the

magnocellular laminae was very similar to that of the

parvocellular laminae, with most of the difference between deprived and undeprived cells being estab- lished by 4 weeks of age (Fig. 1A). Comparisons with

normal animals showed that the initial hypertrophy

of the undeprived magnocellular laminae was less

than that of the parvocellular laminae, and there was

TABLE V

Mean L GN cells areas (ixm 2) for 3 monkeys with monocular closures from birth to between 90 days and 5 years

Lamma Undeprived % Change Deprived % Change % Change Mean + S.D. compared Mean + S.D. compared deprived vs

to normal to normal undeprived

I 321.3+18.2 + 9%* 224.4+36.8 -24%** -30% II 328.4+27.4 +13%* 221.2+38.6 -24%** -33%

III 187.9+15.9 - 4% n.s. 130.4+9.6 -33%*** -31% IV 185.3+20.2 - 3% n.s. 132.7+8.7 -31%*** -28% V 183.9+21.5 - 2% n.s. 123.4+3.1 -34%*** -33% VI 174.1+21.5 - 4% n.s. 119.0+19.1 -35%*** -32%

* P < 0.05. ** P < 0.01.

*** P < 0.001.

62

Fig. 2. The LGN of a normal 273-day-old infant monkey. × 35.

definite shrinkage of cells in the deprived magnocel- lular laminae (Table III). Because magnocellular cells in normal infants are slightly smaller than those in adults, comparison to infant monkeys only rather than the whole group gave a rather greater degree of hypertrophy in laminae I and II (11% and 16%; both P < 0.01) and slightly less shrinkage (11%; P < 0.05 and 6%, n.s.) compared to Table lII, but hypertrophy was still less marked than in the parvo-

cellular laminae. After the initial phase, the magnocellular laminae

did not show the same marked parallel shrinkage as the parvocellular laminae. Following a long survival period cells in the undeprived magnocellular laminae were larger than normal mean values and shrinkage of cells in the deprived magnocellular laminae did not increase after 8 weeks, but possibly decreased slight- ly (Fig. 1B). These changes may be reflections of the

63

Fig. 3. The LGN ipsilateral to the closed eye in a monkey monocularly deprived from birth to 34 days of age. Cells in the undeprived parvocellular laminae IV and VI have undergone hypertrophy, while those in the deprived parvocellular laminae III and V are un- changed in size. x 35.

residual growth seen in the normal magnocel lular

laminae after the first week of life t4.

The differences in the response of magnocel lular

and parvocel lular laminae following monocular clo-

sure at birth are reflected in differences in the rat io of

mean magnocel lular cell area to mean parvocel lular cell area (calculated separa te ly for depr ived and un-

deprived laminae). Over the first 8 weeks this ratio

was lower than normal , reflecting the lesser hyper-

t rophy of cells in the undepr ived magnocel lular lami-

nae and greater shrinkage in depr ived magnocel lular

laminae. The late paral le l shrinkage only affected

parvoceUular laminae and so the ratio in long survival

animals was higher than normal (Table VI).

6a

Fig. 4. The LGN ipsilateral to the closed eye in a monkey monocularly deprived from birth to 406 days of age. Cells in the undeprived parvocellular laminae IV and VI have returned to normal size and those in the deprived laminae have now shrunk, x 35.

These changes in both depr ived and undepr ived

L G N laminae are i l lustrated in Figs. 2 - 4 and 6 -8 .

Comparison of Figs. 3 and 7 which show the L G N of

a monkey closed from birth to 34 days with the nor-

mal L G N of Figs. 2 and 6 shows hyper t rophy of the

undeprived parvocel lular cells with no change in the

deprived cells. By 406 days both the depr ived and un-

deprived parvocel lular cells have undergone a

marked parallel shrinkage, al though the size differ-

ence between them is mainta ined (Figs. 4 and 8, cf.

Figs. 3 and 5).

Monocular deprivation after a period of normal visual experience

If the developing visual system is al lowed a per iod

of normal visual experience before monocular depri-

65

Fig. 5. The LGN ipsilateral to the closed eye in a monkey with late monocular deprivation of 60 days duration started at 275 days of age. Cells in all parvocellular laminae are shrunken, but there is little difference between cells in deprived and undeprived laminae, x 35.

vation is s tar ted the effects of the depr ivat ion are

modified in a way which depends on the dura t ion of

the preceding visual experience. In animals where

closure was de layed until 3 - 4 weeks of age, the dif-

ference between depr ived and undepr ived laminae

developed at a rate similar to that following closure

at birth (Fig. 10) but when closure was de layed until

8 weeks of age or la ter the difference was smaller and

developed more slowly. Closure at 275 days for 2

months produced no visible difference between de-

prived and undepr ived laminae (Figs. 5 and 9). The

results for magnoceUular and parvocel lular laminae

when comparing depr ived and undepr ived laminae

were similar.

E 0

1>.

0 E

×

0

~q E

Z .E z

.=. .=.

.o .o

67

Fig. 8. Higher magnification of cells in the parvoceUular laminae of the LGN of Fig. 4 from a monkey monocularly deprived from birth to 406 days of age. × 90.

Fig. 9. Higher magnification of cells in the parvocellular laminae of the LGN of Fig. 5 from a monkey with late monocular deprivation of 60 days duration started at 275 days of age. x 90.

68

TABLE VI

Ratios of mean magnocellular to mean parvocellular LGN cell area following monocular visual deprivation

Normal ratio: 1.550.

Monocular closure Undeprived Deprived

From birth to between 11 days and 8 weeks 1.380 1.417

From birth to between 3 months and 5 years 1.787 1.777

Closure after 8 weeks of age 1.786 1.806

However, when cell sizes in experimental animals were compared with normal animals the changes were seen to be rather more complex than simply a matter of declining sensitivity with later closure. Un- like closure at birth, monocular closure at 3-4 weeks of age produced little or no hypertrophy of unde- prived parvocellular laminae; the difference between deprived and undeprived laminae was produced by a rapid shrinkage of cells in the deprived parvocellular laminae (Fig. 11). In animals where closure was de- layed until about 8 weeks of age the deprived parvo-

cellular cells still shrank at a rate and to a degree comparable with the effect of closure at 3-4 weeks, but the undeprived parvocellular cells now also

shrank following monocular closure at 8 weeks. This reduced the difference between deprived and unde- prived cells to give the appearance of reduced sensi- tivity (Fig. 11, cf. Fig. 10). Monocular closure at 3-4 months caused rather slower shrinkage of cells in all parvocellular laminae. However, animals closed at 7-9 months of age again showed a rapid shrinkage of both deprived and undeprived parvocellular cells,

undeprived Cells shrinking by approximately 25% in two months of closure and deprived by 30% (Table VII). Closure for 2 months at 10 months, 1 year and

18 months gave progressively decreasing degrees of shrinkage of all parvocellular cells (Fig. 11).

Mean cell areas for all animals closed after 8 weeks of age are given in Table VIII. This shows that the shrinkage of both undeprived and deprived cells rela- tive to the corresponding cells in normal animals is highly significant in all parvocellular laminae. It also shows that cells in the undeprived magnocellular laminae do not undergo significant shrinkage follow- ing late monocular closure. The deprived magnocel- lular cells show only a small degree of shrinkage, with

the result that the difference between deprived and undeprived magnocellular cells is very similar to the difference between deprived and undeprived parvo-

cellular cells. The ratio of mean magnocellular to mean parvo-

cellular cell area in these animals is significantly in- creased (Table VI), confirming that shrinkage after late closure selectively affects the parvocellular lami- nae. The ratios for undeprived and deprived sets of

laminae are very similar to each other, indicating that the difference between the large degree of shrinkage of the deprived parvocellular laminae and the small shrinkage of the deprived magnocellular laminae is

comparable to the shrinkage of the undeprived par-

voceilular laminae in relation to the undeprived mag- nocellular laminae. Thus the selective effect of later closure is very similar in both undeprived and de- prived parvocellular laminae.

The effect o f two months o f monocular closure started

at different ages

In order to make a systematic examination of the effects of increasing age and visual experience on the susceptibility of LGN cells to monocular closure, re- sults for 9 of the monkeys described above have been

analyzed separately. The degree of shrinkage found

after 2 months of closure has been taken to give an in- dication of the sensitivity of the visual system to clo- sure at that age. Comparison of results from deprived laminae with their corresponding undeprived lami- nae showed a rapid drop in sensitivity occurring mainly between 30 and 60 days of age, with a small and reducing effect persisting until at least 18 months of age (Fig. 12). When compared to normal animals,

cells in the magnocellular laminae still followed es- sentially this pattern, with no consistent significant changes occurring in the undeprived magnocellular laminae.

Comparison of results for the parvocellular lami- nae with those from normal animals gives an indica- tion of the time course of the selective late shrinkage effect on these laminae. Following closure at 30 days of age cells in the undeprived laminae showed little change, but by 59 days of age, 2 months of closure produced clear shrinkage. In both these animals the deprived parvocellular cells showed marked shrink- age (Fig. 13), but closure from about 100 days showed a reduced shrinkage of deprived and unde-

x - Approx. age at closure

E

£

~o o!

- 1 0

- 2 0

- 3 0

- 4 0

0

-10

-20

-30

0

-10

-20

-30

0

-10

-20

-30

0

-10

-20

-300!

-10

-20

-30

0

-10

-20

0

. f " . U

,;O 2;o 3;o ~;0 5;o 6 ; o ~ ~ 8 ~ " ~ 9 0 0 Age ( d a y s )

Fig. 10. Graphs showing the mean shrinkage of cells in de- prived parvocellular laminae in relation to cells in correspond- ing undeprived laminae of the same animal for groups of ani- mals with monocular closure started at progressively later ages. Closure at birth and 30 days of age cause a rapid difference in size to develop between cells in deprived and undeprived lami- nae, whereas closure at 60 days of age or later apparently has much less effect.

69

8 .c_ E cJ

E

2

&

u5

%

3O 20

10

0

- 1 0

- 2 0

-30

10

0 - 10

-20

-30

0

-10

-20

-30

0

- 10

-20

-30

0

- 10

-20

- 30

0

- 10

- 20

- 3 0

0

-10

- 20 o;

• - Undeprived

o - Deprived

x Approx age at closure

l / i ;

l I I l I I / z l I W I ~ "

100 200 300 400 500 600 " 900 1000 " 1800 1900

Age (days)

Fig. 11. Graphs showing mean size changes for cells in both de- prived and undeprived parvocellular laminae in relation to cells in corresponding laminae of normal animals for the same ani- mals as Fig. 10. Whereas the difference in size between de- prived and undeprived cells following closure at birth is caused by hypertrophy of undeprived cells, the apparently similar dif- ference caused by closure at 30 days is caused by shrinkage of deprived cells. The apparent loss of sensitivity following clo- sure at 60 days or later when deprived and undeprived cells are compared is seen here to be due in fact to the onset of shrinkage in undeprived cells.

TABLE VII

Mean L G N cell areas (~tm 2) for a monkey with monocular closure for 60 days from day 275 to day 335

Lamina Undeprived % Change Deprived % Change % Change Mean + S.D. compared Mean + S.D. compared deprived vs

to normal to normal undeprived

I 230.3+64.8 -22% 245.8+52.9 -16% II 278.5+55.2 - 4% 221.7+58.4 -24%

III 155.3+35.7 -21% 138.8+24.7 -29% IV 140.6+29.8 -27% 133.0+28.1 -31% V 137.5+22.1 -27% 137.0+25.3 -27% VI 138.6+30.9 -24% 120.9+__20.4 -33%

+ 7% n.s. - 2 0 % * *

- 1 1 % * - 5% n.s.

0% n.s. - 1 3 % *

* P< 0.01. ** P < 0 .001.

70

prived cells. The sensitivity of both deprived and un-

deprived cells then reached a peak at between 200

and 275 days of age, with two months of closure giv-

ing approximately 25% and 30% shrinkages respec- tively. After this age, sensitivity to closure declined

rapidly and progressively with little effect being seen

by 18 months of age.

Dif ferences be tween crossed a n d u n c r o s s e d laminae

Following closure from birth, qualitative examina-

tion of the LGN shows a consistently greater effect in

laminae II, III and V ipsilateral to the closed eye, as has been described previously8, 30. This, however, is

not reflected in the cell size data, as has also been noted previously 30. The size difference between cells in deprived and undeprived parvocellular laminae in

the LGN ipsilateral to the closed eye does not differ

consistently from that in the LGN contralateral to the

closure.

However, in these same animals comparison of de-

prived parvocellular cells with undeprived cells from

the corresponding laminae of the other LGN does

show a consistently greater effect in cells in the un-

crossed laminae (laminae III and V) in closures of

more than 16 days duration (Fig. 14A), although this

is not maintained in the two longest closures (>1 year). When the same data are analyzed in relation

to normal cell sizes, it can be seen that this is due both

to slightly but consistently greater hypertrophy of the cells in undeprived laminae III and V, and to greater

shrinkage of cells in deprived laminae III and V

(Fig. 14B). Cells in both the deprived and unde-

prived uncrossed parvocellular laminae thus respond

more to monocular closure than cells in the crossed

laminae. Since these effects occur in opposite nuclei

they do not contribute to the difference seen qualita-

tively. The mean size of cells in all parvocellular lami-

nae (including both deprived and undeprived) in the LGN ipsilateral to the closure is, however, slightly

smaller on average than the corresponding figure for

the LGN contralateral to the closure.

Of the 6 animals with monocular closure started at

between 3 weeks and 3 months of age, 2 showed very

marked differences in shrinkage between crossed

and uncrossed laminae (e.g. Table IX). In both these

animals cells in the uncrossed deprived and crossed

undeprived parvocellular laminae had shrunk

markedly, whereas cells in the other parvocellular

laminae had not. Both deprived and undeprived par-

vocellular cells in the LGN ipsilateral to the closure were thus smaller than all the parvocellular cells in the contralateral nucleus. The differences between

crossed and uncrossed laminae in these animals were

considerably larger than in any other animals in this

study. In 4 of the 5 animals closed after 6 months of age

and in the 3 closed at about 100 days for long survival

periods, shrinkage of cells in both deprived and unde-

prived parvocellular laminae of the LGN contralater- al to the closed eye exceeded that in the LGN ipsilat-

eral to the closed eye, although the differences were

not large.

DISCUSSION

Previous studies of the effect of visual deprivation

on the sizes of cells in the monkey LGN have de-

scribed changes in cells from deprived laminae using

TABLE VIII

Mean LGN cell areas (lain 2) for 13 monkeys with monocular closures started after 8 weeks o f age

Lamina Undeprived % Change Deprived % Change % Change compared compared deprived vs

Mean + S.D. to normal Mean + S.D. to normal undeprived

I 287.8+40.0 - 2% n.s. 280.7+24.0 - 5% n.s. II 299.3_+21.6 + 3% n.s. 267.7_+39.1 - 8% n.s.

III 172.3+ 18.3 -12%* 158.9_+ 19.3 IV 165.1+16.7 -14%* 152.2+18.0 V 163.4+ 17.6 -13% * 151.4+ 16.8 VI 158.1 + 19.3 -13%* 146.7_+ 16.7

-19%* -21%* -19%* -19%*

- 2 %

-11%

- 8 %

- 8 %

- 7 %

- 7 %

* P < 0 . 0 0 1 .

o/, E

E 10

~. o Q .

c -10

2,o -20 >

-30'

c -40 o

. . . . . . Magnoce/lutar

-- Parvoceltutar

• - U n d e p r i v e d

o - D e p r i v e d

[ I L I L I

100 200 300 400 500 600

Age at ctosure (days)

Fig. 12. Graph showing the effect of two months of monocular deprivation started at progressively later ages measured by comparing sizes of deprived and undeprived cells.

cells in the corresponding undepr ived laminae of the same animal as a controp,8,19,20,30-32, 34. These studies

have shown that if an eye is closed at bir th, cells in the

deprived L G N laminae become smaller than those in

undeprived laminae, and this has been in terpre ted as

shrinkage or failure of growth of cells in the depr ived

° / o

3O (1)

E "E 20' -6 E 10

E

o 0

>

- 1 0 ~

o~ - 2 0 # U

. -30 N co

-40 0

j . . . . . - I

I I I I I I

100 200 300 400 500 600

Age (at c l o s u r e ( d a y s )

Fig. 13. Graph showing the effect of two months of monocular deprivation started at progressively later ages measured by comparing the sizes of deprived and undeprived parvocellular cells with those of normal cells. The shrinkage of both deprived and undeprived parvocellular cells has a peak of sensitivity at between 200 and 300 days.

71

8 o

g

• Uncrossed . . . . . • O'ossed A %

0

-10 ' ~ m

-20 . " ,

-40 [ l 1 i i i L i L i

• Uncrossed ur~eprived B • . . . . • C r o s s e d u n d e p r i v e d

o Ur lE rossL~ : ] d e p r i v e d

o . . . . ~ CrosSed deprived %

40

30

20 % . . . .

10

0 ", . . . . . . . ,

-20 " , ,

2 , 0 10 20 30 40 50 60 70 El 90 100

Age (days)

Fig. 14. Graphs showing differences in response of crossed and uncrossed parvocellular laminae to closure at birth. A: de- prived cells compared to corresponding undeprived cells. For closures of between 16 and 90 days duration the uncrossed de- prived cells (laminae III and V) shrink more in relation to the uncrossed undeprived cells than do the crossed deprived cells (laminae IV and VI) in relation to the crossed undeprived cells. B: deprived and undeprived cells compared to normal cells. For closures of between 16 and 90 days duration cells in the un- crossed undeprived laminae show more hypertrophy and cells in the uncrossed deprived laminae show more shrinkage than cells in the corresponding crossed laminae.

laminae. This size difference appears over the first

few weeks of closure and remains s teady thereaf ter .

If closure is delayed until af ter about 6 weeks of age,

the sensitivity of the L G N as measured by comparing

deprived with undepr ived cells declines fairly rapid-

ly, and if closure is delayed until after about 10

weeks, relatively little change is seen.

When analyzed in this way, the results of the pres-

ent study are in close agreement with and confirm

these previous findings (Figs. 1A and 10). The major

difference between the present and previous work in

the pr imate is that sizes of both depr ived and unde-

prived L G N cells from these exper imenta l animals

have also been compared to sizes of cells from a se-

ries of normal monkeys TM. This has shown that the el-

72

T A B L E IX

Mean L GN cell areas (tim 2) for a monkey with monocular closure for 61 days from day 59

Lamina Undeprived % Change Deprived % Change % Change Mean + S.D. compared Mean _+ S.D. compared deprived vs

to normal to normal undeprived

I 223.1__+52.2 - 2 4 % 242.8_+50.4

II 290.9_+72.7 0 % 204.4_+49.5

I I I 188 .5+28 .5 - 4% 127 .6+19 .2

IV 149 .0+23 .2 - 2 2 % 170 .9+28 .0

V 182 .9+26 .4 - 3% 127 .6+17 .5

VI 133.3+ 16.5 -27% 152.8+ 19.2

- 1 7 % + 9 % n.s.

- 3 0 % - 30%*

- 3 5 % - 32%*

- 1 1 % + 1 5 % *

- 3 2 % - 30%*

- 1 6 % + 1 5 % *

* P < 0.001.

fect of monocular visual deprivation on the LGN is much more complex than had been thought with the undeprived LGN cells, previously used as a control, themselves undergoing major changes in size. Markedly different effects occur at different ages and with different lengths of closure, and either deprived, undeprived or both sets of cells may be mainly af- fected at different times. Previous visual experience

does not simply reduce the sensitivity of the primate LGN to the effects of visual deprivation, but has a major influence on the whole nature of the response of LGN cells to deprivation and so the effects of clo- sure are very different at different ages. Sensitivity to some effects is increased rather than decreased by a period of normal vision, and the LGN is sensitive to visual deprivation for much longer than had pre- viously been thought. This raises a number of impor- tant questions regarding the mechanisms involved in

both normal and abnormal development of the pri- mate visual system, and makes it essential to consider the effect of monocular deprivation on the whole of

the visual system related to both eyes. Technical and statistical aspects of making compa-

risons between normal and experimental animals have been discussed in the previous paper14 and will not be repeated here.

Early effects of closure from birth The major initial effect of monocular closure from

birth is to cause a marked hypertrophy of cells in the undeprived parvocellular laminae of the primate LGN, this hypertrophy being completed by approxi- mately 4 weeks. In contrast, little shrinkage of the deprived parvocellular cells occurs over this same pe- riod. Thus, early deprivation in the primate initially

causes growth of cells in the undeprived parvocellu- lar laminae, rather than shrinkage or failure of growth of cells in the deprived laminae. Although the

possibility that hypertrophy of undeprived cells might occur after monocular closure has previously been considered by a number of authorsS,6, 8,19,37, it

has only previously been found in one study of the cat ~6 and one of the dog 27. In both instances, it oc-

curred in addition to a marked shrinkage of deprived cells. In the monkey the degree of hypertrophy is much greater than in the cat, and it is the dominant early response to monocular closure, with little change occurring in the deprived parvocellular lami- nae. In monkeys with one eye closed from birth to one month of age the difference between the de- prived and undeprived parvocellular laminae is due almost entirely to hypertrophy of cells in the unde- prived laminae.

The early effects of monocular closure from birth on LGN cell size may be compared with correspond- ing data relating to the cortical ocular dominance col- umns 19,20,26,29. At birth, transneuronal autoradiogra-

phy shows that the terminal arborizations of LGN cells representing the two eyes overlap extensively in layer IV of area 17; over the first 3 weeks of normal development these segregate to form the ocular dom- inance stripes. Following monocular deprivation from birth to 3 weeks of age, the undeprived genicu- lo-cortical axons retract less and so the ocular domin- ance stripes related to the undeprived eye are markedly wider than those in a normal 3-week-old monkey20, z9. Unless the axonal arborizations of the undeprived geniculate axons within these wide stripes are less dense than those in normal animals, there must be an overall increase in the total arbori-

zation made by each undeprived geniculo-cortical

axon. The hypertrophy of the undeprived parvocel- lular cells over these first 3 weeks would then corre- spond to an increase in their axonal arborizations in accordance with Guillery's hypothesisS that the sizes of neuronal somata in the LGN are a reflection of the sizes of their axonal arborizations in the visual cortex and possibly also the number of synapses.

What is perhaps more surprising than the hyper- trophy of the undeprived parvocellular cells, is the lack of shrinkage of cells in the deprived parvocellu- lar laminae. This is contrary to most previous inter- pretations, and appears to be at variance with Guil- lery's hypothesis because the deprived ocular domin- ance stripes are described as being very shrunken af- ter 3 weeks of closure20, 29. Physiological studies of the response properties of visual cortical neurons in animals following comparable closures show a marked dominance of responses to the undeprived eye in layer IV1, 20, so that the relative effectiveness of the deprived LGN axons in driving cortical cells is clearly greatly reduced following 3 weeks of closure. There is, therefore, a delay between the effect of mo- nocular deprivation on the physiological responses of cortical cells and its effect on the sizes of LGN neu- rons. It is possible that shrinkage of a cell body fol- lowing shrinkage of its axonal arborization is intrin- sically a slower process than hypertrophy following axonal expansion, and that there is simply a delay be- tween the shrinkage of the geniculo-cortical axons within the ocular dominance columns in the cortex and the shrinkage of the corresponding cell somata in the LGN. However, following closure of an eye started at 3 weeks of age, the deprived parvocellular laminae shrink rapidly and similarly, there is no sug- gestion of a delay in the shrinkage of cells at later ages of closure, so this explanation seems unlikely. An alternative possibility is that at least the pretermi- nal parts of the deprived axonal arborizations in the cortex may remain relatively unchanged after 3 weeks of closure. LeVay et al.20 found unexpectedly increased levels of autoradiographic labeling be- tween the deprived ocular dominance columns fol- lowing 3 weeks of closure from birth. Although they considered that this was probably due to spillover of label in the LGN of very young animals, they con- cluded that 'we have not fully excluded the alterna- tive possibility that the deprived eye's afferents

73

maintain some (nonfunctional) terminals in the open eye's columns'. Blakemore et al. 1 and Swindale et al. 29 have also considered this possibility in relation to the physiological effects of reverse suture in the monkey. Persistence of preterminals may therefore explain the delay in shrinkage of the deprived parvo- cellular cells seen in the present study, and Guillery's hypothesis would then be in accordance with our findings. This possibility is supported by the fact that the labeling of the gaps between the deprived col- umns in the experiments of LeVay et al. 20 had largely disappeared in an animal in which the period of mo- nocular deprivation was increased to 47 days from day 4. The difference between these two survival pe- riods corresponds to the time during which shrinkage of deprived parvocellular cells becomes marked, sug- gesting that the cell shrinkage may be occurring as the preterminals are cleared. The retention of preter- minal axons in the 'wrong' columns may also explain the residual plasticity seen in the cortex of monkeys initially closed at between 3 and 6 weeks, after most of the segregation of the columns has occurred 20.

Hypertrophy occurs following closure at birth, but has not been found following closure started at 3 weeks of age. If, however, following closure at birth, the deprived eye is opened and the undeprived eye closed at 3 weeks of age, the initially deprived parvo- cellular cells do undergo hypertrophy 28. Three weeks of closure have thus preserved the ability of the de- prived cells to hypertrophy when they would not nor- mally be able to do so. This further supports the idea that the deprived eye may maintain nonfunctional preterminals in territory functionally dominated by the initially undeprived eye for several weeks, and suggests that their withdrawal may be postponed by lack of visual input.

The initial effect of closure upon the parvocellular LGN laminae is thus the reverse of what had pre- viously been thought, with deprivation of one eye causing hypertrophy of cells receiving input from the other open eye, and yet having little effect on the cells directly deprived. This has implications for the concept of competition between afferents related to the two eyes which was originally suggested by Wie- sel and HubelaS and developed by Guillery and Stelzner 7, Guillery 5 and Hubel et a1.19. The fact that it is the undeprived parvocellular laminae which show the earliest and greatest response to monocular

74

closure from birth confirms the basic concept of an in- teraction between afferents related to the two eyes. However, at the peak of the hypertrophy the total area of deprived plus undeprived parvocellular cells is about 12% greater than normal, and it would ap- pear that monocular closure allows overlap of de- prived and undeprived thalamo-cortical axonal arbo- rizations to persist for longer, although the densities of the afferents from the two eyes within the area of overlap do not remain equal. In its initial stages, mo- nocular closure thus appears to reduce a competitive interaction normally present between the afferents from the two eyes. The reduction of activity of the deprived afferents seems to allow the undeprived af- ferents to retain a greater than normal territory, while the deprived afferents are not immediately forced out of it.

Later effects of closure from birth Following the initial hypertrophy of the unde-

prived parvocellular cells, a marked shrinkage occurs affecting cells in both the deprived and undeprived parvocellular laminae. After a long survival period the undeprived cells have returned to near normal size, and the deprived cells are markedly smaller. Be- cause of the parallel nature of the later shrinkage no suggestion of this second phase is seen when deprived laminae are compared with undeprived laminae (Fig. 1A, cf. Fig. 1B), and it has not previously been described. It is possible that the hypertrophy is sim- ply a transient response which persists only while the undeprived cells are actively increasing their axonal

arborizations. However, the close similarity between the responses of the deprived and undeprived parvo- cellular cells is striking, and this suggests that the shrinkage of both sets of laminae is part of the same effect and is in some way related to underlying devel- opmental processes. The second phase occurs, al- though there has been no change in the conditions of deprivation. It also occurs even if the closed eye has been reopenediS. It would therefore appear to be a consequence of the abnormalities produced during the initial phase of closure rather than being a direct consequence of the continuing monocular depriva- tion itself. The shrinkage starts at the age at which the relative widths of the cortical ocular dominance stripes are becoming fixed (3-6 weeks 20) and at which the ability to produce changes in the balance

between the LGN laminae is being greatly reduced, which suggests that it is not directly related to the rel- ative expansion and contraction of the ocular domin- ance bands in layer IV. There are, however, a num- ber of striking similarities between the late phase of closure from birth and the effect of late closure started at between 6 and 9 months of age (see below). The late closure effect may be related to changes in layers outside layer IV, and there is also evidence for changes in the supra- and infragranular cortical lay-

ers following long-term early closure. In addition to the changes in physiologically determined ocular dominance which occur outside layer IV, Hubel et al. 19 describe pale bands in layers III and V of a mon- key with a long-term closure started at 2 weeks of age and stained using the Liesgang method, and suggest that this represents a deterioration of descending connections between these layers. It seems reason- able to suggest that if territory in layer IV is not

equally shared between afferents from the two eyes in the initial phase of closure, then the ability of the

cortex to make the correct binocular connections to and within other cortical laminae is impaired, regard- less of the visual input being received at the time.

Monocular closure after a period of normal visual ex- perience

It has generally been considered that a period of normal visual experience prior to monocular depri- vation reduces the sensitivity of LGN cells to changes caused by deprivation. If the LGN is examined quali-

tatively following closure at progressively later ages, the difference between deprived and undeprived laminae produced by closure becomes much less ap- parent, and quantitative comparison of deprived with undeprived cells shows progressively smaller differ- ences in size (Figs. 5 and 10). The main reduction in sensitivity occurs between 1 and 3 months of age, at the time when the ocular dominance columns in the cortex are becoming fixed, and this has been taken to mark the end of the major sensitivity of the primate LGN to visual deprivation33, 36. When comparisons are made between experimental and normal animals, however, a rather different picture emerges. Where- as qualitative examination and quantitative compari- son of deprived and undeprived laminae show little difference between the effects of closure at birth and closure performed at 3 weeks of age, comparison

with normal animals shows that there are major dif- ferences. Whereas closure at birth causes hypertro- phy of undeprived parvocellular laminae and little change in deprived laminae, closure at 3 weeks of age produces a similar difference between deprived and undeprived laminae, but by causing rapid shrinkage of deprived parvocellular cells and no hypertrophy of the undeprived cells. Three weeks of normal devel- opment has caused a major change in the reaction of the LGN to visual deprivation, with both the type of response and the laminae mainly involved changing. It is during this first 3 weeks of normal development that most of the segregation of cortical ocular domin- ance columns occurs, and it seems probable that the loss of the hypertrophy response is related to this.

If closure is delayed until about 60 days of age, when the cortical ocular dominance stripes are largely fixed, previous workers considered that the sensitivi- ty of the LGN to closure was greatly reduced. Com- parison of experimental with normal animals, howev- er, shows that this impression of reduced sensitivity of LGN cells is misleading. When animals closed at 60 days of age are compared with those closed at 3 weeks of age, the response of the deprived cells is the same and there is in fact no change in the sensitivity of the deprived parvocellular cells over this period (Fig. 11). What does change is the response of the undeprived parvocellular cells to closure. From hy- pertrophy at birth and no change at 3 weeks, their re- sponse has now become one of shrinkage and the ap- parent loss of sensitivity of the LGN to closure is due entirely to changes in the response of the undeprived parvocellular cells. Their shrinkage now largely masks the shrinkage of the deprived parvocellular cells (Fig. 10, cf. Fig. 11). The end of the 'sensitive period' in the primate LGN is thus paradoxically brought about by a reversal of the response of the un- deprived parvocellular cells, rather than a change in the sensitivity of the deprived cells. What becomes greatly reduced at the time of fixation of the ocular dominance stripes is only the ability to produce a dif- ference in size between the deprived and undeprived cells in the LGN.

By 3 months the susceptibility of cells in both de- prived and undeprived parvocellular laminae to mo- nocular closure is reduced, but increases again at about 6 months of age. There is then a second period of high sensitivity between 6 and 9 months during

75

which rapid changes in LGN cell size can be brought

about by change in the visual environment• These late changes are comparable both in degree and rate of development with those occurring in the first few weeks of life. They differ from the effects of early clo- sure in that late closure causes shrinkage of both de- prived and undeprived cells, so that the changes are not apparent on qualitative examination, or barely if deprived cells are compared with undeprived cells quantitatively, and that late shrinkage selectively af- fects cells in the parvocellular laminae•

• Comparatively little attention has been given to changes produced by late visual deprivation• In the cortex late monocular closure has little effect on the relative widths of the ocular dominance columns in layer IV as determined by transneuronal autoradio- graphy and injection of HRP into individual LGN laminae17, 20, both of which show the extent of the ax- onal arborizations of LGN cells. However, following long-term late closure there is reduced staining of de- prived ocular dominance columns by the Liesgang re- duced silver method 20, and there are clear changes in layer IV in sections stained for cytochrome oxidase 17. Electrophysiological experiments show little change in the ocular dominance of layer IV cells following late closure1, 20, but have shown clear changes in the ocular dominance of cells in the supra- and infragra- nular cortical layers ~,20,33. It therefore seems rea- sonable to suggest that the late shrinkage seen in par- vocellular LGN cells may be a reflection of changes in the connections from layer IV to these layers, and possibly also of changes at subsequent stages in the intrinsic circuitry of the cortex (cf. Wiesep6). The physiological responses of cells in the supragranular layers appear to be heavily dominated by parvocellu- lar inputs 20 and most cells in them are normally bin- ocularly driven, and this appears to be the stage at which major binocular convergence first occurs in the primate. The importance of simultaneous activation of a cortical cell by input from both eyes in maintain- ing its binocular responses has been discussed by sev- eral authors 18,31,36,39. Late monocular closure may exert its effect not primarily by depriving one eye of stimulation, but because it prevents the cooperation between inputs from the two eyes which is necessary for normal cortical binocular function, and so dis- rupts the development of the intracortical circuitry on which this depends.

76

It was initially something of a puzzle that both de- prived and undeprived LGN cells shrink following late closure of only one eye. If late closure is affecting mainly binocular function this is more explicable, and two findings are of particular interest in this respect. In the cortex late monocular closure causes a reduc- tion of cytochrome oxidase staining not only in de-

prived ocular dominance columns in layer IV but also in the border strips of the undeprived columns, and

this can also be correlated with the changes in the Liesgang staining pattern17. There is therefore mor-

phological evidence that late monocular closure causes changes in undeprived cortical columns, al- though transneuronal autoradiography and HRP in- jections show that this is not simply a change in col- umn width. Recent physiological studies have also emphasized the importance of binocular input to nor- mal cortical function. In the alert monkey a consider- able proportion of neurons in area 17 require a binoc-

ular input to give a maximal response, and a propor- tion of tuned excitatory cells respond only to binocu- lar stimulation 2s. Monocular deprivation will prevent the normal functioning of these cells. Responses of all cells which are binocularly dependent will pre- sumably be abnormal, and so the overall loss to corti- cal function is more than simply the loss of the proc-

essing related to one eye. Since the open eye will not be able to make its usual contribution to binocular processing, it is perhaps not surprising that connec- tions related to it should also be affected.

One of the striking characteristics of the late clo- sure effect is that the sensitivity of LGN cells increas- es up to about 6 months of age before declining from about 9 months. In a detailed quantitative study, O'Kusky and Colonnier 24 have shown a continuing in- crease in the numbers of synapses in the primate visu- al cortex up to at least 6 months of age, and this is most marked in the supragranular layers. It seems likely that the connections which are affected by late closure are still developing at this time. The decline in sensitivity from 9 months onwards may then reflect a progressive consolidation of these connections, corresponding to the loss of sensitivity of the supra- and infragranular layers seen with physiological methods 20.

It is important to make a clear distinction between the effect of a period of normal development on the response to monocular closure and the evolution of

the effects of an early monocular closure interacting with underlying developmental mechanisms. It may therefore be misleading to draw too close a parallel between the changing effects of progressively later

closures and the long-term evolution of the effects of closure from birth. With this proviso, however, the similarities between the late shrinkage phase follow-

ing early closure and the effects of late closure are striking; both affect deprived and undeprived cells al- most equally, both selectively affect cells in the par- vocellular laminae, and both are of similar magni- tude and occur after the initial phase of visual devel- opment. These similarities suggest that the same un- derlying developmental processes are in some way involved. The main difference between them is that in the late phase of closure from birth, cells in de- prived and undeprived laminae are markedly un- equal before shrinkage starts, with the undeprived

parvocellular cells being hypertrophied and the de- prived normal, whereas with late closure both sets of cells start equal. These relative differences in size are largely preserved during the shrinkage so that the fi- nal cell sizes also differ.

If the late changes in the LGN cells are a result of changes in the supra- and infragranular layers of the cortex, how might these changes act on the LGN? Although there is now evidence of direct LGN pro- jections to layer II121.35, in the squirrel monkey at least these do not come from the main geniculate laminae. There is also a sparse geniculate projection to layer VI21, but it seems unlikely that there are enough of these terminals to account for the major effects seen in the LGN. Similarly, although some cells outside layer IV have processes which pass into or through this layer, a major direct geniculate input to them has not been demonstrated23. Several other alternatives may be suggested. Retrograde transneu- ronal effects are well recognized in some systems (see Cowan2), e.g. affecting retinal ganglion cells af- ter visual cortical lesions. If the projections to other layers made by layer IV cells are affected by late clo- sure, this effect may be transmitted retrogradely across the genicuto-cortical synapses ending on these layer IV cells. There are also projections back into layer IV from other cortical laminae, notably layer VI 4,22, and abnormal activity in these may affect the layer IV cells and their geniculate input. Layer VI also projects directly to the LGN and the possibility

77

of a di rect effect via cor t i co-gen icu la te f ibers mus t

also be cons idered . A t p resen t the re is no ev idence to

favour any o n e of these exp lana t ions , and in v iew o f

the mul t ip le i n t e r connec t ions wi th in the visual pa th-

way, it may be ve ry difficult to dis t inguish b e t w e e n

them.

ACKNOWLEDGEMENTS

This work was supported by grants from the Medi- cal Research Council, Royal Society and WeUcome Trust. J.J.S. is a Royal Society Research Fellow.

REFERENCES

1 Blakemore, C., Garey, L. J. and Vital-Durand, F., The physiological effects of monocular deprivation and their re- versal in the monkey's visual cortex, J. Physiol. (Lond.), 283 (1978) 223-262.

2 Cowan, W. M., Anterograde and retrograde transneuronal degeneration in the central and peripheral nervous system. In W. J. H. Nauta and S. O. E. Ebbesson (Eds.), Contem- porary Research Methods in Neuroanatomy, Springer, New York, 1970, pp. 217-251.

3 Garey, L. J. and Vital-Durand, F., Recovery from monoc- ular deprivation in the monkey II. Reversal of morpholog- ical effects in the lateral geniculate nucleus, Proc. roy. Soc. Lond. B, 213 (1981) 425-433.

4 Gilbert, C. D. and Wiesel, T. N., Morphology and intracor- tical projections of functionally characterised neurones in the cat visual cortex, Nature (Lond.), 280 (1979) 120-125.

5 Guillery, R. W., Binocular competition in the control of ge- niculate cell growth, J. comp. NeuroL, 144 (1972) 117-130.

6 Guillery, R. W., Quantitative studies of transneuronal at- rophy in the dorsal lateral geniculate nucleus of cats and kit- tens, J. comp. Neurol., 149 (1973) 423-438.

7 Guillery, R. W. and Stelzner, D. J., The differential effects of unilateral lid closure upon the monocular and binocular segments of the dorsal lateral geniculate nucleus in the cat, J. comp. Neural., 139 (1970) 413-422.

8 Headon, M. P. and Powell, T. P. S., Cellular changes in the lateral geniculate nucleus of infant monkeys after suture of the eyelids, J. Anat., 116 (1973) 135-145.

9 Headon, M. P. and Powell, T. P. S., The effect of bilateral eye closure upon the lateral geniculate nucleus in infant monkeys, Brain Res., 143 (1978) 147-154.

10 Headon, M. P., Sloper, J. J., Hioms, R. W. and Powell, T. P. S., Cell size changes in undeprived laminae of the mon- key lateral geniculate nucleus after monocular closure, Na- ture (Lond.), 281 (1979) 572-574.

11 Headon, M. P., Sloper, J. J., Hiorns, R. W. and Powell, T. P. S., Cell sizes in the lateral geniculate nucleus of normal infant and adult rhesus monkeys, Brain Res., 229 (1981) 183-186.

12 Headon, M. P., Sloper, J. J., Hioms, R. W. and Powell, T. P. S., Shrinkage of calls in undeprived laminae of the mon- key lateral geniculate nucleus following late closure of one eye, Brain Res., 229 (1981) 187-192.

13 Headon, M. P., Sloper, J. J. and Powell, T. P. S., Initial hy- pertrophy of cells in undeprived laminae of the lateral ge- niculate nucleus of the monkey following early monocular visual deprivation, Brain Res., 238 (1982) 439-444.

14 Headon, M. P., Sloper, J. J., Hiorns, R. W. and Powell, T. P. S., Sizes of neurons in the primate lateral geniculate nu- cleus during normal development, Develop. Brain Res., 18 (1985) 51-56.

15 Headon, M. P., Sloper, J. J., Hiorns, R. W. and PoweU, T. P. S., Effect of reopening an eye after a period of monocu- lar deprivation on sizes of neurons in the primate lateral ge- niculate nucleus, Develop. Brain Res., 18 (1985) 79-83.

16 Hickey, T. L., Spear, P. D. and Kratz, K. E., Quantitative studies of cell size in the cat's dorsal lateral geniculate nu- cleus following visual deprivation, J. comp. Neurol., 172 (1977) 265-282.

17 Horton, J. C., Cytochrome oxidase patches: a new cytoar- chitectonic feature of monkey visual cortex, Phil. Trans. Roy. Soc. B, 304 (1984) 199-272.

18 Hubel, D. H. and Wiesel, T. N., Binocular interaction in striate cortex of kittens reared with artificial squint, J. Neu- rophysiol., 28 (1965) 1041-1059.

19 Hubel, D. H., Wiesel, T. N. and Levay, S., Plasticity of oc- ular dominance columns in monkey striate cortex, Phil. Trans. R. Soc. B, 278 (1977) 377-409.

20 LeVay, S., Wiesel, T. N. and Hubel, D. H., The devel- opment of ocular dominance columns in normal and visu- ally deprived monkeys, J. comp. Neurol., 191 (1980) 1-51.

21 Livingstone, M. S. and Hubel, D. H., Thalamic inputs to cytochrome oxidase-rich regions in monkey visual cortex, Proc. nat. Acad. Sci. U.S.A., 79 (1982) 6098-6101.

22 Lund, J. S., Intrinsic organisation of the primate visual cor- tex, area 17, as seen in Golgi preparations. In F. O. Schmitt, F. G. Worden, G. A. Adelman and S. G. Dennis (Eds.), The Organization of the Cerebral Cortex, M.I.T. Press, Cambridge, MA, 1981.

23 Lund, J. S. and Boothe, R. G., Interlaminar connections and pyramidal neuron organisation in the visual cortex, area 17, of the macaque monkey, J. comp. Neurol., 159 (1975) 305-344.

24 O'Kusky, J. and Colonnier, M., Postnatal changes in the number of neurons and synapses in the visual cortex (area 17) of the macaque monkey: a stereological analysis in nor- mal and monocularly deprived animals, J. comp. Neurol., 210 (1982) 291-306.

25 Poggio, G. F. and Talbot, W. H., Mechanisms of static and dynamic stereopsis in foveal cortex of the Rhesus monkey, J. Physiol. (Lond.), 315 (1981) 469-492.

26 Rakic, P., Prenatal genesis of connections subserving oc- ular dominance in the rhesus monkey, Nature (Lond.), 261 (1976) 467-471.

27 Sherman, S. M. and Wilson, J. R., Behavioral and morpho- logical evidence for binocular competition in the postnatal development of the dog's visual system, J. comp. Neurol., 161 (1975) 183-196.

28 Sloper, J. J., Headon, M. P. and Powell, T. P. S., Simulta- neous hypertrophy of cells related to each eye in the lateral geniculate nucleus of the infant monkey following short term reverse suture, Develop. Brain Res., 15 (1984) 295-297.

29 Swindale, N. V., Vital-Durand, F. and Blakemore, C., Re-

78

covery from monocular deprivation in the monkey. III. Re- versal of anatomical effects in the visual cortex, Proc. roy. Soc. Lond. B, 213 (1981) 435-450.

30 Vital-Durand, F., Garey, L. J. and Blakemore, C., Monoc- ular and binocular deprivation in the monkey: morpholog- ical effects and reversibility, Brain Res., 158 (1978) 45-64.

31 Von Noorden, G. K., Histological studies of the visual sys- tem in monkeys with experimental amblyopia, Invest. Oph- thalrnol., 12 (1973) 727-738.

32 Von Noorden, G. K. and Crawford, M. L. J., Morphologic- al and physiological changes in the monkey visual system after short term lid suture, Invest. Ophthalmol. Vis. Sci., 17 (1978) 762-768.

33 Von Noorden, G. K. and Crawford, M. L. J., The sensitive period, Trans. Ophthalmol. Soc. U.K., 99 (1979) 442-446.

34 Von Noorden, G. K. and Middleditch, P. R., Histology of monkey lateral geniculate nucleus after unilateral lid clo- sure and experimental strabismus: further observations, In-

vest. Ophthalmol., 14 (1975) 674-683. 35 Weber, J. T., Huerta, M. F., Kaas. J. H. and Harting. J.

K., The projections of the lateral geniculate nucleus of the squirrel monkey: studies of the interlaminar zones and the S layers, J. comp. Neurol., 213 (1983) 135-145.

36 Wiesel, T. N., Postnatal development of the visual cortex and the influence of environment, Nature (Lond.), 299 (1982) 583-591.

37 Wiesel, T. N. and Hubel, D. H., Effects of visual depriva- tion on morphology and physiology of cells in the cat's later- al geniculate body, J. Neurophysiol., 26 (1963) 978-993.

38 Wiesel, T. N. and Hubel, D. H., Comparison of the effects of unilateral and bilateral eye closure on cortical responses in kittens, J. Neurophysiol., 28 (1965) 1029-1040.

39 Wiesel, T. N. and Hubel, D. H., Ordered arrangement of orientation columns in monkeys lacking visual experience, J. comp. Neurol., 158 (1974) 307-318.