LYSOSOMAL DYSFUNCTIONS ASSOCIATED WITH MUTATIONS
AT MOUSE PIGMENT GENES
EDWARD K. NOVAK
AND
RICHARD T. SWANK
Deprtrtment of Molecular Biology, Roswell Park Memorial Institute,
666 Elm Street, Buffalo, New York 14263
Manuscript received July 5, 1978
Revised copy received August 31, 1978
ABSTRACT
Melanosomes and lysosomes share several structural and biosynthetic properties. Therefore, a large number of mouse pigment mutants were tested to
determine whether genes affecting melanosome structure or function might
also affect the lysosome. Among 31 mouse pigment mutants, six had 1.5- to 2.5fold increased concentrations of kidney p-glucuronidase. Three mutants, pale
ear, pearl and pallid, had a generalized effect on lysosomal enzymes since there
were coordinate increases in kidney P-galactosidase and a-mannosidase. The
effects of these three mutations are lysosome specific since rates of kidney protein synthesis and activities of three nonlysosomal kidney enzymes were normal. Also, the mutants are relatively tissue specific in that all had normal liver
lysomal enzyme concentrations.-A common dysfunction in all three mutants
was a lowered rate of lysosomal enzyme secretion from kidney into urine.
While normal C57BL/6J mice daily secreted 27 to 30% of total kidney
P-glucuronidase and p-galactosidase, secretion of these two enzymes was coordinately depressed to l to e%, 8 to 9% and 4 to 5% of total kidney enzyme
in the pale-ear, pearl and pallid mutants, respectively. Although depressed
lysosomal enzyme secretion is the major pigment mutant alteration, the higher
lysomal enzyme concentrations in pearl and pallid may be partly due to a n
increase in lysosomal enzyme synthesis. In these mutants kidney glucuronidase
to 1.5-fold.-These results suggest that there
synthetic rate was increased 1.4
are several critical genes in mammals that control the biogenesis, processing
and/or function of related classes of subcellular organelles. The mechanism
of action of these genes is amenable to further analysis since they have been
incorporated into congenic inbred strains of mice.
EVERAL facts suggest that mutations affecting the pigment-forming apparatus
in mammals might also cause lysosomal dysfunction. Melanosomes and lysosomes share several structural features and appear to have a similar subcellular
origin. Melanosomes arise from either the Golgi apparatus (TODAand FITZPATRICK 1971) or a specialized region of the smooth endoplasmic reticulum
(MAUL1969; NOVIKOFF,
ALBALA
and BIEMPICA1968). Similarly, in some cell
types, primary lysosomes containing acid phosphatase activity bud from the concave surface of the Golgi apparatus (BAINTONand FARQUHAR
1966); in other
cells they originate from a specialized region of the smooth endoplasmic reticulum called GERL (NOVIKOFF
1976).
Genetics 92: 189-204. May, 1979.
190
E. K. NOVAK A N D
R. T. SWANK
Functionally, the two organelles communicate with the extracellular space and
with each other. Melanosomes are transferred from melanocytes to keratinocytes,
probably by phagocytosis of a portion of the melanocyte cytoplasm (WOLFFand
SCHREINER
1970). Similarly, lysosomes of several cell types are secreted through
the plasma membrane (IGNARRO,
LINT and GEORGE
1974; ZURIERet al. 1974).
I n keratinocytes there appears to be direct fusion of lysosomes and melanosomes
since lysosomal enzyme activity and pigment have been found in the same subcellular organelle (NOVIKOFF,ALBALAand BIEMPICA1968; WOLFFand
SCHREINER
1970). It is, in fact, still an open question as to whether melanosomes
form as and remain a specialized type of lysosome or whether they are distinct
organelles that may interact with lysosomes.
It is plausible, therefore, that genes affecting the biosynthesis or processing of
one organelle might have similar effects on the other. There are, in fact, at least
two precedents for pigment mutations causing altered lysosomal function. The
mouse mutation, beige, an animal model for the human Chediak-Higashi Syndrome, has pigment dilution, giant lysosomes in many tissues and decreased
secretion of several lysosomal enzymes from the kidney (BRANDT,
ELLIOTT
and
SWANK1975). Likewise, the light-ear mutation causes pigment dilution and
decreased secretion of kidney-/?-galactosidase and hexosaminidase (MEISLER
1978).
We have, therefore, extensively surveyed mouse pigment mutants for singlegene effects on lysosome biosynthesis, processing o r function. The mouse is an
ideal mammal to search for such mutations since over 100 pigment mutants have
been described and mapped on the mouse genome (M. C. GREEN1968; SEARLE
1968). Over 50 of these pigment mutants have originated either as spontaneous
mutations in inbred strain C57BL/6J or were subsequently transferred onto
C57BL/6J by repeated backcross matings. As a result, it is possible to survey a
population of mice that are, to a very large extent, genetically identical to the
control C57BL/6J strain except for the chromosomal site of the mutation.
The system we have chosen to test for aberrations in lysosomal function is the
androgen-sensitive mouse kidney. Lysosomal aberrations would be expected to be
amplified in this system since testosterone treatment of female mice causes a 100to 300-fold increase in the rate of synthesis of at least one lysosomal enzyme,
and GANSCHOW
P-glucuronidase, (SWANKand BRANDT
1978; SWANK,PAIGEN
1973) in kidney proximal tubule cells and a concomittant increase in secretion
into urine. The synthesis (MEISLER1978) and secretion (BRANDT,
ELLIOTT
and
SWANK1975; MEISLER1978) of other kidney lysosomal enzymes are also
enhanced by testosterone treatment, though to lesser extents than glucuronidase.
A further advantage of this system is that it is genetically well defined since
single gene mutants affecting the synthesis, structure, intracellular localization
and temporal appearance of glucuronidase have been described (PAIGENet al.
1975; SWANKet al. 1978b).
MATERIALS A N D METHODS
Animals: Mouse pigment mutants were obtained from the Jackson Laboratories, Bar Harbor,
Maine. All mutations originated spontaneously either in the C57BL/6J strain (coisogenic
LYSOSOMAL DYSFUNCTION A N D PIGMENT GENES
191
mutants) or in another strain and subsequently were transferred to C57BL/6J at the Jackson
Laboratory by repeated backcross matings (congenic mutants). Mutants found to have altered
lysomal enzyme levels were later bred at the animal facilities of Roswell Park Memorial Institute to obtain sufficient animals for further characterization. The mice were maintained as
previously described (BRANDT,
ELLIOTand SWANK
1975).
Testosterone induction and tissue preparation: Testosterone pellets (35 mg, prepared by
BHOGIB. SHETHof the Department of Pharmaceutics, University of Tennessee) were implanted
subcutaneously into 8- to 16-week-old female mice for 20 days (BRANDT,
ELLIOTTand SWANK
1975). Homogenates were prepared and stored at -20" (BRANDT,
ELLIOTTand SWANK1975)
except as otherwise specified. To obtain serum, 0.5 to 1.0 ml of mouse blood was clotted for 30
min a t 25" and overnight at 4". The clot was removed and the supernatant collected after centrifugation for ten min at 2,000 rpm.
Enzyme assays and immunotitration of glucuronidase: /3-Glucuronidase (EC. 3.2.1.31) and
p-galactosidase (EC. 3.2.1.23) were assayed by a fluorimetric method using 4-methylumbelliferyl-P-D-glucuronide and 4 methylumbelliferyl-P-D-galactoside, respectively, as substrate
(BRANDT,
ELLIOTTand SWANK1975). a-Mannosidase was assayed using 2.75 mM 4-methylumbellifery-a-D-mannapyranoside
as the substrate. Substrate was added at 0" to tissue homogenate
[previously heated to 56" for ten min to eliminate the cytosol form of the enzyme (DIZLK
1977)] diluted in 0.1 M sodium citrate buffer p H 4.2 containing 1 mM ZnC1, to give a final
volume of 1.0 ml. The mixture was incubated at 37" for five min, after which the reaction was
stopped by adding 0.2 ml of 2.8 M sodium carbonate. Fluorescence product was determined as
in the glucuronidase assay (BRANDT,
ELLIOTT
and SWANK1975).
Alcohol dehydrogenase (EC. 1.1.1.9) was assayed by a spectrophotometric measurement of
NADH produced by ethanol oxidation (BRANDT,ELLIOTTand SWANK1975). D-amino acid
oxidase (EC. 1.4.2.3) was assayed by a spectrophotometric method with D-phenylalanine as
substrate (WELLNERand LICHTENBURG
1971). Arginase (EC. 3.5.31) was assayed spectrophotometrically by determining the rate of urea formation from arginine (SCHIMKE1970). Freshly
prepared homogenates were used for each of these three assays. Glucuronidase was immunotitrated in kidney homogenates and in urine as described by BRANDT,
ELLIOTTand SWANK1975.
Rate of synthesis: The rate of kidney protein and glucuronidase synthesis was measured by
pulsing testosterone-treated mice with radiolabelled leucine, followed by purification of glucuronidase from homogenates with specific anti-mouse glucuronidase immunoglobulin a t 75 to 85 %
yield (CLEVELAND
and SWANK1978). The relative rate of glucuronidase synthesis is defined as
the ratio of counts incorporated into immunopurified glucuronidase divided by counts incorporated into total protein.
Chemicals: Napthol AS-BI p-D-glucuronide, 4 methylumbelliferyl-P-D glucuronide, and
4-methylumbelliferyl p-D-galactoside were obtained from the Sigma Chemical Co., St. Louis,
Missouri, and L(-4,5-3H) leucine (TRK, 170, 54 Ci per mix from Amersham/Searle, Arlington Heights, Illinois. Koch-Light Laboratories Ltd., Colnbrook, England, supplied 4-methylumbelliferyl-a-D mannopyranoside. All other chemicals were reagent grade and obtained from
the Fischer Chemical Co., Rochester, New York.
RESULTS
Survey of mouse pigment mutants: We have surveyed 31 independent pigment
mutants for kidney glucuronidase concentrations after induction of this enzyme
by testosterone (Table 1). These mutations are unique chromosomal sites and
are either coisogenic or congenic with normal strain C57BL/6J. Data o n the
previously characterized beige mouse (BRANDT,ELLIOTT
and SWANK1975:
BRANDTand SWANK1976) have been included for reference purposes here and
elsewhere in this report.
Glucuronidase concentration was 2.5-fold higher in three mutants, pale ear,
pearl and pallid, in addition to the beige mutant. In two other mutants, maroon
E. K. N O V A K A N D R . T. S W A N K
TABLE 1
Kidney P-glucuronidase concentrations
Gene name
Gene
location
-
-
Beige-J
Pale ear
Pearl
Pallid
Maroon
Ruby eye
Anemic light
Dominant spotting
Extreme nonagouti
Himalayan
Slight dilution
Leaden, fuzzy
Viable yellow
White
Albino-2J
Underwhite
Grizzled
Angora
Blotchy
Tanoid
Belted-2J
Lustrous
Slaty
Cordovan-J
Brown-J
Yellow
Recessive yellow
Buff
Mahoganoid
Velvet coat
AI bino-J
Viable dominant spotting
Sepia
White-bellied agouti-J
13
19
13
2
7
19
4
5
2
7
9
1
2
6
7
15
10
5
X
2
15
11
14
4
4
2
8
5
16
15
7
5
1
2
/3 -Glucuronidase
(units/g kidney)
130 I 8.4
330 I 9.7*
313 t 11 *
322 I
:14 *
324 t 13 *
225 f 13 *
210 t 8.9*
184 t 7.92
176 1: 8.7s
172
9.01)
178 i- 8.411
170 t 6.8s
14.9 f 5.0
133 +- 11
153 I 5.2
161 t 16
133 I 9.9
117 1 1 5
161 3- 15
138 1 6.8
151 k 9.9
148 1: 20
1 M I 11
1741- 14
14Q -+ 5.9
149 j, 11
132 j, 11
150 1: 13
154 t 3.6
171 I 1 4
150 1 14
143 2 29
120 1 1.5
138 1 4.2
168 1 1 7
(15)
(15)
(15)
(15)
(15)
(12)
(IO)
(9)
(2)
(2)
(4)
(5)
(3)
(3)
(2)
(2)
(2)
(2)
(4)
(2)
(2)
(6)
(3)
(3)
(6)
(7)
(2)
(4)
(2)
(3)
(2)
(2)
(2)
(5)
(2)
Female mice were treated 20 days with testosterone, Values represent mean +- SEM. Number
of mice tested is given in parenthesis. Entries under gene location refer to the chromosomal site
of the mutation. Angora, lustrous and velvet are mutations affecting hair structure rather than
pigmentation.
* P _< 0.001; $ P 5 0.01; P 5 0.02; '11 P 5 0.05.
and ruby eye, increases of 1.6- to 1.8-fold were noted. Smaller (about 1.3-fold)
but significant increases were noted in a third mutant poup including anemiclight, dominant spotting, extreme nonagouti, slight dilution and Himalayan. In
two cases, anemic-light and dominant spotting, these increases appeared with
only one copy of the mutant gene. Kidney glucuronidase levels in a fourth group
of 23 mutants were not significantly different from normal. No mutant had
193
LYSOSOMAL DYSFUNCTION A N D P I G M E N T GENES
TABLE 2
Kidney lysosomal enzyme concentrations
D-G!ucuronidase
(units/g kidney)
Normal
Beige-J
Pale-ear
Pearl
Pallid
130 i: 8.4
330 & 9.7*
313 & 11 *
3 2 2 3 ~ 1 4*
324 i: 13 *
B-Galactqsidase
(units/g kidney)
(units/g kidney)
wMannosidase
20 f 0.3
42 i: 1.6*
4 4 f 2.3*
32 i: 1.0*
31 f 1.0*
18 f 1.1
37 i: 3.2$
34 i: 1.3*
29 f 1.5$
29 f 3.811
Values are the mean f SEM of five or six mice treated 20 days with testosterone.
* P 5 0.001; $ P 5 0.01; '11 P 5 0.05.
glucuronidase concentrations depressed significantly below the normal level.
The three mutants, pale ear, pearl and pallid, were chosen for further characterization of the abnormally high kidney glucuronidase concentrations found in
this preliminary survey.
Generalized ejjcect on kidney lysosoma2 enzymes: Not only glucuronidase, but
at least two additional kidney lysosomal enzymes, galactosidase and mannosidase, were significantly elevated in activity in all three pigment mutants (Table
2) , suggesting a coordinate mechanism. Galactosidase concentrations were 1.6to 2.2-fold higher in the mutants, and kidney mannosidase activities, 1.6- to 2.1fold higher.
Consistent with a generalized effect of these pigment mutants on lysosomal
enzymes is the fact (data not shown) that all mutants (Table 1) having normal
kidney glucuronidase activity in 20 day testosterone-treated mice likewise had
normal kidney galactosidase and mannosidase concentrations.This was also true
in untreated mice and in mice treated for only seven days with testosterone.
Mutant lysosomal enzyme increase is not testosterone-dependent: Increased
concentrations of kidney lysosomal enzymes in the pale-ear, pearl and pallid
mutants are not testosterone dependent. In untreated females, there were significant increases in all three lysosomal enzymes, with the sole exception of a-mannosidase of pallid mice (Table 3). More importantly, &curonidase was 1.3- to
1.6-fold higher than normal in untreated mutants. Coordinate increases were
TABLE 3
Kidney lysosomal enzyme concentrations in untreated mice
P-G!ucuronidase
(urnts/g kidney)
Normal
Beige-J
Pale-ear
Pearl
Pallid
5.64 i 0.20
8.54 f 0.34*
9.21 f 0.51*
8.70 i 0.25*
7.13 f 0.10*
0-Galactosidase
(umts/g kidney)
21.8 S 0.32
35.7 i 3.0 *
31.2 f 4.1 '11
33.6 f 1.8 *
32.2 f 2.8 $
Values represent the mean -C SEM of eight untreated females.
* P 5 0.001; $ P 5 0.01; I P 5 0.02; Ill P 5 0.05.
e-Mannosidase
(units/g kidney)
11.6 S 0.33
20.3 i: 2.8
21.1 f 1.8 $
17.1 zk 0.59*
12.7 S 0.26
194
E. K. NOVAK A N D R. T. SWANK
TABLE 4
Liuer lysosomal enzyme concentrations
P-Glucuronidase
(units/g liver)
Normal
Beige-J
Pale-ear
Pearl
Pallid
fi-Galactosidase
(units/g liver)
38 t 1.8
36 t 1.1
39 2.9
38 i. 2.0
33 t 2.8
*
a-Mannosidase
(units/g liver)
19 i 3.0
13 i 0.5
19 t 3.6
18 i 2.3
15 t 1.4
9.0 k 1.0
7.2 k 0.4
11.6 f 1.0
8.7 t 0.5
8.7 t 0.5
Values are the meail t SEM of five or six mice treated 20 days with testosterone.
noted for galactosidase (1.5-fold) in all three mutants and for mannosidase (1.5to 1.8-fold) in the pale-ear and pearl mutants. The fact that effects on kidney
lysosomal enzyme concentrations and on secretion of kidney lysosomal enzymes
are not testosterone dependent (see below) is a practical advantage since it
enables study of the mechanism of action of the mutants without long-term
hormone treatment.
Tissue specificity: Concentrations of liver glucuronidase, galactosidase and
mannosidase were not significantly different ( P > 0.05) from normal in any
mutant (Table 4),indicating some tissue specificity.
Likewise, in three out of four mutants, serum lysosomal enzyme content was
unaffected (Table 5 ) . An exception was the pale-ear mutant, which had double
the serum glucuronidase and galactosidase concentrations. In no mutant were
serum levels affected by testosterone treatment.
It is unlikely that the abnormally high level of kidney lysosomal enzymes in
the mutants represents enzyme transported by serum. First, serum lysosomal
enzyme concentrations are not abnormally increased in the beige, pearl and
pallid mutants. Second, total serum glucuronidase levels form a very small fraction (0.03 to 0.1 %) of the kidney glucuronidase content of testosterone-treated
mice. Likewise, in the case of pale-ear mice, which have abnormally high serum
TABLE 5
Serum lysosomal enzyme concentrations
fi-Glucuronidase
(units/ml X 102)
Testosterone
Untreated
treated
5.31 t 0.16
5.06 t 0.36
12.9 i 0.87*
4.97 -C 0.44
4.20 t 0.53
Normal
Beige-J
Pale-ear
Pearl
Pallid
5.70 +- 0.54
6.48 +- 0.78
12.6 +- 0.33*
4.61 4 0.39
5.32 i 0.64
fi-Galactosidase
(units/nil X 102)
Untreated
5.86 -C 0.14
6.43 t 0.49
17.2 t 1.5 $
5.14 t 0.26
4.67 t 0.1911
Testosterone
treated
4.66
5.91
18.0
4.76
5.71
k 0.58
k
0.78
i 1.3 $
+
0.78
t 0.86
Testosterone-treated mice received the hormone for 20 days. Values represent the mean serum
activity for two to four mice. No difference was observed between untreated and testosteronetreated mice ( P 0.05).
* P 5 0.001; $ P 5 0.01; 11 P 5 0.05.
>
195
LYSOSOMAL DYSFUNCTION A N D P I G M E N T GENES
TABLE 6
Kidney nonlysosomal enzyme concentrations
Normal
Beige-J
Pale-ear
Pearl
Pallid
D-amino acid oxidase
(units/g)
Arginase
(units/g)
Alcohol dehydrogenase
375 f 27
394 I- 17
410 f 13
4887 f 42
384 t 17
3660 f 113
3590 f 198
3010 f 2.06
3480 t 232
3620 f 82
422* 47
566 f 131
369 f 63
501
19
450 f 35
(nnits/mg protein)
Fresh homogenates were assayed for D-amino acid oxidase, arginase and alcohol dehydrogenase
in mice treated 20 days with testosterone. Each value is the mean f SEM of six to eight mice.
Values were not significantly different from normal (P 0.05).
>
lysosomal enzyme levels, the excess accumulation of kidney lysosomal enzymes
after testosterone treatment is not accompanied by parallel increases in serum
lysosomal enzyme levels.
Lysosomal enzyme specificity: D-amino acid oxidase (peroxisomal), arginase
(cytosolic) and alcohol dehydrogenase (cytosolic) were assayed in testosteronetreated females (Table 6). There were no significant differences in kidney concentrations of any of these enzymes between normal and mutant mice. Therefore,
in these mutants there is a specific effect on lysosomal enzymes.
These data also indicate that the mutants do not specifically affect testosterone
sensitive enzymes. The above three enzymes were not affected despite the fact
that they, like lysosomal enzymes, are induced (from two- to ten-fold) by longterm testosterone treatment.
Kidney hypertrophy: The androgen-sensitive mouse kidney undergoes a striking hypertrophy of proximal tubule cells after testosterone treatment (DUNN
1948). The hypertrophy in the beige mouse pigment mutant is excessive
(BRANDT,ELLIOTT
and SWANK1975), probably as a result of accumulation of
nonsecreted, residual material in the giant lysosomes typical of this mutation.
Therefore, kidney hypertrophy was analyzed in the other pigment mutants to
determine if analogous processes might be occurring (Table 7). I n untreated
mutants, kidney weights were not significantly different from normal with the
TABLE 7
Kidney hypertrophy
~~
~
Kidney weight (g/2Og body weight)
Untreated
Testosterone-treated
Normal
Beige-J
Pale-ear
Pearl
Pallid
0.220
0.233
0.231
0.243
0.246
t 0.004
f 0.016
-C 0.008
f 0.008
f 0.006$
0.355 f 0.006
0.447 -C 0.009*
0.378 f 0.005$
(4.430 f 0.008'
0.425 & 0.010'
Each value represents 10 to 20 untreated or 20-day testosterone-treatedmice.
'P 5 0.001; $ P 5 0.05. Significance tests compare mutant to normal kidney weight.
196
1
N
E. K. NOVAK A N D R. T. SWANK
64-
0
N^
0
3.2-
3-
1.6-
4
0.8-
U
0
a
2
3
0
Q
OA-
0.2-
U
2
,J
d
4
6
8
1012
RING DIAMETER h)
6.4t
2
4
6
8
1012
RING DIAMETER (mm)
FIGURE
1.-Immuno-quantitation of p-glucuronidase by serial radial diffusion. Aliquots of
P-glucuronidase were added to wells of a plate containing goat anti-mouse glucuronidase antibody in agarose. Ring diameters were measured after 48 hr of incubation at 25" in a moist
chamber. (a) Kidney homogenate, (a) Urine. 0 C57BL/6J Normal,
Pale ear, 0 Pallid,
Pearl.
possible exception of a very slight (10%) increase in the pallid mutant. After
testosterone treatment, however, excessive kidney hypertrophy was noted in all
mutants. The increase varied from a small 10% over normal in the case of pale
ear to an approximately 20% increase in kidney mass in the beige, pearl and
pallid mutations.
Mechanism of increased kidney lysosomal enzyme concentration: The results
of several tests indicate that the major effect of the mutants is to decrease kidney
lysosomal enzyme secretion. A possible explanation for the increased kidney
lysosomal enzyme activity is that the pigment mutations act to affect the catalytic activity of these enzymes without changing their intracellular concentration. However, this is not true, at least in the case of glucuronidase, since it was
found in immunotitration tests by the single radial diffusion method (Figure l a )
that equal units of normal and mutant kidney enzyme had equal immunoreactivity. This was also true for glucuronidase from urine (Figure Ib) .Also, enzyme
activity was additive in mixed normal and mutant kidney extracts or urines.
Another possible mechanism is that the rate of synthesis of lysosomes is higher
in the pigment mutants. Rates of synthesis of kidney glucuronidase were accordingly measured in testosterone-treatedmutant females after radiolabelling with
( 3H)-leucine and purification by specific immunoprecipitation (Table 8). The
relative rate of glucuronidase synthesis did not differ significantly from normal
in the pale-ear and the beige mutants (SWANK
and BRANDT1978). There were,
however, significant increases (40 and 50%, respectively) in the pallid and pearl
mutants. The increase in relative synthetic rate in these mutants was, moreover,
restricted to an increase in radiolabelling of glucuronidase since the rate of
synthesis of total kidney protein was equal to that of normal mice. There was an
apparent small increase (14%) in rate of kidney protein synthesis in pallid mice
in this experiment. However, in four independent experiments rates were equal
in pallid and normal mice.
197
LYSOSOMAL D Y S F U N C T I O N A N D P I G M E N T GENES
TABLE 8
Rates of kidney glucuronidase and protein synthesis
Radioactivity in
glucuronidase
(CPM/g kidney)
(1)/(2) x IO4
7.39 t 0 . a
8.71 t 0.86
11.1 t 0.88s
10.6 t 0.61$
(2)
3630 k 117
3340 rt 117
3960 f 127
4140 rt 78
(1)
28.20 2 110
2930 t 34f3
4380 rt 133
4350 A 275
Normal
Pale-ear
Pearl
Pallid
Relative rate
of glucuronidase
synthesis
Radioactivity in
total cell protein
(CPM/g kidney X
100 pCi 3H-leucine was injected intraperitoneally and kidney homogenates prepared after one
U) days with
testosterone.
$ P 2 0.01; s P 2 0.02.
hour. Each value is the mean t SEM of six to eight individual mice treated
The other likely mechanism for an increased concentration of kidney lysosomal enzymes in the pigment mutants is an alteration in secretion from kidney
into urine. In testosterone-treated mice this is the major mechanism of loss of
kidney lysosomal enzymes. For example, about 30% of both glucuronidase and
galactosidase are daily secreted in normal C57BL/6J mice (Table 9). Measured
in terms of units secreted per day, secretion of glucuronidase was greatly reduced
in beige, pale ear and pallid. There was a coordinately low secretion of the second
lysosomal enzyme galactosidase in these mutants. The defective secretion in these
three mutants became even more apparent when the data were expressed as the
percent of the total kidney lysosomal enzyme content that is secreted daily. While
approximately 30% of both kidney lysosomal enzymes were secreted daily in
normal mice, only 1 to 6% was secreted in the mutants. Again, the coordinate
control of secretion of the two lysosomal enzymes in each mutant was apparent.
Pearl is unusual in that the units of glucuronidase secreted daily approached that
of normal mice ( p > 0.05), although as in the other mutants the galactosidase
secretion was significantly less than normal. The secretory defect in pearl is
evident, however, in that only 8 to 10% of the total kidney content of the two
lysosomal enzymes were secreted daily. Expressing the data as percent total
TABLE 9
Secretion of kidney lysosomal enzymes in testosterone-treated mice
8-Glucuronidase
( units/mouse/day)
Normal
Beige-J
Pale-ear
Pearl
Pallid
13.8 f 1.43
2.6 & O.W*
1.5 t 0.21*
11.1 f 0.73
5.0 rfr 0.54*
% Kidney
8-glucuronidase
8 -Galactosidase
(units/mouse/day)
% Kidney
8-galactosidase
29.9
1.8
1.3
8.0
3.6
1.9 t 0.18
0.43 t 0.04*
0.36 f 0.04.
1.31 t 0.0711
0.73 2 0.04”
26.7
2.3
2.2
9.6
5.7
Protein excretion
(mg/mouse/day)
9.48
7.4
9.3
11.5
10.3
f 1.20
2 0.90
& 0.86
f 1.2
t 0.67
Each metabolism cage contained three mice. Values are the mean daily urinary secretions t
11 P 5 0.05).
Percent kidney content was calculated from the kidney weight (Table 7) and enzyme specific
activity (Table 2 ) .
SEM measured on days 16 to 20 after testosterone implantation. ( * P 5 0.001;
198
E. K. NOVAK A N D
R. T. SWANK
TABLE 10
Secretion
Normal
Beige-J
Pale-ear
Pearl
Pallid
of
P-galactosidase in untreated mice
fl-Galactosidase secretion
units/mouse/day
Kidney
content
0.155 k 0.014
0.014 f 0.003*
0.032 f 0.008$
0.047 k 0.0163
0.062 f 0.019s
3.20
0.19
0.38
0.58
0.78
Four female mice were placed in each metabolism cage and urine was collected for five conP i 0.02.
secutive days and assayed immediately. * P
0.001; $ P
0.01;
<
<
s
kidney content secreted corrects for the fact that there is a 40 to 50% increase
in the rate of glucuronidase synthesis (Table 9) in the pearl and pallid mutants.
The pearl mutation likewise coordinately affected both lysosomal enzymes.
While kidney lysosomal enzyme secretion was greatly depressed in all
mutants, total urinary protein excretion was not affected in any mutant (Table
9). More than 90% of this protein is the so-called mouse urinary protein M U P
(FINLAYSON
et al. 1965; SZOKA
and PAIGEN
1978), which is a low M W protein
synthesized in and secreted from liver in response to testosterone administration
and subsequently rapidly filtered through the kidney glomerulus.
A depressed lysosomal enzyme secretory rate was also evident in mutant
females not treated with testosterone (Table 10). Secretion of galactosidase was
reduced more than ten-fold in untreated normal females as compared to testosterone-treated counterparts. Among the mutants, the units secreted per day was
from 10 to 40% of the normal rate. These differences were magnified by a factor
of approximately two when secretion was calculated as a percent of total kidney
galactosidase content.
DISCUSSION
Mouse pigment mutants are a rich source of defined genes that regulate lysosomal function in a novel manner. Including the five pigment mutants (pale-ear,
pearl, pallid, maroon and ruby eye) described here, plus the previously described
beige (BRANDT, ELLIOTTand SWANK1975) and light-ear (MEISLER1978)
mutants, there are now at least seven pigment genes known to increase significantly the concentrations of mouse kidney lysosomal enzymes. MEISLER(1978)
has likewise reported that kidney galactosidase concentrations are increased in
the pale-ear, pearl and maroon mutants. Smaller increases were noted in five
and LUNDIN
(1977) have found high levels of kidney
other mutants. HAKANSSON
lysosomal enzymes in male buff mutant mice. However, as noted in these studies,
kidney glucuronidase and galactosidase concentrations were reported as normal
in testosterone-treated buff females.
The pale-ear, pearl, pallid, beige (BRANDT,
ELLIOTTand SWANK1975) and
light-ear (MEISLER1978) mutants have in common an abnormally low rate of
secretion of kidney lysosomal enzymes. This accounts for an accumulation of
LYSOSOMAL DYSFUNCTION A N D PIGMENT GENES
199
enzymes in kidney. The percent of total kidney glucuronidase and galactosidase
daily secreted in normal mice was depressed by three- to 15-fold in the mutants
of this study. In these four mutations the effects on kidney levels and secretion
of the two lysosomal enzymes were coordinate, indicating that the mutations
probably affect the entire lysosomal organelle. The normal mechanism of secretion of lysosomal contents from mouse kidney is incompletely defined. It is
known, however, from micrographs of proximal tubule cells that lysosomes are
concentrated in the apical region of the cell near the brush border (BRANDT,
ELLIOTT
and SWANK
1975) and that these lysosomes are engorged with myelinet al. 1975).
like figures that are also visible in the tubular lumen (PAIGEN
The presence of normal quantities of mouse urinary protein in mutant mice
indicates that not all protein secretory processes are affected. Also, in at least
unpublished), light-ear and pale-ear
three mutants, beige (BRANDTand SWANK,
(MEISLER
1978), rates of secretion of amylase from pancreas are normal.
It should be noted that our measure of secretion is the net appearance of lysosomal enzyme in urine. We cannot exclude, therefore, the possibility of more
complex, but equally interesting, mechanisms of action. For example, some mutations could cause an enhanced reabsorption in kidney of previously secreted
kidney lysosomal enzymes. However, it is known that several nonlysosomal
proteins that pass through the kidney tubule lumen are processed identically in
normal and mutant mice. The mutants excrete normal amounts of serumderived, low molecular weight proteins. Also, in the beige mutant we have found
normal rates of absorption in kidney of intravenously administered [lz5I]ribonuclease and [lz5I]horseradish peroxidase (SWANK,
unpublished).
While normal rates of synthesis of kidney glucuronidase were observed in the
pale-ear and beige (SWANK
and BRANDT1978) mutants, two of the mutants,
pearl and pallid, reproducibly had, in addition to the secretion defect, 40 to 50%
increases in kidney glucuronidase synthetic rates. Three other nonlysosomal
kidney enzymes were nevertheless present at normal concentrations, suggesting
that these mutants may specifically affect the biosynthetic rates of lysosomal
enzymes. Further experiments are needed to determine if the mutants simultaneously affect the synthetic rate of many lysosomal enzymes. Such a class of
mutations would be valuable in studies of lysosomal biogenesis. While mammalian mutations affecting the synthesis of individual lysosomal enzymes are
et al. 1975; SWANK
et al. 197813; SWANK,
PAIGEN
and GANSknown (PAIGEN
CHOW 1973), to our knowledge no mutations affecting the rate of synthesis of the
lysosome or other subcellular organelles have been described.
The mutant effects are relatively tissue specific. The fact that liver lysosomal
enzyme levels were not affected is consistent with a secretory effect of the
mutants, since mouse liver, unlike mouse kidney, is not known to secrete a large
proportion of its lysosomal content. Whether the increased lysosomal enzyme
levels in serum of the pale-ear mutant are due to excess soluble enzyme or
originate in blood cells is an interesting, but unresolved, question. This finding
plus the finding that beige is distinctive in the formation of giant lysosomes
(NOVAK,
unpublished) reemphasizes the fact that each pigment mutant is unique.
200
E. K.
NOVAK A N D
R. T. SWANK
The mutant effects in kidney are specific to lysosomal enzymes since the concentrations of three enzymes localized in other subcellular compartments were
normal. This is supported by the abnormal lysosomal secretion in untreated
mutant females, which have normal kidney mass. The fact that there is increased
hypertrophy in all mutants after testosterone treatment likely is caused by an
increased accumulation of material in nonsecreted secondary lysosomes under
hormone-accelerated lysosomal biogenesis and function. The experiments
reported here rule out a general increase in protein synthetic rate as the cause of
the increased hypertrophy in testosterone-treated mutants.
A major advantage of using the pigment mutants is that except for the chromosomal site of the mutation, the mutant strains are essentially identical to normal
strain C57BL/6J. In the case of the beige mutant, the difference could be as small
as a single nucleotide since it arose as a spontaneous (coisogenic) mutant in
normal strain C57BL/6J. The remaining mutants with lysosomal effects arose
in other inbred strains and were subsequently transferred, by investigators at the
Jackson Laboratory, to normal strain C57BL/6J by repeated backcross matings.
Except for the maroon mutation, which has been backcrossed nine times, the
number of backcrosses of the other mutations has been large, varying from 18
(pearl) to 41 (pallid) (PRISCILLA
LANE,personal communication). The proportion of contaminating (non-C57BL/6J) genes after n backcrosses is (0.5)" for
unlinked genes. Therefore, in all mutants more than 99% of genes unlinked to
the pigment gene are of C57BL/6J origin. Contaminating genes linked to the
selected pigment genes are lost at a slower rate depending on the chromosomal
distance between the genes (E. L. GREEN1968). The fact that lysosomal dysfunction is associated with pigment abnormalities in such a large number of
pigment mutants strongly suggests that the pigment genes themselves are
responsible for the lysosomal abnormalities. In support of this conclusion, F.
BERGER(personal communication) has recently found that kidney galactosidase
concentrations are normal among a large number of congenic mutants that did
not affect pigmentation.
The primary effects of the mutants on pigmentation (SEARLE
1968) and lysosomal function at the molecular level are not known. All affected mutants have
diluted pigmentation. The pearl gene ( SARVELLA
1954) dilutes all the main types
of pigment. Abnormally small melanosomes have been detected in pale ear
(LANEand GREEN1967), ruby eye (MOYER
1966) and pallid (THERIAULT
and
HURLEY
1970). Melanosomes in ruby eye have been reported to be altered in
shape (MARKERT
and SILVERS
1956). However, not all pigment mutants with
defective size or shape or melanosomes have lysosomal abnormalities. For
example, normal kidney lysosomal enzyme concentrations were present in the
brown mutant, which has spheroidal rather than ovoid melanosomes (MARKERT
and SILVERS
1956). The pallid mutant has abnormal development of the otoliths
of the inner ear (LYON1951). This defect, but not the pigment dilution, can be
corrected by adding manganese to the diet (ERWAY,
FRASER
and HURLEY
1971) .
By f a r the best characterized pigment mutant is beige (BRANDT
et al. 1978;
LYSOSOMAL DYSFUNCTION A N D PIGMENT GENES
201
WINDHORST
and PADGETT
1973), an animal model for the human ChediakHigashi Syndrome (BLUME and WOLFF1972; RENSHAWet aZ. 1974). Recent
studies of this mutation in humans and animal models have detected abnormalities of intracellular organelles related to the secretion defects we have detected
in other pigment mutants. These include a depressed rate of fusion of phagosomes
with lysosomes (ROOT,ROSENTEALand BALESTRA
1972) and a depressed rate of
secretion of lysosomal enzymes from leukocytes (BOXER
et al. 1976) and platelets
(BOXER
et al. 1977). A possible mechanism for abnormal granule mobility in
cells of beige mice and Chediak-Higashi patients is a lowered polymerization of
microtubules (OLIVER 1976; HINDSand DANES1976). Several microtubulerelated functions including chemotaxis, bactericidal capacity and lysosomal
et at. 1976).
enzyme secretion are restored by treatment with ascorbate (BOXER
It has also been shown that leukocytes from Chediak-Higashi patients have an
abnormally high cyclic AMP concentration (BOXER
et al. 1976), a condition that
is correlated with depressed lysosomal secretion in other systems (IGNARRO,
LINT
et al. 1974).
and GEORGE
1974; ZURIER
The characteristics of the mutants described here are likewise unlike those of
previously described mutants (HERS
and VANHOOF1973; PAIGEN
et al. 1975)
affecting lysosomal function. The combination of defective lysosomal secretion
and abnormalities in other subcellular organelles has not to our knowledge been
described in other lysosomal mutations. These findings reinforce the histochemical and biochemical similarities reported between lysosomes and melanosomes
and suggest that there may be many steps in common in the control of subcellular organelles. Recent studies on bleeding abnormalities in human albinos
with Hermansky-Pudlak syndrome (WITKOPet al. 1973), the beige mouse
(HOLLAND
1976) and the Fawn-hooded rat (TSCHOPP
and ZUCKER
1972) suggest
that similar interrelationships may hold between melanosomes and another subcellular organelle, the platelet-dense granule. A fourth intracellular organelle,
the mast-cell granule, is morphologically abnormal in beige mice (CHI and
LAGUNOFF
1975). It should be possible to use these types of mutants to analyze
separate cellular regulatory processes that have general effects on several cellular
organelles, but are difficult to study by purely biochemical techniques.
Portions of this work have been presented in preliminary form at the Meeting of the Federation of the American Societies for Experimental Biology, April 1977 (NOVAK
and SWANK
1977)
the Symposium on Protein Turnover and Lysosomal Function, Buffalo, New York, August 1977
(SWANKet al. 1978a) and the Laurentian Hormone Conference, Mont Tremblant, Quebec,
et aZ. 1978b).
September 1977 (SWANK
This work would not have been possible without the help of several staff members of the
Jackson Laboratory including EVAEICHER,ELIZABETH
RUSSELL,
PRISCILLA
LANE,EARLGREEN,
MARGARET
GREENand DONALD
BAILEY,who constructed the congenic mutants used in these
studies and made them available to us. We also are indebted to CHARLOTTE
ABRAHAM,
GERALD
JAHREIS and CAROLYN
CLEVELAND
for excellent technical assistance. We thank CYNTHIA
BELL
and NANCY
HORTON
for secretarial aid.
Supported in part by Public Health Semice Grant GM-19521, by National Science Foundation Grant PCM77-24804, and by a grant from the National Foundation March of Dimes. E. K.
NOVAK
was supported in part by Public Health Service Grant GM-07093-03.
202
E. K. NOVAK A N D R. T. S W A N K
LITERATURE CITED
BAINTON,D. F. and M. G. FARQUHAR,
1966 Origin of granules in polymorphonuclear leukocytes. J. Cell Biol. 28: 277-301.
BLUME,R. S. and S. M. WOLFF,1972 The Chediak-Higashi Syndrome: studies in four patients
and a review of the literature. Medicine 51 : 247-280.
BOXER,G. J., H. HOLMSEN,
L. ROBKIN,N. BANG,L. A. BOXER
and R. L. BOEHNER,
1977 Abnormal platelet function in Chediak-Higashi Syndrome. Br. J. Haematol. 35: 521-533.
M. RISTER,H. BESCH,J. ALLENand R. L. BAEHNER
1976 CorBOXER,L. A., A. M. WATANABE,
relation of leukocyte functiomn in Chediak-Higashi Syndrome by ascorbate. New Engl. J.
Med. 295: 1041-1045.
BRANDT,
E. J., R. E. ELLIOTT,and R. T. SWANK,
1975 Defective lysosomal enzyme secretion in
kidneys of Chediak-Higashi (Beige) mice. J. Cell. Biol. 67: 774-788.
BRANDT,
E. J. and R. T. SWANK,1976 The Chediak-Higashi (Beige) mutation in two mouse
strains: allelism and similarity in lysosomal dysfunction. Am. J. Pathol. 82: 573-586.
1978 The murine ChediakBRANDT,
E. J., R. T. SWANK,E. K. NOVAKand M. SKUDLAREK,
Higashi mutation: a review. pp. 223-234. In: Comparative and Developmental Aspects of
Animal Models in Immunity and Disease. Edited by M. GERSHWIN.
Pergamon Press, New
York.
CHI, E. Y. and D. LAGUNOFF,1975 Abnormal mast cell granules in the Beige (ChediakHigashi) mouse. J. Histochem. Cytochem. 23: 117-122.
1978 Regulation of lysosomal enzymes in potassium deficient
CLEVELAND,
C. and R. T. SWANK,
mice. Biochem. J. 170: 249-256.
DIZIK, M., 1977 Characterization of a-mannosidase in inbred mice: Two genes that determine
the extent of sialylation of lysosomal a-mannosidase in mouse liver. Ph.D. thesis, State
University of New York at Buffalo.
DUNN,T.B., 1948 Some observations on normal and pathologic anatomy of the kidney of the
mouse. J. Natl. Cancer Inst. 9: 285-301.
and L. S. HURLEY,
1971 Prevention of congenital otolith defect in
ERWAY,
L. C., A. S. FRASER
pallid mutant mice by manganese supplementation. Genetics 67: 97-108.
J. S., R. ASOFSKY,
M. POTTER
and C. C. RUNNER,1965 Major urinary protein comFINLAYSON,
plex of normal mice: origin. Science 149: 981-982.
GREEN,E. L., 1968 Breeding Systems. pp. 11-22. In: Biology of the Laboratory Mouse. Edited
by E. L. GREEN,Dover Publications, New York.
GREEN,M. C., 1968 Mutant genes and linkages. pp. 87-150. In: Biology of the Laboratory
Mouse. Edited by E. L. GREEN.Dover Publications, New York.
E. M. and L. G. LUNDIN,1977 Effect of a coat color locus on kidney lysosomal
HAKANSSON,
glycosidases in the house mouse. Biochem. Genet. 15: 75-85.
HERS,H. G. and F. VAN HOOF,1973 Lysosomes and Storage Diseases. Academic Press Inc.,
New York,
HINDS,
K. and B. S. DANES,1976 Microtubular defect in Chediak-Higashi Syndrome. Lancet
2: 246.
HOLLAND,
J. M., 1976 Serotonin deficiency and prolonged bleeding time in Beige mice. Proc.
SOC.Exp. Biol. Med. 151 : 32-39.
IGNARRO,
L. J., T. F. LINT and W. J. GEORGE,
1974 Hormonal control of enzyme release from
human neutrophils. J. Exp. Med. 139: 1395-1414.
LANE,P. W. and E. L. GREEN,1967 Pale ear and light ear in the house mouse. J. Heredity
58: 17-20.
LYSOSOMAL D Y S F U N C T I O N A N D P I G M E N T G E N E S
203
LYON,M. F., 1951 Hereditary absence of otoliths in the house mouse. J. Physiol. (London)
114: 410418.
MAUL,G. G., 1969 Golgi-melanosome relationship in human melanoma in vitro. J. Ultrastruct.
Res. 26: 163-176.
MARKERT,
C. L. and W. K. SILVERS,1956 The effects of genotype and cell environment on
melanoblast differentiation in the house mouse. Genetics 41 : 429-450.
MEISLER,M., 1978 Synthesis and secretion of kidney p-galactosidase in mutant le/le mice.
J. Biol. Chem. 253: 3129-3134.
MOYER,F H., 1966 Genetic variations in the fine structure and ontogeny of mouse melanin
granules. Am. Zoologist 6 : 43-66.
NOVAK,E. K. and R. T. SWANK,1977 Altered lysosomal enzyme levels in mouse pigment
mutants. Fed. Proc., Fed. Am. Soc. Exp. Biol. 36: 859.
NOVIKOFF,A. B., 1976 The endoplasmic reticulum: A cytochemists view (a review). Proc.
Natl. Acad. Sci. U.S. 73: 2781-2787.
NOVIKOFF,
A. B., A. ALBALA
and L. BIEMPICA,1968 Ultrastructural and cytochemical studies
on B-16 and Harding-Passey mouse melanomas. The origin of premelanosomes and compound melanosomes. J. Histochem. Cytochem. 14: 299-319.
OLIVER.J. M., 1976 Impaired microtubule function correctable by cyclic GMP and cholinergic
agonists in the Chediak-Higashi Syndrome. Am. J. Pathol. 85: 395412.
PAIGEN,K., R. T. SWANK,S. TOMINO
and R. E. GANSCHOW,
1975 The molecular genetics of
mammalian glucuronidase. J. Cell. Physiol. 85: 379-392.
and G. A. PADGETT,
1974 Leukocyte dysRENSHAW,
H . W., W. C. DAVIS,H. W. FUDENBERG
function in the bovine homologue of the Chediak-Higashi Syndrome of humans. Infect.
Immun. 10: 928-937.
ROOT,R. K., A. S. ROSENTHAL
and D. J. BALESTRA,
1972 Abnormal bactericidal, metabolic and
lysosomal functions of Chediak-Higashi Syndrome leukocytes. J. Clin. Invest. 51 : 64.9-665.
SARVELLA,
P. A., 1954 Pearl, a new spontaneous coat and eye color mutation in the house
mouse. J. Heredity 45: 19-20.
SCHIMKE,R. T., 1970 Arginase (rat liver). pp. 313-317. In: Methods of Enzymology, Vol.
and C. W. TABOR.
Academic Press, New York.
XVII Part A. Edited by H. TABOR
SEARLE,
A. G., 1968 Comparative Genetics of Coat Color in Mnmmals. Academic Press, Inc.,
New York, New York.
SWANK,R. T. and E. J. BRANDT,1978 Turnover of kidney P-glucuronidase in normal and
Chediak-Higashi (Beige) mice. Am. J. Pathol. 92: 755-769.
E. J. BRANDT
and M. SKUDLAREK,
1978a Genetics of Lysosomal funcSWANK,R. T., E. NOVAK,
tions. pp. 251-271. In: Protein Turnover and Lysosomal Function. Edited by D. DOYLEand
H. SEGAL.
Academic Press, New York.
SWANK,R. T., K. PAIGEN,R. DAVEY,V. CHAPMAN,C. LABARCA,
G. WATSON,R. GANSCHOW,
and E. NOVAK,197813 Genetic regulation of mammalian glucuronidase. pp.
E. 5. BRANDT
401436. In: Recent Progress in Hormone Research, Vol. 34. Edited by R. 0. GREEP.Academic Press, New York.
and R. E. GANSCHOW,
1973 Genetic control of glucuronidase inducSWANK,R. T., K. PAIGEN
tion in mice. J. Mol. Biol. 81 : 225-243.
P. R. and K. PAIGEN,
1978 Regulation of mouse urinary protein production by the
by the Mup-A gene. Genetics 90: 597-612.
SZOK4,
THERIAWLT,
L. L. and L. S. HURLEY,1970 Ultrastructure of developing melanosomes in C57
Black and Pallid mice. Dev. Biol. 23: 261-275.
204
E. K. NOVAK A N D R. T. SWANK
TODA,
K. and T. B. FITZPATRICK,
1971 The origin of melanosomes. pp. 265-278. In: Biology
of Normal and Abnormal Melanocytes. Edited by T. KAWAMURA,
T. B. FITZPATRICK
and
M. SEIJI. University of Tokyo Press, Tokyo.
TSCHOPP,
T. B. and M. B. ZUCKER,1972 Hereditary defect in platelet function in rats. Blood
40: 217-226.
WELLNER,
D. and L. A. LICHTENBERG,
1971 Assay of Amino Acid Oxidase. pp. 593-595. In:
Methods of Enzymology, Vol. XVII Pt. B. Edited by H. TABOR
and C. W. TABOR.
Academic
Press, New York.
WINDHORST,
D. B. and G. PADGETT,
1973 The Chediak-Higashi Syndrome and the homologous
trait in animals. J. Invest. Dermatol. 60:529-537.
WITKOP,C. J., C. W. HILL,S. DESNICK,J. K. TRIES,H. L. THORN,
M. JENKINS
and J. G.
WHITE, 1973 Ophthalmologic biochemical, platelet and ultrastructural defects in the
various types of oculocutaneous albinism. J. Invest. Dermatol. 60:443-456.
WOLFF,K. and E. SCHREINER,
1970 Epidermal lysosomes. Arch. Dermatol. 101: 276-286.
S. HOFFSTEIN,
S. KAMMERMAN
and N. H. TAI,1974 Mechanisms
ZURIER,R. B., G. WEISSMAN,
of lysosomal enzyme release from human leukocytes. J. Clin. Invest. 53: 297-309.
Corresponding editor: D. BENNETT