NASA Technical Reports Server (NTRS) 19980002822: Effects of Microgravity or Simulated Launch on Testicular Function in Rats

14 Pages·1.3 MB·English

by  NASA Technical Reports Server (NTRS)

NASA-CR-204998
Effects of microgravity or simulated launch " j
on testicular function in rats
R. P. AMANN, D. R. DEAVER, B. R. ZIRKIN, G. S. GRILLS, W. J. SAPP,
D. N. R. VEERAMACHANENI, J. W. CLEMENS, S. D. BANERJEE, J. FOLMER,
C. M. GRUPPI, D. J. WOLGEMUTH, AND C. S. WILLIAMS
Animal Reproduction and Biotechnology Laboratory, Colorado State University, Fort Collins, Colorado
80523; Department of Dairy and Animal Science, The Pennsylvania State University, University Park,
Pennsylvania 16802; Department of Population Dynamics, The Johns Hopkins University, Baltimore,
? Maryland 21205; Center for Reproductive Sciences and Department of Genetics and Development,
Columbia University College of Physicians and Surgeons, New York, New York 10032;
and Department of Biology, Tuskegee University, Tuskegee, Alabama 36088
AMAba_,R. P.,D. R.DEAVER, B.R.ZIRKIN,G. S.GRILLS, (29) concluded thatspermatogenesis was not affectedin
W.J. SAPP, D. N. R. VEERAMAC_, J.W. CLEMENS, ratsflown 7 or 13 days on COSMOS 1667 orCOSMOS
S.D.BANERJEE,J.FOLMER, C.M. GRUPPI,D.J.WOLGEMUTH, 1887.Serum concentrationsoftestosteronewere low in
ANDC. S. WILLIAMS. Effects ofmicrogravity orsimulated launch ratsafterreturn from spaceflight(11),indicatingthat
on testicular function in rats. J. Appl. Physiol. 73(2}, Suppl.:
endocrine function of the testesmight be alteredby
174S-185S, 1992.--Testes from flight rats on COSMOS 2044
stressfactorsassociatedwithspaceflightoreffectsofmi-
and simulated-launch, vivarium, or caudal-elevation control
crogravity.The reasons forabnormalities of the endo-
rats (5/group) were analyzed by subjective and quantitative
methods. On the basis of observations of fixed tissue, it was crineand possibly the exocrine (spermatogenic) func-
evident that some rats had testicular abnormalities unasso- tionsofthetestesassociatedwith spaceflight,atleastin
ciated with treatment and probably existing when they were rats,areunknown but couldbe relatedto alteredfunc-
assigned randomly to the four treatment groups. Considering tion of the hypothalamus or adenohypophysis (9),al-
rats without preexisting abnormalities, diameter of seminifer- teredfluiddistributioninthebody (22),and/or restricted
ous tubules and numbers of germ cells per tubule cross section bloodand lymph flowwithin thetestis.
were lower (P < 0.05} in flight than in simulated-launch or
Startingwith the assumption thatstressfactorsasso-
vivarium rats. However, ratios of germ cells to each other or to
ciatedwith spaceflight,oreffectsofmicrogravityperse,
Sertoli cells and number of homogenization-resistant sperma-
representsimply a unique setofenvironmental toxins,
tids did not differ from values for simulated-launch or vivarium
we usedapproaches ofreproductivetoxicologyand cellu-
controls. Expression of testis-specific gene products was not
larand molecular biologytoevaluatetesticularfunction
greatly altered by flight. Furthermore, there was no evidence
for production of stress-inducible transcripts of the hsp70 or inratsfrom COSMOS 2044.Sensitivemethods forquan-
hspg0 genes. Concentration of receptors for rat luteinizing hor- tificatioonfalterationsoftesticularfunctionby enumera-
mone in testicular tissue and surface density of smooth endo- tionof germ cells,morphometric analyses,and tech-
plasmic reticulum in Leydig cells were similar in flight and sim- niquesofcellularendocrinologyhave proven tobeinvalu-
ulated-launch rats. However, concentrations of testosterone in able in identifyingreproductive toxins (1-3, 17).As
testiculatrissueorperipheralbloodplasma werereduced(P<
andersta_ding of th_ genet;.c regulation of testicul._r
0.05}inflightratsto<20% ofvaluesforsimulated-launchor
function evolved, it became evident that eukaryotic or-
vivariumcontrolsT.hus spermatogenesiswas essentiallnyor-
realinflightrats,butproductionoftestosteronewas severely ganisms synthesize a subset of gene products in response
i, depressed.Exposuretomicrogravityfor>2 wk mightresultin to environmental stress; these have been termed cellu-
additionaclhanges.Sequelaeofreducedandrogenproduction lar-stress or heat-shock proteins. Some members of this
associatedwithmicrogravityon turnoverofmuscleand bone gene family also are expressed at specific points during
shouldbeconsidered. spermatogenesis in mammals (33), and their expression
is associated with specific germ cell types (34, 35). Thus
testis; spermatogenesis; intratesticular testosterone; Leydig expression of these genes provides molecular markers
cells; cellular-stress genes that can indicate normal spermatogenesis or can reveal
whether a testis is responding to an altered environ-
ment (16).
THERE IS seemingly contradictory evidence on effects of
spaceflight or microgravity on testicular function in METHODS
mammals. Although a slight decrease in number of sper-
Animals and Tissue Collection
matogonia per cross section of seminiferous tubule in
testes from rats flown for 7days on Spacelab 3or 13 days Details of the history, age, and body weight of the rats;
on COSMOS 1887 were reported (23, 28), Serova et al. assignment to four treatment groups (flight, simulated
174S 0161-7567/92 $2.00 Copyright © 1992the American Physiological Society
.--.
f
MICROGRAVITYAND TESTIS FUNCTION 175S
launch, vivarium, and caudal elevation); animal care be- interstitial tissue. Diameters of 25 essentially round
fore and during the experiment; conditions of simulated cross sections through seminiferous tubules and their lu-
or actual launch, flight, and reentry; and timing of tissue mina were measured for each of two to four loci (blocks)
sampling relative to reentry are presented elsewhere representing each testicular pole. For each sta_ VII
(10). Briefly, flight rats were flown on an unmanned cross section (18), the number of Sertoli cells with a dis-
COSMOS satellite for 14days, and tissues were sampled cernible nucleohs, and numbers of nuclei of Aspermato-
for analysis 11h after return to earth. Simulated launch gonia, preleptotene spermatooyt_, pachytene spmmato-
rats were subjected to vibrational and g-force stress but cytes, or step-7 spermatids were recorded. Diameters of
not microgravity. Additional control rats were subjected two nucleoU of Sertoli cells and oftwo nuclei of each type
to caudal elevation (6, 13, 30) to Unwmght the hind limbs of germ veil also were measun_ Concurrently, for each
and simulate effects of thoracic pooling of body fluids. tissue fraction, appearance of_the seminfferous epithe-
All rats were adult males of Czechoslovakian origi_ lium wB classified as normal or having minor, moderate,
For each of five rats (mrs 6-10) in the four groups, or severe degenerative changes. Counts of germ cells
each testis was separated from its epididymis_; both or- were corrected for differences ,in nuclear diameter and
gans were trimmed free of extraneous tissue and section thickness using a modification of Abercrombie's
weighed. The right testis was cut into three pieces; acen- formula (6), and the resulting data were used to calculate
tral portion was placed into acontainer and immersed in numbers of germ cells per Sertoli cell or ratios among
liquid nitrogen (-196°C), and the two poles (ea,'h _p_- g_rm cell _es (2).
senting _-0.25 of the testicular length) were immersed In addition, 2-/zm sections were cut from mlected
into fixative at 4°C. The epididymis also was immersed blocks (abnormal rats were excluded) and stained with
in fixative. The fixative was Triple Fix (24), which con- toluidine blue for enumeration of nuclei of spermato-
tained 1% (wt/vol) paraformaldehyde, 3% (vol/vol) glu- gonia (type A or B spermatogonia) and Sertoli cells in
taraldehyde, 88 mM sucrose, and 2.5 mM difluorodini- essentially round cross sections of stage VI seminiferous
trobenzene in 0.1 M sodium cacodylate buffer. tubules. Tubules were staged on the basis of morphology
Frozen tissues remained in liquid nitrogen or on solid of the spermatogonia (23), with verification of staging
CO_, and fixed tissues in 4°C fixative, until receipt at based on morphological features of spermatids (18). Nu-
Colorado State University. There, frozen tissue was ali- clei in every other section through one, or occasionally
quoted, while at -79°C, for shipment at -79°C to Dr. two, stage VI tubule were counted until 50 cross sections
Deaver for enumeration of homogenization-resistant had been scored. Counts were corrected for nuclear diam-
spermatids and quantification of testosterone and lu- eter and section thickness, after which the mean value
teinizing hormone receptors and to Dr. Wolgemuth for was divided by 2.5 to make the data comparable to those
measurement of expression of cellular-stress genes. for stage VII tubules (sections were 2.0 rather than 0.8
Fixed tissue also was divide& the poles of a testis were gm thick).
randomly designated as A or B, and the outer portion of Efficiency ofsperm production. The number of homoge-
each pole was sliced offand shipped to Dr. Zirkin in0.1 M nization-resistant spermatids per gram of testicular tis-
cacodylate buffer, at 4°C, for morphometric analysis of sue, a measure of the efficiency of sperm production (1),
Leydig cells. The inner portion of each testicular pole was determined by modifying the procedure of Robb et
was retained by Dr. Amann for subjective and objective al. {27). Briefly, frozen tissue was thawed, weighed (131-
evaluations of spermatogenesis and measurement ofsem- 253 rag), and minced with two razor blades. The tissue
iniferous tubule diameter; this tissue later was shared was transferred quantitatively with 2.0 ml buffer (0.5 M
with Dr. Sapp. To enable deductions concerning onset of Tris-HC1, 0.23 M sucrose, 5 mM MgCl_, and 4 mM
alterations in the seminiferous epithelium, we examined NaN 3)into a 7-ml Dounce homogenizer and subjected to
epididymal tissue microscopically to estimate the num- 10strokes with pestle B followed by 10 strokes withpestle
ber of sloughed immature germ cells present among the A. The h_mogenate was centrifuged (800 g for 10mL,_)at
spermatozoa. 4°C. The supernatant was saved {see Leydig Cell Func-
tion), and the pellet was resuspended in 2 ml buffer and
Spermatogenesis homogenized a second time. After centrifugation, the su-
Histological evaluations. Thin slices of fixed tissue were pernatant was added to the previous one, and the pellet
rinsed in 0.1 M cacodylate buffer, immersed 1 h in OsO4 was resuspended in 5-10 ml of 0.145 MNaC1 containing 4
[1% (wt/vol) in 0.1 Mcacodylate buffer], rinsed in buffer, mM NaN3 and 0.05% (vol/vol) Triton X-100 {27). The
number of homogenization-resistant (elongated) sper-
dehydrated through graded ethanols, and embedded in a
mixture of Araldite and PolyBed-812 (hl.64 by wt). For matid nuclei in each suspension was determined bycyto-
two to four blocks per testis pole, 4 × 6-ram sections were meter counts; data were expressed on a tissue wet-weight
cut at 0.8 gm, collected on glass slides, and stained with basis. The resulting values should provide a sensitive
toluidine blue. Each slide was coded with a number that measure of the efficiency of spermatogenesis; but be-
cause the homogenization procedure is different from
precluded identification of treatment or end of the testis
that pioneered by Robb et al. {27},the absolute values are
by observers making subjective or morphometric evalua-
tions. Epididymal tissue was similarly processed. By the not comparable with those of others (2, 27).
use of light microscopy, testicular tissue was subjectively
Gene Expression
evaluated, independently, by two individuals (RPA and
DNR V) expert in spermatogenesis and pathophysiology, Molecular probes. Expression of hsp70 and hspg0 was
for normalcy of both the seminiferous epithelium and the examined. The hsp70 probe is a 1.3-kb HindIII-EcoRI
1768 MICROGRAVITY AND TESTISFUNCTION
cDNA insert from plasmid priMS213 (a gift from Dr. L. sity values forthe amount of mRNA or mRNA hybridiza-
Moran; Ref. 18). This mouse cDNA recognizes high lev- tion signal for that sample. Thus comparisons could be
els of a novel 2.7-kb transcript in spherical and elongat- made among bands within a gel or Northern blot_
ing spermatids in normal mice (34, 35) and also heat-in-
ducible 2.5- and 3.5-kb transcripts in heat-shocked L Ley_ Ca//U/trastructure
mouse celiL Tim hs/_ probe is a 1.1-kb PstI cDNAin- +iFori_h testicular pole (A br._)_._:_ blocks of tissue
sert from the _pHS801 (a gift of Drs. E,_lfz_
and I_ Weber. RM. _). This human cDNA ._,_: ++i_ examined for the five rats "m_ _ant group.
3.2-kb trsnacript 4n mouse testis, expressed
. __: _ s_dons from each block +_ motmted on form-
dantlyiii IZ_ma._matocytesin propnase oz_+ V_a_-coated, 200-mesh grids and s_ed with uranyl ace-
division of mdcsis,_12), and also recognizes a&_b+ tare and lead citrate. Grids were examined with aHitachi
transcriptprasentlniheat-shocked or non-heat-_ed HU-12A electron microscope. To minimize sampling
HeLa cells (15),_..,.,.=+_.+++-+++ +;_++.+' +.+
Leydig cells were initially sel_ at ×1,000 magni-
Ana/yt/c met,hodg:,C,omtrol testis and liver t'_were +
fieation_'atwhich cytoplasmic 0rganelles could not be
obtained from,_mal_._rodents: Sprague-Da_rats
Usually three micrographs, prepared at
(Carom Research. Industries), Swiss Weblike
X20,000 original magnification, were necessary to en-
(Carom Research Industries), and B6D2F1/J,_mice
compass each selected Leydig cell, and three Leydig cells
(Jackson Laboratory). Tissues were quickly plunged into
on each gridwere photographed. Thus 18-30 Leydig cells
liquid nitrogen and stored at -70°C. Total RNA was ex-
were examined per rat.
tracted from tissues by lithium chloride precipitation (4).
Prints were prepared (at ×2.75) so that all stereologi-
Samples were processed in sets, with each set containing
cal determinations (7, 32, 37) were made on micrographs
tissue from one rat in each of the four treatment groups
at a final magnification of ×55,000. Surface densities
from COSMOS 2044 plus samples of testis and liver from
were determined for smooth endoplasmic reticulum
a normal rat and mouse as controls. Similar amounts of
(SER), rough endoplasmic reticulum (RER), outer mito-
RNA from each sample were loaded onto denaturing gels
chondrial membrane (OMM), and inner mitochondrial
(0.8% agarose, 2.2 M formaldehyde), electrophoresed,
membrane (IMM). The 20 )<25-cm stereologic overlay
semi-dry electroblotted to Nytran membrane, and linked
was of a multipurpose test system, containing 112 short
to the membrane by exposure to ultraviolet (UV) light
test lines (each terminating as a line-end point) arrayed
for 5 min. Probes (see above) were labeled by random
as 14 rows of 8 test lines each. The overlay was placed
priming with [s2p]deoxynucleotide triphosphates (8).
over a micrograph and used to determine the surface
[_P]probes were hybridized to mRNA on the filters at
density of each organelle (32) by counting intersections
high stringency in the presence of dextran sulfate (31),
of test lines on the overlay with membranes of the organ-
washed sequentially for 20 min each in 2)<standard sa-
elles. Surface density (Sv) was calculated using the for-
line citrate (SSC) plus 1% (vol/vol) sodium dodecylsul-
mula Sv = 4(I)/(PT)(d), in which ! is the number of in-
fate (SDS) at room temperature, twice in 2)<SSC plus
tersections between test lines and organelle membrane,
0.1% SDS at 65°C, once in SSC plus 0.1% SDS at 65°C,
PT is the number of line-end points over cytoplasm of
once in 0.1)< SSC plus 0.1% SDS at 65°C, and once in
Leydig cells, and d is the length of the test line corrected
0.1× SSC at 65°C (SSC = 0.15 M NaC1 and 0.015 M
for magnification. Repeatability was established by re-
sodium citrate). The integrity and relative abundance of
analysis, several months later, of 50 micrographs each
RNA on the membranes was assessed by UV visualiza-
from the vivarium and flight groups (including both nor-
tion of ethidium bromide incorporated by rRNA and re-
real and abnormal rats). For SER, OMM, and IMM, de-
corded on Polaroid Type-55 film. Molecular weight of
viations from the first counts were 7,4, and 8%.Thus itis
each mRNA transcript was estimated using the 28- and
probable that counting error was <10% for each organ-
18-S bands of rRNA as internal size standards. The 28-S
elle.
rRNA is 4.729 kb in rats and 4.712 kb in mice, and the
18-S rRNA is 1.869 kb in both rats and mice (14, 26).
Leydig Cell Function
Hybridization was detected by exposure of aNorthern
blot to Kodak X-ray film, with an intensifying screen, at The pooled low-speed supernatants prepared from
-70°C for 1-14 d. Both autoradiograms of Northern thawed testicular tissue (see Spermatogenesis) were cen-
blots and negatives of gels were digitized (256-point gray trifuged at 20,000 g for 15 rain at 4°C. The supernatant
scale) and relative amounts of RNA estimated by video was removed and testosterone concentration determined
image analysis with a JAVA system (Jandel Scientific). by radioimmunoassay. Reagents for the liquid-phase as-
An arbitrary standard scale was used to calibrate inten- say, using _I-testosterone-tyrosine methyl ester as the
sity, with a randomly selected area of background set at 0 ligand, were from ICN (Costa Mesa, CA). All samples
and a randomly selected area of high density assigned a were quantified in a single assay and the intra-assay coef-
value of 100. Densitometry values represented the aver- ficient of variation was 11%. The assay was validated for
age intensity over the entire area of a band. To provide a rat testicular tissue by demonstrating that I) increasing
correction factor for unequal sample loading within a gel, amounts of tissue extract displaced t25I-testosterone-
the value for the most intense rRNA band in a gel was TME in a dose-dependent manner parallel to the stan-
divided by the value for the corresponding rRNA in each dard curve; 2) tissue extracts did not bind detectable
sample. This correction factor for relative amount of to- amounts of t2sI-testosterone-TME in the range of dilu-
tal RNA loaded for each sample was used to adjust inten- tions used to assay testicular testosterone; 3) testoster-
MICROGRAVITY AND TESTIS FUNCTION 177S
1.50 (Fig. 2). For analysis of the COSMOS 2044 samples, a
series of duplicate aliquots of the standard membrane
preparation representing 5-20 mg of tissue, or 20-rag ali-
_-_ 1.25 quots representing each sample, were incubated with 6.0
ng *=I-hCG plus 0 or 100 IU hCG, and specific binding
(,,.
quantified as above. All sampl_ were included in one
"1=1
I:o=3 1.00 assay, and the intra-assay "coefficient of variation
was 4.8%.
rn
O
(9
t- Statistical Analyses
_t-===_ 0.75
Data for organ weights, number of spermatogonia in
stage VI seminiferous tubules, .surface density of organ-
0.50 I I I I i J I elles in Leydig cells, concentration of testosterone in
0 1 2 3 4 5 6 blood plasma (Ref. 21; data provided by R. E. Grinde-
land), and concentrations of testosterone or LH recep-
Moss of [1251]hCG tors were analyzed using the SAS General Linear Model
I_O. 1. Mass of human chorionic gonadotropin (hCG) required to (GLM) procedure and a one-way analysis of variance.
saturate receptors forratluteinizing hormone (rLH)in20mgratt_tic- Quantitative data for characteristics of seminiferous tu-
ulartissueinatotavlolumeof350_I;incubated18h at25°C.Data bules and numbers or ratios of germ cells were analyzed
were corrected fornonspecific binding. See text fordetails.
using a two-way analysis of variance considering pole of
the testis (2, random) and treatment group (3 or 4, fixed).
one added toextractswas recoveredinan amount essen-
All four treatment groups were included in analyses of
tiallyequaltothatadded; and 4)valuesobtained forali-
diameters of seminiferous tubules and their lumina, but
quotsofsamples subjectedtochromatography on Sepha-
for other attributes the caudal-elevation rats were ex-
dex LH-20, to isolatetestosterone from potentially
cluded because their seminiferous tubules contained vir-
cross-reactingcompounds, were similarto values for
tually no germ cells. There was no significant difference
nonchromatographic aliquots(r= 0.95;n = 10;slope=
between the A and B poles for any characteristic, except
1.1,and y-intercept= 0.05).
for corrected numbers of preleptotene spermatocytes
The capacityof membrane fragments inthe 20,000g
(considered to be a type I error). If the Fvalue associated
pellettobind human chorionicgonadotropin (hCG) was
with treatment was significant (P < 0.05), Duncan's mul-
used toestimatethe number ofunoccupied receptorsfor
tiple-range test was used to establish which groups dif-
rLH per gram of testiculartissue;presumably rLH re-
fered (P < 0.05).
ceptorsallwere on Leydigcells.A "standard curve" tech-
nique was used,because therewas insufficienttissuefor
RESULTS AND DISCUSSION
a Scatchard analysisofeach sample.A standard testicu-
larmembrane preparationwas prepared,asabove,using General testicular function. Mean testicular weight for
testesfrom 10 Sprague-Dawley rats90-120 days old;it rats in the vivarium group did not differ from those for
must beassumed thataffinityoftherLH receptorissimi- either the flight or simulated-launch groups, but testis
larinAmerican and Czechoslovakian rats.Indeveloping weight was greater (P < 0.05) for rats in the simulated-
the assay,theamount ofhCG necessarytosaturaterLH launch group than those in the flight or caudal-elevation
receptorsinan aliquotofstandardpreparation oftesticu- groups (Table 1). However, some rats in the flight and
larmembranes was determined.Duplicatealiquotsrepre- vivarium groups had small testes, and all rats in the cau-
senting 20 mg testiculartissuewere incubated in 12 ×
75-mm tubes with 1.2,2.4,3.6,4.8,or 6.0ng I_I-hCG, 2o
prepared from NIDDK hCG CR-127 by thechloramine-
T procedure, ina totalof350 #Ibufferfor18 h at25°C.
:E
Tubes were centrifuged30 min at9,000g and the super- n 15
natant decanted; radioactivityof the pelletwas deter-
0
mined. Itwas found (Fig.1)that 4.8ng I_I-hCG was
sufficienttosaturaterLH receptorsinan extractrepre- _ 10
senting 20 mg testiculartissue.To furthervalidatethe D
O
assay,aliquotsof0,5,10,15,or20 mg ofstandard mem- m
brane preparation (in100 #Ibuffer),100 #I of12_I-hCG O
= 5
solutioncontaining6.0ng hCG, and 150/zlofbuffercon-
taining0 or 100 IU (about a 1,000-foldexcess of hCG;
Sigma CG-10) were mixed and incubated for 18 h at
25°C.Tubes were centrifugedand radioactivitywas mea- 01 I I I I
0 5 I0 15 20 25
sured as above; the amount of *_I-hCG specifically
bound was calculatedas binding without excess ligand Moss of Tissue (rag)
minus binding with excess ligand.Specificbinding was FIG.2. Binding of '_I-hCG to receptors for rLHas a function of
directlyproportionaltothemass oftesticulartissuepres- mass oftesticular tissue (0-20 mg/tube) incubated in350_!for 18hat
ent and increased ina linearmanner from 0 to 20 mg 25°C. Data were corrected for nonspecific binding. Seetext fordetails.
178S MICROGRAvrrY AND TESTIS_ON
TASLZ i.Characteristiocfstherighttestis/o_r ratsofCOSMOS 2O44study
8imukted Caudal
Flight Launch Vivamtm _tion
1,416_ 1,048"b 492"
mghtq_mlym_*_, mg SXT* 401' 314" 242"
Seminib_mtulLub_meter_,m 222" 26_ 247e ISS"
Tubulelumendiameter_,m 90* 109' I01_" : 8b_
Normal%, 49* 8S_ 65-"
Earlydel_aeratlon,% 7" 5* 5"
Moderatedegeneration,% 5" 5" 5"
Severedegeneration,% 39. 3b 21" _ nd
StageVIIseminlferoustulmle8 • _ :
Nuclei/tubulecrosssection ..... :_!
_._U 1.41" nd
0.88* 0.84" nd
Pm_ _ 4.40. 6.65b 5.25_b nd
Pschytene spermatocytN 4.24" 7.15" 6.00" nd
Spe_ 18.67" 21.09_ 16.64"_ nd
Ratio (no.ISertoli cell)
Type A _rmatogonia 1.08° 0.42" 0.63" nd
Preleptotene spermatocytes 3.46" 3.14" 3.65" nd
Pschytene spermatocytes 3.40" 3.38" 4.23" nd
Spermatids 10.82" 9.96* 12.14" nd
Ratio
Preleptotene-to-A spermatogenic 9.13" 7.71" 6.44" nd
Pachytene-to-preleptotene 0.98" 1.08" I.I0" nd
Spermatid-to-pachytene 3.2T 2.95" 3.08" nd
Spermatide/g testis, ×106 28.9. 37.8" 33.2" <0.5 b
LH receptors, ng hCG bound/g testis 65.8" 56.8" 63.1" 63.7"
Testosterone, ng/g testis 7.0" 15.1" 64.5b 23.4"
Testosterone,T ng/ml blood plasma 0.27" 1.91b 3.50c 0.51"
Leydig cells, #mZl#m s
Smooth endoplasmic reticulum 7.75* 7.23* 6.07b 7.65b
O.66* 1.06b 0.87_
Rough endoplasmic reticulum 0.59_
Outer mitochondrial membrane 1.43" 1.47" 1.71b 1.48u
Inner mitochondrial membrane 3.85" 3.76" 4.26" 3.94b
Means (n = 5) with the same superscript letter do not differ (P > 0.05). nd, not determined, t Data for testosterone in blood plasma are from
Merrill et al. (21).
dal-elevation group had very small testes. The probable blythecaudal-elevatiognroupshadbeenabnormal when
causes of these differences were revealed by subjective they were assignedto treatmentgroups.By random
histological evaluations of each testis and attached epi- chance,allratsinthesimulated-launcghrouphad nor-
didymis (Fig. 3), and confirmed by quantitative data. realtestes.
One must assume that vivarium rats had not been ex- Lesionsintestesofratswithapreexistinagbnormality
were differenf_rom thoseincaudal-elevatiornats.On
posed to stress factors associated with launch or flight
and were the best representation of the population of thebasisoftesticulawreight,thelesionwasunilaterailn
rats shortly before launch. However, in three vivarium some ratswithpreexistinagbnormalitiesb;utforcaudal-
rats, one or both testes weighed <900 rag, and histologi- elevationratsthelesioninevitablwyas unilateralF.or
cal examinations of three of these four testes (only right caudal-elevatiornats,theentiretestiwsasuniformlyaf-
testes were available) revealed that they were abnormal fected(basedon histologiceaxlaminationsofbothpoles
(see below). Similarly, right testes from two flight rats anddataforefficiencoyfspermproductioninthecentral
weighed <650 mg and had histological abnormalities sim- portionofthetestis)S.eminiferoustubulescontained
ilar to those in abnormal vivarium rats but different from onlySertolciellsplusacomplement ofspermatogonia,a
those characteristic of caudal-elevation rats. Thus the fewprimaryspermatocytes,and occasionalspermatids
cause of lesions in the testes of these two abnormal flight (Fig.3,A andB).Incontrasti,nanabnormaltestisfrom
rats probably was the same as that inducing lesions a flightorvivariumrat,therewas a mixtureofnormal
found in three vivarium rats and was not flight. By this and abnormal seminiferoustubules(Fig.3,D and E);
reasoning, three flight rats (7, 9, and I0), five simulated- andtheextentofinvolvementwasnotuniformthrough-
launch rats (6-I0), and two vivarium rats (7 and 10) did outtheentiretestisO.n thebasisofhistologiceavlalua-
not have marked preexisting abnormalities; weights of tions,testesfrom ratswitha preexistingabnormality
their right testes ranged from 1,200 to 1,500 rag, whereas shouldhave been producingspermatozoa,althoughin
the right testes for the other five rats in these three reducednumbers.These subjectivoebservationswere
groups weighed 450-900 rag. Thus we concluded that confirmedby dataforefficiencoyfspermproductionin
testes of some rats in the flight and vivarium and possi- thecentralportionofeachtesti(s<0.5vs.22× 10esper-
MICROGRAVITY AND TESTIS FUNCTION 179S
H
FIG. 3. Micrographs of testicular (A,B, D, E, G,and H) and epididymal (C, F, andI) tissue. Testicular tissue from a
caudal-elevation rat (A and B) contained only abnormal seminiferous tubules, but a few germ cells were present in
some tubules. These lesions were induced by treatment. Testicular tissue from a vivarium rat with a preexisting
abnormality (Dand E) contained both normal and abnormal seminiferous tubules; abnormal ones contained few germ
cells and Sertoli cells predominated. Distinction of preexisting testicular abnormalities from those induced by caudal
elevation also was evident from content ofcauda epididymidis; in caudal-elevation rats sloughed germ cells rather than
spermatozoa predominated (C), whereas in rats with preexisting lesions there were few sloughed germ cells among
spermatozoa in cauda epididymidis (F). Testicular tissue from flight rats without apreexisting testicular abnormality
was normal in appearance (G) and stage VII seminiferous tubules (H) had normal ratios of germ cells to Sertoli cells
(see text and Table 2); luminal content ofcauda epididymidis also was normal (I). A, D, and Gare at )<62and B, C,E,F.
H, and I are at )<125 magnification; toluidine blue staining.
matids/g testis for caudal-elevation and preexisting ab- and plasma concentrations of testosterone (492 vs. 690
normality rats). Other quantitative differences (P < rag; 185 vs. 228 um diam; 23.4 vs. 60.9 ng/g; and 0.51 vs.
0.05) were evident. Caudal-elevation rats had a smaller 2.85 ng/ml) but a similar concentration of receptors for
right testis than rats with preexisting abnormalities, with rLH in testicular tissue {63.7 vs. 78.8 ng/g). Morphomet-
smaller seminiferous tubules, and lower intratesticular tic characteristics of Leydig cells were similar for both
180,8 MICROGRAVITY AND TESTIS FUNCTION
groups (data not presented). Incaudal-elevation rats, the T_ 2. C_zm_ter/st/cs of the r/g/sttest/s/n rats
epididymis contained numerous germ cells sloughed w/thoutpreez/st/ngabnorma//t/es/n the_,
from the seminiferous epithelium (Fig. 3C), which were dmu/ated-/aundh_ and v/vat/urn
indicative of recent dostruction of the seminfferous epi- O_O,_MOS 2044 _ _ :_;
thelium. Undoubt_. the _ ofcaudal_levation - rats_:
fell into the abdol_ul._ (6,_)L_CfiiW __ Vivarium
guinal canal which _ten_l_rat8 (in__ _....
humans), and were __ e_ to __" ___ :[ 2
1
ature. Epididymidas _of__]¢ _ 1_
ties had few immature germ ceils among the___ _mstor, _m _-_.
(Fig. 3F), which precluded extensive destruction of the _ _ __:_: r,_m w,. _" 106"
seminiferous epithelium within 2 wk before tissue fixa- Sta_ gl se_rous tubule,
5.r 5.57."
tion. In flight rats without preexisting abnorm_ i_ _nia 5.79b
spermatogenesis was normal (]Fig.3, G and H), and the _ _ ,s_ tubules
cauda epididymidis was filled with spermatozoa (Fig. 31). Nucl_/tulmckrosssection
Sertollcells 1.04" 2.15b 1.70"b
The high values for intratesticular concenWations of
Type Aepermatogonia 0.38* 0.88b 0.89_
testosterone and receptors for rLH in _audal-elevation
|'releptotene spermatocytes 3.65" 6.65b 8.10_
rats and rats with preexisting testicular abnormalities Pachytene spermatocytes 3J}2" 7.15b 7.20b
probably reflect atrophy of the seminiferous tubules so Spermati& IL93" 21.09b 20.57 b
that Leydig cells constitute a larger proportion of the Ratlo (no./Sertoli cell)
testis. However, it is possible that these rats had more Type A spermatogonia 0.35" 0.42" 0.51"
Pndeptotene spermatocytes 3.44" 3.14" 3.69"
Leydig cells or Leydig cells that secreted more testos- Pachytene spermatocytes 3.67" 3.38" 4.24"
terone. Spermatids 11.18" 9.96" 12.42"
Only tissues from rats 6-10 in each of the four treat- Ratio
ment groups were available to us for detailed examina- Preleptotene-to-A spermatogenic 10.98" 7.71" 7.43"
Pachytene-to-pre leptotene 1.06" 1.07" 1.17"
tion. However, it is likely that at least two of the other
Spermatid-to-pachytene 3.08_ 2.96" 2.92"
rats (F-3 and V-I) had abnormal testes before assign-
ment to the project, based on testicular weight (<850 vs. Spermatids/g testis, 10s 42.3" 37.8" 38.2"
1,200-1,650 rag). Thus at least 7 of 30 rats (23%) in the LH receptors, ng hCG bound/g testis 55.8" 56.8" 44.0"
flight, simulated-launch, and vivarium groups probably Testosterone, ng/testis 3.81" 15.06_ 20.82 b
were abnormal before imposition of treatments involved Testosterone,_ ng/ml blood plasma 0.23" 1.91_ 1.96 b
in the COSMOS 2044 study. Leydig cells, #m2/#m s
Spermatogenesis. Considering the facts that testes of Smooth endoplasmic reticulum 7.93" 7.23" 5.51 b
several rats probably were abnormal before assignment Rough endoplasmic reticulum 0.53" 0.66" 1.05"
Outermitochondrialmembrane 1.46" 1.47" 1.64"
to treatment, and that groups of five rats/group are of
Innermitochondrialmembrane 3.98" 3.76" 4.06"
marginal size to detect subtle treatment effects (3), dog-
matic conclusions concerning effects on spermatogenesis Rats with a testis <0.90 g and abnormal histological appearance
were considered to have a preexisting abnormality. Data for corrected
of exposure to the stress factors of simulated or actual
number of spermatogonia in 2.0-_m sections for stage VI were divided
launch and reentry, or exposure to microgravity per se, by 2.5 to make them comparable to those for 0.8-#m sections for stage
are impossible. Table 2 summarizes the data after exclu- VII. Means with the same superscript letter do not differ (P > 0.05).
sion of rats presumed to have preexisting testicular ab- ?Data for testosterone in blood plasma are from Merrill et aL(21).
normalities and is considered the best basis for conclu-
sions concerningeffectsofflightH.owever,datafolall chytene spermatocytes to preleptotene spermatocytes
ratsarepresentedinTableIbecauseminorsequelaeon and step-7 spermatids to pachytene spermatocytes were
spermatogenesisoftheunknown agentorgeneticfactor similar, so primary spermatocytes were completing
causingpreexistinagbnormalitieisnsome ratsmay have meiosis. Thus the differences detected may not reflect a
beenundetectedinothers.Consideringallrats,basedon treatment effect.
subjectiveevaluations(Fig.3)andquantitativdeata(Ta- At the penalty of reducing the number of available ani-
ble1),spermatogenesisinflightratsgenerallwyas simi- mals, data also were summarized after excluding the two
lartothatinvivariumrats,butinferiotrothatinsimu- flight and three vivarium rats with obvious preexisting
lated-launcrhats.Thisprobablyreflecttshevariableim- abnormalities in the right testis (Table 2). The reduc-
pactofpreexistinagbnormalitiesincertainrats.Ifthere tions (P < 0.05) inweight ofthe right testis and seminifer-
hadbeenaprofoundtreatmenteffectt,hecorrectednum- ous tubule diameter persisted in flight rats, as did differ-
bersofgermcellsperstageVIIseminiferoutsubulecross ences in numbers of Sertoli cells orgerm cells per seminif-
sectiono,rone ormore oftheratiosbetween a specificerous tubule cross section. However, as for the complete
typeofgerm celland Sertolicellsw,ouldhave been re- data set, the ratios of germ cells to Sertoli cells or among
duced.Althoughsuchdifferencewseredetectedbetween types of germ cells were not affected by treatment. The
theflightand simulated-launchgroupsforpreleptotene number of homogenization-resistant spermatids per
spermatocytesand step-7spermatids,therewas no dif- gram of testis was not affected by treatment and was
ferencefortheratiosB.ecausetheratioofpreleptotene independent of testicular weight (r = -0.34; P > 0.05).
spermatocytestotypeA spermatogoniawasnotreduced, Counts of all spermatogonia in cross sections of stage
spermatogoniawere completingmitosis.Ratiosofpa- VI tubules (by CSW) were consistent with those for type
MICROGRAVITY AND TESTIS FUNCTION 181S
A spermatogonia in stageVII (by DNRV). In stageVI one or a combination of other factorsassociatedwith
crosssections,there were fewer (P < 0.05)spermato- flight.Availabilityoftesticularsamplufrom caudal-ele-
gonia intestesfrom normal flightratsthan intestesof vation rats,which became cryptozchid,enabled distinc-
simulated-launch ornormal vivarium rats(Table 2),al- tionofheateffectsfrom possibleeffot_ iliducedby other
though the number of Sertoli cell nuclei was not signili- aspects of spaceflight. ,_ _:_
cantly different (data not shown). The 5.3% reduction in From the data depicted (Figa 4 ._6_and similar
number of total spermatogonia per cross section of stage data for the other four sets of _'tw6 co.ndtmions
VI seminiferous tubule for flight rats, relative to aimu- _wem reached. _ effects on l_._tis-speclfic
lated-launch rats, was similar to results of Sapp _; ol. gene products hmpT0 amd hsp00 __pmexisting
(28) for rats flown on Spacelab 3 or COSMOS 1887. abnormalities present in some rate--was no ef-
Diameter of seminiferous tubules in normal flight rats fect of spaceflight on normal expr__0_tl-_pecific
was less than in normal simulated-launch or vivarium hsp gene products. Flight, vivaria] simulated-
animals (Table 2), despite apparent normalcy of sper- launch rats without preexisting _normalities
matogenssis. This might reflect adistortion of the semi- (Fig. 4, lanes 1-3) had more of l;_h_r_:-cell-_pecific 2.7-
niferous tubules induced by microgravity, which per- kb hsp70 and 3.2-kb hsp90 transcripts (mRNA) than
sistedthrough the 11h between landing and fxation of flight or vivarium rats with preexisting abnormalities
tissue.Although thereisno known precedent,itispossi- (Fig. 5, ia_les i and 2) or caudal-eleva_on rats (Fig. 4,
blethattherewas a decreaseindiameter and an offset- lane 4, and Fig. 5,lane 3). Reductions of hsp70 and hsp90
tingincreaseinlengthofseminiferoustubules(qualityof transcripts in certain flight and vivarium rats reflected
thematerialprecludeddetermination ofpercentagesemi- corresponding preexisting abnormalities in the seminifer-
niferoustubules,and hence tubulelength),sothatinter- ous epithelium (Fig. 3), including a reduction in the pro-
nal surface area ofthe basement membrane and space portion of germ cells to somatic cells within the testis.
availableforeach Sertolicellwere essentiallyunaltered. The pronounced reduction in amount of germ-cell-speci-
Such changes would explainthedecreasesinnumbers of fic mRNA in caudal-elevation rats, in comparison to rats
Sertolicellsand germ cellsperseminiferoustubulecross with preexisting lesions, supports the morphological dis-
section,whereas theratiosofgerm cellswere normal and tinction (see above) between preexisting abnormalities
number of homogenization-resistant spermatids per and those induced by caudal elevation. Second, we ob-
gram oftestiswas unaffected. served no induction of the heat- or stress-inducible hsp70
Given the totalityofthe qualitativeand quantitative transcripts of 2.5 and 3.5 kb or hsp90 transcript of 2.95 kb
findings(Tables 1 and 2),itisunlikelythatexposure to in any sample. Thus neither spaceflight nor caudal ele-
microgravity for _<2wk had an immediate deleterious vation for <2 wk induced detectable transcription of the
effecton mitosisofspermatogonia,survivalofspermato- heat-inducible hsp70 or hsp90 members of the cellular
cytesthrough meiosis,or maturation of spermatids. If stress protein gene family. It may be productive to exam-
the Sertolicellsor germ cellswere affecteddirectlyby ine patterns of cellular stress-protein gene expression
microgravity,or indirectlyby reduced secretionoftes- after several days of caudal-elevation, while the heat-in-
tosterone,otherhormones, orregulatoryfactors,the le- duced lesion is forming but before substantial sloughing
sion(s)must be eithersubtleor not discernibleafter14 of the seminiferous epithelium. Furthermore, the possi-
ble induction of acellular stress in the testes response to
days offlight.
Gene expression. Yield of total RNA ranged from 39 to long-duration spaceflight remains to be assessed.
600 _g and probably was afunction of the mass of testicu- Leydig cell ultra.structure. In terms of surface density,
lar tissue available. The weight of each sample of testicu- Leydig cells in vivarium rats contained less (P < 0.05)
!at tissue was not determined, because thawing the tissue SER and more (P < 0.05) RER and OMM than Leydig
before extraction might have resulted in degradation of cells in rats from the other three groups (Table 1). Sur-
the RNA. Efforts were made to load similar amounts of face density of IMM did not differ among treatment
RNA (20 _g) on each lane of the gels used for Northern groups.
When data for the caudal-elevation and rats with
analyses, and data later were normalized within each gel
to minimize the effect of differences in the amount of preexistingabnormalitieswere deleted,the only signifi-
RNA loaded. Figures 4 and 5show, respectively, samples cant differenceremaining among treatment groups was
thatforSER (Table 2).Surface density forthiscompo-
of RNA from rats without (Fig. 4) or with (Fig. 5) preex-
isting abnormalities of the right testis. nent was higherinflightorsimulated-launch ratsthan in
vivarium rats.Surface densityofSER averaged 5.5_m2/
The expression of the hsp genes was given high priority
in this study because several members of the hsp70 and #m 3inLeydig cellsofvivarium rats,which isconsistent
hsp90 gene families are expressed in specific germ cells with published data (37)fornormal rats.
(see METHODS). Thus they can serve as molecular Leydig cell function. The number of unoccupied recep-
markers of the presence of these germ cells. Moreover, tors for rLH in the testis, as estimated by binding of hCG,
certain members of the hsp70 and hsp90 gene families was similar for all four treatment groups (Table 1). Nev-
respond to cellular stress and are affected by altered tes- ertheless, it was evident that Leydig cells in flight and
ticular environment, including elevated temperature. A simulated-launch rats produced much less testosterone
stress response in testicular cells could be elicited by a than those in vivarium rats. This was reflected in both
rise in temperature induced by movement of the testes concentration of testosterone in whole testicular tissue
from the scrotum into the body cavity during exposure to and in plasma from trunk blood (Table 1). However,
microgravity or caudal elevation or could be elicited by these conclusions probably were influenced by inclusion
182S MICROGRAVITY AND TESTIS FUNCTION
1 2 3 4 1234 1234
18S >
tm. 4.Expmon of_ stress-protegiennesin
:_tm4Uofrats. TotalRNAwasbolated fromtestes ofro_ 7,
• .o)_,lllmulatedlaunch(/one3),orcaudalelevation(lone#).
forcaudsl-elevatiun rat, in which induced l_ion
A precluded discernmeat of any prasxioting abnormality,
rats were normal be@oretreatment. Total RNA was size
IOO fraetionated bygel electrophoreeis, stained withethidium
bromide,andphotographeudnderultraviol(eUtV)light
{,4)T.he28-and18-SrRNAa aredesi_atedH.ybridiza-
tionoftheriP-labeled/_o(7B0)andhs/_0(C)probesto
so
mRNA after transferto membranes wasdetected byauto-
% llm_t_
radiography. Relativeamount ofhybridization (D)wasde-
6o termined bydensitometric analysis.
40
2o
o
1 2 3 4
LaneNumbers
D
of ratswith preexistingabnormal testes.Considering properly fixed, but rejection of some cells might have
only the normal rats,mean concentrationsoftestoster- resulted in selection of an atypical sample. However, fail-
one intesticulartissueorblood plasma both were lower ure to find a significant, positive correlation between
inflightratsthan ineithersimulated-launch orvivarium surface density of SER and intratesticular testosterone
rats(Table 2).Thus itappears thattestosteroneproduc- concentration probably is not a consequence of having
tionwas reduced inflightratsthathad been normal be- measured surface densities rather than surface area.
foretheexperiment but was similarinsimulated-launch Surface density is highly correlated with the capacity of
and vivarium ratswithout preexistingtesticularabnor- the testis to produce testosterone under conditions of
malities.Values forconcentration ofreceptorsforrLH maximal LH stimulation (36, 37). More plausible expla-
were similarforallthreeofthesegroups (Table 2). nations for our observations are that I) intratesticular
On the basisof data for10 normal ratsinthe flight, testosterone at any instant need not reflect the capacity
simulated-launch, and vivarium groups,thecorrelation of atestis to secrete testosterone, 2) testosterone produc-
between intratesticularconcentrations of testosterone tion in flight rats was considerably less than would be
and receptorsforrLH (r= 0.19)was nonsignificant,al- expected given the concentration of rLH receptors in the
though that between concentrations oftestosteronein tissue and the surface density of SER, 3) Leydig cells
testiculartissueand inblood plasma was significant(r= might have undergone death and degeneration, or 4) we
0.66).Surprisingly,thecorrelationbetween intratesticu- obtained a spurious result associated with the interval
farconcentrationoftestosteroneand surfacedensityof between landing and tissue sampling. No report of in-
smooth endoplasmic reticulumwas negative(r= -0.62; stant-to-instant changes in intratesticular concentra-
P = 0.055),althoughtypicallythesurfaceareaofSER or tion, in relation to surface density of SER, has been pub-
peroxisomes ishighlyand positivelycorrelatedwith tes- lished. Although data for the concentration of LH in
tosteronesecretingcapacity(7,20,36,37). blood plasma from these rats are not available, we specu-
Surface density of organelles(surfacearea per unit late that the problem probably is one of intratesticular
volume of Leydig cell)was determined in thisstudy, fluid dynamics or changes in availability of oxygen and
ratherthan surfaceareaper Leydig cellorpertestis(7). substrates to Leydig cells.
This was done because differentialshrinkage of tissue
GENERAL DISCUSSION
and artifactsassociatedwith imperfectly fixedtissue
precluded enumeration ofLeydig cellspertestis.Leydig Utilityo[thecaudal-elevationmodel.The caudal-eleva-
cellsused todetermine surfacedensityoforganelleswere tion(tailsuspension or hind limbs unweighted) model