permeation of human ovarian tissue with cryoprotective agents in
TRANSCRIPT
Human Reproduction vol.l3 no.2 pp.376-380, 1998
Permeation of human ovarian tissue with cryoprotective agents in preparation for cryopreservation
H.Newton1•6, J.Fisher, J.R.P.Arnold3, D.E.Pegg4, M.J.Faddy5 and R.G.Gosden1
1Centre for Reproduction, Growth and Development, School of Clinical Medicine, University of Leeds, Clarendon Wing, Leeds General Infirmary, 2School of Chemistry, 3Department of Biology, University of Leeds, Leeds LS2 9NS, 4Medical Cryobiology Unit, Biology Department, The University of York, York, UK and 5Department of Mathematics, The University of Queensland, Brisbane, Queensland, Australia
6To whom correspondence should be addressed
The recent improvements in the treatment of cancer by chemo- and radiotherapy have led to a significant increase in the survival rates of patients with malignant disease, but at the expense of distressing side effects. One major problem, especially for younger patients, is that aggressive therapy destroys a significant proportion of the follicular population, which can result in either temporary or permanent infertility. Freeze-banking pieces of ovarian cortex prior to treatment is one strategy for preserving fecundity. When the patient is in remission, fertility could, theoretically, be restored by autografting the thawed tissue at the orthotopic site or by growing isolated follicles to maturity in vitro. Recent studies have found good follicular survival in frozen-thawed human ovarian tissue but to optimize the process an effective cryopreservation method needs to be developed. An essential part of such a technique is to permeate the tissue with a cryoprotectant to minimize ice formation and the extent of this equilibration is an important determinant of post-thaw cellular survival. In the current study, we have investigated the diffusion of four cryoprotective agents into human tissue at both 4°C and 37°C. We have also studied the effect of adding different concentrations of the non-penetrating cryoprotective agent, sucrose, to the freezing media using the release of lactate dehydrogenase as a measure of its protective effect. At 4°C propylene glycol and glycerol penetrated the tissue significantly slower than either ethylene glycol or dimethyl sulphoxide. At the higher temperature of 37°C all four cryoprotectants penetrated at a faster rate, however concern about enhanced toxicity prevents the use of these conditions in practice. Thus, the results suggest that the best method of preparing tissue for freezing is exposure for 30 min to 1.5 M solutions of ethylene glycol or dimethyl sulphoxide at 4°C; this achieved a mean tissue concentration that was almost 80% that of the bathing solution. We also report that the addition of low concentrations of sucrose to the freezing medium does not have a significant protective effect against freezing injury.
376
Key words: cryopreservation/follicle/oocyte/ovary
Introduction Human females are born with a finite store of germ cells represented by primordial follicles in the cortex of the ovary. The number of germ cells reaches a peak of approximately 7 million at mid-gestation and thereafter declines throughout life by the natural processes of ovulation and atresia until the follicular stock falls below a threshold of approximately 1000 and fecundity is lost (Faddy and Gosden, 1995). The treatment of cancer by chemo- and radiotherapy has a highly deleterious effect on the ovary causing a severe depletion of the follicular store. If the treatment is particularly aggressive, or the patient is older than 30 years of age, there is a high possibility of permanent sterility (Apperley and Reddy, 1995; Sanders et al., 1996). The improvements in the prognosis of many young cancer patients, in particular paediatrics, have led to an increased impetus to find an efficient method for the prevention of iatrogenic sterility. One strategy is to conserve the follicular population by cryopreserving strips of ovarian cortical tissue until the patient is in full remission. Fertility can then be restored by returning the tissue to the body either as an orthotopic autograft or, more speculatively, by isolating follicles from the thawed tissue for growth in culture followed by in-vitro fertilization (IVF) (Oktay et al., 1997). Recent studies have shown that healthy follicles are present in human ovarian tissue after freezing and thawing (Hovatta et al., 1996) and that 3 weeks after grafting tissue into immunodeficient mice 44-84% of the follicular population survive, depending on the cryoprotectant used (Newton et al., 1996). These encouraging results were not, however, based on an optimized technique and further work is required if this procedure is to become a clinical reality.
The first requirement for cryopreservation is to permeate the tissue with a compound that moderates the normally lethal effect of ice formation and the build-up of extracellular salts. Many such cryoprotectives (CPA) are known and their mode of action is quite well understood (Nash, 1962; Pegg and Diaper, 1988). The permeation of isolated cells with such CPAs is relatively fast, but diffusion in multicellular tissue systems is much slower, which means the exposure may have to be prolonged to reach adequate concentrations in the centre of the tissue thereby exposing cells near the surface to the risk of excessive toxicity. Consequently it is important to balance equilibration time against the harmful effects of overexposure to CPA prior to cryopreservation and during its removal. The penetration of CPAs has been studied using radioactive tracers
© European Society for Human Reproduction and Embryology Downloaded from https://academic.oup.com/humrep/article-abstract/13/2/376/876041by gueston 09 April 2018
(Karow et al., Clark et al., but more recently
proton nuclear magnetic resonance (NMR) analysis (Abraham
et al., 1988) has been used to measure dimethyl sulphoxide
(DMSO) and (EO) in corneal tissue (Taylor
and Wusteman, communication) and
et al., 1994; Wusteman et ccc,,uuuuc presents a proton spectrum for a given
and since the method is with water and
CPA giving distinct spectra, the concentration of CPA within
the tissue over time can be calculated directly, which makes
it to estimate the necessary equilibration time. In the
we have the rate of penetration of
four commonly used CPAs through human ovarian tissue at
both 4oC and 37oC. A second class of CPA does not permeate cells, and
it includes sugars such as sucrose and polymers such as
hydroxyethyl starch. Their mode of action is less clear than
that of penetrating CPAs and their relative effectiveness is generally lower, but many have indicated a beneficial
effect of supplementing the freezing medium with low concen-
trations of these solutes 197 In the present
the release of the intracellular enzyme lactate dehydrogenase
(LDH) following freezing was used to assay cell damage when
sucrose was included in the preservation medium: this approach
was previously used to optimize a method for the cryopreserv
ation of rabbit renal tissue (Jacobsen et al., 1984).
Materials and methods
Comparison of the rates of penetration o.f four cryoprotective agents NMR spectroscopy
Ovarian tissue was collected from five premenopausal patients aged 32-49 years undergoing total abdominal hysterectomy-oophorectomy or routine gynaecological surgery. The patients gave informed consent to the use of their tissue for research which was approved by the local ethics cmm11ittee of the United Leeds Teaching Hospitals Trust. Slices of ovarian cortex were cut at a thickness of 2 1mn into uniform disks of 4 mm diameter with a biopsy punch (Stiefel, Bucks, UK). The tissue was on ice in Leibovitz L-15 medium and used within 2 h.
EG, DMSO, propylene glycol (PROH) and glycerol (GLY) were used at a concentration of 1.5 M in Leibovitz-Ll5 medium supplemented with 10% fetal calf semm. 1.5 ml of cryoprotective media were pipetted into 1.8 ml cryovials (Nunc, Kamstn1p, Denmark) for equilibration at either 37°C or 4 oc.
Ovarian biopsies were transfened to individual vials and gently rolled ( l Hz) at either 4 oc or 3TC for 5, 10, 20, 40 or 90 min. Following incubation, the tissue was retrieved and surface moisture was removed with filter paper. The tissue samples were then transferred to 0.5 ml Eppendorf tubes and immersed in 400 ~ll of deuterium oxide (D20, 99.9%; Sigma, London, UK). The tubes were sealed and stored for up to 5 months at -20°C The experiment was repeated four times at each time with single pieces of tissue from different patients.
Samples were thawed at room temperature, and left overnight on a 'Spiramix' to ensure equilibration of water and cryoprotectant between the tissue and the D20. Samples of 350 !Jl of the solution were transferred to 5 nm1 glass NMR tubes (Wilmad Glass Co., Buena, NJ, USA),
All spectra were recorded on a GE Omega 500 spectrometer ( 1 H
Cryopn;servation of ovariaill tissmc
frequency 500 MHz). Prior to the time course
inversion recovery solvents at concentrations those in
Spectra were then acquired over 64 transients with 2500Hz in 16 384 data s. The relaxation between transients was 28 s and a 30 pulse width (3 flS duration) was transient was approximately three times the The data were recorded at 25°C, and were the NMR package FELIX® (version L3, Diego, CA, USA). Prior to Fourier transformation the free induction decays (FIDs) were an line factor of 1 Hz. The area under each standard procedures within
Calibration data for each of the agents was obtained by recording the spectra of L5 l\!1 standard solutions to a cryoprotectant to water ratio. These values were measured as DMSO 0.0235; EG 0.0248; PROH 0.0320 and GLY 0.0300. A
deuterium oxide was also to obtain a value for the background water signal for subtraction from all water The
of cryoprotectant as a fraction of was calculated all the test values for each cryoprotective agent.
Assessment (LDH) assay
Ovarian cortical from a further five 25-50 years old were used in the Tissue was cut into 2 nm13 blocks under a dissecting and randomly Five of the groups were transfened into l ml either (i) Leibovitz-LI5 medium, (ii) 1.5 M DMSO, (iii) !.5 lVI DMS0/0.05 M sucrose, (iv) 1.5 M DMS0/0.1 M :;ucrose or (v)
1.5 M DMS0/0.25 M sucrose, ail of which were same medium. Tbe final group was incubated for medium to act as a fresh control.
The vials were rolled (1 Hz) for 30 min penetration of the agent before programmable freezer 10, Planar, Middx, UK). cooled at 2°C/min to -9°C when were seeded u~cm~cm.1 .
continued at 0.3°C/min to -40oc and then at the min to -l40°C Vials were stored in until
Samples were thawed in a water bath at 20°C
for 1-2 min. Tissue was retrieved, washed in medium and placed in individual microtitre wells 150 ~Ll of sterile PBS for overnight incubation at JrC. The tissue was recovered and the wet weight recorded. The LDH levels in the PBS were measured with a UV assay on a BM/Hitachi 717 apparatus at 3~/oC
(Boehringer Mannheim, East Sussex, UK).
Statistical """'"'"'" Regression lines were determined for the NMR
linear least squares method. The LDH data were of variance and Student's t-tesL
Results
time a mean and sn
Downloaded from https://academic.oup.com/humrep/article-abstract/13/2/376/876041by gueston 09 April 2018
!lLNewhm et
10 20 30 40 50
T~mo {m!nu!es)
DMSO EG PROH GLY
· ·DMSO and EG ~ PROH an<.1 GL Y
70 90
The rate of pcl.JcctauLvu
(EG), (PROH) and ovarian tissue at 4 oc. Plotted lines represent the Lhe form = l~a.exp (~bt) ~ (l~a) exp (~ct)
is time, b and c are the peJrle"lratto over time, and ( 1 ~a) are the vrcmo·rtH)f! process over time. Each represents the mean and SD from four c)<AllHJ.l'.Co.
b and c an: are the "''''''~''''''~" of the process At both 4oc and 37oC the data showed variation between nms with some solutes more variation than others: PROH showed the greatest vmiation and EG the and these differences were
< 0.01 at 4oc and P < 0.01 at The rate of of each of the four ,,.,.,,".'''t"'''h
agents at both 4 ac and 37oC was characterized of diffusion followed a second slower
to the rate b> >c in the above The 70% of
for PROH at this
and after min At the same temper-
ofDMSO was not different 1, broken but this
than that of the other After 30 min DMSO and EG had reached a
79%. At 37oC PROH <
2, broken line). After 30 min EG and GLY had reached 81% whilst PROH had reached almost total The parametGr estimates at each Table l
geuase Our results shmv that the smallest amount
the lactate
unfrozen control tissue released LDR 2. and this increased
i.2
/--i l
T t ' 0.8 ,t· t· I 1 l
0.6.
0.4 ' DMSO EG . PROH GLY
-PAOH 0.2 · · · · · DMSO/EG/Gl Y
10 20 30 40 50 60 70 80 90
Tirne (minutes}
Figure 2. The rate of of sulphoxide (DMSO), ethylene (EO), glycol (PROH) and glycerol (GLY)
human ovarian tissue at 3rC. Plotted lines represent the best fit curve of the form y = 1 ~a.exp ( ~bt) ~ (1 ~a) exp ( ~ct) where y is and t is time. Each point represents the mean and SD from four samples.
25
20
15
!f " .~
~ 10
1.6M DMSO 1.5M DMSO 1.5M DMSO 1.5M DMSO Leibovitz Fresh 0.05M sucrose 0.1 M sucrose 0.25M sucrose
Figure 3. The amount of lactate dehydrogenase (LDH) released from human ovarian tissue in one of five different freezing media: L5 M DMSO, 1.5 M DMS0/0.05 IV! sucrose, 1.5 M DMSO/O.l M sucrose, 1.5 M DMS0/0.25 M sucrose and Leibovitz-L15 medium alone, to fresh unfrozen control tissue; n = 23~24 per group.
Table t Parameter estimates
a
DMSO and EG at 4oc 0.7 (0.04) PROH and GLY at 4oC 0.7 (0.04) DMSO, EG and GLY at 3rC 0.78 (0.03) PROH at 37oC
Values in parentheses are SE.
b
0.28 (0.05) 0.19 (0.03) 0.4 (0.07) 0.27 (0.05)
a = proportion of penetTation process over time. b and c = penetration rates over tiine.
c
0.012 (0.004) 0.006 (0.0039) 0.0048 (0.0027) 0.27 (0.05)
DMSO = dimethyl sulphoxide; EG = ethylene glycol; PROH = propylene glycol; GLY = glycerol.
to a mean value of 20.8 ± 2.4 tissue was frozen~thawed in media without a
< The release decreased signific-when 1.5 M DMSO was
included in the medium The further addition of either 0.05 M and l M sucrose decreased LDH release to I 1.8 ± 1.4 and 9.8 ± 1.2 however,
Downloaded from https://academic.oup.com/humrep/article-abstract/13/2/376/876041by gueston 09 April 2018
0.25 M sucrose caused a greater LDH release of 14.6 :±: 1.5 IU/g than freezing in 1.5 M DMSO alone but none of these differences was even though the group sizes were = 0.14, from of variance).
Discussion
In the present experiment we investigated the rates of penetration of four different cryoprotectants and found that at 4oC
glycol and permeate human ovarian tissue more efficiently than either propylene glycol or glycerol. Thomas et al. have recently reported results for the transport ofDMSO at 0-2°C into 3-5 mm thick porcine and human ovarian tissue slices. It was observed that DMSO reached 45% equilibration in 20 min, which compares with about 76% at that time in the 4oc group in our study. This discrepancy most probably is due to the greater thickness and lower temperature used in their investigation. In our previous
of ovarian cryopreservation we found that glycol and dimethyl sulphoxide supported the survival of 84% and 74% of the follicular population after freeze-thawing, whereas 44% and 10% survived in tissue frozen in propylene glycol and glycerol respectively. Thus it would appear that the extent of follicular survival is at least in part determined the speed of CPA permeation and the final tissue concentration achieved. This theory is supported by the work of Candy et al. who murine ovarian tissue to different cryoprotectants for defined periods of time at room temperature before cryopreservation. Histological examination of the tissue on revealed that the dial follicle population was frozen in but that the numbers increased the longer the tissue was to the CPA. Primordial follicle counts taken from the tissue frozen in DMSO, EG and PROH did not differ significantly and did not increase over time from 5-60 min. These results probably reflect the fact that murine ovarian stroma is less dense than that of the human and at room temperature total equilibration is reached in this tissue in less than 5 min.
The effect of temperature on the speed of is also demonstrated in the cunent where the initial rate of of all four CPAs was found to be more rapid at 3rc than at 4 oc. there was little effect of temper-ature on the tissue concentrations of GLY or EG achieved at 30 min. Propylene however, to exhibit a greater dependence but it must be noted that it also displayed the greatest level of variation between runs. In practice the advantage of more rapid diffusion at 37oC would have to be balanced against the probability of toxicity, and this was not in the present study.
The benefit of adding a such as sucrose or mannitol has been demonstrated in both single cells and embryos et al., One suggested mechanism for the beneficial effect may be that they act as an osmotic buffer against the stresses incurred the cells the addition and removal of penetrating
agents. The results of the present study suggest
CryoiJreservation or ovarian tissue
that the addition of low concentrations of sucrose to the media does not
following slight decrease in the release of LDH the addition of 0.1 M or 0.05 M sucrose and thus it n1ay be n'"""'"m
media as an added
to sterilized hosts the ovarian tissue at the s1te
was first in the was frozen and then thawed in a solution insertion the endocrine function and some gave birth to healthy litters of pups
of lmnn~vt~ments in the for many young
cancer patients have altered this situation and there is now increased interest in tissue as a means of
induced automated freezers and more efficient agents have led to follicular survival rates as has been mouse and marmoset ovarian tissue et These can be
the fact that the with ovarian tissue
to 6/l et The human ovary is more fibrous than that of small
animals and ethical and have tended to the study of follicular survival after and auto-grafting. As an model was chosen because its ovaries are to those of the human. When were returned to the animals gave birth encouraging results from the ovine extend the work to human tissue.
showed that
The
present within the factor to determine follicular survival to be the CPA used
et al., In the absence of any CPA the were found to be minute and sterile
Human ovarian tissue has also been incubated in an in-vitro organ culture for up to 15 after cryopreserv-ation. examination of the tissue shovved
two thirds of the follicles had survived freeze-and extended culture et al.,
loss from ovarian tissue the u-.v6 .• u~"" may be due to the formation of intracellular ice or to the increase in the extracellular salt concentration that
Meryman, CPA cells the amount of ice formed at !.ow temperatures as a
379 Downloaded from https://academic.oup.com/humrep/article-abstract/13/2/376/876041by gueston 09 April 2018
H.Newton et al.
result of their colligative properties, thus decreasing the external salt concentration. To date, the protocol used for the cryopreservation of ovarian tissue has been based on the methods used for mature oocytes. Since tissues have the added problems of cellular heterogeneity and the rates of diffusion of CPA in and out (Pegg et al., 1979), it was anticipated that whole ovarian tissue would require changes in technique, including more time for cryoprotectant penetration into the cells and for total washout from the tissue after thawing.
In the present study we have investigated the penetration rates of the four CPAs DMSO, EG, PROH and GLY at both 4oc and 37°C. We decided to use 4oc in order to minimize toxicity and at this temperature we have shown that bathing the tissue for 30 min in a 1.5 M solution of either EG or DMSO produces mean tissue CPA concentrations approaching 80% of that in the bathing medium. In addition we have also found that no statistically significant protection is offered by adding low concentrations of sucrose to the freezing media.
Acknowledgements We gratefully acknowledge the help of Ken Newton with the LDH assays and the Leukaemia Research Fund (London) for supporting this work. J.R.P.A. also acknowledges personal support from the Wellcome Trust.
References Abraham, R.J., Fisher, J. and Loftus, P. (eds) (1988) Introduction to NMR
Spectroscopy. Wiley, Chichester. Apperley, J.F. and Reddy, N. (1995) Mechanisms and management of treatment
related gonadal failure in recipients of high dose chemo-radiotherapy. Blood Rev., 9, 93-116.
Bateson, E.A.J., Busza, A.L., Pegg, D.E. and Taylor, M.J. (1994) Permeation of rabbit common carotid arteries with dimethyl sulfoxide. Cryobiology, 31, 393-397.
Candy, C.J., Wood, M.J. and Whittingham, D.G. (1995) Follicular development in cryopreserved marmoset ovarian tissue after transplantation. Hum. Reprod., 9, 1738-1742.
Candy, C.J., Wood, M.J. and Whittingham, D.G. (1997) Effect of cryoprotectants on the survival of follicles in frozen mouse ovaries. J. Reprod. Fertil., 110, 11-19.
Clark, P., Fahy, G.M. and Karow, A.M. (1984) Factors influencing renal cryopreservation. II. Toxic effects of three cryoprotectants in combination with three vehicle solutions in non frozen rabbit cortical slices. Cryobiology, 21, 274-284.
Faddy, M.J. and Gosden, R.G. (1995) A mathematical model for follicular dynamics in human ovaries. Hum. Reprod., 10, 770-775.
Gosden, R.G., Baird, D.T., Wade, J.C. and Webb, R. (1994) Restoration of fertility to oophorectomised sheep by ovarian autografts stored at -196°C. Hum. Reprod., 9, 597-603.
Gunasena, K.T., Villines, P.M., Critser, E.S. and Critser, J.K. (1997) Live births after autologous transplant of cryopreserved mouse ovaries. Hum. Reprod., 12, 101-106.
Harp, R., Leibach, J., Black, J. et al. (1994) Cryopreservation of murine ovarian tissue. Cryobiology 31, 336-343.
Hovatta, 0., Silye, R., Krausz, T. et al. (1996) Cryopreservation of human ovarian tissue using dimethyl sulphoxide and propanediol-sucrose as cryoprotectants. Hum. Reprod., 11, 1268-1272.
Hovatta, 0., Silye, R., Abir, R. et al. (1997) Extracellular matrix improves survival of both primordial and primary ovarian follicles in long-term culture. Hum. Reprod., 12, 1032-1036.
Jacobsen, I.A., Pegg, D.E., Starklint, H. et al. (1984) Effect of cooling and warming rate on glycerolized rabbit kidneys. Cryobiology, 21, 637-653.
Karow, A.M., Wiggins, S., Carrier, G.O. et al. (1979) Functional preservation of the mammalian kidney. J. Surg. Res., 27, 93-99.
Lovelock, J.E. (1953) The haemolysis of human red blood cells by freezing and thawing. Biochem. Biophys. Acta, 10, 414--426.
380
Mandelbaum, J., Junca, A.M., Plachot, M. et al. (1988) Cryopreservation of human embryos and oocytes. Hum. Reprod., 3, 117-119.
Mazur, P. (1963) Kinetics of water loss from cells at subzero temperatures and the likelihood of intracellular freezing. J. Gen. Phys., 47, 347-369.
Meryman, H.T. (1968) Modified model for the mechanism of freezing injury in erythrocytes. Nature, 218, 333-336.
Meryman, H.T. (1971) Cryoprotective agents. Cryobiology, 8, 173-183. Nash, T. (1962) The chemical constitution of compounds which protect
erythrocytes against freezing damage. J. Gen. Physiol., 46, 167-175. Newton, H., Aubard, Y., Rutherford, A. et al. (1996) Low temperature storage
and grafting of human ovarian tissue. Hum. Reprod., 11, 1487-1491. Oktay, K., Nugent, D., Newton, H. et al. (1997) Isolation and characterisation
of primordial follicles from fresh and cryopreserved human ovarian tissue. Fertil. Steril., 67, 481-486.
Parrott, D .M. V. (1960) The fertility of mice with orthotopic ovarian grafts derived from frozen tissue. J. Reprod. Fertil., 1, 230-241.
Pegg, D.E. and Diaper, M.P. (1988) On the mechanism of injury to slowly frozen erythrocytes. Biophys. ]., 54, 471-488.
Pegg, D.E., Jacobsen, I.A., Armitage, W.J. and Taylor, M.J. (1979) Mechanisms of cryoinjury in organs. In: Pegg, D.E. and Jacobsen, LA. (eds), Organ Preservation 2. Churchill Livingstone, Edinburgh, pp. 132-144.
Sanders, J., Hawley, J., Levy, W. et al. (1996) Pregnancies following highdose cyclophosphamide with or without high-dose busulphan or total body irradiation and bone marrow transplantation. Blood, 87, 3045-3052.
Taylor, M.J. and Busza, A.L. (1992) A convenient, non-invasive method for measuring the kinetics of permeation of dimethyl sulphoxide into isolated corneas using NMR spectroscopy. Cryoletters, 13, 273-282.
Thomas, N., Busza, A., Cooper, A. et al. (1997) Measurement of permeating levels of cryoprotectant during ovarian tissue cryopreservation using 1 H NMR spectroscopy in human and porcine ovaries. Cryoletters, 18, 179-184.
Wusteman, M., Busza, A., Boylan, S. et al. (1995) Ethylene glycol permeation and toxicity in the rabbit common carotid artery. Cryobiology, 32, 428-435.
Received on July 3, 1997; accepted on October 20, 1997
Downloaded from https://academic.oup.com/humrep/article-abstract/13/2/376/876041by gueston 09 April 2018